0% found this document useful (0 votes)

64 views360 pages

A Natural History of Seeing - The Art and Science of Vision - Simon Ings - 2008 - W. W. Norton & Company - 9780393067194 - Anna's Archive

Uploaded by

smdhlybzx

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

64 views360 pages

A Natural History of Seeing - The Art and Science of Vision - Simon Ings - 2008 - W. W. Norton & Company - 9780393067194 - Anna's Archive

Uploaded by

smdhlybzx

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 360

4 ¢ |

~ : ; ~~
:
:
i | , | ven a
te es

A NATURAL
nS TORY Or
ly SEEING
We spend about one-tenth of our
wdking hours completely blind.

Only one percent of what we see

is in focus at any one time.

The blind can be taught to see

through their chests.

n graceful, accessible prose, novelist and sci-

ence writer Simon Ings sets out to solve these
“

and other mysteries of seeing, taking us

through the 600-million-year history of the
eye. With the help of a beguiling mix of illustrated
optical illusions and puzzles, anecdotes, mathemat-
ics, and philosophy, Ings reveals age-old mysteries
from how humans perceive color to Woody Allen’ v7)

ability to raise the inner corners of his eyebrows.

A Natural History of Seeing delves into both th oO

evolution of sight and the evolution of our under-

standing of sight. It gives us the natural science—
the physics of light and the biology of animals and
humans alike—while also addressing Leonardo da
Vinci’s theories of perception in painting and Homer’s
confused and strangely limited sense of color. Pan-
oramic in every sense, it reaches back to the first
seers (and to ancient beliefs that vision is the product
of mysterious optic rays) and forward to the promise
of modern experiments in making robots that see.
NO LONGER
PROPERTY oF PPLp
A NATURAL HISTORY
OF SEEING
ALSO BY SIMON INGS

The Weight of Numbers

Hot-Head
Painkillers
Headlong
Hotwire
City of the Iron Fish
AN AORAck
HISTORY OF
SEEING
The Art and Science of Vision

SIMON INGS

DT
W. W. Norton & Company
New York »« London
Copyright © 2007 by Simon Ings
First American Edition 2008

First published in Great Britain under the title The Eye: A Natural History

All rights reserved

Printed in the United States of America

For information about permission to reproduce selections from this book,

write to Permissions, W. W. Norton & Company, Inc.,
500 Fifth Avenue, New York, NY 10110

For information about special discounts for bulk purchases, please contact
W. W. Norton Special Sales at specialsales®@wwnorton.com or 800-233-4830

Manufacturing by Courier Westford

Production manager: Julia Druskin

Library of Congress Cataloging-in-Publication Data

Ings, Simon.
A natural history of seeing : the art and science of vision /
Simon Ings.—1st American ed.
p- cm.
“First published in Great Britain under the title The Eye: A Natural History.”
Includes bibliographical references and index.
ISBN 978-0-393-06719-4 (hardcover)
1, Vision. 2. Eye. I. Title.
QP475.5.S558 2008
612.8'4—dc22
2008025924

W. W. Norton & Company, Inc.

500 Fifth Avenue, New York, N.Y. 10110
www.wwnorton.com

W. W. Norton & Company Ltd.

Castle House, 75/76 Wells Street, London W1T 3QT

I2i3'
14. 5 627418 940
For Natalie

— new eyes for old

_,. es ay “ si]

‘ : A i. 4 z

Swe Gey

rw iar ite _ 6 Rd fel

sla sayPicea oe Scat ™

' i eae aed eee ht = 7 Pia

hi ats rf _ : 7

} Aigaay si oh 6dir a ni
<= veh ys sy# fe me:a ache ies
ae . - a . , »

es owe 5 fee pan ON

ea a
ae S,
eee Pi
oa 6 :

= es i ; e : ne b> eee

chew tne F LF D date

Bits . SF 6Ree Te
ren We j _ninesey 0 5 ital Pt sats
Rin ~~ ———— tigi~ fe Neem
-- vbife,:
im speaun i nea enaeyoath tet
; irs fracas 4

1 ed areca Pay
CONTENTS

Acknowledgements
PROLOGUE - Youth and age
] The commonwealth of the senses 14
Z The chemistry of vision 52

3 How are eyes possible? ye

4 The adaptable eye 102

) Seeing and thinking 128

6 Theories of vision 154

7 Nervous matter, visually endowed 181

8 Seeing colours 2

2 Unseen colours 244

10 Making eyes to see 264

EPILOGUE - The invisible gorilla 286

a sitet a wie0 wt gitfeo inetparihala tear se ft

Sit! tuSsih cole Ses aauery wi avwt woh Bos errno
abikee +dood ollli shied see! Sarthe he RNa aL HS SERENA HD
Ad girl dost! ob wlbowe eet samaritan Inet x Bk
nr Xesated af So sfsfi pre). Tete ates F an pike 7
treed nel wit hes costed rel! adel ee phammesh coke .
| ensratines seine tans Sse yb eg a) sey ety ined ean ON
Protege eet sept aes Mave Vie dre dl os ahh bide 4 snot “aah
‘eae penn? aD UMA are altbasco pniyor Voi bes gonttllye 4 eat tg
nips baie tessa takey z2ykrwos awk t
hn we pravied evr
fas | ater pri Hd ik Mind poRay Bloup OF eam
wedged 4 veh corti os Vue to eA trent) gate twi
angi waaay) oo) Mapa > Vd A AAR
wn,
taow
Yboos bats 8 tindth cncnt
sola OPIN, sO
s (i ess

2 pt at telorig le ica
Sy eatihte fi 1S 9.Batthstk Aci a arated, Gey eieg 9
et)
~r dhavee} wey SA Megat To peed = yee pe
Boo} Sy yieeaiod and fag £88} agi fsPAL Be ABBES
epee 0Top eo estan eg!fo adit sttbamteenedl HI ie ove
| md ? Yel te Enel 1 ACY ads RAR t
ah nee hice kt i
he 27 Ae Se '.) © ee sein ete
ah grior
trig) Doe OBL £2T ph USS Hes POP ea tea oadoe
26 1) Sage sd pyar treme see,
dauleu'met od. eachaiiiesg a9) ae veonth esBak
gat che go belie: pepe pat
nedes Raper At a i mt r Kare lete

2a ha
gl ie
gapeene a
oP at itaSede
e
fine
Ii a t
Bac ve i gee: it a,

i)eae CeAte ell

il ag
nt
o ae
Parra treat ta(LOOTop
ree nil Bliads xt
peg orsien tye
i
A NATURAL HISTORY
ORS Bel NG
fHOTEL win
rrr io 4 +

Ot “e a en a2
: EB
M PI 4

“a “tes

ri
b ¥
a an r :

ial

i 7

*
PROLOGUE

Youth and age

Natalie, my daughter, began life with one eye orbit placed centrally
in her forehead. It appeared early — barely a week after her conception
— in October 2002 and had things gone awry, there it would have
remained, a wet cyclopean hollow, glowering in the shadow of a
grotesque proboscis that would have grown in place of her nose.*
There are many explanations for what inspired the legend of the
Cyclopes - the one-eyed monsters of classical mythology — invoking
everything from an ancient find of mysterious dwarf elephant skulls,
to the blacksmith’s habit, in the days before protective goggles, of
protecting one eye with a patch. But the most likely inspiration is
also the saddest; occasionally, a human cyclops survives in the womb
long enough to emerge visibly disfigured. Once upon a time, someone
got a good look at what they had aborted - and for ever after wished
they hadn't.
This happened not so long ago. George Gould and Walter Pyle's
indispensable volume Anomalies and Curiosities of Medicine of 1896
includes an account of a woman of thirty-five, the mother of seven
children, ‘who while pregnant was feeding some rabbits, when one
of the animals jumped at her with its eyes “glaring” upon her, causing
a sudden fright:* The child’s mouth and face were small and rabbit-
2 A NATURAL HISTORY OF SEEING

shaped. Instead of a nose, it had ‘a fleshy growth 3/4 inch long by

1/4 inch broad, directed upward at an angle of 45 degrees. Between
the nose and the mouth was an organ resembling an adult eye. Within
it lay ‘two small, imperfect eyes which moved freely while life lasted’
The disfigurement of the face was the outward mark of an
even more colossal failure; the failure of the brain to divide into two
delicately linked halves. So, mercifully, after about ten minutes the
child died. The mother went on to have two more, perfectly
healthy, children.
Human beings arise, not from dust, but from stuff hardly more
edifying: a gelid spittle. How, from this unpromising material,
anything beautiful emerges - let alone anything that comprehends
and communicates something of the world beyond itself - is a
mystery too big to be encompassed by just one version of events.
How we explain how we grow depends largely on fashion. The current
fashion is to talk of rules embedded, like a code, in every cell. But
this, like most shorthand explanations, fosters misunderstanding.
Natalie was not fashioned by rules. Nothing oversaw her growth.
Genes set the conditions of the game, but the game itself built her:
the touch and slide of surfaces, the little chemical kisses, the partings
and the reunions, each part of her tissue communicating chemically
with each neighbouring part as she folded herself into being. This
dance, known to biology as induction, told her brain to split, squeezed
her single orbit into two, and drew two little strings of brain out to
the windows in her skull to make her eyes.
When a newspaper announces that someone has found the gene ‘for’
something or other, the sentiment is not so much wrong as insufficient.
Every gene is ‘for’ a virtually infinite number of things, as vital to the
whole — and as meaningless in itself - as a single key on a piano
keyboard. It is the performance that matters. The chances that a woman
grows a cyclops in her womb because of a genetic defect are very slim.
It is much more likely that the performance was interrupted — that, for
instance, she ate cow cabbage (Veratrum californicum, a notorious
poisoner of hungry sheep and cattle) early in her pregnancy.
Anna, my wife, has never been so hungry that she has been tempted
to eat cow cabbage but necessity did lead us to throw every toxin
known to the construction trade at little Natalie. The beautiful
YOUTH AND AGE 3

ramshackle house we moved into, six months before her birth, chose
the moment of our arrival to give up the will to live. Faced with bust
plumbing and falling ceilings, we had a stark choice: either we shored
the place up, then and there, or we abandoned it. There was no
middle way. So, we knocked down walls and filled the air with horse-
hair and lime. We sawed up MDF panels without wearing masks.
" (In the United States they insist you wear protection before handling
the stuff.) We stripped the lead-based paint from doors, with acid
paste. We sanded boards and varnished them with polyurethane. We
sprayed for woodworm, cockroaches, and larger vermin. With her
incoming nourishment filtered clean by her mother’s placenta, Natalie
was not in the least troubled, and by November her eyes had acquired
all their basic anatomy.
There is light in the womb - enough to need managing. Natalie's
irises — plates of coloured muscle that respond readily to light —- were
still very young, so her eyelids grew and fused together, shielding
her eyes for a while. The following May, with the irises all but fully
grown, the eyelids parted, and Natalie blinked.
By April, the centre of each iris had withered, leaving a little hole
or ‘pupil’ for her to see through. There were still three months to go
before her birth but Natalie could already see.

The first ever recognisably scientific study of human anatomy was the
hs choroid
conjunctiva sclera

suspensory
ligament retina

vitreous humour

ine blood vessels

\optic nerve

The human eye.

4 A NATURAL HISTORY OF SEEING

work of the Greek physician Herophilus (335-280 sc). Herophilus is

credited with one of the first formulations of the scientific method,
and is said to have introduced experiment into medicine: he is reputed
to have vivisected six hundred prisoners in his efforts to come to grips
with the workings of the human body. This vile smear, put about five
hundred years after his death by Christianity’s first hack writer,
Tertullian, is as good a measure as any of the revolutionary impact
made upon our thinking by the medical school of Alexandria.
Herophilus co-founded this school with Erasistratus, and half a millen-
nium later he was still being pilloried as a dangerous iconoclast.
Herophilus’s particular interests were the nervous system and the
eye, and it is probably his work, adopted by Imperial Rome’s most
celebrated anatomist, Claudius Galenus of Pergamum (Galen to English
speakers), that gifted us words like retina, conjunctiva, and cornea.
In describing the eye, Galen (taking his lead from the Alexandrian
anatomists) used an excellent analogy: the layers of the eyeball were,
he said, like layers of clothing - tunicae or tunics. Working inwards, he
described the eyeball just as you might describe what someone is
wearing. The analogy is a good one, because no layer of clothing is
shaped exactly like another, nor is it necessarily made of the same stuff.
My jacket covers my shirt, but not completely — the first couple of shirt
buttons are visible, and my shirt cuffs peep below the jacket’s sleeves.
It is also made of different stuff; it looks different, and hangs differently.
Galen’s analogy is casual, and when he talks about one tunic of the
eye ‘surrounding’ another, we need not take him too literally. The eye
is not a geometrical figure, sealed off from the rest of the body. Like
everything else, it has its exits and its entrances, its folds and frills.
The outermost layer - the white of the eye — is the sclera. It’s a
suggestive choice for English speakers, for whom ‘sclerotic’ means
‘hardened and inflexible’. When Galen’s words were latinized (he
wrote in Greek), new words occasionally arose. Consolidativa was
one, a long-winded alternative to sclera — but it failed to catch on.
Almost every sclera in the animal world is pigmented. In humans
it is white — bluish-white in Natalie’s case, being so thin; yellowish
in my mother’s, as age thickens it year by year. The sclera is made
of collagen — a rather fatuous statement, considering the aston-
ishing variety of forms that collagen can take (four-tenths of our
YOUTH AND AGE 5
body proteins are collagen of one sort or another). Collagen makes
the white of the eye, but it also makes the transparent cornea, the
artist’s ‘window on the soul’, through which we see. The difference
lies in the construction. The cornea, like the white, is solid collagen,
but in it the fibres are laid down with such fine geometrical care
that it admits light.
"Within the sclera lies a middle tunic, the uvea, originally an adjec-
tive meaning grape-like. Open an eyeball out at the pupil, and you
will find that this is one of the most telling and poetic of all anatom-
ical descriptions. The whole inside structure of the eye is veined with
translucent blood vessels, strongly reminiscent of the vessels running
through a peeled grape. Also, the tissue immediately around the
pupil, which feeds and operates the iris, the ring of brightly pigmented
muscle, has a clustered appearance, like a bunch of grapes. Galen
broke the uvea down into several parts, the better to explain what
he saw. One blood-filled, semi-translucent layer, which runs across
the back of the eye, looked to him like a ‘chorion’ - the membrane

The developing eye’s temporary blood supply (the hyaloid),

from Gordon Lynn Walls’s mammoth 1942 volume,
The Vertebrate Eye and Its Adaptive Radiation.
For more on Walls, see Chapter Nine.
6 A NATURAL HISTORY OF SEEING

sac which surrounds the foetus. He dubbed this layer the choroid,
and the name has stuck to this. day.
The choroid is full of blood, feeding the tissues of the eye (much of
the eye consists of tiny blood vessels) and we now know what Galen
could not — that in the foetus, the choroid gives rise to all the other
structures of the uvea. As her months in the uterus passed, the choroids
of Natalie’s eyes waxed and waned. Some became the pretty circles
around each pupil, rings of muscle called irises (meaning ‘rainbows,
another of Galen’s happy neologisms). As it dilates and contracts, the
iris focuses the view, guards against too strong a light, and even signals
feeling, as it widens the pupil to accommodate a loved ones face. Some
of my daughter’s choroid became the muscle that helps her focus,
stretching her crystalline lenses to adjust their power. Some turned
black, coating the back of the eye — rather like the backing-plate of a
camera — with a heavily pigmented layer, one cell thick.
By May 2003, the blood vessels of Natalie’s choroid layer had with-
drawn from her lenses. These no longer required a blood supply, for
they were (but for a single layer, one cell thick) quite dead. Isolated
from neighbouring tissues only five weeks into gestation, afloat in a
watery suspension, the lens is a tissue uniquely isolated from the
rest of the body - an exquisitely arranged meshwork of dead cells
packed with clear proteins called crystallins which, though soluble,
remain virtually unchanged throughout life.
The choroid, as it withdrew from Natalie's lenses, now folded itself
back over the eye's ‘backing plate to form the third and final of
Galen's tunics - the retina. “Retina means mesh or net, and when
Galen used the word, he was referring to the fine layer of blood
vessels lying across the ‘backing plate’ of the eye. For him, the retina
was just a thin, stretched extension of the uvea. Today, we use the
word in quite a different sense, to describe the most mysterious of
the eye's tunics: the layer that sees. It was a layer of which Galen
knew nothing: it was quite invisible to him, and remained undetected
until 1834. To call it a ‘retina’ is, strictly speaking, a misnomer: the
modern ‘retina’ is not at all what Galen was writing about, and is
not remotely like a net. It is a cup of transparent nerve tissue, literally
an outpost of the brain.
+ * *
YOUTH AND AGE 7
Developing nerve cells shake hands with each other indiscriminately.
Their connections acquire a shape and an architecture only when they
are stimulated: the busiest are preserved, while the less travelled lines
wither.
Natalie's retinal cells were twice taught how best to connect: first
_ by dreams, and then by light.
By April, Natalie had begun, now and again, to sleep. Every so
often she closed her eyelids and kept them shut, and when she did,
her eyes trembled, stirring the oxygen-rich liquid, the aqueous
humour, that lies between the iris and the cornea. (The occasional
rapid eye movements that accompany our dreams have at least
one very practical purpose: they feed the front of the eye.) The
trembling woke the light-sensitive cells of Natalie's retinas. Stimulated,
her retinal cells fired at random, preserving and strengthening their
connections.* Even before she saw, Natalie went through the motions
of dreaming, and those motions taught the cells in her retina to hold
hands.
Prepared by dreams, Natalie's retinal cells took their next lessons
from light. Even before birth, the body is no stranger to illumination.
Flesh itself lights up a little, every time a nerve fires.» Perhaps this
familiarity with light is why any young nerve cell, transported to the
retina — the body’s most light-sensitive surface - will learn to see,
just as every seeing nerve cell, moved elsewhere, adapts to its new
environment.®
The womb is not dark: it is easily penetrated by light from the
outside. From November 2002 to July 2003, the month of her birth,
Natalie’s retinas seized what news they could through the amniotic
murk. They adapted to darkness and to blur. One layer of nerves grew
into highly light-sensitive cells, rods, superb harvesters of light. At birth,
Natalie was well on the way to acquiring good nocturnal vision.
Babies see well at night, but in the glare of day, they are all but
blind. A sunny day is a million times brighter than a night-time
nursery, and a completly different form of vision is needed to handle
such a glare - a form Natalie had not yet acquired. (It is one of
the ironies of birth that it fills our world with light and blinds us
in the process.)
Exposed to the light, Natalie’s eyes had not so much to ‘adapt’ to
8 A NATURAL HISTORY OF SEEING

the brightness of day, as acquire a new way of seeing. In her retinas

lay another set of nerve cells, cones. Cones, when they are fully grown,
look very different to rods. Indeed, these two kinds of photo-sensitive
cell are only distantly related; the distinction between them had
arisen in the eyes of jawless fish swimming in Devonian seas around
400 million years ago.” At birth, Natalie’s cones were little different
from ordinary nerve cells, but once exposed to the glare of day they
began to change. Natalie will be six before her cones are fully grown,
packed tightly into the fovea, the tiny circle on the retina where
images are focused and light explodes with colour.

This first year of my daughter’s life is visually simple. A few familiar

faces. Eyes and smiles. The rest is a blur, slowly resolving as her
lenses — dead and round as the lenses of the fish that were her distant
ancestors — acquire the slimline shape proper to a land mammal.
Though dead, lenses grow: layers of protein are laid on like the skins
of an onion, changing their profile and their power. It is a process
as natural and unstoppable as the growth of hair and nails. Even
now, as my daughter's lenses take on their optimum shape, they are
growing thicker, less permeable, and imperceptibly more stiff. This
constant accretion of material will eventually spoil her sight, as it is
about to spoil mine and as it has, or will, spoil yours. The eye focuses
on objects at different distances by means of a ring of muscle, the
ciliary muscle, around the lens. As our lenses harden, the harder the
ciliary muscle has to work to wrench them to the right shape. The
decline in lens flexibility is smooth and gradual, and we rarely notice
it until the process begins to affect the ease with which we can read
print.
I look at Natalie through lenses that are forty years old. In a few
years, these words will be a blur to me; my lenses will not bend to
accommodate them. I will moan, crack some mordant joke about
my age, and go to the optician, who will sell me some glasses -
lenses hung on wires before the eyes, strangest of all nature’s strange
variations on the theme of vision. The top half of each lens will
allow me to see distant objects, and the lower part will help me
see close ones. Benjamin Franklin invented bifocals — glasses whose
lenses are, effectively, several lenses in one. This is one of those
YOUTH AND AGE 9
half-interesting factoids which bring the ageing spectacle-wearer
no comfort whatsoever.
Natalie may be short-sighted. If she is, it doesn’t matter. A third
of all infants are born myopic. The retina — which is to say, the brain
— has the problem in hand. It hungers for a clear view, and since it
directs the growth of the eyeball, it will make what corrections it
can. By the time she is six, we will know whether her aquatic ancestry
is so strong that it spoils her sight, leaving her lenses too round and
powerful to see the world with clarity. If it is, she will wear glasses
to correct her ‘short sight.

For the first couple of months after she was born, as she tried to
figure out how to look at the world, Natalie used to cross her eyes
a lot. Once she began to pick things up, her hands began to teach
her eyes how to see. Her eyes imagined there were two identical
blocks before her; her fingers found only one. Her eyes, vanquished,
uncrossed.
By ten months old, Natalie sees the world pretty much as it is. She
knows what is near, and what is far away. I see her thinking: “How do
I reach that? She stretches out her arm to me. I am sitting on the
other side of the room. She knows I am far away, but she doesn’t know
how far. She cannot crawl yet, so ‘far away’ to her simply means ‘out
of reach.
Seeing her reach out; I can't help but smile. She returns the smile
and reaches out again. I lean forward and reach out for her. We're
too far away to touch, but this is our game. We reach out, and smile,
and this will be a kind of touch. Mind will brush mind. Impossible
to imagine that she thinks any differently; that we are playing different
games; that she has no idea, yet, about mind; about what I am, or
what she is.
Anna comes into the room. Anna is my wife, Natalie's mother, the
woman who gave Natalie her wonderful grey eyes. Natalie turns to
her mother. Which is to say, she looks into her mother’s eyes; looks
to see where Anna’ eyes are directed. She follows Annas gaze. Anna
is looking at me. So Natalie turns back to me, smiles, and reaches
out.
We are playing with her. We are getting reactions and for us, these
10 A NATURAL HISTORY OF SEEING

reactions make a human kind of sense. But for Natalie, our game is
pure geometry. Things within reach and things out of reach. Limbs
draw lines through space; eyes follow the lines. For Natalie, this is
not a game about people; it is a game about space.
There is a lot for her to learn. Between a game of lines and a game
of minds there is a great mountain of knowledge to be scaled, and
if you look into her eyes you can see her, climbing furiously. She is
emerging, bit by bit, month after month, out of the mathematics of
rote existence and into the heady, complex spaces of language,
symbolic logic, and, at last, an understanding of others. In the begin-
ning my daughter’s eyes were blank: windows on to an unlit room.
These days, they shine. Each day, they react and express more. Most
of the time, they smile; when she cries, they shine with a terrible
grey light.

This is a book about the nature of the eye. It is about all the eyes
that are, and ever have been, and may yet be. It is about how human
eyes see the world, and how other eyes see it. It is about what happens
to the world when it is looked at, and about what happens to us
when we look at each other. It is about evolution, chemistry, optics,
colour, psychology, anthropology, and consciousness. It is about what
we know, and it is about how we came to know it. So this is also a
book about personal ambition, folly, failure, confusion, and language.
This has not been done before. There are books of zoology, books
about evolution, books of vision theory, and psychology, optics, medi-
cine, and biochemistry. There are atlases, and textbooks, privately
published labours of love, papers, bulletin boards, polemics. The
knowledge is there. The language in which it is expressed is some-
times exquisite. But no one, to date, has ever attempted to reach into
each and every one of these territories to account for the eye. No
one has put the eye at the centre of a sprawling and epic story.
I soon found out why. Like Laurence Sterne’s hapless autobiogra-
pher Tristram Shandy - whose life is lived more quickly than it can
be written down - I watched with growing horror as the pile of
unread research material grew, month after month, to dwarf the
material I had already mastered. My own daughter’s eyes are an
admonition: I have watched as she has acquired colour vision,
YOUTH AND AGE 11

empathy, complex attention, simple ball skills and even reading — all
in the time it has taken me to barely scratch the surface of these
subjects. And of course I had to range far beyond these human
matters to fulfil my brief - to tackle other eyes, and other ways of
seeing, and ask how all these eyes came to be, and wonder what they
might become.
How to organise such diverse material? Two solutions presented
themselves - as obvious as they were useless. I could describe how
eyes came to be, tracing their evolution and development from their
biochemical origins to their modern forms. And how dull that would
have been: to plunge my reader-victims first into the toils of organic
chemistry, then into the niceties of evolutionary interpretation,
without a word said about why the story of the eye was at all worth-
while or interesting. This would hardly be a story at all, because it
would have to fan out, as the eye has fanned out over evolutionary
time, into myriad forms, countless small tales of variation and inno-
vation. It would have resembled a failed experiment in hypertext
fiction — lots of promise and great openings, but no satisfaction, no
ending, no shape.
The obvious (and equally wrong) alternative was to write a ruth-
lessly simplified history of how certain men and women came to
understand the eye. The trouble is, the eye is not one story. It is many.
To preserve the chronology, every paragraph of the book would have
had to begin with ‘Meanwhile . . ? or ‘Coincidentally . . ’
My solution is, I concede, a little strange. It is a reflection of my
own journey through the literature of the eye - a tale of wonder,
and confusion, and flashes of understanding.
Chapter One stakes out the territory. Can a single account, however
casual and anecdotal, really hold such a vast subject together? Why
would we want to relate the four-eyed world of the fish Anableps
anableps to the meaningful glances of the higher primates; or the micro-
scopic eyespots of algae to the hallucinations artificially generated in
the brains of human volunteers by the fields of a trans-cranial magnetic
stimulator? The answer is in the chapter title: vision belongs to a
‘commonwealth of the senses. It is not a unique and privileged sense,
but an exquisite manifestation of a drive common to all living things
— vegetable as well as animal — to harness light. In Chapters Two to
12 A NATURAL HISTORY OF SEEING

Five, I trace the evolution of that commonwealth of the senses, and

describe how different eyes came to fulfil different survival roles for
different animals. The story is told episodically, from the point of
view of the men and women who uncovered the eye's long history,
and ends with an account of the oddities of our own human eyes.
Could it be that thinking arises as a response to seeing? Without
eyes, would minds exist at all? Since written records began, philoso-
phers have wrestled to equate the mechanics of the eye with the
experience of vision. As our understanding of the eye has deepened,
so their writings have shaped our investigations, helping and
hindering in equal measure. Chapter Six looks at our theories of
vision — again, through the lives and words of the people doing the
theorising — while the rest of the book, except the last chapter, looks
at how the problematic partnership of theory and discovery has
tackled one of the more intractable problems of vision: understanding
the retina.
Chapter Ten brings us up to date, by considering the future of
vision. What does it mean to build an eye? At what point do we say
that a machine genuinely sees? What extra senses might we give our
own eyes and one day, will our eyes do more than see?
If this mix of history, science and anecdote has resulted in a
mongrel of a book = well, so much the better. It is not meant to be
a textbook, nor a history in the usual sense (a quick glance at the
index will reveal, to the informed eye, some truly outrageous omis-
sions). I have tried to give the reader a glimpse of what vision feels
like as a subject of study and wonder; to give the artists and scien-
tists, merchants and priests, adventurers and dilettantes who stum-
bled upon, and over, and through these mysteries, a voice. Wherever
possible, I have let them speak in their own words.

I have dedicated this book to my daughter. Her eyes inspired it, and
I hope that when she is older, she will glance through it, and tell me
what she thinks of it. By then, she and I will be looking at each other
through eyes that perform equally well.
But very soon after, the tables will turn.
The fat in the orbits of my eyes will shrink. The whites of my eyes
will yellow. Probably my irises will acquire a curious, milky blue ring,
YOUTH AND AGE 13

an arcus senilis made of cholesterol - nothing to worry about, not

a sign of any imbalance, ‘just an age thing. My brown irises will
bleach, and if I live long enough they may even return to the baby-
blue of my infancy.
Already, the gel that filled the back of my eyes at birth is partly
liquid now - a process that began when I was three. With its deli-
cate, ill-understood structure melted quite away — ‘vitreous humour
indeed! — the little flecks of collagen gathering inside my eyes are
free to float hither and thither, like tiny spiders crawling across the
panes of my pupils, a nuisance that will never go away.
If I reach seventy, there is an even chance that my lenses, exposed
to light for so many years, will grow milky with cataract; a one in
twenty chance that my retinal pigment epithelium, the retina’s by
then exhausted gardener, will fail; that blood vessels — seeking, in
their clumsy way, to repair the damage - will worm their way into
the retina, throwing everything awry, distorting what I see, and
robbing me of sight entirely, from the centre out. (This macular
degeneration is what I really fear. It is the largest cause of blindness
for people over fifty-five; I used to smoke, and that makes me espe-
cially prone.) And there is a one in forty chance that my eyes, held
in shape by the pressure of the liquid inside them, will clog and fill
and swell, damaging my optic nerve; glaucoma: a treatable condi-
tion but often, in the elderly, relentless.
Maybe none of this will happen. (There will probably not be time:
neither side of my family is particularly long-lived, and there is no
reason why I should be an exception.) How well will I see my
daughter, when she is the age I am now? The answer is worth remem-
bering, a ray of light to pierce this gloomy talk of age and illness.
Age withers us. But our eyes - so delicate, so fragile — resist. Even
as the rods in my retinas start to die, nerves will regroup to make
the most of their failing information source. At eighty, my eyes will
still see; if not with an adult’s clarity, then at least with a child’s more
blocky, pixellated vision.
If I live to a good age, I will see my daughter for the last time as
well as a three-year-old sees — as well as my daughter sees me now.®
ONE

The commonwealth of the senses

Helen Adams Keller, daughter of a former captain in the Confed-

erate Army, was born in 1880 in Tuscumbia, a small town in north-
west Alabama. At nineteen months old, she came down with a fever
and fell into a coma. When she awoke, she was blind. Her hearing
deteriorated soon afterwards. What her illness was no one knows
for sure, though meningitis seems likely.
Thanks to her own efforts and the extraordinary dedication of her
tutor, Anne Sullivan, herself blind, Helen Keller went to college, wrote
about a dozen books, met a dozen US presidents and travelled the
world. She also found time to develop a lively appreciation of the
plastic arts. ‘I sometimes wonder if the hand is not more sensitive to
the beauties of sculpture than the eye; she wrote. ‘I should think the
wonderful rhythmical flow of lines and curves could be more subtly
felt than seen. Be this as it may, I know that I can feel the heart-throbs
of the ancient Greeks in their marble gods and goddesses."
In what sense was Helen Keller not seeing the statue?

If you ignore a wasp, it wont sting you. I know this sounds like
an old wives’ tale, but for Natalie’s aunt Florence - a precocious
ten-year-old — hosting wasps has become something of a party
THE COMMONWEALTH OF THE SENSES 15

trick. Come the summer they gather around her in swarms. I’ve
seen them walk across her eyelids. She sits very still, and she never
gets stung.
Wasps have evolved a coarse-grained vision that enables them to
map food sources on the wing. The curving routes they take past the
picnic table are fly-bys: they're mapping where the lunch is. The down-
side of this curious style of vision is that the wasp cannot detect rela-
tive movement. As far as the wasp is concerned, everything is stationary.
Move, and you confuse it. Run about screaming, waving a rolled up
newspaper, and you vanish. To say this puts the wasp at a disadvan-
tage is putting it mildly: why else do you think it evolved a sting?
The wasp cannot see things. It can only map them. So in what
sense does the wasp have ‘eyes’?

This chapter is entitled “The commonwealth of the senses’ to convey

two different but parallel ideas.
First, the eye is not a lonely miracle. It is a sense organ, similar
in application to an ear, or a nose. Of course there are differences
between seeing and hearing and smelling, but there are common
features too, and these are important. If we look at vision as one of
a family of senses, we can more easily understand where the eye
came from, how it evolved, and how it works. This idea is explored
in the first part of this chapter.
Next, I want to show that eyes are themselves a commonwealth: not
one thing, but many, and supporting many different ways of seeing.
Human vision is a suite of abilities. Occasionally, following an
accident or illness, one of these abilities fails. Moving objects cease
to move smoothly; they disappear and reappear, apparently at
random. Friends and relations become unrecognisable, replaced by
impostors; everyday objects become objects of profound mystery.
All the colour drains out of the world.
From the range of visual abilities we take for granted, many animals
get by quite happily and successfully with just one or two. There are
animals with light-meter eyes, sextant eyes, range-finding eyes, and
eyes which exist solely to detect movement. The visual limitations of
these animals never show up in the normal course of events, but only
following some drastic change in their environment. A captive toad,
16 A NATURAL HISTORY OF SEEING

though its gaoler thoughtfully provides it with dead flies, will starve
to death waiting for a live fly to zip past its eyes. The native birds of
Hawaii, lacking natural predators, ignore fast-moving objects; year by
year, they are being transformed into brightly coloured road kill.
Rather than just conduct a ‘magical mystery tour’ of different sorts
of eyes, I have organised the second part of this chapter according
to how they handle light. All eyes have to handle light, and light
obeys the same laws everywhere. This means that even the most
unlikely eyes will share some common features.
One of the less obvious, but extremely important, common features
of eyes is the way they move, either by themselves or as part of a
moveable head or body. This deserves, and gets, its own section in
this chapter.
I end with a section that boasts the grand (if not grandiloquent)
title “The “problem” of consciousness’. We are aware of what we see;
we appreciate the view. The humble mussel, on the other hand, has
eyes that are scarcely more sophisticated than the motion detectors
that turn on porch lights. It has neither the eyes nor the nous to appre-
ciate the view. What about cats? Do they ‘appreciate the view? Are
they conscious of what they see? Are they ‘conscious’ — in the usual
sense of the word — at all?
This is not a book about consciousness (there are quite enough
of those already). But, very briefly, I try to explain why I think it is
legitimate to write about how other animals see and experience their
world, and to do so without the annoying scare-marks - inverted
commas and the rest - that litter more circumspect accounts.

1 — Seeing and touching

The eye owes its existence to the light. Out of indifferent animal
organs the light produces an organ to correspond to itself . . .
J.W. von Goethe

Imagine that Helen Keller were restored to us. She and I are wandering
around a gallery of statues. We are having a friendly competition. I
am going out of my way to prove that sight is the monarch of the
THE COMMONWEALTH OF THE SENSES 17
senses, quite different in kind and impossible to emulate or even
describe. Keller, on the contrary, is out to prove that sight is not so
special; just one of a family of closely related senses, and that one
sénse can stand in, quite easily, for another.
One handsome sculpture attracts our attention. (Perhaps it is by the
seventeenth-century sculptor, Jean Gonnelli. Born in Tuscany, Gonnelli
went blind at twenty, long before he executed his best work. He is
reputed to have modelled his portrait of Pope Urban VIII using touch.)
When I look at the sculpture, I sense the presence of a three-
dimensional object. When Keller touches it, so does she. I see these
three dimensions by processing the subtle inconsistencies between
the two-dimensional image projected on to my left retina and the
two-dimensional image projected on to my right retina. I also make
use of my familiarity with the way shadows work, and the way
light behaves when it strikes an object. Keller experiences its three-
dimensionality through touch, and correlates these sensations
through her sense of how her fingers are oriented with respect to
each other (‘proprioception).
You might say that from where I stand, I see the whole scultpure
in one go — a single glance — whilst Keller senses the object only where
it touches her skin. But you could just as easily reverse the emphasis,
and point out that Keller can curl her fingers around the object,
whereas, to perceive it from different angles, I have to move my body.
And while I might like to think that I can take in an object ‘at a
glance; in reality my eyes are never still. Every third of a second they
jolt or ‘saccade; moving my gaze from one part of the object to another.
My ‘single glance’ is a multitude of little fixations, not unlike the
twitching of an insect’s antennae, or a mouses whiskers - or Keller's
busy fingers, come to that. (A classic experiment in eye tracking by
the Russian psychologist Alfred Yarbus reveals how we run our eyes
over a face.’)
I look at this object from a distance: to touch it, Keller stands
within reach of it. If I step back, I can still see the statue. If Keller
steps back, she loses sensory contact.
True enough, but is this really a qualitative difference between
sight and touch? Each time I step back, I see the statue in less detail.
Because my eyes are designed to find and exaggerate edges and lines
18 A NATURAL HISTORY OF SEEING

Alfred Yarbus’s record of eye movements while

studying a picture of a girl's face.

- so that objects stand out from the mote and blur of the world -
this loss of detail is hard to appreciate. (Look up at the moon, and
its features appear razor-sharp: impossible to infer from this view
that these edges are actually wildernesses of shattered regolith.) The
further away I move, the less I see.
When Keller lets go of the statue, her sense of it is entirely lost,
instantly. Nothing passes from the object to her hands, the way light
passes from the object to my eyes. But what if it did? We can easily
imagine a technology that transmits tactile information about an
object to a pair of mechanical gloves. This has long been a common-
place of science fiction, and there are plenty of (admittedly, fairly
ungainly) examples on the market. So, this difference between touch
and sight is pretty trivial: merely a question of when a sense can or
cannot be used. Technology has been moving these frontiers about
for a very long time. (Excuse me while I polish my spectacles.) My
leading Keller out of reach of the statue is, philosophically speaking,
no more significant than Keller turning out the lights. Just because
I’m plunged in the dark doesn't mean I cannot see.
Are there any differences left?
THE COMMONWEALTH OF THE SENSES 19

Certainly. Keller cannot see colours - I can. This is a real and

profound difference, as is my inability to tell, just from looking, what
temperature the sculpture is, what consistency, how heavy it is, or
how malleable.
Of course there are differences between touch and sight. But, Keller
counters, the differences are not the point. It’s the similarities that
are important — the degree to which my eyes and her skin enable
us to experience the same things in similar ways.
And she is right.
Since the early 1970s, Paul Bach-y-Rita has been building pros-
thetic eyes for the blind: not false eyes, not glass eyes, but fully working
organs. Bach-y-Rita - a specialist in rehabilitation medicine at the
University of Wisconsin-Madison — has helped the blind to see.
His eyes do not look like eyes. The earliest models looked like
clothing. Bach-y-Rita’s vests are worn either across the stomach or
across the back. Sewn into the material are 256 mechanical vibra-
tors (nicknamed ‘tactors’ because, when they're activated, the subject
can feel their touch). A computer worn at the hip receives pixellated
images from an ultra-low resolution video camera, worn on a pair
of spectacles, and translates these images into mechanical vibrations,
via the tactors. The upshot is a kind of Braille, or Pin-Art, vision.
Bach-y-Rita’s subjects reported that after a couple of hours they were
no longer aware of the tingling sensations generated by the tactors.
They were able to navigate between obstacles, and, eventually, to recog-
nise faces. When the ‘view changed — because they moved, or because
something before them moved, they reacted appropriately. If someone
screwed up a piece of paper and threw it at them, they would duck.
Even more suggestive is an experiment, reported by Daniel
Dennett, in which a researcher, without warning, manipulated a
zoom button on a volunteer’s camera, making it seem as though he
were hurtling forward. The volunteer raised his hands to protect his
face. But the volunteer's vest was strapped to his back.
The artificial sense Paul Bach-y-Rita bestowed upon his blind
volunteers not only works like vision — it feels like vision. It seems
that the mind is not overly fussy where it gets its sensory information
from. What matters is what ‘shape’ the information takes. If visual
information is received through the skin of your back, it only takes
20 A NATURAL HISTORY OF SEEING

your brain a couple of hours to start seeing through your back. If

your back starts itching, you won't mistake the itch for a flash of
light. The ‘shape’ of an itch is different to the ‘shape’ of a face, and
the brain knows how to deal with each.
The senses have become specialised over evolutionary time, but they
are never entirely compartmentalised. Consider these photographs, taken
by Jerome J. Wolken of Carnegie Mellon University in Pennsylvania.
They capture what happens when a rod from a frog's eye detects light.
Rods come in two parts: a fairly normal-looking cell body, and a column
made up of thousands of discs containing a pigment, rhodopsin. When
exposed to white light, the pigment column expands, like a slinky, to
twice its length, with no increase in width. In the dark, it contracts
back. The rods behave just like muscle cells - and for good reason. In
many functional respects, they are muscle cells. A muscle fibre expands
and contracts in response to electrical stimulation. The retinal rod, too,
responds to an electrical signal; but one that comes, not from a nerve,
but from a biochemical reaction to light. Imagine what a working retina
must look like on a cellular scale; a vast automated Pin-Art machine,
turning light to texture. At
this scale, we can easily see
the. functional similarities
between the eye and one of
Paul Bach-y-Rita’s vests.

If you swim in the sea, sooner

or later you're going to end
up looking like a fish. There
are only a handful of really
good ideas in nature and,
whatever their genetic past,
living things tend to evolve
into one of just those few,
very efficient, forms. This is
convergent evolution.
The mechanics of vision: a rod cell If you hunt, youre going
from a frog’s retina expands when to develop eyes. There are
exposed to light. very few really good forms
of

THE COMMONWEALTH OF THE SENSES oo

sclera

optic nerve

cornea iris lens visual cells

nerve fibres

cartilagenous capsule

Convergent evolution: though unrelated, the vertebrate eye (top)

and the cephalopod eye have acquired many similar structures.

an eye can take, so we should not be surprised when quite unrelated

species evolve eyes that look remarkably alike. The eyes of cephalopods
(squids, octopuses and the like) and the eyes of humans are entirely
unrelated. Human eyes develop from specialised patches of skin;
cephalopod eyes are outgrowths of nervous tissue. Human retinas
look inwards, pointing their photoreceptors towards the back of the
eye; cephalopod retinas look out, pointing their photoreceptors
towards the light. Yet in striking ways, human eyes and cephalopod
eyes are virtually identical. Both work like cameras: a lens, behind
an aperture, focuses inverted images on to a photosensitive layer at
22) A NATURAL HISTORY OF SEEING

the back of the eye. Both have irises to control the depth of field,
and both have lids to shield them from too much light.
Like Paul Bach-y-Rita’s seeing-blind volunteers, the star-nosed
mole Condylura cristata ‘sees through touch. Its nose has evolved
into a mobile fleshy organ about a centimetre across. It is exquis-
itely sensitive, with five times as many nerves as run through the
human hand. Its 25,000 ‘tactors’ (to stretch a metaphor) are arranged
across twenty-two ‘fingers, so that the mole’s nose is most sensitive
at its centre. The distribution of nerve endings bears a more than
passing resemblance to the retina of a mammalian eye.‘
Condylura is not the only burrowing animal to have invented a
touch-based ‘vision. Kenneth Catania, the man who soaked his feet
in the bogs of North America to uncover the habits of the star-nosed
mole, has also studied the naked mole rat Heterocephalus glaber,
possibly the world’s least prepossessing mammal, and certainly the
closest anything warm and furry has ever come to impersonating
the termite. The naked mole rat’s looks are as ghastly as its habits —
in place of an insect’s antennae, it has grown its teeth forward like
bayonets. But they are not just weapons; Catania’s studies suggest
the blind mole rat’s teeth are as sensitive as a star-nosed mole’ nose.
Sight is one of those few really good ideas abroad in the world,
and an ability to see is something that many very different animals
acquire.

2 — Handling light

How many different kinds of eye are there? Most eyes fall into
certain broad categories: there are single-chambered eyes like ours;
insects and crustacea have complex eyes, whilst many animals get
by with very primitive eyes that lack even the rudiments of lenses.
Outside these broad categories there are numerous oddities: the
tiny (3 mm-long) crustacean Copilia scans the world line by line,
its retinas in constant motion.’ A close relative of the starfish, the
brittlestar Ophiocoma wendtii, is one huge complex eye, its whole
surface punctured by little eyespots linked by nerve bundles
running just under the skin.
ri

THE COMMONWEALTH OF THE SENSES mae}

The differences between these eyes can be dizzying, and yet all of
them have evolved in response to the way light strikes our spinning,
weather-racked planet. If we look behind their differences, we will
uncover their common principles.
The following four sketches - in no particular order - describe
the ways in which light sculpts the eye.

True eyes gather light. The largest eye ever found belonged to a giant
squid, Architeuthis dux, a monster even by the standards of its own
species, found alive and stranded by three fisherman in Timble Tickle
Bay, Newfoundland, in 1878. Before they chopped it up for dogmeat,
they measured it: its tentacles were ten and a half metres long, and
its eye was forty centimetres in diameter — as wide as my computer
screen. Giant squid are predators, moving vertically through the water
in search of prey — mainly other squid, although they have been
known to make a light snack of baby whales. (Adult sperm whales
repay the compliment; giant squid form at least eighty per cent of
their diet.) Like most ocean species, giant squid cruise at a preferred
depth, a kilometre below the surface, the very point where sunlight
gutters out, leaving the ocean in perpetual night. This is the reason
for the giant squid’s gigantic eyes: it needs them so it can see its
tentacles in the dark. (They aren't muscular, and the giant squid has
to dart forward to swallow the prey its tentacles have captured.)
The giant squid is a visual predator, operating at the margins of
sunlight. Its big eye, with its correspondingly big pupil, gathers more
light than would a small eye.
A kilometre below the ‘cruising height’ of the giant squid, even
further from the reach of sunlight, lurks an even more daunting sea
monster, Mesonychoteuthis hamiltoni, dubbed the ‘colossal squid’ in
2003 after a young specimen was caught in the Antarctic. No intact
adult specimens have ever been recovered, and no eyes have survived
sufficiently undamaged for close study. But the beaks of the elusive
adults have turned up in the stomachs of sperm whales, and using
them as a measure, Steve O’Shea, senior research fellow at Auckland
University of Technology, predicts that a fully grown colossal squid
would have a torso (or ‘mantle’) about four metres long. After his
autopsy of the young Mesonychoteuthis, O’Shea wrote to one corre-
24 A NATURAL HISTORY OF SEEING

spondent: ‘I can assure you that the eyes of Mesonychoteuthis are

larger than anything known for [the giant squid] Architeuthis.°
The eyes of abyssal monsters like Mesonychoteuthis and Archi-
teuthis gather all the light they can, but animals exposed to ordinary
sunlight have carefully to regulate their sensitivity, and shield them-
selves from too much light. Daylight on Earth is a million times
stronger than dim starlight, and night rapidly follows day. Eyes must
not only feed on light; they must respond to extremes of glut and
starvation.
In nocturnal animals like owls and opossums, the opening of the
eye — the pupil - is almost as wide as the eye itself, and any expo-
sure to direct sunlight would be utterly blinding. To protect their
eyes from damage, and to adapt quickly to changes in light level,
many eyes have evolved pupils that can shut out the glare at an
instant’s notice. A cat's iris is slit vertically, and geckos have slit pupils
that are notched to create a series of pinholes: both muscular arrange-
ments can respond more quickly and more dramatically than can
the human iris. Our pupils, evolved for daylight living, dilate to four
times their constricted width when the light grows dim, granting a
sixteen-fold increase in their light-gathering powers.

Find a friend, and stretch a rope out taut between you. Give the rope
a flick. A wave will travel down its length.
That's light.
Flick your end of the rope up and down, and the wave will travel
vertically. Flick the rope from side to side, and the wave will travel
horizontally. In the same way, a wave of light can travel in any plane.
In the vacuum of space, there is no plane along which light waves
preferentially travel. However, when light hits the Earth - when it
passes through a cloud, or reflects off an ocean - the plane along
which a light wave is travelling can make a difference. Light travel-
ling in one plane will go one way; light travelling in another plane
will go another; light in yet another plane will be absorbed.
To picture how light is filtered in this way — or ‘polarised’ - find a
picket fence. Stand on one side of it, stand your friend on the other,
and stretch the rope through a gap in the pickets. Flick your end
of the rope up and down; the wave will travel vertically along the
4

THE COMMONWEALTH OF THE SENSES 25

rope, through the gap in the fence. Flick the rope from side to side;
the wave will travel horizontally until it reaches the fence, where it
hits the pickets and dies. Polarised light is made up of waves vibrating
in one plane, and it is made by filtering out all the light waves travelling
in other planes.
- Every eye on Earth detects light using rhodopsin, though this is
sometimes supplemented by other pigments. A molecule of
rhodopsin changes shape when it is hit by a wave of light. But there
is a catch: the light has to hit the molecule at a certain angle — along
a certain plane - for the shape change to occur. The pernicketiness
of rhodopsin molecules makes no difference to human vision, as
our seeing cells arrange their pigment molecules every which way.
The likelihood is that a wave of light entering the human eye, on
whatever plane, will hit a correctly orientated rhodopsin molecule
eventually.
The compound eyes of insects and crustaceans, however, are effec-
tively colonies of many hundreds or even thousands of tiny eyes
called ommatidia. Conditions in a compound eye are invariably
cramped, so each ommatidium stacks up its pigments in a neat space-
saving column. This means that fully three-quarters of its pigment
molecules lie on the same plane. This tidy arrangement means that
the pigments react strongly to light travelling in one plane, and hardly
at all to light travelling in another. On first acquaintance, this design
seems disadvantageous. What if the insect stumbled into an area lit
by the ‘wrong sort of polarised light? It would be rendered virtually
blind.
Happily, such dramatic effects do not occur in nature. What does
happen is that the Earth’s atmosphere itself acts as a weak polarising
filter. This is useful: even on an overcast day, animals sensitive to
polarised light can detect a column of light, passing directly from
the sun to their eyes. They know where the sun is in the sky, at every
moment, without having to look directly at it. Bees, ants and many
other animals use their polarised vision to navigate: they steer by
the sun. (Nocturnal insects like the dung beetle, not to be outdone,
steer by moonlight.’)
Water, and other non-metallic reflecting surfaces, also polarise
light. At a certain angle, all light reflected off water is polarised on
26 A NATURAL HISTORY OF SEEING

a plane parallel to the water's surface; some water bugs, including

the water boatman Notonecta, use this light to spot new water sources.
Underwater light is even more strongly polarised, and many aquatic
animals are aware of this. Polarised light reflecting off an object will
have a characteristic pattern; using these patterns to identify prey
makes a great deal of sense in the ocean, where the colours of things
shift markedly with small changes in depth. Even transparent animals
like jellyfish cast polarisation ‘shadows.
Some of the cephalopods - octopuses, squid, cuttlefish and the
like — may use polarised light to communicate. By subtly ruffling
the tiny platelets that make patches of their skin iridescent, cuttlefish
can not only alter the colour of their skin, but also its polarisation
pattern, providing them with a secret communications channel, invis-
ible to predators.*
Although polarisation vision has been harnessed in many different
ways, in most cases the evolutionary pressure to adapt to polarised light
may run the other way, towards eyes that are blind to polarised
light. Polarisation can be confusing. The polarisation of reflected light,
for example from a leaf, depends on the position of the sun, the angle
of the leaf, even upon whether it has rained or not. Insects that use
polarised light for navigation generally restrict this vision to the upward-
looking portion of their eyes. Elsewhere, the ommatidia (the elements,
corresponding to a small simple eye, that make up the compound eye
of an insect) are twisted along their length, so that each layer of
photopigment is pointed in a slightly different direction to the layers
above and below. The honeybee relies on polarised light to navigate,
but most of its 5,500 ommatidia are twisted through 180 degrees, and
are thus robbed of any sensitivity to polarised light.
Although it is sometimes said that humans have no sense of the
polarisation of light, this ignores the fact that we are a tool-making
species. For several centuries humans have made tools that show us
how light is polarised.
Raudulf’s Saga, a short family tale preserved in an early four-
teenth-century manuscript, describes the visit of the Norwegian king
Olafur (Saint Olaf, died 1030) to a rich and wise farmer, Raudulfur.
Late one night, one of Raudulf’s sons, Sigurdur, boasts that he can
‘discern the motion of heavenly bodies . . . although I do not see
¥
4

THE COMMONWEALTH OF THE SENSES 27

[them]; a neat trick which enables him to ‘discern all hours both day
and night.
“This is a great skill? says King Olafur, and he’s not kidding. In the
morning, he puts the boy’s talent to the test. “The weather was thick
and snowy, as Sigurdur had predicted . .. The King made people
look out and they could nowhere see a clear sky. Then he asked
Sigurdur where the sun was. Sigurdur gave an unambiguous reply:
Thus far, we are comfortably in the land of the tall, if not the fairy,
tale. Sigurdur’s supernatural talents are too good to be true. What
happens next, though, must give us pause, because it turns out the
king has a sure-fire way of checking whether Sigurdur is right.
“Then the king made them fetch the sunstone, and held it up and
saw where light radiated from the stone and thus directly verified
Sigurdur’s prediction’
The account could not be more matter-of-fact. The king’s ‘solar
stone’ is no fantasy. The storyteller presents it, without fanfare or
fuss, as just another piece of everyday royal kit; as familiar as a watch,
as utilitarian as a compass.°
The king’s sunstone may have been a piece of the mineral cordierite;
pebbles of the stuff are scattered along Norway’s coast. Using a neatly
cleaved crystal of cordierite, it is easy to tell the direction of light
polarization: the stone’s colour will change from blue to light yellow
when it is pointed towards the sun. Or the stone may have been Iceland
spar, an optically pure form of calcite. How wonderful if it were: the
eyes of trilobites — the earliest eyes we know of, five hundred and fifty
million years old — have lenses made of the same stuff.

The sun emits energy at different wavelengths, and visible light is

just a tiny part of the solar spectrum. Reading across that spectrum,
from long wavelengths to short ones, we begin among low-energy,
long-wavelength radio waves. Passing through the infra-red, we come
upon low-energy visible red light. Once through the colours of the
visible spectrum, we watch as high-energy visible blue light fades
into the ultraviolet, and this itself quickly gives way to short-wave-
length, high-energy gamma rays.
Visible light is the part of the electro-magnetic spectrum that
penetrates most easily to the planet's surface. Most of the energy we
28 A NATURAL HISTORY OF SEEING

receive from the sun arrives as visible light, peaking at around 500
nanometres (0.0002 inches), the wavelength of blue-green light.
Every animal sees a slightly different spectrum, depending on
what use it makes of particular wavelengths of light. At the long-
wavelength end, some fish and butterflies see a little into the infra-
red, which gives them extra visual sensation at dawn and dusk. But
the infra-red band is hard for the eye to harness, and no eye can see
very far into it. Infra-red is absorbed by the Earth's atmosphere, water
vapour, and the water surrounding living cells. Infra-red wavelengths
are long, which accentuates their wave-like properties. They tend to
bend (or diffract) around corners, like waves around a sea wall.
Images made in infra-red light are blurred; load infra-red film into
your camera, and you will find that a human face will always come
out as a white blank: a hot surface, entirely lacking in definition. This
is why animals that use the infra-red spectrum use it for target-
finding and navigation, rather than the examination or manipula-
tion of objects. The infra-red-sensing pits near a snake's eye enable
it to locate its prey, while some beetles navigate towards food-rich
forest fires by detecting the infra-red energy of the blaze.
Humans can see a sufficiently powerful infra-red source, as the
vision pioneer George Wald discovered while developing infra-red
viewing devices for the US Army Board of Engineers during the
Second World War. (The devices were not particularly sensitive, so
they built prototype infra-red searchlights, so powerful they trig-
gered a dim red glow in the eye and a distinct and worrying warming
of the skin.'°) In the real world, however, infra-red vision has little
to offer the human species, a primate whose main visual behaviours
require precise, focused vision.
Insects, birds, fish and mice see shorter wavelengths than humans,
well into the ultraviolet. Many flowers boast striking patterns, only
visible in ultraviolet light, to attract pollinating insects. Even with
their superb visual acuity and excellent colour sense, extending well
into the ultraviolet, kestrels find it hard to spot the drab voles which
are their favourite food. Happily for the kestrels, however, voles
communicate by leaving trails of urine - indeed, they pee almost
continuously - and vole urine reflects ultraviolet light. For kestrels,
hunting for voles is simply a matter of following the arrows.
THE COMMONWEALTH OF THE SENSES 29
Ultraviolet light, like infra-red light, is largely blocked by the Earth’s
atmosphere. But it is extremely energetic - as anyone who forgets
their sun cream on the beach can testify. It is energetic enough to
trigger photoreceptor cells sensitive to far longer wavelengths. The
challenge for the ultraviolet-adapted eye is to develop a system that
doesn't simply fire randomly and everywhere the moment sunshine
hits it.
No big animal sees far into the ultraviolet. The larger the eye, the
more light it can take in, and there must be a point at which the
potential damage ultraviolet light can do when focused outweighs
its usefulness. Many birds and insects have evolved to see ultraviolet
wavelengths, but they live for only a short time, dying before the
damage becomes significant. Large mammals have much longer lives,
and their exposure over years could destroy their eyes’ photopigments
and turn their lenses cloudy.
Mammals that are active at dawn and dusk require less shielding
against ultraviolet light, and benefit more from good vision in blue
and near ultraviolet light. (Any rider will tell you that horses can
pick their way through the dusk with an ease that borders on the
uncanny.) Animals which are active in daylight have evolved many
ingenious systems for shielding their eyes against ultraviolet light.
Squirrel eyes have yellow-tinted lenses, while the photoreceptors of
fish, birds and turtles are tipped with coloured oil droplets, which
may also play a role in their perception of colour. The particular,
and unique defence evolved by primates, including human beings,
involves a yellow pigment that absorbs ultraviolet light almost
completely. This ‘macular pigment’ covers the whole fovea, which is
why it turns a bright lemon-yellow when the retina is exposed to
the air - an effect which first startled the physician and inventor
Samuel Soemmering in 1795.
By far the most effective ultraviolet filter in the human eye is the
lens. Our lenses are extremely efficient filters of ultraviolet light,
reflecting our daylight habits, but with the lenses removed, human
eyes can perceive ultraviolet wavelengths, something which, though
barely wondered over, must have been apparent to ophthalmic
patients for at least two thousand years. In the first century AD, Roman
doctors routinely displaced and removed irreversibly swollen and
30 A NATURAL HISTORY OF SEEING

clouded lenses from the eyes of their patients. The condition they
were treating, cataract, is still with us, and still irreversible. Since
1947, it has been possible to replace the lens with a plastic substitute.
Before artificial lenses were available, however, those who had their
lenses removed by surgery found that they could see into the ultra-
violet; blues were clearer and richer, and ultraviolet light, energetically
triggering every photoreceptor it hit, was a blueish-white wash.
Because most wavelengths are visible to most animals, we can be
forgiven for thinking that most animals see the same colours, but
many animals have very poor colour vision. Colour is just a way of
distinguishing objects from each other. Where daylight is more or
less the same colour all the time, objects retain their native colours,
and the need for complex colour perception is quite low. In 2000,
the Taiwanese neurophysiologist Chuan-Chin Chiao showed that
daylight, filtered through a forest canopy, changes colour only very
slightly.** No wonder that mammals, with very few exceptions, make
do quite happily with ‘dichromatic vision: in human terms, they are
colour-blind.
For some forest-dwelling foragers, impoverished colour vision
may confer a survival advantage, since — for reasons I shall explore
in Chapter Nine — it goes hand in hand with a greater sensitivity to
lustre. A camouflaged object in dappled light is immediately noticed
by a colour-blind person: its distinctive sheen gives it away.
Marmosets may be natural camouflage-breakers; the females enjoy
three-colour vision quite similar to our own, while the males are
born with two-colour vision — and it is the males who are better at
spotting food sources.”
In environments where the colour of daylight fluctuates a great
deal, colours vary hugely, and several dimensions of colour vision
are a real asset. The colourworld of reef fish is one such, very different
from our own, where the bright blues and yellows of many reef fish
function as camouflage.’? Water absorbs sunlight, robbing it one by
one of wavelengths, the deeper one goes. These changes are quick
and dramatic: reds, oranges and yellows vanish from human percep-
tion within about ten metres. For the same reason, the ‘blueing’ of
objects with distance is far more marked underwater than on land.
To keep tabs on the coloration of their surroundings, some reef fish
,”

THE COMMONWEALTH OF THE SENSES a0)

have up to twice as many different colour receptors in the eye as do

humans.
But the record for the most dimensions of colour vision is held,
not by reef fish, but by brightly coloured stomatopoda or mantis
shrimps. Mantis shrimps are not shrimps. Some are thirty centimetres
long. They will attack and attempt to eat just about anything, and
have been known to break a diver’s fingers. A blow from the claws
and head of some species is almost as powerful as that from a .22
calibre bullet. In 1998, a green mantis shrimp put its head through
the half-centimetre thick glass of its tank at the Yarmouth Sea-life
Centre.'+ Look into its eyes, and there are more surprises still. Some
species of mantis shrimp boast sixteen kinds of narrow-band colour
detectors. Behavioural experiments have shown that these provide
them with true colour perception. The evolutionary pressure on the
mantis shrimp to develop such sophisticated visual apparatus appears
to have been two-fold. It operates in the shallows, where colours
vary drastically with small changes in depth, and it relies on displays
and dances to govern its behaviour with other mantis shrimps. A
colour-blind mantis shrimp will not be able to side-step the aggressive
approaches of rivals; even more seriously, it will not be able to locate
and engage a mate.

Light reveals nothing about how far it has travelled. A ray of light
arriving from a distant star makes no more fanfare on its arrival
than does the glint reflected off the rim of a cup. How, then, are we
to perceive depth?
William Molyneux, the Anglo-Irish politician and physicist, put
his finger on the central difficulty in his work, Dioptrika Nova.
Published in 1692, this was — contrary to the impression given by
the title — the first book on optics written in English. ‘For distance
of itself,’ he writes, ‘is not to be perceived; for ’tis a line (or length)
presented to our eye with its end toward us, which must therefore
be only a point .. ? To put it another way, we only ever see light
when it hits us. We cannot stand to the side of light and measure
the distance it has travelled. So how can we tell, just by looking, how
far away things are?
To measure the distance of a near object, we might use muscular
20) A NATURAL HISTORY OF SEEING

information from the eyeballs, as the eyes converge slightly to focus

on a near object. This seemed like a plausible explanation for how
we judge distances of up to six metres. But even if this mechanism
existed (it doesnt), this still left the problem of how we judge the
distance of faraway objects. This had to involve a different mecha-
nism, because when we look at objects further than about six metres
away, both eyes are looking straight ahead. But there seemed to be
no difficulty in explaining our depth perception. Since the Renais-
sance, painters had demonstrated, time and time again, how well the
eye infers three-dimensional forms from environmental clues like
shading, perspective, and occlusion (the way things get in the way
of each other). Some of these lessons are part of folklore: the grass
is always greener on the other side of the fence, because individual
blades of grass are more clearly seen from the side than from above.
A field of grass hides the earth better when looked at from the side.
David Rittenhouse (1732-1796) was a clockmaker by trade and
an astronomer by inclination. Self-educated, he built the first astro-
nomical telescope on American soil, and with it in 1768 he discov-
ered the atmosphere of Venus, independently of its earlier discovery
by the Russian Mikhail Lomonosov in 1761. He also came up with
+
,

THE COMMONWEALTH OF THE SENSES 33

the optical illusion above (charmingly dubbed the ‘muffin par illu-
sion by the vision researcher Donald Hoffman). Which circles are
muffins, which are empty dishes? The answer depends on which
way up you hold the book. Rittenhouse’s illusion reveals how much
our sense of three-dimensional form relies upon the way our envi-
ronment behaves. We assume that everything is lit from above, and
this is how we interpret the shaded circles.
Leonardo da Vinci wrote very well about this kind of thing. His
subjects included shadows, the way the eye handles detail, and the
way the colours of things shift blue-wards the further away they are.
He also makes a stab at explaining stereopsis: the indefinable ‘is-
ness’ three-dimensional objects acquire when looked at with two
eyes:

. a painting, though conducted with the greatest art and

finished to the last perfection, both with regard to its contours,
its lights, its shadows and its colours, can never show a relievo
equal to that of the natural objects, unless these be viewed at
a distance and with a single eye . . . because a painted figure
intercepts all the space behind its apparent place, so as to
preclude the eyes from the sight of every part of the imaginary
ground behind it.*®

This explanation is not quite sufficient. It assumes that every object

is small enough, or near enough, to occlude slightly different parts
of space in the view of each eye. This is by no means always the
case. Neither can he explain why we see a three-dimensional figure
whenever we fuse stereoscopic drawings.
The trouble with Leonardo's explanation is that it relies solely on
optics. It took nearly two hundred years before someone realised
that optics alone could not explain the effect. That someone was
Charles Wheatstone (1802-1875), inventor of the concertina,
inventor of the telegraph, and Victorian polymath. In 1838, Wheat-
stone subtly revised Leonardo's argument. Out of one eye, Wheatstone
says, a good enough painting can be confused with a real scene. But
look at it through two eyes, and the illusion is broken. Why? Because
‘When the painting and the object are seen with both eyes, in the
34 A NATURAL HISTORY OF SEEING

case of the painting two similar pictures are projected on the retinas;
in the case of the solid object, the two pictures are dissimilar . . .””
The stereoscopic drawings that head the chapters of this book are
from Wheatstone’s original paper. Each features a pair of dissimilar
drawings of the same object. Go cross-eyed, to fuse the paired images
into one, and you will see a single three-dimensional figure. ‘I have
employed only outline figures; Wheatstone explained, ‘for had either
shading or colouring been introduced it might be supposed that the
effect was wholly or in part due to these circumstances, whereas by
leaving them out of consideration no room is left to doubt that the
entire effect of relief is owing to the simultaneous perception of the
two monocular projections, one on each retina.
Soon afterwards, Wheatstone smartened up the demonstration
equipment he had thrown together for his Royal Society talk of 1838,
and invented the stereoscope, ‘an instrument which will enable any
person to observe all the phenomena in question with the greatest
ease and certainty.
Anyone old enough to remember looking through red plastic
ViewMasters as a child knows intimately what a stereoscope can do.
Two photographs of the same scene, taken five centimetres apart
(the average distance between human eyes), are mounted in an optical
device that helps the viewer fuse these images into one. ‘The result
is a single, captivating three-dimensional image. (Strictly speaking,
you dont need a stereoscope to see the image; go cross-eyed until
the photographs overlap, and the images will fuse by themselves.)
One curious by-product of the stereoscope is a niggling sensation
of vertigo. Nobody has eyes exactly five centimetres apart. Slight
differences in stereopsis between the viewer and the apparatus can
exaggerate the 3-D effect of the photograph. One gets the (perfectly
justified) feeling that one is looking out of somebody else's eyes.
Wheatstone followed with the pseudoscope which, by means of
prisms or mirrors, transposed the optical images of an object or
stereogram as they were normally brought to the eyes. It never caught
on: plunging into a slightly alien three-dimensional world was one
thing, but having it turned inside out was quite another.
Wheatstone’s stereoscope demonstrates that when we fuse our left-
eye and right-eye views of the world, the tiny discrepancies between
v¢
THE COMMONWEALTH OF THE SENSES 35

them give us a sense of depth. What he could not do at the time was
explain how these tiny discrepancies are spotted, measured and inter-
preted by the eye and brain. How do we stitch these slightly dissimilar
views together in reliable ways? When you walk into a thick and
tangled forest, how do you know there are, say, four trees a few feet
away from you, and not three trees, further away? Confronted by four
trees in a row, and with no foreknowledge of how the images in one
eye match up with the images in the other, you can make sixteen
perfectly sensible assumptions about where the trees are. Twelve of
them will be wrong. What if you were faced with ten trees? There are
a hundred possible places the trees could be — that's ninety opportunities
for error. Now think about how many trees there are in a wood. No
wonder people thought — even as late as the mid-sixties — that envi-
ronmental cues like shading, perspective and occlusion were the main
engines of depth perception, and that stereopsis was a mere curio.
Bela Julesz (1928-2003) changed all that. In 1953, Julesz, a PhD
36 A NATURAL HISTORY OF SEEING

student at the Hungarian Academy of Science, embarked on a thesis

analysing the encoding of different televised images. It was a labour of
love — apart from anything else, television had yet to arrive in Hungary.
The thesis was never published: in 1956, Russian tanks drove into
Hungary. Julesz, armed with a copy of Plato’s Apologia of Socrates
(his favourite book), fled with his wife from Hungary to the United
States. He went to Bell Laboratories in New Jersey, where he worked
as a radar engineer. Later he joined a group dedicated to ‘reducing
visual information in pictures without perceptual deterioration: Julesz
thought that some familiarity with human vision would be useful,
and so he started to study psychology.
What he read startled him: apparently, depth perception was a
nebulous and complicated process that ran the gamut of the brain’s
machinery, from memory to symbolic logic. This was news to Julesz
who, as a former radar engineer, knew damn well that all you needed
was two eyes: After all, in order to break camouflage in aerial recon-
naissance, one would view aerial images (taken from two different
positions) through a stereoscope, and the camouflaged target would
jump out in vivid depth.** Julesz realised that, to prove his point, he
needed to show that even perfect camouflage will fail the stereoscope
test. So he created the random dot stereogram. Take a look at Julesz’
original 1959 figure. In both pictures, there is a simple shape lurking
among the dots. In neither can you make out the figure, because it
is perfectly camouflaged: a white figure, covered in random dots,
against a white background, covered in random dots.
Go cross-eyed, fusing the squares into a single square. Now do
you see the figure?
Random dot stereograms are easy to make. Although they have always
THE COMMONWEALTH OF THE SENSES 37
been generated on computers (so no need for glue, scissors and obses-
sive neatness) the process is simple. Take two sheets of paper printed
over with the same pattern of random dots, stack them together, and
cut out a shape.’? Take two more identical sheets, and lay one shape
over each sheet. Adjust the shapes so that both lie on exactly the same
spot on their sheets. Then displace one figure horizontally by a few
millimetres. Go cross-eyed, fusing the pictures, and the shape will appear.
The left- and right-hand pictures are equivalent to left-eye and
right-eye views of the same scene. Depending on which way you
displaced the shape — towards or away from each other — the shape
will either leap out at you, or appear as a hole in the image. Either
way, the shape will reveal itself. As Julesz said, ‘in real life, there is
no ideal camouflage:
If, confronted by ten trees, we can make ninety wrong guesses
about their number and position, how can we possibly successfully
fuse the hundreds of dots that make up a random dot stereogram?
The answer is straightforward: trees are big, complex, gnarled objects.
Every square inch is a mass of surfaces, shadows, occlusions — a mass
of dots, in fact. Stereopsis works by processing the dots.
The views of both eyes are mapped over a single sheet of neurons
in the visual cortex of the brain. When both views are laid over each
other, a special class of neurons spots the inconsistencies: a spark in
the left eye doesn’t quite overlap with a spark in the right eye; a
smear in the right eye doesn't extend quite as far as the same smear
in the left eye. We know where the trees are, because the scene before
us has already been sifted for all those tell-tale little disparities
between the left and right eyes. We see where the trees are because
we already know where to look.
An apology: some of you won't have been able to see anything in
the stereogram. About one in twenty people will have little or no
experience of anything discussed in the last few pages — because
they are stereoblind. Many do not know about their handicap. Stere-
opsis is subtle, and in its absence, environmental cues are enough
to see us through most everyday situations.
The conundrum of the observer in the forest, confused by a handful
of trees, only holds while the observer remains absolutely still. The
moment the observer moves, the precise location of the trees becomes
38 A NATURAL HISTORY OF SEEING

easy to judge, thanks to another technique for depth perception,

parallax. Rock your head from side to side, and the way objects seem
to move relative to each other will give you extremely accurate infor-
mation about their relative distances. Parallax is used throughout
the animal kingdom. Grazing animals generally have eyes mounted
on the sides of their heads, to give them all-round vision. Because
their fields of view barely overlap, stereopsis plays no part in their
depth perception. Instead, they use parallax to judge the approach
of predators. Predators use parallax, too. The praying mantis rocks
its head from side to side to judge the distance of its prey.
Compared to parallax, stereopsis is a relatively rare and unusual
form of depth perception; as far as we can tell, only a handful of
lucky primates enjoy it. In 1995, towards the end of a career which
revolutionised our understanding of the early stages of vision, Julesz
came up with a just-so story to explain why primates might have
acquired this extra, super-subtle sense of depth. Since there is no
camouflage in 3-D, stereopsis probably evolved first in our insectiv-
orous lemur ancestors, enabling them to spot insects that freeze,
blending into the foliage at the first sign of danger. In other words,
the wonderful ‘is-ness’ of the three-dimensional world may be sensory
overspill from a quite different primate adaptation: the ability to see
through camouflage. This would certainly explain why some insects
have embarked on a hugely expensive counter-counter-tactic, re-
making their bodies as twigs and leaves.

3 — The moving eye

Vision itself is a dynamic process. There is little in the world

that stands still, at least not as imaged in our retinas, for our
eyes are always moving. The visual system is almost exclu-
sively organized to detect change and motion.
Haldan K. Hartline’

Our eyes provide us with a picture of the world. It is an odd, magical

picture, better than any film or video. In a moving image, everything
is set afloat upon the picture frame, and the camera must exert an
THE “COMMONWEALTH OF THE SENSES 39
iron discipline if things are to retain their proper relationships. Move
the camera too quickly, and the world lurches. Zip the camera from
object to object to object, and space crumples. The eye, however, is
free to explore the space it inhabits. It flicks, glances, pans, tilts — all
in, literally, the blink of an eye. Yet spatial relationships are preserved.
Stationary objects always appear stationary, and moving objects
~ always appear to move.
The picture provided by our eyes is magical in other ways. It
highlights the things we are interested in, enhances them in count-
less ways, and always puts the object of the greatest interest in the
centre of the frame. None the less, this picture is inferior to the
pictures provided by a camera, since virtually none of it is in focus.
And, unlike a camera, it is unreliable. A camera records what is
really out there, but the eye doesn't. In the picture provided by our
eyes, whole objects can fail to register. An entire event can come
and go, and leave not a trace. A cynic might say that the magic of
such a picture lies in the fact that we never really notice how bad
it is.
The picture we receive through our eyes is odd. It is magical.
Which is to say: it is not a picture.
Although we think of vision in terms of images, images are not
the be-all and end-all of vision. The natural world is full of eyes that
cannot generate images. Where, then, is the common ground between
our eyes and others? Is the story of the eye one story, as I have
claimed, or is it many?
One way to understand how all eyes are related is to put aside,
for the moment, the different experiences of vision, and instead look
at what eyes actually do.
All eyes move. Either they come with their own musculature, or
they are mounted on moveable heads or bodies. The significance of
eye movement was slow to dawn, but today it is the focus of the
very latest researches into perception and attention. At first, only
human eye movement seemed interesting. Now, as we learn about
the eye movements of very different species, we are beginning to
uncover certain universal principles, on which depend many different
forms of vision.
40 A NATURAL HISTORY OF SEEING

For the staid and sedentary professors of the University of Prague,

medical graduate Jan Purkinje (1787-1869) must have cut an
unnerving figure. His rooms were furnished with whirligigs and
swings. Regularly, he would spin himself into a nauseous stupor.
Purkinje, born to a family of Czech peasants, began his adult life
as a teacher in a religious order. This did not last long; the ‘contin-
uous slavery to the superiors whose lives and dignity did not always
come up to my expectations’ soon led him to shed his monk's habit;
by his own account, he walked the three hundred kilometres to Prague,
where he supported his philosophical studies by private tutoring.
Purkinje was interested in the scientific value of his own experi-
ence. He believed that the results he obtained from his own body
were more useful than any he might obtain from animals, or from
fatally ill patients. When he began to study the physiology of vision,
he took a variety of drugs, in a series of risky experiments that were
looked on askance by his peers. Quite what they thought of his other
abiding fascination is not recorded: Purkinje was particularly

Jan Purkinje in a spin.

+
;

THE* COMMONWEALTH OF THE SENSES 41

interested in a new method of pacifying the insane which had come

out of Britain in the 1810s, ‘which is now become of so much
consequence in the cure of maniacs of almost every description’ The
best treatment for mania, apparently, was to spin the maniac into
stupefied and wobbly submission.
‘The debility arising from swinging is never to be dreaded? declared
" Joseph Cox, the inventor of the spinning chair: ‘it is generally accom-
panied by sleep and a sense of fatigue, while the slumbers thus procured
differ as much from those induced by opiates, as the rest of the hardy
sons of labour from that of the pampered intemperate debauchee’”*
Cox had a point: at a time when sedation was a hit-and-miss affair,
accomplished by all manner of means -— vomiting, purging, bleeding,
digitalis, bathing, blisters, camphor, sedatives, stimulants - rotation in
a chair was certainly one way of calming people down.
Purkinje, who spent no little effort exploring the links between
movement and vision, built an optical toy which screened looped
animations; one of his ‘movies’ shows him taking his home-built
rotating chair for a spin. He used it to induce vertigo, stretching the
abilities of his eyes to reveal - in their exhaustion and failure - the
reflex that stabilises vision.
Purkinje understood that, first and foremost, the eye moves to
compensate for the body’s movements. A moving eye keeps the world
still. It does so by means of a reflex, which Purkinje described with
admirable exactitude. Briefly, Purkinje’s reflex — the ‘vestibulo-ocular
reflex’ — is triggered whenever our body’s movements disturb the
delicate hairs growing in the liquid-filled canals of the inner ear.
Nervous responses from the hairs tell us how we are moving in
space, and this information, which helps us balance, also triggers
involuntary movements of our eyes. Thus, stationary objects continue
to appear stationary, even when we are moving. If, for example, I
look at something while moving my head, the vestibulo-ocular reflex
compensates by moving my eyes in the opposite direction, cancelling
out the head movement.
Spinning sloshes the fluid about in our ears, upsetting our
balance. This loss of balance in turn confuses our vestibulo-ocular
reflex. We not only feel as though we are falling - it really looks
as though we are falling, as the world zooms past our eyes; as if
42 A NATURAL HISTORY OF SEEING

gravity had changed direction, pulling us, not downward, but to

the side.
Purkinje’s work was based on the idea that optical illusions reveal
the underlying mechanisms of vision. Extending this insight to the
world of the senses generally, he evolved a biological science that
was not purely mechanical, but made room for feeling and - even-
tually, in the hands of his successors - mind. Purkinje’s successes
were won in the teeth of often savage opposition. Hostility was
directed both at the man himself (as a Bohemian, his ethnic back-
ground was a big hindrance to him) and at his working methods.
What were his peers to make of this supposedly serious scholar,
stumbling about the corridors, his eyes bleared by belladonna (a
cosmetic eyedrop historically favoured by young women because of
the way it dilated and paralysed their pupils)? When he introduced
demonstrations and practical laboratory work into his teaching, this
was simply intolerable, and there were calls for him to be stripped
of his professorship.
Purkinje survived largely thanks to a well-placed friend, the
German poet and polymath Johann Wolfgang von Goethe (1749-
1832). A gifted natural philosopher, Goethe had also attempted to
formulate a science concerned with subjective experience, and wrote
extensively on colour perception, with mixed results. With Goethe’
backing, Purkinje was able to establish intellectual credentials for
his new, physiological science. Much of his own work, however, was
forgotten. For years after his death, he was better remembered for
discovering the forensic potential of fingerprints, and for purifying
blood plasma, than for his greatest scientific work: the anatomy of
optical illusions. This work, which had appeared under such provoca-
tive titles as Contributions to the knowledge of vision from a subjective
perspective and New subjective reports about vision had to be redis-
covered, after his death, by a new generation of psychologists.
His shade haunts almost every chapter of this book.

Most animals care very little for the substance of things. They are
much more interested in where things are going. If it moves, it matters.
Moving targets may be threats, meals, or mates. The earliest image-
forming eyes evolved, not to detect objects, but to detect their motion.
<
,

THE’ COMMONWEALTH OF THE SENSES 43

A toad catches flying insects with its tongue; if by mistake it swallows
a seed, a leaf, a piece of lint, it simply turns its stomach inside out
and licks the offending particle off with its tongue.
Most animals live lives of unutterable regularity, and their eyes
have evolved simply to warn them of interruptions to their usual
., Programme. Rabbit eyes are remarkably attuned to motion, and this
makes rabbits incredibly efficient at spotting predators. Otherwise,
they are purblind: they live in a perpetual fog.
Humans are foragers; we take a more than usual interest in what
things are. But even our eyes are tuned, first and foremost, to motion.
In 1875 the Viennese physiologist Sigmund Exner showed that two
brief, stationary flashes, provided they are not too far away from
each other, are seen as a single object in motion. This habit of fusing
stationary dots into moving objects makes a great deal of sense in
nature, where prey and predators disappear and reappear constantly,
as they move through grass, run behind trees, and peer around rocks.
But the power of the phenomenon (called the phi phenomenon)
will perhaps best be demonstrated the moment you set this book
down and turn on the television. Every film and television
programme ever made depends on phi. Both display still images
quickly enough for our eyes to read them as a single moving image.
This is not ‘persistence of vision, which is simply the eye’s inability
to tell a steady light from one that flickers faster than fifty times a
second. In bright light, the eye does better; detecting flicker up to
one hundred times a second. This is one of the reasons the front
row of a cinema isn’t popular: from here, bathed in light, and with
much of the image located in the movement-sensitive periphery of
vision, the eye is constantly aware of the flicker rate of the film.
When the flicker is obvious — not just snatching at the corners of
your eyeballs, but there in front of you, a part of the cinema expe-
rience — it very quickly ceases to matter. Watch a silent film at its
proper speed, on old equipment, and the image flickers about eighteen
times a second. But, thanks to the phi phenomenon, the illusion of
a single, moving image is still preserved, and the film is still eminently
watchable. The phi phenomenon is quite happy to stitch together
stills into moving images, even when they move quite slowly.
* + *
44 A NATURAL HISTORY OF SEEING

Take a moment to scan your surroundings: turn your head slowly

from side to side. You'll notice that however smooth your head move-
ments are, your eye ‘snaps’ from location to location. It’s trying to
keep whatever you are looking at steadily in view.
Why, when they compensate for our movements, do our eyes insist
on snapping from position to position? Try moving your eyes across
this sentence at a slow, even speed. Impossible, yes? The eye stutters.
The only way you can remotely manage to move your eyes smoothly
is by letting the page go out of focus — which rather defeats the point.
If the eye’s involuntary movements exist to compensate for our body
movements, why can't the eye move as smoothly as the body does?
Actually, it can: slip a pen behind this book, so that the top pokes
above the page. Move the pen smoothly back and forth. Follow it
with your eyes. Suddenly, it’s like someone has oiled the mechanism;
the eyes move smoothly back and forth. This is the ‘optokinetic reflex.
It is yet another visual stabiliser, one that Purkinje did not have the
tools to unpick, since it uses visual information itself to move the
eyes and stabilise the view upon the retina. It is the biological equiv-
alent of the ‘judder control’ on digital video cameras.
Hold your eyes still, and the image of a moving target is quite
likely to flash by individual photoreceptors in milliseconds. If it moves
past very quickly, your photoreceptors wont have a chance to respond,
and you wont see the target at all. Even if it's moving quite slowly,
the target is likely to be blurred. This is why humans fixate so strongly
on objects moving across their field of view - if they didnt, they
wouldnt be able to see them. If you concentrate on a moving object,
the eye will turn smoothly in an attempt to track it. The more you
are interested in what you are looking at, the more it will try to hold
on. But there is a limit: move your pen too fast, and the image will
blur. The moment that happens, the eye will give up its ‘tracking
shot’ and revert to ‘angle shots. It will saccade, snapping from location
to location, capturing a selection of ‘stills.
In French, le saccade describes the way a sail snaps in the wind.
(It was a Frenchman, the eminent vision researcher Louis Emile
Javal, who named these sudden, involuntary movements of the
eyeball.) There is nothing subtle about a saccade; nothing deli-
cate. The movement is sudden, ballistic, and powerful enough that
THE’ COMMONWEALTH OF THE SENSES 45
the eyeball commonly overshoots its target; a barrage of smaller
correcting saccades follows any large eye movement.
What if our eyes did not saccade? What if they were entirely
Passive, so that we simply drank in the scene before us? What would
we see? To get at the answer, it is not enough simply to stand and
_ Stare. Even when you are looking intently at something, your eyes
are drifting and trembling three or four times a second. To still these
tiny movements (called microsaccades) the eye has to be paralysed.
There are a couple of ways of disabling the eye. One method I
absolutely would not recommend was dreamt up by the man gener-
ally regarded as Purkinje’s intellectual successor, the physicist and
philosopher Ernst Mach. As a student in Vienna in the 1850s Mach,
unable to afford experiments in pure physics, caught the experimental
psychology bug. His youthful enthusiasm for the fertile new zone
he saw emerging between physics, psychology and physiology is
evident from this experiment: he thought it would be a good idea
to stuff putty under his eyelids to stop his eyes from saccading. The
trouble with invasive methods like these —- aside from the obvious
dangers — is that the sheer discomfort and oddness of the procedure
gets in the way of decent observation, as happened in Mach’s case.
A much more elegant approach became possible once plastics were
developed and contact lenses became feasible. In the early 1950s, two
research teams (one from Brown University in Rhode Island, and
one from Reading University in England) stumbled independently
on a way to stabilise images on the retina. They fastened tiny spotlights
to the contact lenses their volunteers wore. It didn’t matter how much
the volunteers moved their eyes: the little spots of light would always
be stabilised at exactly the same places on their retinas. The results
could not have been more spectacular: after about a second of this
curious, frozen vision, the volunteers lost sight of the lights.”
The eye exists to detect movement. Any image, perfectly stabilised
on the retina, vanishes. Our eyes cannot see stationary objects, and
must tremble constantly to bring them into view.

Three pairs of muscles move the eye, but nobody quite knows how.
The superior rectus and inferior rectus (labelled D and E in David
Hosack’s exquisite illustration; see colour section) move the eye up and
46 A NATURAL HISTORY OF SEEING

down; the medial rectus and lateral rectus (C and A) move it from
side to side. However, even after 450 years of anatomical research,
beginning with Vesalius and Falloppio, there is still no general agree-
ment on what the third pair of muscles is for. This pair, the superior
oblique (B) and the inferior oblique (not shown), only came to notice
some hundred years after the other two pairs had been described.
In theory, the obliques should, via a system of pulleys, spin the eye
about its axis — which is ridiculous enough. Making a model that
can reveal their true function is far from easy. The earliest mechanical
models treated the eye as a perfect sphere, and used strings for
muscles. They did not reveal very much. For a start, all the muscles
work in concert, so even when the whole problem is reduced to a
matter of ball and string, the mechanics are horrendous. More seri-
ously, muscles are not strings. They change their behaviour as they
are stretched, and this means that they behave differently every time
the eye looks in a different direction.
The puzzle of how the eye muscles pull the eye into position is
more than a mere mechanical curiosity. Misalignment of the eyes is
a treatable condition, and the more information we have about how
the muscles are supposed to work, the better our corrective treatments
become. With the advent of computers, it seemed as though the
problem would be solved. David A. Robinson, professor of biomedical
engineering at Johns Hopkins University in Baltimore, attempted the
first computer simulation, factoring in points of muscle origin and
insertion, normal elasticity and muscle tone, the checks imposed by
ligaments and other surrounding tissues, and much else. When he
ran the program, the computer promptly turned its eyes to face the
back of its head.*? Robinson persevered. His model is good enough
these days to be used all over the world to plan eye surgery. No
sooner did the moving eye begin to make sense, however, than a
new mystery arose: a human eye that did not move at all.
In 1995 John Findlay, from the University of Durham, and Iain
Gilchrist, in Bristol - men who had built their careers on research
into eye movement and vision — were disconcerted to hear from an
undergraduate, A.I., who claimed not to be able to move her eyes.*4
A. had volunteered to take part in a study into people with squint
— a persistent misalignment of the eyeballs. Her own squint was, she
THE COMMONWEALTH OF THE SENSES 47
said, very mild, as it had been corrected by surgery when she was
twelve. And there was one other thing she felt she should share with
the researchers because it might disqualify her from the study: she
had no eye muscles.
According to the surgeons who had operated to correct her squint,
the muscles were so atrophied that they could never have worked.
And this, for Gilchrist and Findlay, was no small source of worry.
According to their theory of active vision — a model to explain why
the eye has to move in order to see ~ A.I’s eyes should not have
been able to function. And yet here she was: a student at university,
writing letters, reading books, getting through life perfectly well and
suffering from no apparent visual handicap. She did not drive, nor
care for sport, but these life choices did not seem driven by any
pressing or obvious visual defect. Though her eyes could not saccade,
she could read normally. A speed of 257 words per minute wouldn't
win her any speed-reading contests, but it was well within the range
considered normal.
However, there was something about her: she moved like a bird.
She twitched. She cocked and bobbed her head. The more closely
her head movements were studied, the more bird-like they turned
out to be. And for a very good reason: birds saccade mainly by
moving their heads.
Not every eye in nature is moveable. Many - most - animals have
eyes that are fixed permanently in place. Invertebrates often have
eyes that cannot move, welded to a hard exoskeleton. A few get
around this problem by locating their eyes on the end of long stalks.
The jumping spider has come up with a different solution: it moves
its retina, looking out through the lens of its eye the way a pilot
looks through the glass canopy of his cockpit. For most insects, a
fixed eye is not a problem: they are so tiny, they can simply swivel
their whole bodies to change the view. Small animals are very
powerful relative to their weight, so having to turn around is no real
imposition. Airborne insects can simply swing about in space to
capture any view they want.
The larger you get, the weaker you are, relative to your size. So larger
animals move less of themselves to change the direction of their gaze.
All vertebrates have eyes that are at least a little bit moveable. The
48 A NATURAL HISTORY OF SEEING

smaller ones — birds, for example - make little use of this ability,
and prefer to move their heads. Anything larger moves its eyes a
great deal, and turns its head only when strictly necessary.
It appeared that A.I., unable since infancy to move her eyes, had
learned to saccade the way a bird does - by moving her head.
More strange has been the discovery that a bird’s head movements
— and even an insect’s body movements - exactly parallel the move-
ments of our own eyes. Birds and flies use the same basic strategies of
smooth pursuit and saccade-and-fixate that we do. Indeed, these
strategies are proving to be commonplace throughout the animal world.
Smooth pursuit makes sense when tracking a target, whether
youre an insect, a bird, or a human being. Equally, by cutting up the
visual continuum into a series of stills, the saccading eye (or head,
or whole body) minimises motion blur, reduces disorientation, and
more reliably reveals moving objects against a still background. The
apparent eccentricities of human eye movement have turned out to
be near-universal strategies for seeing.

4 - The ‘problem’ of consciousness

D.B-s head was a bomb, ready to explode. The malformed blood

vessels in his brain could rupture at any moment, and destroy the
delicate tissues surrounding them. There was no time to lose.
In 1986 a surgical operation removed the offending tissue from
D.B’s brain. The price he paid for this life-saving surgery was high:
by cutting into his right primary visual cortex, the surgeons inevitably
damaged his vision. He could still see out of both eyes - but the left
halves of both views were now invisible to him.
To appreciate how this could happen, we need to understand some-
thing about how we integrate the output of two eyes into a single
view of the world. Take a playing card, tear it lengthways in two, and
turn each piece upside-down. To fit the two halves together again,
they have to be swapped around. Rays of light passing from the air
into the dense, wet material of our eyes are refracted, or bent, so
that the images that fall upon our retinas are upside-down. Like the
two halves of the playing card, the images from the left and right
THE COMMONWEALTH OF THE SENSES 49
eyes have to be swapped around before they can be brought together
again at the primary visual cortex. This cross-over is accomplished
before the optic nerves reach the brain proper, at a ‘junction box;
the optic chiasm.
Because human eyes face forwards, there is a considerable cross-
. over between the views of the left and right eyes, so it is not necessary
for the optic nerves to swap sides entirely. Instead, nerve fibres
carrying the left-most fields of view from both eyes are drawn
together at the chiasm and continue to the right-hand side of the
primary visual cortex. Nerve fibres carrying the right-most field of
view of both eyes supply the left-hand side of the cortex. The result,
in the primary visual cortex, is a single, inverted view of the world.
By damaging his right primary visual cortex, the surgeons were
damaging D.B-s view of the left-hand side of space. That was what
D.B.s doctor, Larry Weiskrantz, assumed. Then one day, D.B’s ophthal-
mologist, Mike Sanders, leaned into the left-hand side of D.B’s field
of view during a test — and D.B. reached out to him.
Could D.B. see the hand? Certainly not. D.B. was a conscientious
subject, and together he and Sanders had very precisely mapped the
area of his blindness. But if D.B. couldn't see, who — or what — had
guided his hand? Weiskrantz and Sanders set about testing the quality
of D.B.s blindness. When they asked him whether a stick held in his
blind area was horizontal or vertical, D.B. could not see the stick,
and could only guess. The odd thing was, D.B’s correct guesses were
more frequent than his mistakes. The difference between correct
guesses and incorrect guesses was large — much larger than could
be explained by chance. D.B. could not see the stick - but some-
thing within him could sense it.
What if they waved the stick around? Suddenly, D.B’s testimony
changed. ‘D.B. “sees” in response to a vigorously moving stimulus,
but he does not see it as a coherent moving object, but instead reports
complex patterns of “waves” .*5 As reported by Weiskrantz, a slowly-
moving stick gives D.B. ‘a kind of “feeling” that something is there’
There's something a bit odd about the inverted commas Weiskrantz
has placed around ‘see’ It’s hardly surprising that D.B., half-blinded
by his operation, was unable fully to appreciate a coherent image of
a stick waving about on his blind side. What he did appreciate was
50 A NATURAL HISTORY OF SEEING

a general perception of movement, rather as we might glimpse, ‘out

of the corner of our eye, a person or object passing through our
peripheral vision. It would be nice to be able to say that some visual
pathways are conscious, and some are unconscious; that conscious-
ness was some kind of mysterious juice that flows along some of the
brain’s channels and not along others. But D.B’s experience shows us
that things are not so simple. Weiskrantz’s accounts show us a man
who, given the right stimulus, becomes aware of being able to see.
Many eyes see in this limited fashion. Just because the tiny
planktonic feeder Pontella has three lenses in each eye, this doesnt
mean it sees the world in incredible detail. On the contrary, these
three lenses have evolved to project light, very carefully, on to just
six photoreceptors. If this is vision, my car alarm can see.
On the other hand, the study of some animal eyes has raised the
unsettling possibility that these animals might be much cleverer than
we thought they were.
Thanks to the work of the American psychobiologist Roger Sperry,
in the 1960s, we are familiar enough with the idea that the left and
right sides of our brain handle different tasks. Our ‘right brains’ are
reputedly more visual and intuitive, whilst our left brains are more
verbal and analytical. Until very recently, we thought this speciali-
sation was unique; a sign of superior, human intelligence. We were
wrong: in 2000, an Italian team, headed by Angelo Bisazza, found
that mosquito fish (Gambusia holbrooki) prefer to check out likely
mates with their left eye and possible predators with their right.”
Evolving from an early visual specialisation, which helped our fishy
ancestors recognise members of their own species, handedness
appears now to be a common characteristic of all vertebrates.
However, when it comes to foxing our assumptions about intel-
ligence, nothing comes close to the box jellyfish Chironex fleckeri.
The venomous Chironex (its sting kills within four minutes) has eight
camera-type eyes like our own, eight slit-shaped eyes and eight lens-
less pit eyes. That’s two dozen eyes, connected, not to a brain, but to
a simple nervous circuit. The eyes appear to be feeding the jellyfish
information it hasn't the wit to use.
Is Chironex, with its fancy eyes, like the elk, whose impressive
antlers are too big and heavy for its head? Is it caught up in some
“
,

THE COMMONWEALTH OF THE SENSES 51

strange, highly specialised arms race that favours sophisticated eyes
for their own sake? As we learn more about the behaviour of the
box jellyfish (they are active, agile swimmers and the only jellyfish
which copulate), an even more unlikely possibility gains credence. It
seems that Chironex is thinking. Understanding how such brains might
_work — brains so unlike our own - is one of our great contemporary
challenges.
TWO

The chemistry of vision

Natalie eats butter like it’s going out of fashion. Bread, for her, is a
butter delivery system. Salted butter will not do: there is too much
water in it. What she wants is unsalted butter - the high-fat stuff
that only the French manage to stay thin on. We give her slices of
butter the way other people give their children slices of apple. She
is an active child, and if we can encourage her to stay that way, she
can have all the delicious, fatty, vitamin-rich butter she can cram
down her throat.
Today, 254 million pre-school age children are deficient in
vitamin A: just under half the children in Africa and just under
three-quarters of the children of south-east Asia. At least a million
child deaths each year are linked to vitamin A deficiency. Of the
survivors, between a quarter and half a million go blind. The
testimony of one field doctor, H.A. Oomen, speaking at a conference
on vitamin deficiencies in 1958, is — fifty years on - still frighteningly
topical:

The over and over repeated experience of discovering a child,

recently blinded, in the arms of the mother; having to tell her that
I now could do nothing more to save its eyesight; remembering
‘THE CHEMISTRY OF VISION 53
that I could have done so with a few spoonfuls of cod-liver oil
some days ago; these things still enter my nightmares.’

There is a chemistry to vision, and unpicking that chemistry

changed the world. Most importantly for us, it opened up a new
way of studying evolution. Most importantly for the health of chil-
dren everywhere, it revealed vital facts about metabolism and
nutrition. It inspired developments in forensic science.
It even got people killed.

1 — The tell-tale eye

‘My boy, I wish you to witness an experiment? He drew from

its case a powerful microscope of French make.
‘What on earth are you going to do, sir?’
The doctor's brilliant eyes flashed with a mystic light as he
replied: ‘Find the fiend who did this crime - and then we
will hang him on a gallows so high that all men from the
rivers to the ends of the earth shall see and feel and know
the might of an unconquerable race of men?
Thomas Dixon Jr, The Clansman (1905)

Odile Gaijean was a martyr to her son’s fecklessness. During the

1980s Odile, a resident of eastern Alsace, ran up huge debts to bail
her son out of a financially disastrous stay in Dakar, the capital of
Senegal. When he turned up on her doorstep and announced that
he was married, his mother was less than impressed. Then his wife
- a former prostitute, Fatou Sarre - turned up.
Domestic bliss it wasn't. The feud between wife and mother-in-
law began with strong words. Soon, they were trading blows. One
day Odile, beside herself, picked up a knife. Fatou, not to be outdone,
picked up a hammer.
Fatou won.
When it was done, she dragged her mother-in-law out of the house
and into the barn. Afraid of what might happen if her husband found
out it was she who had killed his mother, Fatou set about destroying
54 A NATURAL HISTORY OF SEEING

the evidence of her crime. In March 1990, on trial for murder, Fatou
explained to the court that she did not dare leave Odile’s dying vision
intact, for it would surely show her brandishing the hammer.
Which was why she gouged out her mother-in-law’s eyes.
Three years later, there was a similar case, involving a hapless
Cameroonian housebreaker. The press detected a pattern. Was France
witnessing a reawakening of ancient ‘Cameroonian legends’ and
‘African beliefs’?
No. On further interrogation, it turned out that Fatou’s ‘beliefs’ -
such as they were — dated back to a Bollywood film she had seen
one night in Dakar. It was a whodunnit, loosely based, not on ‘ancient
legends’ (Cameroonian or otherwise) but on a much more recent
source — European detective fiction.’

The short story, Claire Lenoir, caused a minor sensation on its publication
in Paris in 1867. Huysmans refers to it in his novel A Rebours, and
twenty years later, there was still enough mileage in the idea for the
author, Villiers de PIsle-Adam, to expand it into a successful novel.
Claire Lenoir is the wife of Césaire, an old friend of the story’s narrator,
Tribulat Bonhomet. On his way to see them, Bonhomet befriends Henry
Clifton, a young English naval lieutenant who has taken a posting in
the South Seas to cure himself of an ill-fated love affair with a married
woman (who bears more than a passing resemblance to Claire).
Later - just before he meets the Lenoirs - Bonhomet stumbles
over a news report:

L’Académie des Sciences de Paris has stated the authenticity of

certain surprising facts. It can be asserted that the animals
destined to our nourishment - such as sheep, lambs, horses
and cats — conserve in their eyes, after the butcher’s death stroke,
the impression of the objects they have seen before they die. It
is a photograph of pavements, stalls, gutters, of vague figures,
among which one almost always distinguishes that of the man
who has slaughtered them .. .3

It is an ill omen and, sure enough, things rapidly take a turn for the
worse. His friend Césaire falls desperately ill. Somehow Césaire has
THE CHEMISTRY OF VISION 55
discovered his wife has been unfaithful; on his death-bed, he swears
a terrible oath of revenge.
A year later, news arrives that Henry Clifton has been brutally
murdered by a vampiric assailant. Césaire’s widow Claire takes to
her bed, driven out of her mind by persistent nightmares. Bonhomet
is with her when she dies and, noticing a blurry image lingering in
her eyes, he examines them with the aid of his trusty ophthalmoscope.
What he sees:

no language, dead or alive ... could, under the sun and under the
moon, express in its unimaginable horror . . . I saw the skies, the
far-off floods, a great rock, the night and the stars! And upright,
on the rock, larger in height than the living, a man . . lifted with
one hand, towards the abyss, a bloody head, with dripping hair
... and, in the severed head, the features, frightfully obscured, of
the young man I had known, Sir Henry Clifton, the lost lieutenant!

It is strange, at this remove, to imagine the popular curiosity generated

by the ophthalmoscope; a device, combining magnifying lenses and
a light-source, with which one can view the inside of the living
eyeball. The nearest contemporary equivalent would, I suppose, be
the craze, a few years ago, for the ‘magic eye’ posters that conceal
three-dimensional figures in a computer-generated pattern of appar-
ently random blobs. They too, as we have seen, have their serious
scientific side, and now that the fashion for them is past, like the
ophthalmoscope, they have largely slipped out of the public gaze,
back into the somewhat rarefied world of vision science.
The way de l’Isle-Adam collides the two great scientific crazes of
his day - ophthalmology and photography - is a splendid example
of the fantastic at work. The eye is the soul’s camera, and the ophthal-
moscope reaches in to spy upon the soul!
If the idea seems trite today, it has only become so through repe-
tition. For de l'Isle-Adam’s readers, Claire Lenoir addressed the bodily
anxieties generated by the era's huge advances in surgery and imaging,
as surely and as disconcertingly as the the films of David Cronenberg
address the bodily anxieties of our own, increasingly anonymous
and ‘virtualised’ day.
56 A NATURAL HISTORY OF SEEING

Exposed, the body reveals its affinity with mechanism: the eyes
are cameras, and to the well-equipped observer they reveal the secrets
of the heart as surely as a daguerreotype reveals the interior of the
sitter’s drawing room. This is a great opportunity, but a risky one, at
least in the uncanny world of Claire Lenoir. At some level, humans
are not machines; they are messy, unpredictable, generators of night-
mares. You can look, but you may not like what you find:

And Science, the old queen-sovereign with clear eyes, with perhaps
too disinterested a logic, with her infamous embrace, sneered in
my ear that she was not, she also, more than a lure of the Unknown
that spies on us and waits for us — inexorable, implacable!

Such was the public’s fascination for the new-fangled excitements

of forensic science, that the fiction which followed Claire Lenoir
become steadily less like ghost stories, and more like police proce-
durals. In 1897, Jules Claretie wrote L’Accusateur, in which the eye
of a murdered diplomat, dissected by the police, reveals the blurred
image of the face the victim saw as he died.* Jules Verne’s Les Freres
Kip (1902) sits even more comfortably within the detective genre in
its tale of innocent brothers, arrested for a murder they did not
commit, released at the eleventh hour on the evidence of a retinal
photograph. Three years later, retinal photography revealed the iden-
tity of a rapist in Thomas Dixon Jr’s jaw-droppingly racist novel The
Clansman - the basis for D.W. Griffith’s film The Birth of a Nation.
Philosophers and scientists could no more resist the analogy
between vision and photography than could the reading public. In
1859, Charles Darwin's The Origin of Species explained how, given
large numbers, time, and a testing environment, living things acquire
an efficient, apparently designed form. In acquiring vision, had nature
anticipated the very solutions hit upon by those pioneers of photog-
raphy, Wedgwood, Niépce and Daguerre?
For the retina to behave like a camera’ plate, it had to contain some
light-sensitive chemical, analogous to the silver nitrate film covering
the glass slides on which the earliest photographs - daguerreotypes -
were captured. The presence of a reddish pigment in the retinas of
frogs and squid was established in 1851 by the German anatomist
THE CHEMISTRY OF VISION 57

Wilhelm Kiihne.

Heinrich Miller (1820-1864). In 1876, his brilliant young compatriot,

Franz Christian Boll (1849-1879), narrowed the pigment’s source
down to one particular group of cells — the ‘rods. The unusual shape
of these cells, coupled with the way they were packed so closely together
across the retina, was a clear indication that they possessed an optical
function. Were these the cells that saw?
Boll noticed that, when a frog died, the reddish colour of its retinas
bleached away, so that after about a minute, they appeared colourless.
At first he thought this was a consequence of the frog’s death. But then
he discovered that even dead frogs retained healthy reddish retinas for
up to twenty-four hours, provided they were kept in the dark. If light
bleached a pigment stored in the ‘rod’ cells of the frog retina, might that
chemical event not convey the impression of light to the frog’s brain?
In 1879, Boll's experiments were cut short: he was only thirty
when he died, a victim of the tuberculosis that had blighted his
otherwise glittering professional life. But he had done enough to
convince his peers at the Berlin Academy that “This change in the
58 A NATURAL HISTORY OF SEEING

outer segments of the rods forms indisputably a part of the process

of vision.’®
Among Boll’s admirers was Wilhelm ‘Willy’ Kithne (183 7-1900).
A powerful figure in the field of physiology, Kihne had recently
succeeded Hermann von Helmholtz as professor of physiology at
Heidelberg. Kiihne took up Boll’s discoveries with ‘fiery zeal.
Boll had called his pigment ‘visual red’; Kiihne reckoned it was
more purple, and changed the name accordingly. Otherwise, Kiihne
agreed with Boll —- and acknowledged his debt - on almost every
important point. If the retina was the eye’s ‘photographic plate, then
Boll was surely right, and ‘visual purple’ was the eye's equivalent of
the camera’s silver nitrate. Kihne hoped to complete Boll’s work by
extracting evidence from the retina itself.
Working with the simplest equipment in a darkroom, Kiihne found
a way to maintain retinas so that he could study them closely. In the
dark, the dissected retinas remained purple. Light bleached them,
but not instantly. Kihne observed distinct stages in the bleaching
process, from purple to orange, yellow, and buff, until the retina
became entirely transparent.
Fifty years passed before any new knowledge was added to the
meticulous description of visual purple assembled by Kiihne over
two busy years. Indeed, the pile of unanswerable questions Kihne
had amassed concerning the stuff might have obscured his actual
achievement, had he not gone to extraordinary and repeated lengths
to publicise his central finding: that the eye behaves like a camera.
On 16 November 1880, in the nearby town of Bruchsal, a young
man was beheaded by guillotine. The body was taken to Heidelberg,
where Kiihne, in his capacity as professor of physiology, was waiting.
In a gloomy room, its few windows screened with red and yellow
glass, Kihne dissected the dead boy’s eyes. Ten minutes later, he
showed his colleagues a sharp pattern on the surface of the left retina.
This, Kihne told the company, was an optogram: a dying vision,
preserved as a pattern in that delicate and unstable pigment, visual
purple. The trouble was, Kiihne’s sketch failed to match any object
that may have been visible to the felon as he died.
We know now that Kiihne had been confronted - and bested —
by a curious specialism of the human eye. Its area of focused vision
£
od
THE CHEMISTRY OE VISION 59

A rabbits dying vision, captured by Kiihne’s most successful optogram.

— the fovea — is tiny in comparison to the size of the whole retina.

Even more puzzlingly, the fovea has very few rods. Kiithne had better
success recording the optograms of frogs and rabbits. This account
of Kiithne’s early and modestly successful ‘optography’ is by the
biochemist George Wald - himself a Nobel prize-winner for his
studies of visual pigments. (Wald refers to visual purple by its
currently preferred name, rhodopsin.)

One of Kihne’s early optograms was made as follows. An albino

rabbit was fastened with its head facing a barred window. From
this position the rabbit could see only a gray and clouded sky.
The animal's head was covered for several minutes with a cloth
to adapt its eyes to the dark, that is to let rhodopsin accumulate
in its rods. Then the animal was exposed for three minutes to
the light. It was immediately decapitated, the eye removed and
cut open along the equator, and the rear half of the eyeball
containing the retina laid in a solution of alum for fixation. The
next day Kiihne saw, printed upon the retina in bleached and
unaltered rhodopsin, a picture of the window with the clear
pattern of its bars.°

The key word in this passage, of course, is fixation. A living retina

has no interest in fixing images. Far from capturing perfect stills, it
60 A NATURAL HISTORY OF SEEING

offers up a constantly changing view. As Kiihne himself wrote, ‘the

retina behaves not merely like a photographic plate, but like an entire
photographic workshop, in which the workman continually renews
the plate by laying on new light-sensitive material, while simultane-
ously erasing the old image.
It is impossible now to say whether Kiihne’s macabre autopsy on
the felon from Bruchsal was a rational and necessary extension of
his laboratory work, or an aberrant piece of showmanship. His work,
and the work of his predecessor Franz Boll, had been widely
published in magazines across the globe. In Britain, Fortnightly
Nature, The Athenaeum and The Nineteenth Century ran articles, as
did Harper’s Weekly in America. And we should not under-estimate
the pressure Kthne and other researchers were under to extend the
frontiers of forensic science. Since 1857, when the first, probably
apocryphal story of optography appeared in Notes and Queries,
curiosity about the possibilities of forensic photography regularly
found its way into court transcripts. By 1869, the French Society of
Forensic Medicine was so concerned that the courts might start
admitting ‘optographs’ as evidence in murder trials, it asked Dr
Maxime Vernois to conduct a scientific study. Vernois set about his
task with enthusiasm, violently’ slaughtering seventeen experimental
animals ‘under bright lights. They died in vain:

It is impossible to find upon the retina of a victim the portrait

of its murderer or the representation of whatever object or physical
trait that presented itself to its eyes at the time of death.’

Despite the failures recorded by Vernois and Kiihne, enthusiasm for

the idea persisted, and spiced up genuine murder investigations. In
1888 Walter Dew - a London policeman later famous for catching
the murderer Dr Hawley Harvey Crippen — witnessed ‘the most
gruesome memory of the whole of my Police career’ when he entered
13 Millers Court in Whitechapel, where lay the disordered remains
of Mary Kelly, a victim of Jack the Ripper. ‘Several photographs of
the eyes were taken by expert photographers with the latest type
cameras, Dew remembered in his autobiography, in the ‘forlorn hope’
that an image of her killer was retained on the retina.’
THE CHEMISTRY OF VISION 61

Is Dew’s memory reliable? Perhaps not: but the idea was certainly
on everyones mind. At the inquest of Annie Chapman, another of
Jack's victims, the jury foreman asked police surgeon Dr George
Bagster Phillips, “Was any photograph of the eyes of the deceased
taken, in case they should retain any impression of the murderer?’
_Thave no particular opinion upon that point myself; Phillips replied.
‘I was asked about it very early in the enquiry and gave my opinion
that the operation would be useless, especially in this case’
There is, as far as I have been able to ascertain, only one, extremely
dubious, example of a conviction attained by an optogram. It is from
the Sunday Express and is quoted in Veronique Campion Vincent’s
invaluable paper “The tell-tale eye; published in the journal Folklore in
1999. It seems that readers of the edition published on 4 July 1925
thrilled to the tale of Fritz Angerstein, a resident of Limberg in Germany,
who killed eight members of his household with a hatchet. The police,
seeking a quick confession, told the court that they had taken a photo-
graph of the retinas of Angerstein’s gardener, which revealed ‘a picture
of Angerstein with his raised arm gripping a hatchet? Hearing this,
Angerstein threw in the towel and confessed to the killings.
And the police photograph? Most likely, it was simply a tale, spun
to demoralise a gullible defendant. It was never made public.

2 —- A question of diet

Nutrition has often been the subject of conjectures and

ingenious hypotheses but our actual knowledge is so insufficient
that their only use is to try to satisfy our imagination. If
we could arrive at some more exact facts they could well
have applications in medicine.
Francois Magendie®

Robert Boyle (1627-1691), the fourteenth child of the fabulously wealthy

Earl of Cork, never needed — and never bothered to acquire - gainful
employment of any kind. Instead, he became one of seventeenth-century
Europe’ leading intellectuals. Boyle was a key advocate of experiment
as a path to truth. ‘Our Boyle; his friend Henry Oldenburg wrote,
62 A NATURAL HISTORY OF SEEING

with wry wit, ‘is one of those who are distrustful enough of their
reasoning to wish that the phenomena should agree with it. But
there was another side to Boyle, which could not resist a good tale
or a wild claim, provided it was well-expressed and intelligently
defended. These tall stories pepper his writings and correspondence,
testifying not only to the sharpness of his intellect but also to the
breadth of his imagination.
Consider this interview with Major Edward Halsall. (Boyle grants
Halsall anonymity in his account, but we can be fairly sure it was
him, thanks to the researches of the historian Michael Hunter, who
brought this story to my attention.) Halsall was a royalist who was
imprisoned during the Cromwell era for his part in the murder of
Anthony Ascham, the Commonwealth’s envoy to Spain, in 1650.
Halsall, ‘being accused at Madrid of a state crime, had irons put
upon his feet, and was kept for twenty months in a room, which
though otherwise not unfurnished was, by reason of its darkness,
reckoned among the dungeons, for there was no window to it? During
that time ‘he saw no light at all, save when the gaolers came from
time to time with candles to bring him wine and other provisions,
and see that he made no attempt to escape’ It took Halsall’s eyes
seven months fully to adjust to the dark, and even then he saw things
only dimly. By the end of his imprisonment, however, Halsall’s night
vision had grown prodigious: “ .. he could see the mice that used to
feed upon his leavings plainly run up and down the room and ‘he
could see well enough to make a mousetrap with his cup.”
Halsall claimed that, had he been given books and paper, he would
have been able to see well enough to read and write. Unlikely as this
is — the discrimination of print requires considerably more light than
the discrimination of objects - everything else about his account
seems eminently plausible.
It is extremely unlikely that anyone with access to this book -
including me — has the remotest experience of the real capabilities of
night vision. We live our lives doused in light, and we think of our
night vision as somewhat second-rate, compared to the night vision of
cats or foxes or owls. But although we lack certain exotic adaptations
that can aid nocturnal vision, we can still detect light that is a billionth
the strength of daylight. This is sensitive enough that, in good conditions,
THE CHEMISTRY OF VISION 63

we could see the flame of a single candle seventeen miles away. Very
few species could do better. (It is our relatively poor hearing and sense
of smell that lets us down: nocturnal species rely very much on these
senses to supplement their vision.)

ne so many of our remarkable physical and sensory abilities, night

vision is really appreciated only when it goes wrong. Night blindness
is perhaps the most venerable of all eye diseases; it has excited
medical opinion for at least as long as cataract.
We can be reasonably certain that ‘night blindness’ has been known
in Egypt, India and China since records began. In all three cultures,
the cure seems to have been roasted or fried animal liver. This excites
today’s historians, since we know liver is a good source of vitamin
A, and vitamin A plays a central role in vision.
The Berkeley nutritionist and historian George Wolf is prepared to
stake a claim for the ancient Egyptians, who — according to the Ebers
papyrus of 1520 BC — treated an inability to see at night by squeezing
the juices from an animal's liver straight into the patient's eyes.'! Since
the liver stores about ninety per cent of the body’s vitamin A supply,
this gruesome-sounding procedure would work. Any liquid of that
sort passing down the naso-lacrimal duct would be absorbed into the
blood stream, as the makers of eye-drops well know. To this day, both
night blindness and painfully dry eyes (the two conditions go hand
in hand) are treated with eye-baths of fish liver oil.
Wolf’s case is not altogether sound, because the Ebers papyrus —
the earliest medical text we know of - lists illnesses only by name.
Is ‘night blindness’ a lack of night vision, or the inability to see except
at night (photophobia)? Or is it neither? There is no way to be sure.
Wolf’s detractors go further, pointing out that in early medical texts,
fried or roasted animal liver seems to be the cure for everything.
They also point out the risk of reading magical texts through medical
spectacles. The appearance of the liver is affected by numerous
illnesses and conditions, and this variable appearance has made it
especially useful to soothsayers throughout the ages, who descried,
in that lump of protean flesh, everything from the outcome of an
eye infection to the propitious date for a war.
If the Egyptians did know how to cure night blindness, the knowl-
64 A NATURAL HISTORY OF SEEING

edge was lost: Aulus Cornelius Celsus, the Roman encyclopedist, born
25 BC (about 1,500 years after:the Ebers papyrus was written), makes
no mention of liver or fish oil in his otherwise spot-on description of
chronically dry eyes. Also, he makes no mention of night blindness:

There is a kind of dry inflammation of the eyes called by the

Greeks xerophthalmia. The eyes neither swell nor run, but are
nonetheless red and heavy and painful and at night the lids get
stuck together by very troublesome rheum . .”

It all sounds innocuous enough. Celsus - an encyclopedist, rather

than a medical man - saved his readers from the more unpalatable
details of the diseases he described. To get a full measure of the seri-
ousness of xerophthalmia, we do better to turn to a source closer to
home. This is Colonel Robert Henry Elliot (1864-1936), writing of
his experiences as an English ophthalmologist in India:

Never, whilst memory lasts, can one obliterate the mental pictures
of those pitiful little bundles of marasmic, apathetic humanity, lying
in the arms of gaunt women, on whose faces and forms famine
had laid its devastating hands. Their faint, feeble, fretful wails ring
still in one’s ears today, summoning up visions of wasted stick-like
limbs, of distended abdomens, of dry, inelastic, scurfy, scaly skins,
of hair scanty, brittle and dry, and of sightless desiccated eyes . . .

From Celsus’s dry eyes to the full-blown deterioration of the cornea

and, ultimately, the ghastly systemic failures described so movingly
by Colonel Elliot, the distance is great. It is hard to believe that they
are, in essence, degrees of the same condition. Even more bizarre is
the notion that this condition has nothing to do with disease, nor
with starvation. What other explanation can there be for the failure,
first of vision, and then of the whole body?
In 1816, in an attempt to understand the biological value of certain
foodstuffs, Francois Magendie - a child of revolutionary Paris and a
former physician - set out to observe the effects of a restricted diet.
He was particularly interested in what role nitrogen has to play in diges-
tion. The answer he got back, after ten years of painstaking work, was
,a

THE CHEMISTRY OF VISION 65

none at all:‘As so often in research; Magendie wrote, with what bitterness

we may imagine, ‘unexpected results had contradicted every reasonable
expectation: But in the pursuit of this knowledge, Magendie had stum-
bled upon a striking, if unpleasant, discovery: he had found that he
was able to starve his experimental dogs to death on diets that should,
on the face of it, have given them all the energy they needed for life.
By his own account, Magendie ‘placed a small dog about three
years old upon a diet exclusively of pure refined sugar with distilled
water for drink; he had both ad libitum’ By the third week the animal,
already weakened, lost its appetite, and developed small ulcers in
the centre of each cornea. The ulcers spread, then the corneas lique-
fied. Shortly afterwards, the dog died.
Magendie tried other nutritious foods. “Everyone knows that dogs
can live very well on bread alone, he confidently asserted — but when
he put this to the test, he found that ‘a dog does not live above fifty
days: The most calorific foods in Magendie'’s pantry - wheat gluten,
starches, sugar, olive oil - were not enough for life. This was totally
unexpected. There was something missing - something available
only as part of a varied diet - but what?
It was the Japanese doctor Masamichi Mori (1860-1932) who
drew serious medical attention to the problem. In his 1904 paper -
the first account of a xerophthalmia epidemic and the first really
good large-scale study ever made of dietary deficiency — Mori details
the plight of 1,400 children between the ages of two and five. Some
had night blindness; others suffered dry and ulcerated eyes. Some
had eyes that had completely disintegrated. Some were dying. They
were suffering from xerophthalmia, known to Mori as ‘hikan.
Mori’s study of hikan is an exemplary piece of medical detective
work. Local lore said that inappropriately early weaning could bring
on the condition; and sure enough, rural women tended to withdraw
the breast much earlier than women who lived in towns, where hikan
was relatively rare. At the same time, Mori noticed that the children
of fishermen seemed strangely immune to the condition, even though
they were weaned earlier than urban children.
Mori concluded that animal oils — and fish oils, especially — were
essential to preserve health. The more fats you ate, the less likely you
were to suffer from hikan.
66 A NATURAL HISTORY OF SEEING

As the nineteenth century waned and turned we find many

insightful descriptions of thecondition; among Russian peasants
during severe Lenten fasts, children in a Paris orphanage, Russian
soldiers in the Crimea, waifs and strays in Berlin, and even among
the soldiers of the Confederate army in the United States, where the
condition was blamed on the ‘meagre diet, absence of vegetables and
vegetable oils, and other depressing influences of a soldier's life.
The trouble was knowing what to do about it. The great public
health advances ushered in by Pasteur’s germ theory had drawn atten-
tion away from questions of diet. The experience of the German doctor
S. Kuschbert is typical: in 1884, he noticed night blindness and dry
eyes in twenty-five of the orphan children with whom he was working.
On further investigation, he discovered that the ulcers on the chil-
dren's corneas contained bacteria. He drew the obvious wrong conclu-
sion: that the children had caught one of Pasteur’s famous germs.
Even allowing that some conditions were caused by a bad diet —
rickets and scurvy being the best known - no one had isolated the
factors that protected the body against them. These dietary factors
had to be tiny — so tiny as to be ephemeral - and how could a body
be so dependent on mere traces of something?
Incredulity ran so high that it took the evidence of a world war
to overcome it. The maritime blockade of continental Europe during
the First World War led to malnutrition in several countries. Analysed
by physicians and nutritionists, these cases held the clue to the
greatest medical advance since Pasteur: the discovery of vitamins.
In 1917, Carl E. Bloch (1872-1952) was the physician in charge of
the paediatric clinic of the University of Copenhagen and of a children’s
home which cared for eighty-six infants under the age of two. The home
had two buildings. The first cared for newborns and the sick. The
other had two rooms, each housing healthy babies. As you would expect,
the healthy babies enjoyed a more varied diet than the sick ones.
The outbreak, when it came, was sudden and devastating. Bloch
reported forty cases of xerophthalmia among the children under his
care. ‘The children were sensitive to light, and in the early stages some
were night blind. Dryness of the conjunctiva followed, spreading to the
cornea. He also reported ‘a viscous discharge from the eyes’ All the cases
occurred between May and June 1917, among the healthy children.
*

-THE CHEMISTRY OF VISION 67

How could children on a varied diet fall victim to a dietary illness

to which already sick children, fed on a narrow diet of gruel and
whole milk, were immune? Bloch found the answer in the matron’s
meticulous records: following a minor attack of diarrhoea among
the healthy children, she had swapped the traditional beer-and-bread
soup, made with whole milk, for an oatmeal gruel with rusks.
And there it was: whole milk. Children deprived of whole milk
succumbed to xerophthalmia. The others, even the ill ones, were immune.
Bloch wasted no time. Along with whole milk, he fed the xero-
phthalmic children other fats and oils. Cod-liver oil turned out to be
especially effective: even apparently hopeless cases recovered within a
week. The xerophthalmia never returned.A year later, when the German
submarine blockade drove the prices of butter and whole milk well
beyond the reach of the poor, the Danish government introduced butter
rationing; every adult and child could obtain 250 grammes of the stuff
every week. Xerophthalmia, which was never rare in Demark and which
had reached epidemic proportions during the war years, virtually
vanished overnight. Only when hostilities ended and butter was no
longer rationed did it make a brief and short-lived resurgence.
The cluster of related compounds which travel under the banner
‘vitamin A’ were finally isolated between 1912 and 1914 by two inde-
pendent American teams. In the twenty years that followed, thirty
clinical trials of vitamin A established its importance, not just for
eyesight, but for warding off a host of infections. Vitamin A, the
‘anti-infective vitamin, was vigorously promoted by doctors and phar-
maceutical companies. Not until 1947 was a way found to synthesise
it; the inter-war generation had to resign themselves to daily doses
of cod-liver oil, whose ghastly taste was often disguised with treacle.

Many scientists acquire their sense of vocation early. George Wald

(1906-1997) was not one of them. Born in New York, the youngest
, had always been good with his hands
Wald
son of Polish immigrants
~ he and his friends built a simple radio so they could listen to-ball
games — but a visit to an engineering firm where a friend of his
father worked convinced him that engineering was not for him. After
taking a vaudeville act around local Jewish community centres, Wald
decided that his showmanship marked him out as a born lawyer. As
68 A NATURAL HISTORY OF SEEING

a pre-law student at New

York University, he soon
learned otherwise.
Eventually, he fell into
zoology — and never looked
back.
By 1932, the twenty-six-
year-old postgraduate was
on the first leg of an extra-
ordinary tour of German
universities, pursuing the
mysterious relation between
vitamin A and vision.
The adventure began
auspiciously. Arriving in
George Wald. Berlin with his wife Frances,
Wald went to work with
Otto Warburg — a Nobel laureate, no less. Wald and Warburg knew
that vitamin A played some role in vision because a lack of vitamin
A brought on night blindness. But how direct was that link? Wald’s
first job was to confirm the results of a dramatic experiment performed
by the Yale ophthalmologist Arthur Meyer Yudkin, who had fed the
freeze-dried retinas of slaughtered farm animals to vitamin A-deficient
rats. These rats recovered as quickly as rats given cod-liver oil - a
long-established source of vitamin A. The conclusion — that retinas
contain vitamin A — was confirmed by Wald’s chemical tests.
Next, Wald went to Zurich, to the laboratory of the Russian-born
chemist Paul Karrer, to find out whether vitamin A was a component
of the retinal pigment, ‘visual purple’ or rhodopsin, that Boll and
Kithne had already identified. Wald and Karrer collected retinas from
cattle, sheep, and pigs, extracted them with solvents, and found
vitamin A in all of them.
The final leg of Wald’s journey — to the Kaiser Wilhelm Institute for
Medical Research in Heidelberg - was meant to be the crowning event
of his tour. There, he was to meet Otto Meyerhof, an expert in the
biochemistry of muscle tissue. But Adolf Hitler had come into power
only months before, in January 1933, and the Nazi authorities had
-THE CHEMISTRY OF VISION 69

begun to remove some of the finest minds in the country from their
academic posts. In Heidelberg, the Institute was putting up a fine show
of resistance; its president and director shared an unerring knack for
losing official letters. Nevertheless, the local Nazi party was agitating
for the removal of Meyerhof, who was, like Wald, a Jew.
Even as Wald was pushing paper around Meyerhof’s laboratory,
‘ waiting for him and his team to return from holiday, Wald’s backers,
the National Research Council, made jittery by the deteriorating
political situation, told him to cut short his trip and head home. At
the same time, a delivery arrived. Someone must have got the dates
wrong — the building was virtually deserted and here, stacked in the
hall, were crates and crates of frogs. Three hundred frogs.
It was an opportunity George Wald seized with both hands. If he
was not to work with Meyerhof, he could take advantage of the great
man’s extraordinarily well-equipped laboratory. He set about re-
creating the findings of Kiihne, who had seen the purplish rhodopsin
of frog retinas pass through distinct colour shifts on its way from
purple to clear. First, light bleached ‘visual purple’ to something Kiihne
had dubbed ‘visual yellow. On further exposure, visual yellow then
became transparent: Kiihne called this stage ‘visual white. Wald
performed chemical tests to find out what ‘visual white’ was made of.
It was pure vitamin A.
What of ‘visual yellow’? Visual yellow was very similar to vitamin
A;Wald found, but with one telling difference: in the dark, it converted
readily back into rhodopsin. He had begun to uncover the biochem-
istry of vision.
Wald’s account of the visual cycle goes something like this: the
pigment rhodopsin consists of a membrane protein (an ‘opsin’)
connected to a light-reactive component, ‘retinal. When light hits
the rhodopsin pigment, the retinal changes shape and the pigment
changes colour, from purple to yellow. In the dark, the retinal flips
back to its former shape, and the pigment turns purple again. Over
time, the retinal decays. Extra vitamin A is needed to support the
re-combination of retinal and opsin into rhodopsin, and if the body
is deficient in vitamin A, night blindness follows.
Safely back at Harvard, Wald contemplated the life's work now
stretching ahead of him. The questions his work threw up fell under
70 A NATURAL HISTORY OF SEEING

two broad headings. First, how prevalent was rhodopsin? The eyes
of frogs, sheep, cows, pigs and people all contained rhodopsin. Was
this a coincidence, or was rhodopsin truly universal? Second, how
important was rhodopsin to vision? Maybe rhodopsin was just one
of many different pigments used by the eye. Maybe pigments were
just the easiest to spot of several different visual mechanisms. The
number of possible complexities was enormous.
It took Wald, his wife and co-worker Frances, his long-time collab-
orator Paul Brown and untold others who passed through their labo-
ratory about twenty years to work through all the possibilities. In
that time, they found virtually nothing to complicate the original
picture: every single eye, with no exceptions, uses rhodopsin to see.
Some animals, it is true, do supplement rhodopsin with additional
pigments. None the less, every animal makes use of rhodopsin.
Rhodopsin is a universal visual pigment.
Complications did set in concerning rhodopsin itself. Wald and
his co-workers found that it comes in many different flavours.
Minuscule genetic differences in the opsin membrane can transform
the behaviour of the whole molecule, making it react most strongly
to quite different wavelengths of light. This is what powers colour
vision in the great majority of animals.
So it was that George Wald and his collaborators - in particular
his second wife, Ruth Hubbard - led the post-war revolution that
changed biology from a cellular to a molecular science. In his Nobel
lecture of 1967 he explained:

That early Wanderjahr in the laboratories of three Nobel laure-

ates — Warburg, Karrer, Meyerhof — opened a new life for me:
the life with molecules. From then on it has been a constant
going back and forth between organisms and their molecules
- extracting the molecules from the organisms, to find what
they are and how they behave, returning to the organisms to
find in their responses and behavior the greatly amplified
expression of those molecules."

Molecules, like animals, have a story to tell. Molecules, too, have a

history.
THREE

How are eyes possible?

1 — Building blocks

Woods Hole, a town in Cape Cod, Massachusetts, has been a Mecca

for America’s marine scientists for over a century. Here, in the mid-
1930s, at the town’s Marine Biological Laboratory, George Wald
pioneered a new way of looking at the story of life on Earth: rather
than trace the history of whole animals, why not trace the history
of their ingredients? Why not trace the history of their molecules?
Wald’s work at Woods Hole threw up as tidy an example of such a
history as one could wish for. He discovered that amphibians and
salt-water fishes both use a common form of rhodopsin — the same
form used by the rods in our own retinas. Fresh-water fishes, on the
contrary, use a form of rhodopsin capable of detecting longer wave-
lengths of light. (Water absorbs long-wavelength light, which is why,
the deeper you swim, the more blue everything becomes. The light
dappling the shallows where most fresh-water fish swim has a longer
average wavelength than light that has perhaps had to penetrate half
a kilometre of ocean.) Salmon, which spawn in fresh water, use
72 A NATURAL HISTORY OF SEEING

fresh-water rhodopsin; other species, which spawn in salt water, use

salt-water rhodopsin. Tadpoles, Wald found, swap salt-water rhodopsin
for fresh-water rhodopsin during their metamorphosis into frogs.
Wald concluded that salt-water rhodopsin is older than fresh-
water, and that fresh-water rhodopsin has evolved from the salt-
water form. When he wrote up these and similar findings in
provocative articles with titles like “The Origin of Optical Activity,
they opened the door to a whole new field of evolutionary biology.
In 1990, an ambitious project was launched to compare the sequence
of letters that spell out the genes for all the different kinds of rhodopsin
then known. Genetic sequences had by then been established for four
fruit-fly opsins, one opsin from an octopus, all four opsins from humans
(one from the rod cells, and three from our three different types of
cone) and one rod opsin each from chickens, sheep, cows, and mice.
The results showed that these different opsins all descended from a
common ancestor. In the 600 million years it had taken for animals
as diverse as fruit flies and chickens and people to arise from that
shared ancestor, rhodopsin —- a molecule common to them all — had
undergone only the most trivial changes.
These findings are impressive, but not unexpected. In 1882, the
German physiologist Theodor Engelmann speculated that, for it to
occur in the eyes of virtually every animal, rhodopsin must have
existed at a very early stage of evolution, most probably before eyes
themselves had arisen. Engelmann’ conjecture arose out of his work
with photosynthesis: he had discovered that some bacteria are able
to make food from water and the carbon dioxide in the air, using
the energy from sunlight. Nearly a century later, in the late 1970s,
Engelmann’ hunch paid off: a form of rhodopsin, harnessed variously
as a defence against ultraviolet light, and as a mechanism of photo-
synthesis, was found in bacteria.
Rhodopsin’s use in vision, which concerns us here, is a story that
begins surprisingly early in the evolution of life. The ability to detect
not just the presence of light, but also its intensity, wavelength and
the direction it is coming from, is present among humble bacteria,
fungi, algae, and even the single-celled protozoa.
One of these, Euglena gracilis, an algal flagellate, uses its photore-
ceptive pigments to acquire the energy to drive its whip-like tail. This
‘HOW ARE EYES POSSIBLE? 73
‘propeller’ is constructed by the complex folding of the cell membrane,
and is arranged to drive Euglena towards the light. When Euglena
enters sufficiently bright conditions, it begins to photosynthesise.
When the light declines; the mechanisms of photosynthesis break
down and Euglena survives by eating, making it truly both a plant
_and an animal. Though Euglena contains rhodopsin, its tail is driven
by a different set of pigments, the flavins. Flavins are as widespread
in nature as rhodopsin and have an analogous evolutionary history.
They crop up in fungi, algae and plants and appear to have inde-
pendently arisen at least three times. In the higher animals, ourselves
included, flavins control the metabolism of the retina; they, too, have
found a role in vision.
Euglenas pigment globules have no particular arrangement, and
though they power a directional tail, they themselves resemble nothing
more sophisticated than an ‘umbrella or light filter. The multiple
pigment layers inside the alga Chlamydomonas, on the other hand,
suggest a complex optical structure. The layers may work as reflectors
to contain and concentrate the light, increasing
the likelihood that it will react with the pigment.
Is this the world’s most primitive retina? The
protozoan Erythropsis boasts an even more
sophisticated eyespot, topping its highly
ordered crystalline structure with a large, trans-
parent, spherical lens.
The closer we come to vision proper, the
more sophisticated the behaviour of an animal
— even a single-celled animal - can be. In 1975,
the Italian researchers Ester Piccinni and
Pietro Omodeo were astounded to discover a
group of algae which used their simple eyes
to hunt for food. Lacking anything
approaching a nervous system, these tiny

Gordon Lynn Wallss idealised rod cell; its main

body and pigment stack are connected by a tail-
like filament.
74 A NATURAL HISTORY OF SEEING

creatures could spot prey and entangle it with explosively released

threads.
Evolution cannot un-invent. The basic building blocks of complex
structures like the eye were available long before they were recruited
for sophisticated tasks like vision. Human lenses are derived from
proteins that protect bacteria from rapid rises in temperature. The
flexible guanine mirrors which make a cat's eye glow in the dark also
gas-proof the swim bladders of fish. The red and yellow pigments in
the eyes of insects also colour the wings of butterflies. We have noted
the slinky-like behaviour shared by retinal rods and muscle cells; it
may be worth adding that the filament connecting the two halves of
the retinal rod cell is built exactly the same way as Euglena’s tail.
Eyes have been around for at least 538 million years — but their
constituent elements are more ancient still.

2 - The third eye

Light hits a rod in my eye, and bleaches the pigment within. A

cascade of chemical events shuts down the rod’s ‘ion channels’ -
proteins that pass an electrical current around the cell. With its
ion channels shut down, production of a chemical, glutamate, is
stopped. This sudden chemical ‘silence’ excites a neighbouring nerve
cell and sets it chattering. News passes from neighbour to neighbour,
from nerve cell to nerve cell: the beginning of vision.
How curious, that light should excite our eyes by turning off our
photoreceptors. Not every eye works this way: invertebrate photore-
ceptors grow excited in the light. And if it’s variety you're after, then
there are all manner of light-detecting mechanisms — ones that have
nothing to do with vision - that harness quite different chemical
reactions. It was from these light-detecting mechanisms that vision
evolved, for living things - with eyes or without, animal or vegetable
— have always been superbly attuned to light.
Most living things get by without eyes and settle for a simpler,
cheaper light sense. The ‘phytochromes in plants, which are sensitive
to red and infra-red light, control how leaves and flowers fold and
open and turn to the sun. They also control the plant’s germination,
‘HOW ARE EYES POSSIBLE? 75
growth and flowering. Many animals - including animals with well-
developed image-forming eyes - use a generalised light sense to
regulate their metabolic rhythms and trigger important behaviours.
The burrowing sea anemone Calamactis praelogus bends towards
the light. The flatworm Convoluta roscoffensis emerges from the sandy
_Shallows at low tide so its symbiotic algae can photosynthesise. Pass
a shadow across the sea urchin Diadema setosum and it will withdraw,
waving its spines in a complex defensive pattern.
This simple movement towards or away from light has evolved, in
higher animals, into an ambitious scheme of migrations and timed
mating cycles. Salamanders and frogs navigate by a dermal light sense,
suggesting that their skins are sensitive to polarised light. A generalised
dermal light sense controls the flight rhythms and reproductive
systems of birds. Pigeon chicks, hooded so that they cannot see, will
still greet the light by wagging their heads and stretching their legs.
The sea hare, a kind of sea slug, uses light to regulate its pattern of
feeding and sleeping. However, its sensitivity to light resides not in its
skin, but in its nerve fibres. Some animals, like leeches and starfish,
have skins which are generally light-sensitive. Others, like the horseshoe
crab, have a number of photosensitive sites which regulate their ‘body
clocks. The more organised and centralised the animal’s nervous
system, the more likely it is that a general dermal light sense will be
supplemented or replaced with a more evolved structure, located in
or near the brain. This is the famous pineal body, a third eye present
in all vertebrates, including humans. The first recorded discussions of
the pineal organ in humans come from Hindu writings of around the
sixteenth century, whose writers believed that the pineal was a form
of eye. They were not wrong, though not until 1964 was it conclu-
sively demonstrated that light penetrates the human brain. Until then
there had been a tendency, among early evolutionists, to treat the
human pineal ‘gland’ as a functionless left-over from our reptile past
— as vestigial and uninteresting as the appendix. Its unassuming appear-
ance no doubt played a part: this tiny, pine-cone-shaped organ is only
six millimetres long, and weighs just one-tenth of a gramme.
The pineal body’s role in controlling the vertebrate body clock
was discovered by a slightly roundabout route. In 1911 Karl von
Frisch — who went on to win a Nobel Prize for decoding the language
76 A NATURAL HISTORY OF SEEING

of bees - was studying for his University Teaching Certificate at the

University of Munich. Thanks in no small part to the example of
his illustrious uncle, Sigmund Exner (the man who first unpicked
the workings of the insect eye) von Frisch was fascinated by how
other animals perceive the world. In particular, he was curious to
know how animals react to light. How, for instance, does the Euro-
pean minnow ‘know to adapt its coloration to remain camouflaged
in shallow, sun-dappled water? To understand this mechanism, von
Frisch shone a light on the head of a blinded European minnow.
The minnow’s skin, chameleon-like, began to pale - just as it would
when a sighted minnow swam into sunlight. By isolating the light-
sensitive region on the fish’s head, von Frisch was able to show that
the fish's changes in coloration were triggered by its pineal. Even if
the fish were blinded, its pineal body still ‘saw.
The link between the activities of the pineal body and changes in
skin coloration, though interesting, seemed to have no universal
significance until, in 1958, Aaron Lerner isolated a substance from
the pineal bodies of cows and studied its effects on known skin
pigments. Because it caused granules of the pigment melanin to
contract, he called his newly isolated pineal extract ‘melatonin.
In turn, the significance of melatonin’s presence in the pineal organ
was not fully understood until 1974, and came about largely thanks
to the work of Julius Axelrod (1912-2004), an unpromising medical
student who spent his early career in New York, testing vitamin
supplements in milk.
‘Successful scientists are generally recognised at a young age. They
go to the best schools on scholarships, receive their postdoctoral
training fellowships at prestigious laboratories, and publish early;
Axelrod wrote in 1988. ‘None of this happened to me?!
Axelrod had no illusions about his career. ‘I expected that I would
remain in the Laboratory of Industrial Hygiene for the rest of my
working life. It was not a bad job, the work was moderately interesting,
and the salary was adequate: But he had not counted on the friendship
and mentorship of Dr Bernard Brodie, a researcher at Goldwater
Memorial Hospital. Encouraged by Brodie, Axelrod, though he had
no formal qualifications, established himself in pharmacological
research, studying analgesic drugs and the effects of caffeine and
HOW ARE EYES POSSIBLE? 7

Julius Axelrod.

ephedrine. By the mid-fifties, Axelrod was recognised as a pioneer

in the study of drug metabolism and addiction, and was invited to
set up his own laboratory at the National Institute of Mental Health.
There, Axelrod and his co-workers began to unpick the chemical
signals by which nerve cells communicate with each other - work
which earned Axelrod a Nobel Prize in 1970.
Axelrod was drawn to the pineal because of the chemical provenance
of that curious molecule melatonin. The nucleus of the melatonin
molecule is serotonin, one of life's most useful building blocks. (It turns
up in cephalopods and amphibians, in the pineals, retinas and associated
retinal tissues of all vertebrates, in figs, plums, bananas .. .) Serotonin,
because of its structural resemblance to LSD, was then believed to
trigger psychoses. Axelrod found this assumption was incorrect. At the
same time, he discovered that the pineal, far from being vestigial, was
a biochemically active organ. But what was it for? With his colleagues
Herbert Weissbach, Richard Wurtman and Sol Snyder, Axelrod revealed
that the serotonin-melatonin cycle is a kind of biological clock,
regulating both our daily cycles of sleep and metabolism and also the
reproductive cycles of oestrus and menstruation. Further, the clock’s
multiple rhythms were synchronised by changes in light level. The
pineal set itself by detecting daylight. It was, if not exactly an eye, at
least a light detector of considerable sophistication.
78 A NATURAL HISTORY OF SEEING

The closer we look at them, the more similarities we find between

the eye and the pineal. Architecturally they are quite different, but
at the level of fine structure, there are some arresting parallels. The
pineal photoreceptor cell is structurally similar to the cone cell.
Genetic studies have since shown that cones are more ancient than
rods; that rods evolved out of cones.” Might cones have evolved from
pineal receptors? In a sense, it doesn't matter. Their similarity may
simply be another, eloquent product of convergent evolution. The
significant point is that pineal receptors and cones use virtually iden-
tical opsins, and the biochemical processes that follow their exposure
to light are virtually interchangeable.
The eye is not an isolated miracle. It is simply one particular —
undeniably spectacular — species of light-sensing organ.

3 — eyeless in Basel

To suppose that the eye, with all its inimitable contrivances

for adjusting the focus to different distances, for admitting
different amounts of light, and for the correction of spherical
and chromatic aberration, could have been formed by
natural selection, seems, I freely confess, absurd in the
highest possible degree.

So Charles Darwin begins his uphill task: persuading the readers of

The Origin of Species that even ‘organs of extreme perfection and
complication are not the product of a rational creator, but of a blind
mechanical-historical process: evolution.
Why is Darwin so obviously girding himself for battle? Why is
the eye's evolution so hard to credit?
Darwin expects his theory to be challenged on two fronts. For
some, there does not seem to have been enough time for anything
so perfect as the eye to have arisen by chance. Arguably the most
elegantly presented statement of incredulity comes from the
Reverend William Paley’s book Natural Theology published over
half a century earlier, in 1802: ‘Were there no example in the world
of contrivance except that of the eye, it would be alone sufficient
“ HOW ARE EYES POSSIBLE? 79
to support the conclusion which we draw from it, as to the neces-
sity of an intelligent Creator?4
For others, it seems far-fetched that a patch of skin could ever
evolve into an eye following the principles Darwin laid down.
According to Darwin's theory, the present form of a species could
- only have arisen because every little intermediate stage of develop-
ment conferred some survival benefit upon living individuals. But
how can anything so complex as an eye have arisen by stages?
To these established sources of incredulity, we can now add a
third. Consider the number of times the eye has arisen in nature
(estimates wobble wildly between forty and sixty). Consider the range
of different tissues, pigments and structures that go to make an eye,
not to mention the range of behaviours that come packaged with
the ability to see. When we find all these factors coming together in
the same ways, not just once, but again and again, we cannot help
but wonder whether even 538 million years gives evolution enough
time for so many elaborate conjuring tricks.
No less dedicated an evolutionary biologist than Walter Gehring
felt the need to voice this worry, in his book The Homeobox Story:
‘Because the evolution of the prototype eye, at a stage before selection
can exert its effect, must be a rare event, the independent evolution
of so many prototypes represents a serious problem that is difficult
to reconcile with Darwin's theory.®
When an eye begins to distinguish images, the evolutionary pres-
sure to see better than your neighbours will be colossal. But what is
the evolutionary pressure on an ‘eye’ that is no more than a ribbon
of pigment in a single cell? Gehring has no problem with the idea
that the eye evolved. His problem is with the idea that up to sixty
versions could have evolved independently. Even the most primitive
eye is made of complex materials - materials that have their own
evolutionary tale to tell. The coming-together of all these ingredi-
ents had to be, in Gehring’s words, ‘a rare event. It was not driven
by evolutionary pressures anything like as powerful as those which
pit visual predator against visual predator in a race to develop keener
and keener eyesight. So, Gehring believes that the eye arose only
once, and all its many and varied versions are cousins.
* * *
80 A NATURAL HISTORY OF SEEING

As I write this, a housefly has just landed on the remains of the cake
I was eating. It’s looking at the cake with immovable compound eyes,
which dominate its head. I am looking at it through squishy vertebrate
eyes that still carry the tang of ancient oceans. I can just about imagine
that the fly and I are related. We both have symmetrical body plans;
we both have heads, and mouths, and anuses. But there the similarities
end and a staggering list of differences begins. Even our skeletons bear
no resemblance: I keep mine on the inside, while the fly is encased in
articulated armour. Then there are our eyes. How can the bulbous,
faceted horrors of a fly’s eye be related to my exquisite, camera-like
structures?
Flies, cats, dogs and people all start life as a single cell. The differ-
ences emerge over time, but at the start, we all look alike. Early in
development, there are basic structural similarities among all animals.
Heads, mouths, guts; if we have limbs, they extend from our trunks.
If the life we see around us today began in one place, at one time,
then living things must share a common ancestor. At the very earliest
stages of development, we resemble that ancestor, if not exactly, then
at least in the sense that we possess those features which we have
inherited, and which we all share. The German populariser of evolu-
tionary theory, Ernst Haeckel (1834-1919), made much of the likely
similarity between an animal's foetal form and the shape of its ancestors.
Taken too literally, the idea —- known by its (not very catchy) catchphrase
‘ontogeny recapitulates phylogeny — can lead the gullible palaeontologist
down all sorts of garden paths. Still, it’s a good rule of thumb.
Studying animals at the earliest stages of their development gives
us clues about human ancestry, and our relationship to other animals.
If animals came out differently each time, there would be very little
point in being an embryologist. If your dog had puppies, and one
puppy had four legs, one had six legs, and one had two legs growing
out of its head in place of its eyes, you would not think much of
those who spent their professional lives looking for a hidden order
in nature. At the other extreme, one would be unlikely to sign up
for a biological science course if humans were all the animal life the
planet had to offer; if humans bore humans exactly like the humans
before them, and the fossil record, however far back you searched,
contained layer upon layer upon layer of human remains. (In such
' HOW ARE EYES POSSIBLE? 81

a world, foetuses would presumably look like miniature men and

women. Even that first cell would be human-shaped, with little arm-
and leg-like projections . . .)
Embryology is worth doing because in the real world, animals
make more or less exact copies of themselves. Dogs make offspring
_. that are like dogs, and people make people - most of the time. The
flexibility that allows animals to change over time — so that several
new species may arise from a common ancestor — needs there to be
an outside chance that what emerges from the egg or the womb is
not like its parents. Once in a blue moon, dogs do have six legs.
Choctaw County Historical Museum in Alabama has one in a jar.
And now and again, newborn humans turn out to be not very
human.
Many women produce monsters: they just never know it. Human
infants whose body plan is in disarray rarely make it out of the
womb alive; really gross errors crop up so early in the development
cycle that the foetus is spontaneously aborted long before the mother
is aware of being pregnant. Over half of all pregnancies go disas-
trously wrong - we just never notice them.°
Much of the study of early development concerns what happens
when the ‘hidden order of nature’ goes awry. Why are six-fingered
children so rare - and why do they happen at all? By studying the
exceptions one can, with a following wind, begin to work out the under-
lying rules that govern an animal's development.
Walter Gehring claims that his career in molecular biology began
the day his uncle sent him a mysterious box. Inside the box was a
pile of brown, rather unprepossessing butterfly pupae. Young Walter,
following the instructions in his uncle's letter, put the box up in the
attic - and then forgot about it. The following spring, stumbling over
the box again, he opened it. The butterflies had just emerged. Gehring
stared in wonder: what had turned an ugly pupa into a beautiful
butterfly? “These mysteries; says Gehring, ‘have haunted me for my
entire life’
‘With New Fly, Science Outdoes Hollywood, screamed the front
page of The New York Times, some fifty years later, in 1995.” The Times
was excited by experiments Gehring had been conducting on fruit
flies. Gehring, in pursuit of the secrets of development, had managed
82 A NATURAL HISTORY OF SEEING

to grow eyes on flies’ knees, in

place of antennae, and even
on their wings. The popular
response to the news was as
boisterous as it was predictable.
Victor Frankenstein, it seemed,
was alive and well and living in
Basel. One of Gehring’s own
photographs, showing a fly
with extra eyes, was used on a
poster in a campaign opposing
gene technology.
This was a bit rich. Fora start,
Gehring’s experiments were not
strictly news. In 1935 a former
Walter Gehring in 1964.
bookseller, Hans Spemann, won
a Nobel prize for getting eyes to grow on the bellies of amphibia. And
the only reason either man was able to accomplish such gruesome feats
of magic is because nature itself constantly throws up sports of this
sort. One of the photographs in Richard Dawkinss splendid account
of evolution, Climbing Mount Improbable, is of a frog whose eyes have
grown, not out from the top of its skull, but down, into the roof of its
mouth; it was discovered, hale and hearty, in an Ontario garden.
Thirty years after the publication of Darwin's The Origin of Species,
the embryologist William Bateson gave a name to the mechanism by
which animals acquire one organ or extremity in place of another:
homeosis, the change of something into the likeness of something else.
(Bateson was fond of coining neologisms. ‘Genetics’ is another of his.)
Homeosis can express itself either as a birth defect, or during repair
to a damaged or missing organ. When a lobster loses an antenna, another
antenna grows in its place. After more than a couple of amputations,
the luckless crustacean loses the plot, and grows a leg there instead.
Spemann’s fervent hope was that, by understanding homeotic muta-
tion, he would come to understand what an animal’s body plan (or
‘organiser centre’) looked like and what it was made of. Gehring’s first
experiments — which bore more than a passing resemblance to those
of Spemann - revealed that fruit fly embryos are arranged very simply.
“HOW ARE EYES POSSIBLE? 83
Gehring’s descriptions read more like the patents for toys than studies
of real animals. The embryonic fruit fly is a set of discs, called ‘imaginal
discs, because each carries a genetic blueprint of what it wants to be.
If the top disc wants to be a head, you can guarantee that the bottom
disc wants to be a thorax. Rearrange the discs, and you interfere in
. predictable ways with the body plan of the fly.
From these experiments, Gehring was able to isolate the genes
responsible for giving each disc its destiny. These ancient genes didn’t
directly make anything you could point to. Rather, they turned on
other genes, which turned on others, thus shaping the development
of large, ordered structures; a cascade of ever more numerous, ever
more specific instructions, triggered by one broad ‘executive’ order.
When not given witty monikers by whimsical geneticists (the ‘Harry
Potter’ puberty gene, anyone?) genes are named after the conditions
and abnormalities that led to their discovery. This, inevitably, means
that their name commemorates what happens when they don’t work.
The gene that triggers the development of eyes in fruit flies — discovered
in 1915 — is called ‘eyeless. Although Gehring experimented with eyeless,
he had no particular interest in eyes, and it was a fluke finding by a
junior colleague that triggered the work for which he is best known.
In 1993 Rebecca Quiring was hunting for the body-plan genes of
fruit flies - the ancient genes which tell the imaginal discs of the
foetal fly what form to take. On the assumption that certain proteins,
called transcription factors, bind selectively to specific genes, Quiring
cloned a known protein to use as a ‘lure’ But the protein insisted on
binding only to a gene - eyeless - that they already knew about. It
seemed a pity for Quiring to just throw away all her work, so Gehring
suggested that she submit the genetic sequence of her protein to the
European Molecular Biology Laboratory's database in Heidelberg. If
she had inadvertently cloned eyeless, well, that was a result of sorts,
and it could at least count towards her PhD.
The acres of paper the Heidelberg computer spewed out in reply
to Quiring’s submission suggested nothing so much as a printer
glitch. It took a little while to figure out what this mass of data
actually was - and when Quiring did, she skipped down the corridors
in high glee. Certainly, her protein corresponded to the fruit fly’s
eyeless gene. But the Heidelberg computer, not content with making
84 A NATURAL HISTORY OF SEEING

one match, had found others: smalleye, a gene from a mouse; and
aniridia, a gene from a human.being. All three are essential to the
proper development of eyes. A mouse with a missing smalleye gene
is born with severe facial disfigurements, including undeveloped eyes
and sealed nostrils. Humans with a missing aniridia gene have
missing or malformed irises, cataracts, glaucoma, squint, and (sugges-
tively, given the role of smalleye in mice) a loss of the sense of smell.
Were eyeless, smalleye and aniridia just different names for the
same gene? It was not hard to believe that mice and humans might
share a gene for eye development. Men and mice are both mammals,
and have a fairly recent common ancestor. What really stretched
credulity was that the gene responsible for mammal eyes is also
responsible for the compound eyes of fruit flies. The common
ancestor of flies and humans lived about five hundred million years
ago. Snails are more closely related to flies than we are. Could a
single control gene really trigger the development of such different
eyes as the faceted compound eyes of a fly and the wet cameras of
the human being?
Gehring’s team had to demonstrate that the eyeless gene was the one
gene ultimately responsible for the development of the fruit fly’s
compound eyes. So Gehring returned to the work he had conducted
thirty years before on the ‘imaginal discs’ that make up a fruit fly
embryo. By turning on the eyeless gene in different imaginal discs, the
team induced ectopic eyes — eyes that grew where they did not belong:
on wings, and knees, and in place of antennas. One fly had fourteen
extra eyes which, although it could not see through them - they were
not connected to the brain - were sensitive to light.
What would happen if they replaced the fly’s native eyeless gene
with smalleye? .
Since that first burst of print-out from Heidelberg, a slew of eyeless-
like genes in other species had come to the department's notice,
including genes in squid, sea-squirts and zebra fish. All were essen-
tially the same, with a little local variation. (This is why - to run
ahead of ourselves for a second ~ modern accounts tend to drop
the species-specific names and lump all these genes under one conve-
nient, if unmemorable, label: Pax6.) So, despite The New York Times,
no one seriously expected anything freakish. (A mouse eye on the
“HOW ARE EYES POSSIBLE? 85
head of a fly; quivering nostrils in place of mouthparts .. .) Smalleye
and eyeless are about ninety-four per cent identical.
On the other hand, no one could be sure how well smalleye would
stand in for its insect cousin. Early results were very unpromising.
Then, the day before they planned to give up, Gehring’s collaborators,
_ Georg Halder and Patrick Callaerts, found the first spots of red pigment
on a fly’s legs. A week later, the first facets of an insect eye had begun
to form. Soon the flies boasted spectacular working eyes on their legs,
their antennae, and their wings. The mouse gene smalleye had directed
the creation of perfect compound eyes.
In their paper in Science in March 1995, Gehring, Halder and
Callaerts declared: “The observation that mammals and insects, which
have evolved separately for more than 500 million years, share the
same master control gene for eye morphogenesis indicates that the
genetic control mechanisms of development are much more universal
than anticipated:*
A year later, Gehring went much further when, in response to a
critical letter in Science, he announced that his work went ‘against
the dogma of eye evolution that can be found in most textbooks’
This claim - and in particular the use of the word dogma - was,
frankly, inflammatory. Gehring must have thought attack was the
best form of defence, for his target could not have been more highly
esteemed. Ernst Mayr (1904-2005), considered by many the founding
father of evolutionary biology, was the man who had first come up
with the idea that the eye had evolved independently many times.
According to Gehring’s letter, Mayr’s notion was the new orthodoxy
— and he was the man to demolish it.
Setting aside its choice of language, Gehring’s grandiose claim is
founded on the rather dangerous assumption that he has found the
gene ‘responsible’ for eye development. At first glance this seems
undeniable - Gehring’s ability to induce ectopic eyes in fruit flies
using a smalleye gene from a mouse demonstrates as much.
But what does it actually mean when we say that a gene is ‘responsible
for something? Genes are not an army of microscopic labourers. They
don't build organised structures, the way construction workers erect
buildings. Genes squat inside cells. Genes that are turned off do nothing;
those that are turned on control the production of different proteins.
86 A NATURAL HISTORY OF SEEING

And that’s all. An animal's life can be seen as a ‘working out’ of the
billions of little directives issued by genes in every one of the animal's
cells, but these directives are dependent on circumstances. Every cell
contains instructions on how to build a head but, thanks to induc-
tion (the ‘dance of development’ in which different foetal tissues touch
and communicate chemically with each other) the instruction ‘build
a head’ is heeded only once.
The point is well made in the very letter to Science that elicited
such a strong reaction from Gehring. W. Joe Dickinson and Jon Seger,
from the University of Utah, point out that the eyeless gene is turned
on in plenty of cells that have nothing to do with the eye. Indeed,
if the gene is turned on in all cells, ‘this does not convert the entire
embryo into eye structures’? Dickinson and Seger conclude: * . . we
take this to imply that [eyeless] originally served some basic devel-
opmental process other than eye induction:?
Many a squid would agree: the squid Pax6 gene controls the devel-
opment of its tentacles, suggesting that the gene has a role in the
formation of many different kinds of sensory organ. Certain species
of blind worm use a version of eyeless to control the shape they
acquire as they grow. Every living thing has a genetic vocabulary to
write in. As Dickinson and Seger’s letter reminded the readers of
Science, a gene called hedgehog crops up in the wing development
of flies and birds - structures as markedly different in their way as
insect and mouse eyes. This does not mean that the fly’s wing and
the bird’s wing are functionally ‘related.
‘What we claimed, said Mayr, ‘and it’s as correct as ever, is that
the eyes themselves, the photoreceptive organs, developed indepen-
dently at least forty times. This is not in the slightest touched by
finding that there are genes used in making eyes that existed long
before eyes. You should go to the species that have no eyes but have
this gene and find out what it’s doing’*°
Genetic studies like Gehring’s have revealed a rich pre-history for
the eye, without which its emergence and re-emergence, forty times
over, would be inexplicable. They tell us what tools were used to build
the first eye, and roughly when those tools first became available.
But they cannot tell us when or how the first eye arose, or which
animals first saw.
“HOW ARE EYES POSSIBLE? 87

4 — The shoeshiner’s tale

... man in thinking greatly errs particularly when inquiring

after cause and effect; the two together constitute the
indissoluble phenomenon.
J.W. von Goethe”?

Louis Jacques-Mande Daguerre was born in a small town on the

outskirts of Paris in 1787. He did not invent photography. In the
mid-nineteenth century, there were many competing processes for
recording and fixing photographic images, and historians of photog-
raphy salute many pioneers. The architect, stage designer and illu-
sionist Daguerre has the distinction of being the man who sold
photography to the French government. The government then freely
gave the process to the world, but other, better processes were already
abroad, and the gesture fell flat.
Photographs taken by Daguerre’s method - daguerreotypes - are
the earliest photographs we have, because Daguerre cracked the
vexing problem of how to stop the images from fading. His other big
achievement was to reduce the photographic exposure time from some
eight hours to only thirty minutes. This improvement made it possible
to take photography out of the laboratory and capture reasonably
faithful real-world scenes.
His 1838 daguerreotype of the boulevard du Temple, for instance,
required an exposure time of only ten or twenty minutes and, conse-
quently, the play of light and shadows on the trees and buildings is
remarkably life-like. We can tell, at a glance, that this photograph
was taken in the middle of the day. But where are the people? The
street, one of the busiest in Paris, must have been heaving with them.
But where are they?
With an exposure time of tens of minutes, Daguerre knew he
would not be able to capture the human business of this busy Parisian
street. Indeed, the novelty of the image — an empty boulevard, indeed!
— was probably uppermost in his mind when he framed the shot.
88 A NATURAL HISTORY OF SEEING

But human life is not altogether excluded. In the bottom left of the
picture is a shadowy figure --a man having his boots polished.
As a German magazine of 1839 observed, he ‘must have held himself
extremely still for he can be very clearly seen, in contrast to the shoeshine
man, whose ceaseless movement causes him to appear completely blurred
and imprecise’ Imagine we had no newspaper accounts, no diaries, no
histories to work from. Imagine we only had Daguerre'ss photograph.
What would it tell us about the boulevard du Temple? What would we
conclude from our study of this snapshot in time? Would it be
unreasonable to conclude that Paris had been virtually abandoned,
haunted only by a handful of tall gentlemen and wraith-like bootblacks?
Fossils are evolution’s daguerreotypes, recording — in their peculiar,
mysterious, and eccentric ways — the progress of life on this planet.
They are our most valuable source of information about life’s history;
and yet, very often, they lead us horribly astray.

In a letter to an American colleague, Charles Darwin famously declared

that thinking about the evolution of the eye gave him a ‘cold shudder.
These occasional concessions to incredulity are a favourite motif in
Darwin's writing (elsewhere, he says that thinking about the tails of
peacocks made him physically sick). He uses this piece from his
rhetorical armoury when he introduces readers of The Origin of Species
to evolution - a process of nature which is ‘insuperable to our
imagination because of its sheer scale, ubiquity and (in most cases)
extreme slowness.
No one looking at a countryside scene can ‘picture evolution in
action, any more than anyone, looking at the star-lit sky, can ‘picture’
the scale of the galaxy. Darwin took biology out of the anthropo-
morphic world of mythical and biblical parable and into the astro-
nomically scaled world of evolutionary time. We should expect a
shudder or two.
What really had Charles Darwin on the run was the lack of fossil
evidence.
For evolution to be truly a force in nature, it had to have given rise
to all life, developing it by minuscule increments. In Darwin's day,
however, there was no evidence of the existence of life earlier than the
Cambrian period. Animals had simply appeared around 540 million
“HOW ARE EYES POSSIBLE? 89
years ago (or 540Ma, to use the scientific shorthand). Even more
worrying for Darwin and his supporters, these first life forms were not
remotely primitive. They had claws, jaws, teeth, tentacles — and eyes.
Older rocks had been identified and studied. Earlier ages had been
mapped by geologists. But none of these older rocks contained fossils.
_ On the evidence of the rocks, none of these earlier times had boasted
life. It was as though, far from emerging slowly from microscopic
animalcules and slime, life had simply exploded 543 Ma, fully formed
and ready for action.
For many of Darwin's generation, the evidence for the instantaneous
appearance of life on Earth during the Cambrian period offered the
possibility of a reconciliation between evolution and divine creation.
The most articulate stab at a reconciliation came from William Buckland,
professor of geology at Oxford. Some twenty years before Darwin's The
Origin of Species was published, Buckland pointed to the Cambrian
explosion as a possible, though distinctly un-biblical, creation event.
Buckland’s opposition to Darwin's theory of evolution seems ironic
now. It was he who wrote up the first description of a dinosaur fossil,
the ‘Great Fossil Lizard of Stonesfield’; he who found evidence of
mammals living in the‘age of reptiles’; he whose grasp of the principle
of ‘survival of the fittest’ made for some memorable lectures. (One
of his students recalled how Buckland ‘ . . rushed, skull in hand, at
the first undergraduate on the front bench - and shouted, “What
rules the world?” The youth, terrified, threw himself against the next
back seat, and answered not a word. He rushed then on me, pointing
the hyena full in my face - “What rules the world?” “Haven't an idea,”
I answered. “The stomach, sir,’ he cried.)
Buckland’s opposition to evolution was not wilful or blinkered,
and at the time, the weight of evidence lay on his side.
Much of the fossil record of life on Earth remained to be discov-
ered, as both Buckland and Darwin knew perfectly well. Even the
terminology they used was tentative and uncertain, which is why
Darwin never writes about the Cambrian period itself, but about the
‘Silurian, a word that has since shifted its meaning to refer to a
completely different slice of geological time. Writing of the sudden
appearance of the complex, woodlouse-like trilobites in the fossil
record, Darwin had no doubt that ‘all the Silurian trilobites have
90 A NATURAL HISTORY OF SEEING

descended from some one crustacean which must have lived long
before the Silurian’ :
Neither Darwin, Buckland nor any of the early champions and
opponents of his disquieting theory lived to see this expectation
confirmed. During the second half of the nineteenth century there
were scattered reports of fossil finds in pre-Cambrian strata. But
Darwin's theories came to be accepted, not so much because of the
fossil evidence, but primarily because they were able to explain
circumstantially the diversity of modern life.
Not until 1946 did an Australian mining geologist happen upon
the first well-preserved pre-Cambrian fossils. Exploring the Ediacara
Hills, a range of mountains north of the city of Adelaide, the mining
geologist Reginald C. Sprigg found fossils of what looked like soft-
bodied organisms. Some resembled jellyfish; others were more like
shellfish, or worms.
Whether the Ediacaran faunas were an odd, non-animal trial run’
at multicellular life or an early, eyeless and toothless period of animal
evolution, their existence has done little to erode the mystery of the
Cambrian explosion. The plain fact is that within five million years,
between 543 and 538Ma, life on Earth transformed completely. The
way animals were organised, how they lived and how they behaved,
underwent a massive sea-change, from blind drifting and casual grazing
to visually guided predation, defence, camouflage and evasion. It all
happened so quickly that the fossil record — laid down excruciatingly
slowly over millennia - can no more record its progress than Louis
Daguerre’s camera, pointed down the boulevard du Temple in 1838,
could have captured its busy traffic of carriages and people.

What lessons should we draw from the Cambrian explosion? For the
American palaeontologist Stephen Jay Gould (1941-2002), writing in
the late 1970s, the sudden appearance of so many odd and arresting
animals suggested ‘a unique time of organic flexibility, before major
developmental pathways became irrevocably set. Maybe animals didn't
evolve slowly at all. Maybe they evolved intermittently, quickly, and wildly.
The theory of ‘punctuated equilibrium; promulgated by Gould and
his colleague Niles Eldredge, implied an as-yet-unidentified genetic
mechanism capable of triggering a Mardi Gras of novel forms.
“HOW ARE EYES POSSIBLE? 91

Genetics, however, conspicuously failed to throw up any evidence

of this mechanism.
The first animals appeared about 750Ma. We can be fairly sure of
this, because we have a good idea of how often complex proteins
such as haemoglobin mutate. If we count up the number of genetic
_ differences between the haemoglobin used by one species and the
haemoglobin used by another, we can estimate the age of their
common ancestor. (This neat trick was inspired by George Wald’s
work on rhodopsin.) By counting up the differences between the
most unrelated species, we arrive at the age of the oldest-ever animal
- the common ancestor of all animals everywhere - and we find
that it lived 750Ma. Gould’s conjecture thus finds itself in a heap of
trouble, because 750Ma is two hundred million years before animals
strode into view in the Cambrian.
Since the kinds of genetic innovation that made animals possible
occurred two hundred million years before the Cambrian explosion,
they can hardly be said to have ‘caused it. The only genetic event we
know of that might have triggered an event as extraordinary as the
Cambrian explosion was the first appearance of Gehring’s body-plan
genes — Paxé6 and the like. (One imagines these newly minted genes,
like demented Christmas elves, stringing together segmented
novelties like toy trains and tossing them into warm Cambrian seas.)
But as Gehring’s work has shown, Pax6 and its fellows are shared
by species as profoundly unrelated as humans and houseflies. These
genes are therefore as old as our oldest common ancestor — too old
to have triggered the explosion.
Once Gehring’s body-plan genes had been shown to pre-date the
Cambrian, Gould’s conjecture was left high and dry - a soft target
for its critics.’
If there is no genetic mechanism to explain the Cambrian explosion,
there is no shortage of other plausible explanations. Perhaps an extinc-
tion event cleared the way for novel forms of life. The Earth has expe-
rienced at least six major extinction events — from asteroid strikes to
plagues of clever apes. The ‘ape plague’ story is still being written, but
the outcome of the other five events was always the same: a period
of rapid evolutionary development among the species that survived.
The trouble is, the Cambrian explosion was not preceded by an
92 A NATURAL HISTORY OF SEEING

extinction event. In some areas, it is true, the pre-Cambrian fauna

(called Ediacaran, after the hills in which their fossils were first
discovered) died out before the Cambrian fauna took over, but in other
places, Ediacaran life seems to have been gently edged out by Cambrian
newcomers over millions of years. An increase in oxygen levels, around
545Ma, coincided with the Cambrian event, but oxygen levels have
rocketed and plunged on numerous occasions during Earth's history,
with surprisingly little effect on the biosphere. Maybe there were more
shallow seas than before, as sea levels rose, drowning the Cambrian
continent’s barren lowlands (there was, of course, no life on land at
this time — or no life to speak of; an algal colony here or there, perhaps).
But why should an increase in shallow seas make everything grow so
big, and acquire such novel forms?
These are all good ideas, and there are plenty more. But none of
them is complete or compelling.

Part of the problem is to do with the sort of explanation we are

looking for. It is surprisingly difficult to talk about what ‘caused’ the
Cambrian explosion, because every cause is the effect of a preceding
cause. Where do we draw the line?
Three things we can be certain of about the Cambrian explosion:
animals got big, they got hard (hard enough to be preserved in the
fossil record), and they got clever.
We do not know why animals grew bigger in the Cambrian, but
we do know that getting bigger isn't hard to do. When the dinosaurs
died, mammals blew up like balloons to fill the suddenly empty
niches. Increasing size may explain why Cambrian animals first began
to develop hard parts. Hard surfaces provide purchase for muscles,
so that movement becomes more powerful, directional and coherent.
The so-called ‘small shelly fauna’ that are the first Cambrian fossils
are sometimes the skeletal remains of whole creatures, and sometimes
the individual leaves of a ‘chain mail’ covering much larger animals.
The use of the phrase ‘chain mail’ is apposite, because alongside
evidence of hard parts comes evidence of predation: some of these
shells have holes bored through them.
The Cambrian was a violent time. The fossilised eggs of bristle
worms and jellyfish contain foetuses so well developed, they are almost
“HOW ARE EYES POSSIBLE? 93

adult; an expensive reproductive strategy, employed only when newly

hatched young have to be ready for everything the world has to
throw at them.
What was out there? For a start, there was Anomalocaris, a metre-
long stalk-eyed nightmare with a mouth like a camera iris. We know
it ate animals, because they are preserved inside Anomalocaris’s
_ fossilised guts. Then there were the trilobites. These came in all shapes
and sizes and they were around for a very long time: about 300
million years. Some trilobites were bottom feeders, some were fed
by bacterial mats living in their gills, and some were hunters. The
very first trilobite species we know of had eyes. And not blurry
squitty eyes, but gorgeous, faceted, compound eyes, exquisitely
constructed and assembled to a design that has never been repeated,
for all that eyes have arisen at least forty times on Earth.
With size and armour came cleverness. We can measure an
animal’s cleverness by studying its tracks. The fossilised tracks of
Ediacaran animals are touchingly simple; the tracks of complacent
grazers, little fleshy lawnmowers, trimming back the algal mats of
primaeval seas. There are burrows, but they are shallow and often
drilled horizontally into a sloping bank, and their walls are neither
lined nor armoured, offering little in the way of defence. Indeed,
there seems to be nothing to defend against. If there were predators,
then they were opportunists, incapacitating or killing whatever
hapless creature happened to cross their path. This lackadaisical style
of predation could not have inspired the evasion strategies we find
among animals today. Speed, strength, size, hardened shelters, and
hard skeletons are all excellent defences against predation, but there
is no evidence for any of them among the Ediacaran fauna.
Cambrian tracks tell a very different story. The surface is perfo-
rated with vertical burrows drilled into the sediment. Monstrous
tracks weave back and forth across the sea bed. Pits in the rock mark
the frantic excavations of rapacious hunters. These are not the tracks
of harmless, fearless grazers — these are the spoor of conflicts, battles,
competitions for resources; of the desperate measures animals take
to eat and not be eaten.
What triggered these extraordinary physical and behavioural
novelties?
94 A NATURAL HISTORY OF SEEING

When animal life consisted solely of tiny, millimetre-long worms

and planktonic drifters, there was no need for vision, for there was
no animal big enough to see. Visible light (including the ultraviolet)
varies in wavelength between .35 and .8 microns — just less than the
width of the smaller bacteria. Such animals are big enough to block
the light - but only by a whisker. Euglena gracilis is 50 microns long.
For Euglena, there can be no vision. Light detection, certainly. But
not vision. Vision requires more than an occasional nudge from the
crest of a light wave. It requires a deluge, heavy enough to reveal
patterns of shine and shade. Euglena is only sixty times larger than
the wavelength of red light; it is just too small.
Just as there is a minimum size for a true, image-forming eye, there
is a minimum level of complexity. Vision requires a nervous system
sufficiently complex to harness the dance of light within the eye. Why
animals should have acquired such complexity is still a mystery. If an
animal had a rudimentary light sense, perhaps an increase in size would
have led directly to an increase in nervous complexity. As the eye grew
bigger, and its light-gathering capacity increased, vision would have
improved, maturing, over generations, from mere light detection into
an ever-more suggestive view of the world. Provided with this extra
information, the nervous system connected to this proto-eye would
have evolved a growing sensitivity to rapid changes in light level, sugges-
tive of the approach of a mate, or a predator. No one knows exactly
why animals grew bigger in the Cambrian, but competition triggered
by the evolution of vision would certainly have been influential enough
to drive an increase in body size.
By the same token, we do not know why animals chose, 543Ma,
to acquire teeth, legs, grasping forelimbs, evasion tactics, mining
skills, colours, bristles, shells, mirrors, and all the other paraphernalia
of modern life, but we do know the reason they acquired them. Like
Adam and Eve caught naked and defenceless in the Garden, Life
became aware, for the very first time, that it was being watched.
There's no point having legs if you can’t see where youre going.
(The other senses will give you some idea of your environment, but
only vision will give you the instantaneous information that makes
agile movements possible.) Teeth and armour, too, are the product
of an arms race that could not have got started without vision. When
‘HOW ARE. EYES POSSIBLE? 95

the chances of getting snared by a predator are little more than

random, there's no reason to evolve heavy, expensive defences against
misfortune. But once the predator can see you, and pursue you, the
chances of falling victim to its attentions are astronomically increased
and a suit of armour becomes a worthwhile investment.
: One alternative is simple inedibility. No doubt, Ediacaran animals
~ employed venoms and toxins for their personal protection. They are
cheap and easy to make, and just the sort of budget weapon a soft-
bodied animal could afford. But faced with animals who prepare their
food first and complain to the chef later, like a metre-long Anomalocaris,
or a trilobite, which although only a few centimetres long, has spiny
legs that double up as meat shredders, toxins alone are not enough.
What our prey animal needs is spines. Lots of spines. Marella splen-
dens, a large (two-centimetre-long) blind bottom feeder is so spiny, it
resembles a stripped skeleton more than a living animal. It was the
Volvo of the Cambrian shallows, tootling complacently about in its
safety cage. Other creatures were more direct, growing hedgehog-like
spines, and even blades. Many trilobites are fantastically spiny; others
were able to roll up into an armoured ball so exquisitely fashioned that,
in some cases, it came with a safety lock; the more you tried to prise
the thing open from the outside, the more secure the fastening became.
Armour is a last-ditch defence; it would be far better to go about
one’s business without interference. To avoid a blind predator, one
need only keep quiet, move slowly and try not to be smelly. To avoid
a sighted predator, one has to do something, and the most obvious
thing to do is hide. There is a type of Cambrian sandstone nicknamed
‘pipe rock’ by geologists because of the number of vertical burrows
puncturing it. It looks like fossilised crumpet.
Burrowing has its limitations. If you spend all day hiding in your
hole, where is your food going to come from? The birth of vision
heralded ever-more-subtle strategies for concealment; ways of hiding
that did not involve burrows, and permitted animals to move -
however cautiously — through their environment. Animals evolved
to disguise their own nature; to appear inedible, or dangerous, too
large or too awkward to swallow. Some hid the fact that they were
food at all. Camouflage, warning colours, mimicry, deceit, or
complex behaviour could, of course, also be harnessed by
96 A NATURAL HISTORY OF SEEING

predators. They too could disguise themselves, so that their prey

was lured towards them. Since most animals were both prey and
predator, quite sophisticated and subtle disguises were bound to
evolve. The sheer complexity of the Cambrian continues to reveal
itself, and the most startling recent finds, at the time of writing, are
those reported by Andrew Parker, a zoologist whose interest in the
fossil record grew out of his studies of structural colour.
Structural colour occurs when a lighted surface throws off multiple
reflections. These reflections are generated so close to each other that
the light waves interfere; some wavelengths are completely damped,
while others shine with redoubled force. There are two ways of gener-
ating structural colour: either light is reflected off a corrugated surface
whose corrugations are fine enough to generate interference effects
(such as a peacock’s feather or a compact disc), or light passes through
multiple translucent layers, each of which reflects a small portion of
the total incoming light (like a fly’s wing or a roll of clingfilm).
A well-organised surface texture of a certain fineness is likely to
generate structural colour, so we should not be surprised to find
them in nature. Yet we need to explain why animals should openly
exhibit such colours. Pilots flying over the Brazilian rainforest have
reported seeing brief flashes as blue morpho butterflies flap their
beautiful, structurally blue wings above the forest canopy. Any animal
sending out such a beacon of colour - brighter by far than any
pigment - must do so out of some pressing need.
Structural colours make excellent signalling devices - something
Parker discovered for himself in 1995 when, in Watson’s Bay in Sydney
Harbour, he caught two seed shrimps in courtship. The idea that
seed shrimps have a courtship ritual at all seemed, at the time, a
little far-fetched. These humble crustaceans are tiny, ranging from
just a hundred microns to a few millimetres in diameter. And yet,
as Parker watched, the male flashed a dazzling blue light. Seconds
later, the pair were mating.
This extraordinary discovery led Parker on a scientific treasure hunt
that ended among the fossils of the Burgess Shale in the Canadian
Rockies, the world’s richest and most celebrated Cambrian fossil bed,
discovered by Charles Walcott in 1907. From the sunny shallows of
Watsons Bay to the 538-million-year-old shallows of the Cambrian
‘HOW ARE EYES POSSIBLE? 97

seems quite a leap — but the humble seed shrimp had covered most
of the distance. Seed shrimps have been around for about 250 million
years. Though it is possible that the seed shrimp acquired its signalling
apparatus recently, it is equally possible - even likely, given how little
the animal had changed over millions of years —- that seed shrimps
have been flashing seductive lights at each other since their inception.
Parker was on the trail of the earliest colours in nature. Pigments, alas,
do not survive the fossilisation process. But structural colours do. It
had already been established that lamp shells, living 350Ma, boasted
multi-layer reflectors as iridescent as those of a butterfly’s wing.
With the brashness that is his trademark, Parker looked for equiv-
alent structures on some of the best-known fossils of the Burgess
Shale: Wiwaxia corrugata (imagine an animated knife-block with the
knives put in blade-uppermost); Canadia spinosa (much like a modern
bristle-worm); and the walking safety-cage Marella splendens. To his
delight, Parker was able to show that all three animals boasted structural
colours. They were not merely colourful — they were shiny.
Why would a potential meal like Canadia need to shine? Parker
argues that it is inconceivable that these structural colours were an
accident. Animals this conspicuous would have perished in an eye-
blink, were their colours not some form of generally understood
warning. Were Cambrian predators sophisticated enough to be
disconcerted by flashing lights? Were they capable of learning that
sparkly creatures tend to stick in the craw?
Parker’s discovery of colour in the Cambrian suggests a dangerous,
fast-paced, colourful environment. It also suggests that Cambrian
animals possessed a modicum of cleverness; the shape, texture and
colour of things meant something to them.
According to Parker’s recent book In the Blink of an Eye, the advent
of vision powered the Cambrian explosion.’? He even points to a
specific eye as the culprit - that of a species of predatory trilobite.
This is, to say the least, unfortunate: the earliest trilobites don't appear
in the fossil record until three and a half million years after.the
explosion they are meant to trigger.“* Otherwise, Parker’s general
idea — that the advent of vision triggered a massive explosion of
species 543Ma — seems self-evident, not to say fatuous.
For an account of how the arrival of vision triggered natural
98 A NATURAL HISTORY OF SEEING

variety, we do not need Parker’s account, enjoyable as it is. We can

turn to the ecological writings of Canadian atheist Grant Allen,
writing in 1879:

Insects produce flowers. Flowers produce the colour-sense in

insects. The colour-sense produces a taste for colour. The taste
for colour produces butterflies and brilliant beetles. Birds and
mammals produce fruits. Fruits produce a taste for colour in
birds and mammals. The taste for colour produces the external
hues of hummingbirds, parrots, and monkeys. Man's frugivorous
ancestry produces in him a similar taste; and that taste produces
the various final results of the chromatic arts.’

Visual predation did not ‘cause’ the Cambrian explosion; visual preda-
tion, and the adaptational pressure it placed upon living things, was
the Cambrian explosion. The question that needs answering — the
question nobody has the answer to, and the one which inspired
Stephen Jay Gould’s brave conjecture about a mysterious ‘genetic
event’ - is what made sophisticated nervous systems possible in the

light-sensitive cells number of 1%

transparent cells improvements
dark pigment
0 RRP CCT OT
1829

176
350,000 generations

1553

1033

The results of Dan-Eric Nilsson and Susanne Pelger’s seminal 1994

computer model of eye evolution. Their fish-like eye would evolve
too fast for the process to be reliably recorded in the fossil record.
/HOW ARE EYES POSSIBLE? 99
first place? Until animals got big and complex, there would have
been no acute, image-forming eyes, for no animal would have been
capable of seeing.

Given that small genetic changes can lead to large alterations of

physique, it may be that very little time indeed is required for a
“ small, slug-like Ediacaran animal to develop a really quite startling
Cambrian motley, complete with legs, claws, teeth, a suit of many
colours, eyes and a complex nervous system.
How long does it take for at least the outward form of an eye to
evolve? Two Swedish scientists, Susanne Pelger and Dan-Eric Nilsson,
made a creditable first stab at the problem in 1994, designing a computer
simulation of how a fish’s eye could evolve from a flat, photosensitive
surface. Even in the absence of any knowledge about the mutation rates
and life cycles of real fish, this experiment was still useful, because it
showed how a really good eye might evolve from a really bad one,
smoothly, without awkward or useless intervening stages.
Almost every book on my research shelf trots out the famous
diagram of their findings — reproduced here, yet again, with numbing
inevitability. But its elegance earns it another outing. With startling
visual clarity, it reveals how an eye may evolve, and how seemingly
complex structures like a lens can arise from small, graduated changes
in tissue density.
The rules Nilsson and Pelger set for their model were impressively
few. They made one central assumption - that good eyes are better than
bad ones — and otherwise let their model develop as it would. Each
new generation, one part of the ‘eye’ was permitted randomly to alter
its refractive index — the degree to which it bent light — by one per cent.
The model was also allowed random changes of shape, again by small
increments. If the new version was optically superior to its ‘parent, it
was permitted to generate ‘children: If optically inferior, it ‘died.
The result was decisive. Starting with a flat eyespot, the model
deformed to form a shallow pit, marginally improving the eyespot's
acuity. The dimple deepened until it was as deep as it was wide, like
the eye of a flatworm. Then the edges of the pit began to close over
to form an aperture.
The amount of available light dictates how narrow the aperture
100 A NATURAL HISTORY OF SEEING

of this ‘dimple eye’ will be. The narrower the hole, the better the
whole structure will work, casting an inverted image on the back
wall of the eye. If the hole is too narrow, very little light will enter
and the image, though sharp, will be too dim to see. The only way
to improve an eye of this sort is to develop a lens. The stuff of which
the lens is made is, in most cases, already present. Some natural ‘pit’
eyes are empty; the nautilus has an eye of this sort. But Nilsson and
Pelger’s eye very quickly came to resemble the eye of a snail, whose
pit eyes are full of a more or less homogeneous jelly. (It is not hard
to imagine how this jelly is formed, since all sensitive tissues that
are exposed to the elements have some form of mucosal protection.)
Eventually, random changes in refractive index through the jelly
created a perfectly round optical lens.
Of course, in that word ‘eventually’ lurks all manner of hard ques-
tions. Having shown that an eye could evolve, were Nilsson and Pelger
able to say how long it would take to evolve? Indeed they were. Happily
we know something about mutation rates, about how reliably these
mutations will show up in individuals, and about how much indi-
vidual variation there is in the population. By picking extremely
pessimistic values for these and other variables, Nilsson and Pelger
calculated how many generations it would require for their fish to
stumble upon the 1,829 single-percentage improvements it takes to
make an eye. The answer stunned everyone: 400,000 generations.
Assume our fish breeds once a year (another pessimistic estimate).
That's fewer than half a million years for an eye to evolve from an
eyespot. How would such an event be reflected in the geological record?
It wouldn't be reflected at all.
Geologically speaking, 400,000 years is no time at all. Even if we
were, by some miracle, to find the needle in the haystack, the half-
way stage of our fish’s ocular evolution, how would we possibly know
that was what it was? The rocks won't tell us that one fish lived a
few hundred thousand years earlier than another. Geological time
just can't be measured with such fineness. Our new-found fossil
might simply be another species of fish, with less sophisticated eyes.
Computers were really not all that good in 1995. Ten years on,
we could write much better models of all sorts of aspects of eye
evolution, from the biochemistry and architecture of photocells to
‘HOW ARE EYES POSSIBLE? 101

the stitching of retinas. But none of this work is likely to have the
same headline-grabbing kudos as Nilsson and Pelger’s original paper,
which captures the dramatic evolution of whole eyeballs.
Nilsson and Pelger’s model, like all models, depends on certain
assumptions; not least, that before vision developed, patches of an
animal's skin were - or could become — photosensitive. It assumes, in
~ other words, a pre-history for eyesight: a period, who knows how long,
in which animals acquired the neural sophistication to respond to a
new-fangled sense. That pre-history has yet to be written. Perhaps it,
too, is the story of a few eye-blinks of geological time. Or it might be
an excruciatingly casual tale, taking two hundred million years or so
between the emergence of animals and the eventual, tardy, but unde-
niably dramatic acquisition of image-forming eyes in the Cambrian.
Animals did perfectly well without eyes for two hundred million
years. They ate algae. They slumped. They pulsed. But then there
were eyes — and the fun began.
FOUR

The adaptable eye

1 — ‘A basically disastrous design’

‘The apt; writes Edgar Rice Burroughs, in his science fantasy Warlord
of Mars, ‘was our most consistent and dangerous foe’

It is a huge, white-furred creature with six limbs . . . Its two

huge eyes inspired my greatest curiosity. They extend in two vast,
oval patches from the center of the top of the cranium down
either side of the head to below the roots of the horns, so that
these weapons really grow out from the lower part of the eyes,
which are composed of several thousand ocelli each.

Insect eyes — huge, unblinking, impersonal - are one of the

more visceral and chilling pieces of visual repertoire available to
writers of science fiction and the makers of horror films. And no
wonder: our own eyes, with their bright whites, rainbow irises,
responsive pupils, brows and lashes, not to mention their wide racial
variation, have evolved to communicate and carry meaning. Our
THE ADAPTABLE EYE 103

Robert Hooke’s startling portrait of a fly is the first high-

magnification image ever drawn of the compound eye.

eyes are windows on the soul. How can we fail to shudder at an eye
that cannot weep, turn, or even blink?
There is no reason why natural selection might not, in some corner
of the universe, have thrown up an apt-like monster with large,
bulbous, faceted eyes. Stranger sports have stalked our own planet.
If a snow-elk can get away with antlers far too heavy for its head,
why should there not be an apt-like creature somewhere, looking
out at the world through equally bulky, bulbous eyes?
It is not too hard to imagine a world where every sighted animal,
however big or small, sees through compound eyes. Indeed, for about
85 million years, directly after the Cambrian explosion, compound eyes
enjoyed a monopoly of image-forming vision. Some of them were quite
big: the eyes of the larger sorts of trilobite grew to widths of several
centimetres.
104 A NATURAL HISTORY OF SEEING

The trilobites were an extraordinarily varied class of animal, ranging

from microscopic to the size of a dog. There were stalk-eyed trilobites,
trilobites with bulbous visual arrays that dominated the head, eyes so
large they met in the middle, eyes hidden away in streamlined recesses;
even squitty single facets, the remains of eyes that were degenerating,
abandoned as the animal adopted a lifestyle - under the sediment or
floating, plankton-like, in deep ocean - where vision was no longer
an asset. In 2003, a Moroccan fossil dealer sold trilobite expert Richard
Fortey, who works for the Natural History Museum in London, a fossil
species new to science which Fortey christened Erbenochile. Its eyes
look like the speaker stacks of a particularly bad-taste hi-fi system.
They came with an integral eye-shade, so Erbenochile must have
operated in the shallows during the day. Like the rest of its kind, it
could not look up, and it did not have to: there were no free-swim-
ming predators willing or able to swoop upon it.
It is fortunate for our understanding of how the compound eye
evolved that larval trilobites have left fossil traces. Just as foetal infants,
as they develop, roughly recapitulate the evolutionary history of their
species, so larval trilobites reveal something about their evolution.
The youngest trilobite larvae have widely scattered eyelets pointing
this way and that. These rough arrays of light detectors hardly deserve
the name ‘eyes. Over successive moultings, these arrays compact and
re-order themselves into magnificent, regular compound eyes.
Trilobite eyelets were topped with lenses made of crystal. This
extraordinary design of eye vanished with the last trilobite, and the
crystal eye has never been reinvented. The mineral employed in the
trilobite eye was calcite - the same mineral that makes chalk, lime-
stone and marble. In its purest state, calcite is transparent. When lots
of tiny microscopic surfaces are compacted — the way fossil fragments
are compounded in chalk - light is refracted and reflected to such
a degree that the material appears bright white. When calcite grows
slowly, layer by layer, as a crystal, it is clearer than glass. There are
several forms a calcite crystal can take; a magnificent example of
one, Iceland spar, stands in a case on the first floor of the Natural
History Museum in London. Look through the sides of this crystal
and you will see double: the crystal divides light into two beams.
Another property of calcite is the way it admits light through its
THE ADAPTABLE EYE 105

length. Light striking the top of the crystal at an angle is refracted

to the sides. Only light which shines straight in is carried down its
length. This was the secret of trilobite vision.
The chief mechanical challenge for an eye is to gather light from
different parts of the scene to form an image. A camera-type eye,
_ like our own, has an easy evolutionary path. Once the eye has formed
a pit, closing the walls will produce a pin-hole which casts inverted
images on the back wall of the eye. Filling the eye with clear material
that focuses the light will improve the image. A lens arising from
that clear material will, by moving or deforming, add flexibility to
the eye's optical performance. By now the eye works very well and
is (for reasons we will come to later) quite large. Having more than
a few such eyes would be wasteful and greedy. Humans have two.
Some fish have four. Spiders, being greedy devils, have between six
and eight, but only two are image-forming.
A compound eye, as it evolves, has no such intuitive path. Consider
the choice a primitive array of eyespots has to make at the beginning
of its evolutionary journey. So as not to see the same point in space
multiple times, the array must point its eyespots in different direc-
tions, so that they can report on the light level in different regions
of space. A flat plate of eyespots has to deform, and there are two
ways it can do this: it can form a dimple, or a pimple. Human eyes
are the descendants of dimples. Compound eyes are the descendants
of pimples. To begin with, both solutions seem equally good. Only
later do pimple eyes reveal their inadequacy and by then it is far too
late. Evolution cannot take back its mistakes.
A pimple can never acquire a single, simple focusing mechanism.
Eyespots covering a pimple can form an image only by increasing their
numbers and arranging themselves ever more neatly and ever more
tightly together. Each eyespot, observing a point in space adjacent to
the point observed by its neighbour, provides the nascent compound
eye with one piece of information (a ‘pixel’), and the eye sees a picture
made up of as many pixels as there are eyespots. It is an ingenious
form of vision, but it has one problem. With no device to focus the
light, how can the eye order the image it is trying to capture?
‘The eyespots themselves, to narrow their field of view, evolve into
exquisite and complex structures called ommatidia. Architecturally,
106 A NATURAL HISTORY OF SEEING

each ommatidium is a true eye — and a true camera-type eye, at that.

In engineering terms, this is a disastrously wasteful replication of
effort. (One is reminded of those commercial high-rise buildings in
Bombay where the communal air conditioning has broken down
and the corridors are crammed with the ducts from countless private
units — one for each room.)
An ommatidium consists of a tube, often shielded with pigment
and filled with half a dozen or so elongated photosensitive cells
packed in next to each other and pressed together at the centre. This
photosensitive core is called a rhabdom. The tube is topped by a
clear, curved protective coating, which acts as a lens to gather light
from just the right part of space, so that each ommatidium has a
field of view which abuts its neighbours’ without overlapping. The
trilobite made a happy choice in calcite - a crystal which passes
through its length only that light which hits it head-on.
Although the calcite trick has never been repeated, subsequent inno-
vations have addressed the problem of how to keep each ommatidium
focused on one narrow point in space. Pigment shields, graded lenses,
light-guides, photoreceptors: the paraphernalia of compound vision
still set taxing problems for vision researchers. The pity of it is, of
course, that no matter how exquisite an individual ommatidium gets,
its allotted task is a dull one: to capture just a single pixel of the view.
Growing tens of thousands of little, exquisitely fashioned eyes to
acquire a rough, pixellated view of the world seems a ridiculously long-
winded way of acquiring vision. Dan-Eric Nilsson once remarked of
compound eyes that ‘it is only a small exaggeration to say that evolution
seems to be fighting a desperate battle to improve a basically disastrous
design.* To improve the visual resolution of the compound eye, two
things have to happen: first, there need to be more ommatidia, gener-
ating more pixels. Second, the pixels have to get smaller: that is, each
ommatidium will have to capture an ever-narrower field of view. And
this is surprisingly difficult to do, on two counts.
Let's take the easier one first. A lens held at arm’s length gathers
light from a small patch of space: it has a narrow field of view. Move
the lens closer to the eye, and its field of view widens. By the same
logic, you can narrow an eye’s field of view by increasing the space
between its lens and its photoreceptors.
THE ADAPTABLE EYE 107

corneal lens
lens cylinder

- pigment cell

rhabdom

receptor cell

ganglion layer

nerve fibres

A basically disastrous design’: the apposition-type compound eye.

Human photoreceptors receive light focused through a lens that

lies many millimeters away, which means that each receptor is exposed
to light entering from a very narrow field of view; about 0.007 degrees.
A honeybee'’s photoreceptor, on the other hand, receives light from a
lens that’s barely 100 microns away, which means it is exposed to light
from a much wider field of view — about one degree. That’s an area
about the size of your thumbnail at the end of an extended arm. A
view made of pixels that size would, as the optical instrument maker
Henry Mallock observed in 1894, ‘give a picture about as good as if
executed in rather coarse wool-work and viewed at a distance of a
foot: The honeybee’ eyes are among the best image-forming compound
eyes we know of, and the chief reason for their (relative) excellence
is the distance there is between an ommatidium’s lens and its photo-
receptors. As a consequence, the bee's eyes are so bulbous they domi-
nate its head. Were they any larger, it would fall out of the sky.
It is instructive to imagine how big a compound eye would have to
be to see as well as a human eye. First, we would have to lengthen each
ommatidium by several feet. A delightful and much-reproduced cartoon
by the German vision scientist Kuno Kirschfeld substitutes our eyes
with compound eyes of equal resolution. It reveals why Burroughs’s
fearsome apt, if it exists anywhere in the universe, is as purblind as Mr
Magoo. If it saw as well as we do, it would have eyes wider than its
body is tall and would never be able to raise its fearsome head from
the cave floor. Since huge compound eyes are (to say the least)
108 A NATURAL HISTORY OF SEEING

impractical, evolutionary pressure

has tended to favour miniaturisation.
(Our apt would probably have
squinty little eyes, so small as to be
barely noticeable.) And miniatur-
isation raises the second, more
complicated difficulty, of narrowing
the ommatidium’s field of view.
Given two equally long omma-
tidia of different widths, the
narrower will have the narrower
field of view. But there are severe
limits to just how slim an omma-
tidium can be. Each ommatidium
contains, not just one, but half a
dozen or more photoreceptors,
which cannot be shed willy-nilly
because each one is tuned to a
different wavelength of light. This is
the basis for colour vision, and
Kuno Kirschfeld’s fantastical colour vision is especially
compound eye boasts human- important to small invertebrates,
like acuity, but at a ludicrous many of whom (the honeybee
physical cost. included) see more colours than we
do. The receptors themselves can
become narrower, it is true, and the receptors in modern compound
eyes are squeezed to their limit: they are virtually no wider than the
light waves they are supposed to detect. Many insects tend to see
shorter wavelengths of light than vertebrates do, probably so that they
can keep their receptors as narrow as possible. But this still leaves the
compound eye looking ungainly and bulbous.
The complications set in when we turn to the very narrowest of
natural ommatidia - ommatidia so narrow that they run into diffi-
culties with the nature of light.
Punch a small hole through a sheet of paper and shine a light
through it on to another sheet. You will see a bright disc where the
light shines through the hole, surrounded by a small, faint corona.
THE ADAPTABLE EYE 109

Just as sea waves curve around the walls of a harbour, light waves
curve around the edges of an aperture. Repeat the experiment, but
make the hole even smaller; the bright circle is smaller, obviously,
but the corona is correspondingly larger. The smaller the aperture,
the greater the interference from waves curving around the aperture’s
_edges. This is the phenomenon of diffraction. The smaller a lens gets
- and the lenses in compound eyes are just a few times larger than
the wavelength of light itself - the fuzzier the image becomes.
It hardly matters whether light entering a single ommatidium is
blurred. The ommatidium is only a light detector. The trouble is,
when youre as small as an ommatidium, the blurring can be so bad
that the photoreceptors find it difficult to detect any light. Light
waves can be blurred so much that light captured by one ommatidium
can travel through the barrel of another, only to be detected by the
photoreceptors of a third.
Some species of insect get around this problem by having
pigmented shields around every ommatidium. These don’t have to
be permanent; at night, some insects can withdraw these shields to
improve their night vision.
Other species have hit upon more elaborate solutions. Some do
without lenses altogether, opting for elongated ‘lens cylinders’ with
subtle ray-bending properties. (The horseshoe crab has such an eye,
and for years investigators agonised over its smooth, lensless corneas.)
Butterfly eyes use lenses and lens cylinders. Elsewhere, we find eyes
whose ommatidia harness refraction to trap their light. Refraction is
the name given to the way light bends when it encounters a boundary
between media of different densities; it is why a pencil, dipped into a
glass of water, appears to bend. Two factors determine how much the
light alters course when it passes from one medium to another. The
first is the angle at which it hits the boundary. If the angle is very
shallow, the light is not refracted at all, but reflected. This is why panes
of glass ‘catch the light; making them visible even though they are trans-
parent. The second determining factor is the difference in the densities
encountered at the boundary. In water, glass does not catch the light
— indeed, it becomes virtually invisible - because water is much closer
to the density of glass and thus the refraction effect is much weaker.
The ‘pram bug’ Phronima has found a remarkable use for refraction.
110 A NATURAL HISTORY OF SEEING

A tiny, deep-sea crustacean, Phronima has adopted transparency as its

camouflage, but it still has the problem of how to disguise its eyes. Its
solution: to distort them beyond all recognition. Phronimas eyes are
half a centimetre long — a collosal investment for an animal less than
two centimetres from from top to tail. Rays of light entering its omma-
tidia pass through transparent fibres so narrow that the light cannot
hit the sides at a sharp enough angle to escape. This effect, called ‘total
internal reflection, will be familiar to anyone whos ever owned one of
those shaggy fibre-optic lamps. Phronima has fibre-optic eyes.
Why have more compound eyes not adopted fibre-optics? It
sounds like an ideal solution to the miniaturisation problem. Most
ommatidia do have modest light-guiding properties, but there is a
problem: diffraction imposes an iron limit on how narrow each ‘fibre’
can be. Light waves travelling along an optic fibre set up standing
waves, creating points of utter stillness and points of maximum vibra-
tion. The narrower the receptor gets, the more ‘dead-zones’ it has,
until eventually most of the energy of the light wave exists outside
the receptor. The narrower the prison becomes, the weaker it gets,
until light escapes its bonds completely.

The compound eye is an outlandish and ungainly object; not so much

sculpted by evolution as wrenched and twisted. But if compound eyes
are so poor, why do most animals still plump for them?
Animals with compound eyes are generally small. The smaller you
are, the stronger you are, relative to your weight. Compound eyes
may look ungainly, but a small animal is not going to be much incon-
venienced by having big eyes (and there is one obvious advantage
of having two colossal bulbous eyes - they let you see in every
possible direction, with virtually no blind spots).
Being small has other consequences. If you are very small, you
are unlikely ever to need to see very clearly, because small things
that are far off are probably of no interest. An approaching predator
is likely to be much bigger than you, and easily spotted from a
distance, even with quite blurry vision. Prey, on the other hand, is
likely to be very close by. When a praying mantis is hungry, it simply
reaches out its arms for its next meal. Anything of interest to an
insect is going to make such a large image in its eye - whether
THE ADAPTABLE EYE isl

because of its nearness or its sheer size - that high-resolution vision

is just a waste of effort.
Humans, being large foragers, tend to fetishise the importance
of being able to spot small things. Most other animals are less
bothered about images, and concern themselves more with tracking
_movement. This is another reason why compound vision is a popular
evolutionary choice: even the crudest compound eye handles move-
ment well. The very fineness of human eyesight means that moving
images blur rather badly, whereas insects, with their crude, pixel-
lated vision, are extremely resistant to motion blur. When your visual
world consists of a series of fist-sized pixels, as does the wasp’s,
there's very little detail to lose. Wasps are among the very few animals
whose eyes can see as clearly when they're moving as when they’re
still (that is, not very clearly at all). They have evolved eyes that are
coarse-grained all over, so that they can reconnoitre food sources
at speed.
Wasps are not the only species whose vision seems radar-like in
its workings. Because compound eyes are not very acute, predatory
insects tend not to try and spot their prey against masses of complex
vegetation. Instead, they get below their targets, to silhouette them
against the sky. Many insects see the sky above them with astonishing
acuity. The eye of the dragonfly Anax junius - with over 28,500
ommatidia, the densest compound eye of all — boasts a narrow band
of very acute vision. Far from caring about images however, Anax
junius watches for movement through eyes that scan the sky like a
radar.
To see what compound eyes are really capable of, we need look
no further than the aerial acrobatics of airborne insects. All the
compound eye's small but telling strengths come together in flight.
Honeybees, as we've seen, have blurry vision, yet they fly with remark-
able élan, pulling off tricks that put the best stunt pilots to shame.
A honeybee’s brain is the size of a sesame seed. How do they manage
to fly so well? How do they manage to land?
If, when you come in to land, you keep the texture of things
moving evenly across your field of view, you will automatically reduce
speed as you descend closer to the ground. If you keep things moving
in straight lines, you will automatically descend at a constant angle.
112 A NATURAL HISTORY OF SEEING

You don’t need to know your airspeed or your height. All you have
to do is keep your eyes open. >.
In the thirties and forties, G.C. Grindley (known to everyone as
‘C’) was one of the brightest stars in the field of animal behaviour.
Declared unfit for service at the outbreak of World War II, Grindley
— known for his interest in vision and motion - was recruited by
the Flying Personnel Selection Committee to study how pilots land
their aircraft. He accompanied an instructor, Peter May, on a
training flight. May’s description of how to land a plane was
admirably down-to-earth: “You look at the point where you want
to land and fly towards it till the ground explodes around it. Then
you flatten out.?
Back on the ground, Grindley tried to realise May's description
in mathematics, and came up with the equations of ‘optic flow. Today,
at the All-Weather Bee-flight Facility at the Australian National
University in Canberra, those same equations are supporting studies
of how bees fly.
Airborne insects judge distances using optic flow. If the bee can't
see any texture, it assumes it is flying in open air, and flies quickly.
(This is why bees become disoriented indoors and tend to bounce
off painted walls: they cannot distinguish between plain surfaces and
open sky.) Once the bee notices texture it slows down, and tries to
keep the texture moving steadily across its field of view. The closer
the texture, the more quickly it moves across the bee's field of view,
and the more slowly the bee flies. If images start to move very fast
on one side, the bee veers away. A bee comes in to land the same
way Peter May did; it descends at an constant angle, then levels out
at the moment when the images directly ahead ‘explode out’ towards
the sides.
The pattern of optic flow is echoed in the eyes of most flying insects,
whose forward-facing ommatidia are narrower and more compact than
the ommatidia that look to the sides. This means their forward vision
is most acute, while views to the sides are deliberately coarse-grained,
to capture the fast-flowing landscape. By dividing the visual scene into
substantial pixels of varying width, flying insects have acquired eyes
optimised to detect optic flow; that is, optimised for flight.
+ * +
THE ADAPTABLE EYE ES

ee eo SS?

Optic flow: the arrows represent the speed and direction of

texture flowing past the photoreceptors of the eye.

Another reason for the ubiquity of compound eyes may be the sheer
difficulty involved in trying to change a compound eye to a single-
chambered eye without at any stage damaging one’s eyesight; the
compound eye may simply be too specialised to unpick. Mind you,
spiders and scorpions seem to have succeeded, abandoning compound
eyes in favour of single-chamber eyes around twenty million years
ago. True spiders have eight eyes. Their high-resolution eyes — the
best of them as large and as powerful as the eyes of small rodents -
point forwards, while the others scout for movements off to the sides.
In web-building spiders, the peripheral eyes act as navigation devices.
Why have other invertebrates not followed the spider’s example?
Some have tried. The first was a mysterious form of trilobite called
Phacops. Fifty million years after the emergence of the first trilobites,
the phacopids made themselves a new sort of eye. With it, they came
to dominate the whole trilobite class.
The individual ommatidia of compound eyes capture a single pixel
of the view. The more ommatidia there are, the better the eye can
see. But Phacops rewrote the rule book: it lost lenses. However, the
114 A NATURAL HISTORY OF SEEING

hundred or so lenses it retained grew much bigger — some are nearly

a millimetre across — and each sat in a small depression, dividing it
from its neighbours.
When these eyes were first studied, it was assumed that they were
a degenerate form. One typical path for evolution to take is to bring
young animals to sexual maturity earlier and earlier. This means that
adult bodies can eventually acquire foetal characteristics. In the 1960s
it was assumed that phacopid trilobites had re-adopted their foetal
eyes; abandoning compound eyes for scattered eyespots — good for
detecting movement, but little else.
But why, if these eyes were degenerate, were the individual lenses
so large? At first it was thought that they were large so as to gather
light from dim environments. But the figures made no sense. The
eyes were actually too good: a single ommatidium boasting a lens
like this would be blinded by starlight. It seems much more likely
that each lens served a retina made up of many ommatidia. And
why were these lenses spaced out so precisely on an hexagonal
pattern, with a neatness that put their compound cousins to shame?
(A compound eye normally contains a good number of small irreg-
ularities.)
In 1972, Kenneth Towe - a palaeobiologist at the Smithsonian
Museum in Washington - managed to take a photograph (of the
city’s FBI headquarters, no less) through the lens of a Phacops eye.
Towe'’s photographs showed clear inverted chunks of the building.
The images were too clear to be incidental to the eye’s main purpose,
and made a nonsense of the idea that these were primitive eyes.
Working out how the images could be so clear required the assistance
of a nuclear physicist, Riccardo Levi-Setti. He discovered that the
lenses in Towe's trilobite eye were not pure calcite: a layer of magne-
sium ran through the middle, correcting their optics and minimising
blur. These lenses were more than light collectors. They were supposed
to capture images.
The trilobite could not possibly have used inverted chunks of the
visual scene in their raw state. It must have turned each of these
chunks the right way up in the eye and stitched them together. And
in that process, it may possibly have acquired another extraordinary
optical talent: depth perception. The lenses are arranged so very
THE ADAPTABLE EYE 115
carefully, and yet their views overlap. Was Phacops capable of
triangulating the distance of objects from these overlapping views?
If so, this was a truly extraordinary visual system, giving Phacops
a perception of depth that puts our own, merely binocular, vision to
shame. This is too good an idea to leave languishing in the text-
books, and has recently inspired a number of projects in artificial
vision. Artificial Phacops eyes may yet provide self-steering robots
with eyes that can navigate the canyons of Mars, building up detailed
maps as they go.3
Even supposing that Phacops wasn't quite so ambitious (and it is
hard to see why it would be), this hardly matters. Whatever uses it
had for vision, it saw images, and it had to stitch these images
together. It follows, therefore, that Phacops would have needed a
retina several neural layers deep to process its images. This, in turn,
says volumes about what kind of nervous system it must have had.
Phacops was no dummy.

There were other ‘experiments, as the compound eye was driven by

natural selection to improve its limited design. Chief among them
is the ‘superposition’ compound eye, whose mysterious workings were
unpicked by the Austrian physiologist, Sigmund Exner.
Exner was born in 1846, to a wealthy and socially prominent
Viennese family. He was extraordinarily fortunate in his education;
his tutors, Ernst Briicke and Hermann von Helmholtz, were the two
greatest names in nineteenth-century physiology. From 1871, as an
assistant in Professor Briicke’s renowned physiological laboratory
at the University of Vienna, Exner was the mentor of the young
Sigmund Freud. Both speculated on the nature of the connection
between the mind and the nervous system, and Exner’s book on
the subject provided the framework for Freud’s early thinking about
the brain.
Exner’s fascination with natural history encompassed the nuts
and bolts, as well as the big, imponderable questions. A keen hiker,
Exner once observed a buzzard riding a thermal. To understand
how the bird managed to rise in the air without moving its wings,
he built an experimental buzzard from feathers, clockwork and card-
board. And he once delivered a delightfully oddball lecture which
116 A NATURAL HISTORY OF SEEING

attempted to reconcile aerodynamics with the artistic presentation

of flying and floating human figures; angels, putti and the like. With
a show of perfect sincerity, he stated that putti - the winged heads
gracing the corners of many a Renaissance painted ceiling — weigh
about two grammes and travel at a speed of around two metres a
second.
The marriage of engineering nous and a nutty imagination
served Exner well throughout his career. Who else, at a time when
sound recordings were scratched in wax, would have created the
world’s first sound archive (the Phonogrammarchiv in Vienna),
laying down recordings which are still playable today? Who else
could have written a study of compound eyes as insightful as his
1891 monograph; a text so clear and authoritative that it remains
a standard, last reprinted in 1989? This work is more than a text-
book: in it, Exner solved one of the most intractable problems in
vision — a class of compound eye which seems to refute the laws
of optics.
Some compound eyes do not produce multiple images. Their
lenses, pressed promiscuously together to form a dome over an eye
that is virtually hollow, produce, upon a uniform retina not so dissim-
ilar to our own, a single, upright, image.
Fireflies, beetles, moths, and some some species of plankton have

Two forms of compound vision: the individual eyelets of an

apposition eye (left) each view a different region of space. Each
facet of a superposition eye (right) bends light selectively on to
different photoreceptors. Because it harnesses more of the available
light, a superposition eye is better at seeing at dusk and dawn.
THE ADAPTABLE EYE 7

turned their compound eyes into single-chambered eyes, called ‘super-

position eyes, because the images from each lens overlap, producing
a single coherent image. It took Sigmund Exner most of the 1880s
to work out the horrendous mathematics of such a composite upright
image. In doing so, he made a remarkable discovery. The lenses were
Not behaving like lenses: rather than focusing beams of light, they
were redirecting them. The individual lenses bent light rays so that
they met at a single ‘focus’ point where one coherent upright image
was formed.
How could lenses behave this way? Two convex lenses, carefully
arranged, can invert and re-invert an image, effectively redirecting
the light; but there were no such arrangements in the eyes of beetles
or moths. Even more puzzlingly, the surfaces of the lenses just didn’t
seem curved enough to converge light rays in the normal way -
never mind contort them in ways unknown to optics. Exner realised
that variations in refractive power inside the lenses must be
performing the inversion and re-inversion. The lenses were not simple
affairs but ‘lens cylinders’ with complex layers.
With no means of testing his idea, Exner set about the calculations
that would explain the behaviour of these lens cylinders. A hundred
years passed before equipment was invented that could test his
conjectures, but these experiments, conducted in 1979 and 1989,
confirmed his figures to the smallest detail.
Many superposition eyes see just as well as apposition eyes; the
best see as well as a bee. And, with up to three hundred optical
elements contributing to the image, the superposition eye is hugely
sensitive — up to a thousand times as sensitive as an apposition eye
— and ideal for seeing in the dark. It is so good, at least one species,
the nocturnal hawkmoth Deilephila elpenor, can distinguish colours
in starlight.°
New forms of compound vision continue to emerge. In 1976,
Kuno Kirschfeld discovered that flies and a handful of crustaceans
use overlapping views and some complex retinal knitting to sample
each point in space seven times. This seven-fold increase in light-
detecting ability is great for dusk and dawn vision. Twelve years later,
Dan-Eric Nilsson discovered a swimming crab with a unique form
of compound vision, involving lenses, wave-guides, and parabolic
118 A NATURAL HISTORY OF SEEING

How flies see the world: in a form of vision dubbed neural

superposition, photoreceptors in different ommatidia pool
their signals at the level of the lamina (the insect retina)
to improve vision in poor light.

mirrors. Michael Land of Sussex University has the distinction of

having worked out the optics of the lobster eye, which uses cuboid
mirrors in place of lenses. (A distinction he shares with Klaus Vogt
of Freiburg University - unbeknownst to each other, both had been
working on the same problem.) There will be other discoveries yet.

2 — Leather, water and jelly

In 2003, Anna Gislén, of Lund University in Sweden, grabbed one

of vision research’s more exotic projects: she went snorkelling around
Thailand's biggest and most fertile shallow coral reef.
In tow she had eighteen Western children — volunteers from the
40,000 holidaymakers who come to the Mu Ko Surin National Park
each year — and seventeen children from the Moken, a local nomad
people. Gislén was testing the truth of a remark by a senior colleague:
that the Moken can see clearly underwater.
Human vision is lousy in water, because most of the focusing
power of the human eye is achieved by its curved cornea: light passing
from air into an eyeball is refracted strongly. Light passing from
THE ADAPTABLE EYE 119

water into the eyeball, on the other hand, is hardly refracted at all,
and the human lens is not nearly strong enough to make up for the
shortfall in focusing power.
But someone forgot to tell the Moken, who ply the Burmese arch-
ipelago and Thailand’s western coast. While the menfolk spear fish,
the children spend their days diving for clams and sea cucumbers:
‘When I first saw them, I was instantly struck by their familiarity
with water, Gislén told the Australian television company, ABC.
‘Though it may sound like a cliché, they looked a bit like a school
of little fish swimming around: Simple eye tests — gauging the chil-
dren’ sharpness of vision by asking them to distinguish different
patterns underwater — quickly gave experimental weight to the folk
tale: the Moken children could see fine detail underwater more than
twice as well as the Europeans.
The Moken were not superhuman — but they were using the eye
to its limits. While the pupils of the eyes of the European children
expanded underwater, in response to the dimness of the light, the
pupils of Moken eyes shrank to their smallest possible diameter -
smaller than European pupils ever go. Though this considerably
reduced the amount of light available, it improved acuity no end.
Moken children also used their lenses more than European children,
squishing them to the limit of human performance.
Gislén is cautious about the implications of her research. “This
behaviour, says her report, ‘is clearly an adaptive strategy. But after
centuries of this kind of life, the Moken people's unique visual ability
may just possibly have found its way into their genome. Survival
pressures are, after all, the reason why genetic variations are selected
and preserved, why communities change and vary, and why, ulti-
mately, new species arise.

The arrival of eyed vertebrates, some eighty-five million years after

the Cambrian explosion, introduced an optical innovation so
profound, it may be considered a wholesale reinvention of the busi-
ness of vision. Although mechanically similar to the simple, single-
chambered eyes of molluscs and crabs, the vertebrate eye harnessed
the available materials in entirely new ways. Its lenses were made
out of crystallins, incredibly stable proteins that provided flexibility,
120 A NATURAL HISTORY OF SEEING

Spherical aberration: the periphery of a lens focuses light more

strongly than its centre, making a blurred image (shaded oval).

transparency and excellent ultraviolet protection for eyes that could

grow bigger, and see the world better, than ever before.
The larger you get, the less power you have, relative to your weight.
Ants can carry five times their own body weight in their teeth; humans
most certainly cannot. The larger you are, the more likely it is, then,
that you feed on animals smaller than you. By the same token, your
predators are likely to be a lot larger than you, and correspondingly
faster, so it is important to spot them from further away. A compound
eye is useless at fine detail and long-range imaging. A vertebrate eye
is excellent at both. Had it not been for the vertebrate eye, living things
would probably have remained lobster-sized for ever.
Even the smallest vertebrate eye boasts a lens that is far, far larger
than any single lens in a compound eye. Its sheer size resolves the
diffraction problem straight away, because the area of focused light
is so large that its accompanying diffraction rings are virtually unde-
tectable. But large lenses have two other weaknesses: spherical aber-
ration and chromatic aberration.
In spherical aberration, the light hitting the edge of a lens is
refracted so strongly that it comes into focus in front of the rest of
the image. One solution is to make the outside of the lens of weaker
optical stuff than the inside. The physicist Heinrich Matthiessen
THE ADAPTABLE EYE 121

blue
white red

blue

Chromatic aberration: a lens focuses short-wavelength

(blue) light more strongly than long-wavelength (red) light.
The figure reflects roughly how the human eye handles
chromatic aberration: red light is brought to focus at the
fovea, while blue light blurs across a wider area.

(1830-1906) worked out the optics of the ‘graded index’ lenses of

fish, and thereby laid the groundwork for Exner’s discovery of lens
cylinders and other optical exotica.
The first vertebrates to venture on to dry land would necessarily
have been extremely short-sighted. Fish, having gone to all the trouble
of evolving powerful, spherical lenses, were now entering an environ-
ment where a lens was more trouble than it was worth. Light passing
from the air into such an eyeball is so effectively refracted that the lens
overcompensates, bringing the image into focus much too far in front
of the retina. This is why the lenses of land-based eyes have lost much
of their curvature. It is one of nature's happy accidents that weakening
the lens by drastically slimming it down is one solution to the problem
of spherical aberration. Nowhere, except at the very edges, can light
hit the lens at an angle acute enough to blur the image. This is just as
well because — as we've already seen with our own eyes — the peripheral
image through a slim-line lens isn't that great to start with.
The other major optical problem of large lenses is chromatic aber-
ration. To put it simply, lenses bring the blues of a scene into focus in
front of the reds. The kind of distracting mess this generates can be
ip) A NATURAL HISTORY OF SEEING

best appreciated by looking through a pair of toy binoculars: its cheap

lenses, cast out of an homogeneous plastic, cannot correct for chromatic
aberration, and vividly coloured haloes appear around objects,
disrupting their outlines and confusing the perception of depth. The
effect is pretty, but you wouldn't want to have to live with it.
Early telescope makers soon learned how to bring all wavelengths
into focus at the same point; using a combination of different materials,
they were able to produce achromatic lenses. Oddly, although the lenses
of eyes are laid down bit by bit, which resolves spherical aberration, no
vertebrate lens has ever really addressed the problem of chromatic aber-
ration. Some vertebrates have resolved the problem by adapting the
retina: the squirrel’s retina is corrugated, so that different parts capture
differently coloured focused views. Most other vertebrates don't go to
this much trouble. They have smooth retinas, in which the cones are
arranged in concentric fields to capture light of different wavelengths.
The human retina has evolved the fovea: a small flattened disc,
the very centre of which is crammed with just two types of our
three colour-detecting cone cells. Greenish and yellowish light thus
comes into focus at the centre of the fovea with nice precision. Our
third, blue-detecting cone type is not found in the fovea but is
scattered among the rods filling the rest of the retina. This is for the
very good reason that blue light comes into focus a good half-
millimetre in front of the retina, eventually reaching it in a weak,
blurred wash. Thus, blue light is always out of focus in the human
eye, which may explain why, when we look at a painting, the blues
and greens seem to recede, while the yellows and reds leap out. It
also rather begs the question of how blue objects in our field of view
have sharply defined edges at all? The degree to which the retina
has to reprocess the mess of different lights inside the eye, accentu-
ating lines, boundaries and forms, is the subject of Chapter Seven.

Some of the largest eyes in nature belonged to ichthyosaurs. These

marine contemporaries of the dinosaurs were themselves descended
from land animals, making them the Jurassic equivalent of dolphins.
Like dolphins, ichthyosaurs came to the surface to breathe and gave
birth to live young. The chances are that their table manners were
very like a modern dolphin’s: their long thin jaws would have been
THE ADAPTABLE EYE 123

excellent at capturing fast, agile prey such as large fish and squid.
Some early ichthyosaurs had teeth adapted for crushing shellfish.
Of the eighty known species of ichthyosaur, the largest was
Temnodontosaurus platyodon; more than 15 metres long, Temnodon-
tosaurus resembled a great white shark more than a dolphin and,
hot surprisingly, its eye - at about thirty centimetres across — is the
- largest known vertebrate eye.
A more familiar ichthyosaur is Ophthalmosaurus, named, with
good reason, for its eyes. Although smaller in absolute terms than
the eyes of Temnodontosaurus they were nevertheless the largest in
relation to body size. We know a great deal about how Ophthal-
mosaurus saw because of a characteristic of vertebrate eyes not often
mentioned: they have bones.
If you are of a certain age you may, when you look the in the mirror,
spot an off-white ring around your cornea. If you can, you are looking
at pure cholesterol. Where humans lay down cholesterol — within the
sclera and around the cornea — most vertebrates lay down bone. For
some reason, mammals and crocodiles have lost their sclerotic bones,
but every other vertebrate has them, in some form or other. The most
complete sclerotic rings we know of — hefty, doughnut-shaped bones,
like a fossilised pineapple slice - belonged to the ichthyosaurs.
When a fish-shaped object moves, the front part is pushed by the
surrounding water while the flanks are pulled. An eye that’s relatively
small in relation to body size can be positioned where the forces are
about equal. Ichthyosaurs, with their huge eyes, didn't have this
option. They reduced the stress on their eyes as much as possible by
evolving particularly flat corneas, so that their eyes were relatively
flush with their skin, but this made their eyeballs an odd shape.
So, the ichthyosaur’s shallow eyeballs needed scaffolding. This is
very good news to palaeontologists, since fossilised sclerotic rings
provide them with reliable measurements, not only revealing how
old an ichthyosaur specimen was when it died, but also how much
light its eyes could gather. Ophthalmosaurus is well-named: it could
see in the dark - and were it a land animal, the obvious conclusion
would be that it was nocturnal. For an air-breathing aquatic animal,
this is very unlikely; Ophthalmosaurus would have to have slept on the
ocean surface in broad daylight. It is much more likely that it was
124 A NATURAL HISTORY OF SEEING

a deep diver, hunting for prey in the dark waters more than half a
kilometre below the surface.
This is still not enough to explain why it had such large eyes.
Modern seals have relatively small eyes, and they hunt just as deeply
for their food. No one has yet explained with certainty why
ichthyosaurs boasted such large eyes. Perhaps it says more about their
prey than it does about ichthyosaurs themselves: the smaller and more
manoevrable their prey, the larger their eyes would need to be, to see
and track a meal at such depth.

The vertebrate eye evolved in the ocean, and when it moved on to

dry: land, it took a little of the ocean with it. It had no choice: the
tissues from which the vertebrate eye is made need to stay wet to
survive. Even while it was still in the water, the vertebrate eye had
acquired a number of protective devices, any and all of which were
capable of protecting and cleaning the eye once it left the seas.
These were not ‘eyelids’ as such; the water-dwelling eye has no
need for the wiper-blade of a moving eyelid. Rather, they were spec-
tacles, examples of which can be found throughout the natural world,
on land as well as in the seas.
The corneas of lampreys exude transparent membranes: ideally,
these lie against the cornea proper, making them quite hard to spot.
If they are damaged, they can peel away, with no damage to the
cornea beneath. Bottom-feeding fish have their eyes permanently
shut, hidden beneath a layer of translucent skin that protects them
against sand and debris. Some fish with bulging eyes wear translucent
skins over their eyes, to improve their streamlining. Snakes and other
reptiles are descended from animals that had eyelids more or less
like ours. Confronted with constantly blowing sand, however, their
eyelids have fused over and become transparent.
For most land animals, cleaning the eye is as important as protecting
it; hence moving, semi-rigid eyelids. Look into the near corner of your
eye: the little pinkish fold there is the last remaining nubbin of your
third eyelid. Many terrestrial vertebrates have hung on to these inner
eyelids — called nictitating membranes — and their uses are legion.
Beavers and manatees use them as a barrier to reduce inflammation
when they go swimming; seals peel them back while they swim, but
THE ADAPTABLE EYE 125

slide them over their eyes on land to clean off sand and other debris;
the aardvark uses them to shield its eyes from its termite prey; birds
of prey use them to protect their eyes from their chicks while they are
feeding them, and in polar bears, the nictitating membranes protect
the eyes from snow blindness.
_ So much for our wet, delicate and almost unbearably vulnerable
eyes: looked at from an evolutionary perspective, the vertebrate eye
is one of life’s most resilient adaptations, capable of resolving the
world with exquisite clarity at depth, in desert storms and Antarctic
white-outs, under the brightest of bright lights, and in near-blackout.

Retinas are sculpted by the kind of light they are exposed to. Blurred
light encourages visual abilities that do not rely on clear images: areas
devoted to the detection of movement, and to providing good all-round
vision in dim light. Well-focused light favours visual abilities that make
use of plentiful light and a fine-grained view of the world. This is why
the photoreceptors of the vertebrate retina have evolved and specialised
into rods and cones. Rods can be triggered by light coming from all
directions; their resolution is poor, but they are very good at gathering
light. Cones are different; like little optical fibres, cones gather only the
light that hits them straight on, and cannot be so easily triggered by
stray light. Only well-focused light, entering the eye through the centre
of the pupil, will stimulate them. This makes them useless in the dark,
but very good at seeing the world in detail during the day.
For rabbits, the whole horizon is in reasonably good focus, and
they become aware of potential predators the moment they appear.
They barely have to move their eyes to get a bead on potential threats,
and their eyes perform very few saccades. They do not have foveas,
because they do not need them. Neither do cats, even though their
eyes face forwards and they have to judge distances with some accu-
racy. Cats pay more attention to the horizon than to a specific focal
point. Like the retinas of rabbits, cat retinas boast ‘visual streaks’:
lines of enhanced acuity, aligned to the horizon.
Seabirds are among the many animals that have both a visual
streak and an area of good central vision (not nearly as specialised
as a primate fovea) called the area centralis. Birds of prey have a
fully fledged fovea; some birds ave two, which enables them to
126 A NATURAL HISTORY OF SEEING

watch what they’re eating at the same time as theyre watching the
world.
The vertebrate retina has adapted over time to serve some truly
bizarre visual requirements. My favourite is the surface-feeding fish
Anableps anableps, a four-eyed fish from Central and South America
which greatly entertains my daughter when we go to our local
aquarium. It scuds along the surface looking for flies: thanks to its
double eyes, it can see clearly in and out of the water.
Human eyes are no less specialised, and no less odd. Only animals
who forage in a three-dimensional environment, for example, the
forest canopy, can afford a retina that is not in some way tuned to
the horizon. In humans, who have returned to the plain, this hori-
zonless style of vision, far from falling away, has been taken to a
pitch of extraordinary specialisation. Only a handful of birds of prey
have foveas with finer acuity than our own - and they have to detect
prey hundreds of metres away on the wing. Why are humans, of all
land-dwelling vertebrates, virtually eagle-eyed?
This account from 1896 holds the clue:

An expert tool juggler in one of the great English needle facto-

ries, in a recent test of skill, performed one of the most delicate
mechanical feats imaginable. He took a common sewing needle
of medium size (length 1 5/8 inches) and drilled a hole through
its entire length from eye to point - the opening being just
large enough to admit the passage of a very fine hair. Another
workman in a watch-factory of the United States drilled a hole
through a hair of his beard and ran a fiber of silk through it.®

Without hawk-eyed vision, we would be unable*to perform close

work. We would not be able to make tools. We would not be able to
arrange and organise our environment. We would not be able to
make fine distinctions between objects, spaces, species, peoples, and
individuals. The very richness of our social world, full of manipu-
latable things, rather than fleeting, corner-of-the-eye impressions,
has driven the human eye towards high acuity vision, and continues
to do so. Since the 1280s, humans have been grinding lenses and
sticking them on wires in front of their eyes. Now we stick the lenses
THE ADAPTABLE EYE 127

on to our eyes: last weekend I went for a dive wearing soft plastic
self-balancing contact lenses, which correct for the slightly uneven
shape of my eyes’ natural lenses (a common condition, astigmatism).
When I came out, I plucked those two tiny, barely visible miracles
of materials science off my corneas and tossed them into the ocean:
they were only disposables.
~ We look through telescopes and microscopes. The more foolhardy
of us allow our corneas to be sculpted by lasers and spinning blades,
on the promise that we might get to see the world in finer and finer
detail. There is much evidence that, in our hunger for good tool-
making vision, we are steadily blinding our species, and becoming,
as a consequence, more and more reliant on technology to prop up
our ailing sight. Myopia rates are rising alarmingly. In 1996, about
sixty per cent of American twenty-three-to thirty-four-year-olds were
short-sighted, compared with about twenty per cent of people over
sixty-five. In Asia, things are much worse. The Singapore National
Eye Centre estimates that more than eighty per cent of the country’s
eighteen-year-old men are myopic.’
This matters. In the developed world, degenerative myopia is a
leading cause of blindness. People who (like me) wear lenses stronger
than six dioptres are more susceptible to glaucoma and more likely
to suffer a detached retina. Although susceptibility to myopia can
be inherited, the environment is most to blame. The bald fact is that
illiterate populations need fewer spectacles. In rural Nepal and the
upper Amazon, only one in a hundred is short-sighted. There is even
evidence that children grow more short-sighted in term-time than
during the summer holidays. Many epidemiologists blame the current
myopia explosion on families (families are a favourite soft target for
this sort of hell-in-a-handcart punditry): children should be getting
out more, and playing on their PlayStations less!
No one ever suggests closing down the schools.
FIVE

Seeing and thinking

About a year ago, I was sitting in a café when an exhausted-looking

woman sat down beside me, took her baby daughter out of her pram,
and set her down on the chair beside her. The baby - and she really
was only a baby - climbed off the chair, dropped neatly to the floor,
and started running around the room.
Impressed, I asked her mother how old she was. “Ten months, she
replied. “REBECGA, MIND THE...)
Rebecca bounced off the leg of a nearby table, giggled, turned,
and hurled herself at a wall. Thanks to her prodigious ability to walk
- run — before she was even a year old, Rebecca demonstrated an
important fact about human vision to me: for the first year or so, it
is distinctly unreliable.
During its first month of life, a baby’s eyes can do little more
than protect themselves. An object moving on a collision course
with its eyes will make the baby blink. At three months, the baby
uses the motion of objects to work out where their edges are; a
month later, the baby will use both motion and stereo vision to
work out how the edges relate to each other, beginning, for the first
time, to see in three dimensions. At seven months, a baby can use
shading, perspective, and occlusion (the way some objects obscure
,a

* SEEING AND THINKING 129

others) to perceive depth and shape. Now - and only now - is the
child is old enough to recognise familiar objects by sight alone. At
twelve months, they begin to name things.

This last point, impressive as it is, may seem irrelevant to the business
of vision, for surely there is no obvious or direct connection between
“vision and language?
The cognitive and computer scientist Donald Hoffman thinks
there is. He has studied how we divide the world into parts, and how
we give those parts names, uncovering a set of ‘rules’ by which we
divide surfaces and lines into things and parts of things, and he is
finding ways to express these rules in numbers. He explains: “The
way we carve up the world verbally is not arbitrary; it depends in
part on how we carve it up visually. And the way we carve up the
world visually is not arbitrary; it depends in part on fundamental
principles of mathematics:
How should we interpret Hoffman’s work? He has come up with
a workable model of how we see things. But that does not mean our
eyes really work like that. The American psychologist James Jerome
Gibson (1904-1979), whose most influential papers were published
in the 1960s, pooh-poohed the idea that perception required any
kind of calculation. According to Gibson, the mind does not need
to assemble the visual world out of single points of light. It does not
need means (rules) to achieve ends (vision). All it has to do is
‘resonate to the many salient patterns light contains.
My guess is that somewhere between the maddening vagueness
of Gibson’s ‘ecological school of vision and the maddening prescrip-
tiveness of the computational school (for which Donald Hoffman is
a recent and able exponent) there exists a good description of how
seeing is related to thinking. I have neither the space nor the talent
to do the job myself. I can do no more than conduct a whistle-stop
tour of the mysteries and problems surrounding the seemingly
mundane act of ‘seeing something, and suggest ways in which seeing,
thinking and language are related.
The first part of this chapter, ‘Seeing things, looks at what is, even
among vertebrates, a rare talent: the ability to divide the world into
this and that. How are the same objects spotted and recognised from
130 A NATURAL HISTORY OF SEEING

very different angles, under many different lights? The second part,
‘Seeing and remembering, looks specifically at human vision, which
has taken the business of identifying objects to a curious level of
specialisation. The final part, Reading each other’ looks at the eye's
role in communication. The several species on Earth that employ
language all have excellent spatial vision (they have something to
talk about) and are all social (they have someone to talk to). I tell
the story of efforts to unpick the relationship between good vision
and social living in our own species — and why that relationship may
have encouraged the development of language.

1 — Seeing things

Konrad Lorenz (1903-1989) discovered his vocation early. Each

evening, when he was a small child, one or other of his parents would
read to him from The Wonderful Adventures of Nils, a children’s book
by the Nobel prize-winning novelist Selma Lagerl6f. It tells the story
of Nils Holgersson, a boy turned by magic into an elf. Happily for
science of animal behaviour, elves ride on the backs of wild geese:
‘T yearned to become a wild goose; Lorenz remembered, on receiving
his own Nobel prize. “ .. on realizing that this was impossible, I
desperately wanted to have one and, when this also proved impossible,
I settled for having domestic ducks’
From these charming, humble beginnings, and from the‘shattering
disillusionment’ that accompanied his adult study of the literature
of animal behaviour (‘None of these people knew animals’, he
remarked. ‘None of them was an expert’) Lorenz constructed a new
science: the study of animal behaviour.
A particular interest was the way newly hatched birds ‘imprint
on the first big moving object in their visual world. Unlike humans,
who seem content to wobble blearily about for their first year, newly
hatched goslings wire up their visual worlds with impressive urgency.
Haste, of course, can lead to mistakes, and for the goslings who
imprinted on Lorenz, the scientist became ‘mum (and more - the
same animals later found him sexually attractive).
In 1936, Lorenz met Nikolaas Tinbergen, the restless younger son
SEEING AND THINKING 131

of Dutch intellectuals. Tinbergen, with a catalogue of mediocre school

reports and lax university attendance behind him, was approaching
his thirties, trying to turn his childhood interest in rambling and
nature study (anything to get out of the classroom) to some account.
His eccentric and original approach to biological problems, and espe-
cially his insistence on the value of field study, found a ready audience
“in Lorenz. Ostensibly, their relationship was one of master and
pupil, but Tinbergen’ irrepressible urge to check Lorenz's hunches
by experiment soon elicited, Tinbergen later remembered, ‘an almost
childish admiration’ in the senior partner.
It is significant, incidentally, that Lorenz and Tinbergen favoured
birds for their studies. Birds are easier to understand than mammals
because, they, like us, depend largely on their eyes for social contact;
most other mammals think through their noses. The biologist Sir
Julian Huxley once remarked that if we were olfactory animals, there
would be no bird watchers: we would have mammal-smelling soci-
eties instead.
Tinbergen set out to find the simplest stimuli that would provoke
a response in the birds under Lorenz's care. If he dangled a scrap of

Top: baby blackbirds recognise a mother blackbird

in this arrangement of circles.

Bottom: in this shadow-play suited to the visual proclivities of

ducks and chickens, a hawk (left) harries a goose (right).
132 A NATURAL HISTORY OF SEEING

black cloth from his fingers in front of his jackdaws, they reacted as
though one of their number was being attacked. Young blackbirds,
presented with two discs of different sizes (one about a third of the
diameter of the other) treated the arrangement as a parent. Chickens
and ducks distinguished hawks and geese as crosses: the goose has
a cross-piece at the back; hawks have a cross-piece at the front.
We may smirk, and think that these animals were easily fooled,
but if we do, we miss the true glory of Tinbergen’s discoveries. The
world is not full of flying crosses or carefully proportioned black
discs but it is stuffed with parents, siblings, predators, mates, and
prey. To recognise these things from every possible angle, under
every possible light, animals make generalised models. Our experi-
ments fool them only because we've presented them with a model
that closely resembles their mental model.
We can fool ourselves in exactly the same way. The Necker cube
(below, left) is fun because it ‘flips: The figure has no shading or occlu-
sion, so there are two ways of interpreting how it lies in three-dimen-
sional space. Is that a near edge, or a far one? Is that surface seen from
above, or below? The mind cannot decide; it can only juggle the alter-
natives. This is, quite frankly, weird, not least because both answers
are quite spectacularly wrong. The right answer is obvious: the figure
lies flat on the paper. We might know that the figure is flat, but we
find it extremely difficult to see that it is flat. Only when we tweak it
(below right) so that its near and far corners are perfectly aligned -
a coincidence too great ever to occur with any frequency in the real
/ SEEING AND THINKING 133

world — do our eyes accept the obvious: the figure is two-dimensional.

On a busy, changeably lit planet, where objects approach us from
all angles, set against goodness knows what backgrounds, seeing involves
making assumptions. The more limited your surroundings and
behaviour, the less you need interrogate the world, and the larger your
assumptions can be. Brains are ruinously expensive to run, so the fewer
brains an animal requires, the better. A hedgehog can attempt to seduce
a toilet brush, and its long-term breeding chances are not much affected.
For a human being, a mistake on this scale would be catastrophic. We
are a social animal, and we expect more from our mates.
We are still capable of error. To see is to surmise, and error is always
possible. Konrad Lorenz recalled one such mistake during his Nobel
lecture: ‘Once I mistook a mill for a sternwheel steamer. A vessel was
anchored on the banks of the Danube near Budapest. It had a little
smoking funnel and at its stern an enormous slowly turning paddle-
wheel.: This was no different in kind from the mistakes made by
Tinbergen’s baby blackbirds. Seeing things involves making assump-
tions and that requires having an idea of what they might be.
It may not be much of an idea. The pathway from retinal pattern
to idea and back to visual impression is such a well-trodden one
that we're no more aware of it than we are of the beating of our
hearts. As Hermann von Helmholtz wrote: “The psychic activities
that lead us to infer that there in front of us at a certain place there
is a certain object of a certain character, are generally not conscious
activities, but unconscious ones.” An idea it is, none the less - and
this raises an interesting and by no means easy question: at what
point does seeing become thinking?

2 — Seeing and remembering

Different animals have different fields of view. Some fish have an excel-
lent vertical field of view, and are able to observe the world both above
and below. Some birds and insects have eyes positioned so that they
can see all the way around; they have a 360° horizontal field of view.
The horizontal field of view enjoyed by humans is half that — about
180°. Of that, the central 140° are observed by both eyes.
134 A NATURAL HISTORY OF SEEING

When I stick my thumb out, at arm's length, its width covers about
two degrees of my visual field. My eyes can bring an area slightly
less than that into perfect focus.
Just one degree away from the very centre of my vision, I see
things only half as well. Five degrees off centre, my ‘visual acuity’ is
quartered. Beyond that five-degree radius, I can no longer be sure
of what I am seeing. At twenty degrees from the foveal centre, my
visual acuity falls below the common legal standard for blindness.
Our eyes bring little bits of the visual world to our attention, and
from these shards we build our world.
Even as I write this, knowing it to be true, I balk at its oddness.
I look up from the keyboard and I turn my head, just once. In that
glance, I see a world of vivid colour, a world of three dimensions,
of solid, textured forms. I see a world of objects: sofas and screens,
floorboards, a pair of sunglasses on the edge of my desk; a reading
lamp. Many of these objects are made of parts, and I see these parts,
and how they are fitted together: the lamp base, the switch, the bulb.
Outside my window there is a tree, its every leaf clear and distinct.
A car passes on the street beyond the tree, and I know, even from
this brief glance, what sort of car it was, what colour it was, and
even, roughly, how quickly it was moving.
I get out of my seat. The car is gone, but the lamp is here: I can
touch it. The bare floorboards are solid beneath my feet, as I saw
they would be. The wooden lattice-work screen that belonged to my
wife's mother casts speckles of light on the bare polished floor. It is
a bright day, and if I put my hand in front of the screen, the light
makes little pools of warmth on my skin. Everything is as I saw it.
Everything confirms the astonishing accuracy of my casual glance.
Impossible to imagine that I did not actually see any of this. Impos-
sible to imagine that I made it all up. If I made it all up, how could
I have got so much of it right? How can this illusion be so accurate
as to withstand the test of exploration? If I really can’t tell, from my
peripheral vision, that my sunglasses are perched precariously on a
corner of my desk, how come, when I turn my head to study them,
there they are, right where I thought they were? How come they are,
indeed, sunglasses, and not some other, quite unexpected object?
Is there, somewhere in my head, a model of the world; a mental
‘SEEING AND THINKING 135

model so accurate it includes my sunglasses? Maybe, without my

knowing it, every little degree-wide glimpse of the world contributes
to a vast mental model. Maybe that is what my model is - a composite
of little glimpses, memorised with perfect fidelity. The early pioneers
of robotics thought so, and built their robot eyes and visual systems
accordingly. But most of their robots never moved, and most of the
‘ones that did move, fell over. The others walked into walls. Slowly.
The computational demands required to hold and move about in a
pin-point accurate, three-dimensional model of visual space are not
just high (putting them beyond the capabilities of robots) they are
outrageously high, far exceeding the capacity of our brains.
Yet mice, with their little mouse brains, do not fall over the furni-
ture; birds do not fly into brick walls.
How do they manage it? How do we manage it?

It is 1997; the British artist, Emma Kay, is writing the Bible.

“The disciples were Matthew, Mark, Luke, John “The Baptist’, Peter
“The Rock’, Paul, Simon, James, and three others’
She is sitting at a big, white desk. Her manuscript, written in long-
hand, rests in front of her. It’s a slim volume: by the time she is
finished, it will only be about 7,000 words long. This is the Bible
according to Emma Kay: rather shorter than the original, and not
terribly accurate. ‘I had a vague recollection that Elijah was fed by
ravens, she explained to Rosie Millard for The Tastemakers, a book
describing the Brit Art scene in 2002. ‘T had absolutely no idea why
or how or where it fitted in, but it had to go in somehow.*
The following year, Emma Kay wrote the complete works of
Shakespeare; the year after that, she wrote a history of the world,
which has since been published under the title Worldview. It’s
80,000 words long - as long as the average novel — and it’s a
complete failure. Which is the point. Emma Kay doesn't use refer-
ence books. She relies on her own memory. (This is why her work
tells us far more about her — her childhood, her upbringing, her
interests, her personality — than it does about the Bible, or Shake-
speare, or the history of the world.) Emma Kay’s work asks the
question: how much do we really know about anything? Her work
is certainly amusing. It is also humbling. Spawn of cheerful atheists,
136 A NATURAL HISTORY OF SEEING

I'd never heard of Elijah’s ravens. What would my Bible look like?
How do we know what we know? Do we know something because
we remember it, down to every last detail? Or do we know it because
we can lay hands on it - because we know how to know it?
Why hold all the world in mind, when we can simply flick our
gaze from thing to thing? Why remember, in exquisite detail, which
objects are which, and where, and in what relation, when our eyes
can rake the real scene in moments, giving us, every third of a second,
a nugget of fresh information?
Did I know my sunglasses were perched precariously on the edge
of my desk? Even as I asked the question, my eyes were interrogating
that corner of space. They got there before me. Even as I asked the
question, my glasses were in view.
Perhaps the most elegant scientific demonstration of the eye's
‘search function was devised by Dana Ballard. Ballard is one of a
new generation of robot builders. Robots have come a long way from
their pioneering days. They no longer carry models of space around
in their tiny, painfully crammed ‘brains. Instead, they mimic as closely
as possible the behaviours of real animals. Their eyes move, inter-
rogating the world for the information they need.
Ballard wanted to know what information his robots needed to see,
how the information should be stored, and for how long. On the assump-
tion that 538 million years of evolution must be good for something,
he began by looking at how human eyes behave when engaged in simple
tasks. He asked a group of volunteers to copy a geometrical pattern,
using a set of coloured blocks. Looking closely, this is what he saw:
First, his volunteers glanced from the pattern in front of them to
the tray of loose blocks. Next, they picked up a block of a certain
colour. Then, they glanced at the pattern again. Finally, they fitted
the block they had selected into the position required to make a
copy of the pattern.
This seems just a bit complicated: if the eyes see a block of a
certain colour in a certain position, why cant both pieces of infor-
mation be retained in one go? Why does the eye make one set of
movements to choose a block of a particular colour, and another,
virtually identical set, to establish where it should go?
The answer seems to be that it is easier for the eye to move than
“SEEING AND THINKING 137
it is for the mind to remember. Why fill your head with transitory
and trivial information about coloured blocks, when the eye can
simply glance at them, again and again? The eyes are our search
engines, cheaper and faster than memory.

After a lifetime of pen-pushing for the local electricity board, quite

out of the blue, my father acquired an enthusiasm that would domi-
nate the rest of his life. One day, when he was at a loose end, he
volunteered to help fetch and carry for a local archaeological team
who were excavating the layout of some medieval gardens. A line of
Roman foundations and an Anglo-Saxon burial later, he was hooked.
And this is how Icame to spend the weekends of my early adolescence
traipsing across freshly ploughed farmland in the rain.
Field-walking is the least glamorous, most miserable activity in
archaeology. At a dead march, head bowed, eyes fixed to the mud
and grit beneath your feet, you plod, endlessly, for hours, watching,
waiting, for something, anything, to catch your eye. We walked the
chalklands of the South Downs, watching for Samian ware, for the
cores of worked flints; if we were lucky, a scraper or two.
Flint tools were a highly developed, very precisely worked technology.
The raw material was carefully selected and highly prized. Flints and
flint tools were traded across Europe. A flint scraper found in East
Sussex might have been struck from a core imported from France.
So here am I, in 1977, twelve years old, galumphing blindly over
the remains of this rich and subtle past, and I’m picking up anything.
Everything. Stones. Bits of brightly coloured earth. I haven't got a
clue. Flints fracture in the cold: morning frosts shave splinters off
dark, greasy, glassy flint and scatter them for nerdy twelve-year-olds
to cut their frozen fingers on. Dad shrugs. ‘Frost damage, he says,
handing the flake back to me. Stupidly, I toss it aside. As we work
back across the field, I will spot it again, and make the same mistake.
(It's the hope that kills you.)
Then, after a few weeks, something strange happens. Is that the
worked core of a flint? Yes. Is that a piece of pot rim? Yes. They are
starting to reveal themselves, emerging from the surrounding chaos.
They are making themselves known.
They are catching my eye.
138 A NATURAL HISTORY OF SEEING

Twenty-five years on, I am sitting in the office of Michael Land,

the man Richard Dawkins dubbed ‘the King Midas of eye research.
I can hardly write a book about eyesight and not meet the man who
discovered some of the more unusual eyes in nature; who unpicked
the workings of the lobster eye, notorious for apparently contradicting
the laws of optics; who was one of the first to see saccadic patterns
in the flight-paths of small airborne insects; and who, with Dan-Eric
Nilsson, wrote Animal Eyes, the bible of biological vision.
Quite how we have ended up talking about orchids, I’m not sure.
In addition to his own work - never mind the help he gives other
researchers, keeping up a lively correspondence with everybody who
is anybody in the field of vision research — Michael Land finds time
to explore English woodlands for rare orchids.
Like the pot-sherds of my childhood, Land’s orchids are hard for
a novice to spot. It is only over time that they begin to stand out,
and draw the eye.
What has happened in Michael Land’s head, that he can spot a
rare orchid at twenty paces?
Come to think of it, what has happened in your head, that you
are able to read this sentence?
When we read, our eyes fixate on between just twenty and seventy
per cent of the words. Who or what tells our eyes which words to
choose? Where does that knowledge reside? How is it made, where
is it stored, and how is it accessed?
Land, who has always found field-work more congenial than lab
work, is looking for the answers to these questions in everyday situations.
Typical of his hands-on, artisanal approach to his subject, his experimental
gear is home-made. ‘Where was I going to find thirty thousand quid
to buy one of these from the States?’ he asks, as he-displays - with no
little pride - an eye-tracking device of his own invention. This cost
£1,000 to make and colleagues from richer climes (Land works in a
cupboard-sized office in the University of Sussex) can never understand
where he’s hidden all the computerised bits. That's because there aren't
any. Land uses a tiny mirror to split the images recorded by a small
head-mounted videocamera. The movements of the wearer's eyeball
fill one half of the video image; their view of the world appears on the
other. Land and his colleagues correlate these images by hand.
“SEEING AND THINKING 139
These cottage-industry records are some of the most revealing
visual documents of eye behaviour ever produced, and allowed eye-
movement research to leave the laboratory and enter the real world.
The quotidian character of his findings has earned him headlines.
It was Land who discovered that, when a ball is bowled at a cricketer,
the cricketer takes his eye off the ball and looks instead at where the
ball is about to bounce. Similarly, when negotiating a turn, a driver’s
eye will provide motor information to her arms almost a second before
any muscular movement is required. It seems, from Land’s studies, that
half a second of visual information is held in a ‘buffer’ of some kind.
The eyes stay one step ahead of the body, dealing with the next view,
the next task, the next set of predictions and calculations, while the
body relies on the ‘buffer. This raises the odd but compelling idea that
the ‘present moment, as we experience it, has a measurable duration.
We operate in the world, not as it is, but as it existed half a second ago.
How can the eye stay half a second ahead of us? How does it
know where to go?
Throughout our lives, we learn to see things afresh, adapt to new
experiences and spot novel objects; be they pot-sherds, orchids, flying
cricket balls, or fast-approaching bends in the road. But it has always
been assumed that this ‘perceptual learning’ is an advanced business,
involving many areas of the brain. Land’s recordings show that, on
the contrary, this ‘learning’ must be an excessively simple affair, that
enables the eye to recognise and respond to salient features of the
scene long before they are ever brought (if indeed they are ever
brought) to our conscious attention.
When we say that ‘something catches our eye; we are being less
fanciful than we might imagine. The eye is not our servant. It is our
ambassador. In the tug-of-war between Self and World, the eye is
the red ribbon at the centre of the rope.
Michael Land fell in love with orchids and wanted to spot them
in their natural habitats, and this desire set in train a process of
learning. But it is the plants themselves — their colours, their appear-
ance in different lights, their textures, the opacity of their petals -
that have taught him to spot an orchid in an English wood.
Our eyes interrogate the world, and through our eyes, the world
announces itself.
140 A NATURAL HISTORY OF SEEING

3 — Reading each other

The eyes of men converse as much as their tongues.

Ralph Waldo Emerson

Last year, while visiting New York, I made time to visit the FAO
Schwarz toy store. I returned home with a box of interconnecting
plastic shapes called Zolo. As you fit these shapes together, you find
yourself making little creatures.
The shapes aren't anatomical. You don't say, as you tip the pieces
out: “That’s an arm, that’s a torso — look! an ear’ Yet after a few minutes’
play, and almost in spite of yourself, you will find in your hands a
cute, crazy little alien thing with an expression and a character all
its Own.
The reasoning is simple. Among the shapes are a handful of small
white balls, each with a discreet black dot. These lifeless things are
eyes — though they bear precious little resemblance to any natural
eye. White balls with black dots: somehow they look like eyes, even
before you attach them to other pieces of Zolo. Added to whatever
shape you happen to be building, they make the thing - however
random - come instantly to life. That curly swizzle-stick becomes
a tongue; that red ball, a nose.
Why does this happen? What gives this simple shape - a dot
within a circle - such power over our imaginations?

In a world where every animal is looking for something to eat, the

first object a sighted animal must learn to recognise is the eye itself.
Being looked at matters. If something's going to fight you, or eat you,
or try to mate with you, it's going to be looking at you.
Stare at a hog-nosed snake from a distance of about a metre, and
it will play dead. Turn away, and it will move. This behaviour — tonic
immobility - seems rather futile to us, with our object-centred,
stereoscopic gaze. But since most animals watch for movement and
care little for stationary objects, the strategy works well in most cases.
ee

“SEEING AND THINKING 141

Lizards, chickens, blue crabs and ducks all exhibit this behaviour,
and its traces are still strong in us - as anyone whos ever played
‘grandmother's footsteps’ as a child can testify.
Every eye is on the watch for other eyes. Whether it wants to be
or not, the eye is never neutral. It does not simply observe: it commu-
nicates.
Long before the cave artists of the Ardéche, in southern France,
turned their cave walls into galleries of bison, horse and mammoth,
the process of natural selection had thrown up the planet's first
artworks. The eyes adorning a peacocks tail, the eyes on a hawk
moth caterpillar, the eyespots on the tail fin of the four-eyed butterfly-
fish Chaetodon capistratus — these are, in every sense, pictures, works
of trompe [ceil, there to confound and impress the onlooker.
Few animals risk making a feature of their real eyes. The banded
butterflyfish Chaetodon striatus goes so far as to hide them within
a black bandanna. The ‘whites’ of most vertebrate eyes are not white
at all; many develop an all-over dark pigment that effectively conceals
where, precisely, they are directing their gaze. Only an extremely
social animal would make its eyes more noticeable than they had to
be. Humans are exceptional in having bright ‘whites’ to their eyes
precisely so that an individual can tell where its fellows are looking.
In a world where every eye is primed to spot every other eye, this
is a risky adaptation. What benefit can it possibly bestow?
Answering this apparently simple question involves a story so
convoluted, I can only beg the reader’s patience.

To begin at the beginning: 1957 was a bad year to be a rat. In laboratory

after laboratory, experimental rats were collapsing in convulsions, and
nobody could work out why. The problem attracted the particular
attention of the zoologist Michael Chance, one of a team of bright
young researchers brought to Birmingham University by Joel Elkes,
generally regarded as the father of experimental psychiatry.
Contrary to the clichés of the animal liberation lobby, lab assistants
tend to be sentimental creatures and grow very attached to the
animals in their care; certainly the rats Chance examined were healthy
and well tended. And yet the instant an assistant unwittingly jangled
his or her keys — over they went, in spectacular spasms. Something
142 A NATURAL HISTORY OF SEEING

was making otherwise healthy rats behave as ifthey were convulsing.

Realising this, Chance very quickly spotted what was wrong. In the
wild, a startled rat runs away and hides. Even a pet rat generally has
a sleeping box, where it can conceal itself. No one had thought it
necessary to give laboratory rats sleeping, boxes.
The moment sleeping boxes were installed in the cages, the rats
stopped convulsing. When Chance jangled his keys, they did what
rats normally do — they ran away and hid.‘
At the time, little was known about the emotional lives of animals.
The dominant psychological doctrine of the time, behaviourism,
regarded the ‘inner states’ of animals and people as having little scien-
tific value. They could only be by-products of reflex behaviours,
fortuitously generated phenomena which played no useful role.
But if rats were simple switching systems, swapping behaviours
the way a toy robot changes directions in a maze, why were stressed
rats such lousy subjects? Chance showed that animals kept in condi-
tions closer to their natural environment produced better experi-
mental results, while animals stressed beyond endurance not only
behaved abnormally; they lived abnormally, slept, ate and moved
abnormally, and any measure of their condition during an experiment
was more likely to be a product of stress than anything to do with
the experiment.
For Chance, the study had two consequences. It drew him into
pioneering work to improve the living conditions of animals — espe-
cially primates - kept for laboratory study, and it led him to think
more about fear.
Every animal, faced with a deadly threat, resorts to some sort of
random behaviour. Many animals rely almost entirely on unpre-
dictability: rather than head straight for cover, a rabbit pursued by
a fox will bob and weave in a chaotic zigzag. Confused animals make
poor predators, so unpredictability is a very good defence, cheap to
develop and easy to deploy. (When they convulse, rats become diffi-
cult to catch and hold. A cat that has not been taught by its mother
how to kill its prey will try to hold a mouse in its jaws; the mouse
convulses and almost always gets away.)
The puzzle comes when we look at our own unpredictable behav-
iours. And we have them, in abundance. We lash out. We react in
SEEING AND THINKING 143
unpredictable ways. We fool, and we cheat, and we lie. The odd thing
is, we behave like this all the time. We are the moodiest animal on
the planet, deploying, with great finesse and subtlety, every waking
moment, behaviours that in other animals are crude, last-ditch, and
above all, short-lived bursts of anomalous behaviour.
- Except in primate societies, conflicts never persist in nature, and
squabbles between animals of the same species rarely become serious
enough to trigger truly unpredictable behaviours.
Chance, together with another Joel Elkes protégé, Allan Mead, was
the first to realise that humans are different: they regularly use unpre-
dictability. In the light of Chance’s study of lab rats, this observation
had daunting implications. It meant that primates had acclimatised
to, and learned to master, a condition of perpetual terror. This was
a kind of evolutionary pressure that no one had ever considered
before. Was terror a driver of human evolution?

We tell our children, in school and in our books and science

programmes, that human intelligence arose because we had to develop
tools to survive, and language arose because we had to communicate
and co-operate with each other. No doubt these things are true -
but are they enough to explain human intelligence? Many primates
use tools, and communicate well. The more we learn about animal
behaviour in the wild, the more instances of language and tool use
we find. Why did human language and tool use become so different,
in kind and scale, from the language and tool use of other animals?
A veteran of forty years of lemur research, Alison Jolly first visited
the island of Madagascar in 1963, after studying at Cornell and Yale.
Ask her why she chose the lemur for her special study, and her
answer is simple: ‘It’s cute and it lives a long, long way away from
Yale?
Back in 1966, the lemurs presented something of a puzzle. They
lacked monkey-like intelligence - they were as daft as brushes - and
yet they lived in big, monkey-sized communities. It had been thought
that a rise in intelligence would lead to group living. The example
of the lemurs suggested that things were the other way about: that
group living could arise even among dullards; only later would it
tend to select for intelligence.
144 A NATURAL HISTORY OF SEEING

Alison Jolly’s insight made perfect sense to Michael Chance. If

social living drove primates to acquire intelligence, then intelligence
was a response to a single, over-riding aspect of social living: the
terror of it.
In a primate community, everyone has to behave correctly towards
everyone else. Age, sex, temperament and position in the hierarchy
all affect how each individual behaves, and how others must behave
back. Some measure of the true horror of this kind of life - which
seems, after fifty million years of evolution, perfectly pleasant, even
desirable to us - can be glimpsed in the wild. Among apes, baboons,
or chimpanzees, most males are hunched and nervous, radiating
insecurity. When a subordinate male Macaque monkey approaches
a dominant male, it takes along a baby to protect itself from an
unprovoked attack. If that sort of behaviour seems merely ‘clever’;
imagine your reaction if a human were to do the same thing. The
manipulative behaviours that gave us manners, ethics, co-operation
and the rest are the same behaviours that enable hostage-taking,
kidnap and the use of human shields. The bad and the good are
inseparable: they are all, at bottom, behaviours designed to keep one’s
head above water in society. We are, to use an outdated term, ‘tool-
maker man’ — but the first tools we learned-to wield were each other.
Because they are unable to anticipate, moment to moment, the
behaviours of their fellows, primates watch each other all the time.
A primate who wants to get ahead in the troop doesn't necessarily
have to be the strongest, the biggest, or even the most well liked; it
simply needs to attract attention. The primatologist, Jane Goodall,
Alison Jolly’s contemporary, came across one male among her chim-
panzee troop who rose to social prominence by pure self-promotion,
by rolling empty oil drums down a hill while whooping and drum-
ming.
A primate in authority is, by definition, a primate that commands
visual attention. A young female mountain gorilla studied by another
primatologist, Diane Fossey, stared for hours every day into her troop
leader's face. If he tossed a glance her way, she shuddered with plea-
sure. Among the higher primates, the way unpredictable behaviour
is deployed acquires a Machiavellian, ‘treat em mean and keep ‘em
keen’ sophistication. The dominant animal’s ‘subjects’ look to their
SEEING AND THINKING 145

leader for cues. Moment to moment, they do not know whether they
are to be punished or rewarded. They do not know whether to sit
or stand. They do not know what their master feels, so — simulating
what they hope is their master’s mood - they watch attentively for
the smallest change of expression.
We instinctively know that a look can wound, and a glance can
save. Now we think we know why: it is a power that has been amassed
by the eye over the fifty million years we have spent in Hell; if, like
Jean-Paul Sartre, you incline to the view that Hell is other people.

In a letter to the magazine New Scientist in August 1991, a reader,

Harry Miller, recalled taking a party of blind children to London Zoo:

I had the idea that I might allow the children to feel and cuddle
the baby chimps, learning about their hair, hands, toes and so
on, by touch. The experiment, however, proved to be a disaster.
As soon as the tiny chimps saw the blind children they stared
at their eyes - or where their eyes should have been - and
immediately went into typical chimpanzee attack postures, their
hair standing upright all over their bodies, their huge mobile
lips pouting and grimacing, while they jumped up and down
on all fours uttering screams and barks that rose in crescendo
... 1 ushered out the children mumbling whatever excuses I
could think of.°

The very casualness of our language is testament to the social impor-

tance of the eye. ‘His glance just melted me, we say. “They looked
daggers at me. ‘She shot me one of those looks: When Miller’s blind
children failed to play by their visual rules, the chimpanzees of
London Zoo expressed their affront in ways both shocking and
violent.
Of the thirty-plus regions of the primate brain dedicated to vision,
several specialise in spotting social signals. In these regions, there
are brain cells primed to spot faces. Some of these cells respond
most strongly when the head is pointed in a certain direction; others
grow excited when someone looks askance, so that the eyes are
pointing in a direction different to the head. This battery of gaze
146 A NATURAL HISTORY OF SEEING

detectors may have evolved in response to the way the faces of great
apes have flattened over time..As binocular vision became more
important and a good sense of smell became superfluous, primates
lost their muzzles. Watching where someone’s muzzle is pointing is
a simple business: tracking a relatively flat face requires more sophis-
tication. Adopting an upright posture would have made gaze detec-
tion even more problematic for humans, further loosening the
relationship between gaze direction and posture.
A quick glance back at Alfred Yarbus’s record of how the eye
studies a portrait - on page 18 - reveals how well our faces are
designed to reveal our emotions. The eye in Yarbus’s study had difh-
culty investigating evenly illuminated areas, and could only leap from
one area of high contrast to another. This meant that the eye was
drawn to those very areas of the face which expressed the widest
range of emotions: the lips, the eyes, and the eyebrows.
Armed with gaze detectors, and displaying strong visual markings
— eyebrows, high cheekbones and bright white sclera - humans are
ideally equipped to develop a further, even more sophisticated
survival tool, that takes social thinking to a new and exciting level.
This is empathy: the knowledge that others operate in much the
same way that you do. If you are looking, they are looking. If you
can see the ball, they can see the ball. If you show an interest in the
banana, they, too, will be drawn to the banana.
Two-month-old human babies pay more attention to the eyes than
to any other part of their mother’s face. Six-month-old children spend
two to three times as long looking at faces that are looking at them.
From about eighteen months of age, human children, engaged in
some activity or other, begin checking on the level of attention they
are arousing in others. A glance which moves from an object to a
parent's face and back to the object asks a question: ‘Are you seeing
this?’ It also prompts a response, as the adult’s attention is drawn by
the child’s glance. Thus reinforced, checking behaviour rapidly
develops into pointing behaviour. The question, ‘Are you seeing this?”
becomes a request, ‘Look at this!’ The way in which we recruit each
other's attention is central to the way we learn language as infants.
‘Banana!’ I might announce, pointing to a banana, but if you don't
know what the pointing is for, you'll never know what a ‘banana is.
“SEEING AND THINKING 147

Although we are undisputed masters of this behaviour, it is by

no means unique to humans. Baboons use quick, furtive glances to
recruit helpers when they're being challenged by other baboons,
and regularly fool each other by pretending to spot non-existent
threats. Whether any primates other than humans really appreciate
that their fellows have internal lives is open to question, because
no one has quite worked out an experiment that would reveal the
answer, One way or another.
It is hard to say when human children acquire this knowledge.
Simon Baron-Cohen of Cambridge University believes this under-
standing is not innate, but develops over time, prompted and rein-
forced by the pattern of back-and-forth glances that pass between
the child and others. ‘I think that long before kids are interested in
trying to manipulate your attention, or follow your direction of atten-
tion, they are fascinated by the eyes; he says. ‘A lot of their behaviour
may not be mentalistic:”
The experience of people with autism seems to support this.
Autistic people understand ‘the facts of vision. They know how to
hide something from another person's view. They understand the
rules of geometry; that an object can be visible to one person while
being hidden from another. But in spite of this, they have difficulty
equating gaze-direction with seeing. They are unable to model
another person's internal life; they cannot therefore know that another
person actually experiences what he or she sees. A baby of twelve
to sixteen months knows to check that its mother is looking at it
before it points. Autistic children fail to check, if indeed they point
at all, and some, seeing another person point, will miss the meaning
of the gesture altogether and focus on the tip of the extended finger.
From the behaviour of autistic people, we learn something impor-
tant, and heart-rending: being human is a skill we teach each other;
and we teach it, first of all, through our eyes.

We do not merely look at faces. We read them. An adult primate’s

facial expressions will guide the behaviour of its offspring, especially
if the offspring is nervous or uncertain. (Next time you see a small
child freeze at the top of a playground slide, see if you can spot its
anxious-looking parent.)
148 A NATURAL HISTORY OF SEEING

Here is an inverted picture of the former British prime minister

Margaret Thatcher. Even though the picture is inverted, it's clear
enough that she is smiling. Now turn the book around: ghastly, isn't
it? Thatcher’s mouth is stretched into a zombie grimace, and her eyes
are the eyes of a corpse.
We have Peter Thompson of York University to thank for this
chilling illusion. In 1980, Thompson manipulated this photograph
by cutting out her mouth and eyes and pasting them back on the
picture the wrong way round.* You didn't take much notice of that
when you looked at the inverted face, because the meaning was clear
enough: you saw a smiling mouth, the tell-tale creases at the corners
of the eyes, and were content. When you turned the book around,
the head was the right way up, but now the former prime minister's
expression was inverted. With its meaning pi es away, the human
face is monstrous.
Paul Ekman, a consultant to both the Dalai Lama and parts of
the US Defense Department, is an acknowledged expert in reading
faces. He is very careful about who he works with. Gorbachev wanted
him to train his secret service; he declined. A US president wanted
to learn how better to mislead his adversaries in summit meetings,
and received the same answer.‘Ido not teach people how to be better
liars, Ekman says. The increasing technological anonymity of human
society is doing a fine job of that already, without his help.
Wraparound vision: the compound eye ofthe Antarctic krill
Euphausia superba.

Smothered in tiny crystal eyelets, a brittlestar’s carapace doubles

as a primitive compound eye.
The compound eyes of shrimps, prawns, lobsters and crayfish have mirrored boxes in place of
lenses. Left: the eye of adecapod shrimp. Right: close-up ofthe surface of a crayfish eye.

Sigmund Exner's
photograph — the first
experiment of its kind —
is focused, not through
a lens, but through the
optics of the superpositio
compound eye of a firefly.
- om Es ee
The mantis shrimp crams up to sixteen different colour receptors into the horizontal
bands running across its eyes. ce

David Brewster's graph ofthe intensity of light at different wavelengths.

Unwittingly, he has revealed the relative sensitivities of human cones.
Blackeyed susans (Rudbeckia hirta) photographed in visible light (left) and in ultraviolet (right).
Pollinating insects are often drawn to patterns our eyes are not equipped to see.

The camouflaged eye: the banded butterflyfish Chaetodon striatus.

Round pupils like ours make optical sense, but other shapes abound, especially among fish and
squid. Camouflage is part of the reason, but there may be undiscovered visual advantages to
each shape. Clockwise from top left: blue-spotted ray; epaulet shark; harbour cuttlefish; pale
gecko; yellow flounder; spiny-tailed gecko.
A trilobite with a sunshade: Erbenochile, shielded from the glare of its sunlit habitat, enjoyed
wraparound vision but, like other trilobites, it could not look up. It did not need to;
there were no predators around to swoop down on it.
Vision through touch.
Above: a star-nosed
mole’s nose boasts a
central ‘acute’ region, like
a fovea. Right: a naked
mole rat acquires spatial
information from its
sensitive teeth.
The nerve cells of the retina, drawn by Santiago Ramon y Cajal. Cajal’s description
of anervous system made up of distinct, individual cells of different types won him
a Nobel prize in 1906.
A section through the visual cortex stained by the Golgi method fixes entire
nerve cells, but only a tiny fraction ofthe cells so treated respond to the
process. The result is a beautiful, radically simplified image of how individual
cells connect to form a single nervous system.

The musculature of the eyeball according to the nineteenth-century

American physician David Hosack.
Views through an ophthalmoscope. Top: the right eye of a
brunette Italian woman, 36 years of age. The fovea is a barely
visible yellowish disc, free of blood vessels, to the left of the
nerve head. Bottom: the left eye of a grey cat. These exquisite
paintings by Annette Burgess, published in 1934, capture ana-
tomical detail better than any photograph, and even record the
play of light as it glances off the cat's tapetum.
The loneliest tissue in the human body: the lens
of an eye, suspended and connected to its ciliary A biological sundial: the pineal organ of the
muscle by thousands oftiny elastic ‘guy wires’. desert spiny lizard.
The ectopic eyes on the legs of this fruit fly were formed by
artificially activating the fly's eyeless gene.

The compound eyes on the antennae ofthis fruit fly were triggered
by the gene smalleye, from a mouse.
John Dalton (1766-1844) Thomas Young (1773 —1829)
portrait by Benjamin R. Faulkner. portrait by Henry P. Briggs.

Sir Charles Wheatstone (1802-1875) Santiago Ramon y Cajal (left) playing chess in
portrait by Charles Martin. 1898; a photograph taken by one ofhis children.
a b é

Three contrast effects (the squares in the centre of each pair are identical). Whether detecting
brightness (a), hue (b) or intensity (c), the eye responds to contrasts, not absolute values.

Edwin Land prepares an

experimental ‘mondrian’
to test the robustness of
colour vision.
Because it exaggerates the boundaries between Ernst Mach’s even grey bands, the eye
zenerates an illusion of fluted columns (a). The illusion is almost as strong using bars of
differing saturation (b) and hue (c).

Look at it long enough, and this

complementary-coloured Union
Jack produces a true-colour
afterimage. See page 226 for an
explanation.
i al

Adelbert Ames's ‘full-size monocular distorted room’. An exterior shot reveals

the architectural trickery behind the illusion.
“SEEING AND THINKING 149
Ekman was brought up in Newark, New Jersey; at school, he was
a year ahead of the novelist Philip Roth. Thrown out of high school
for talking back to his teachers, the fifteen-year-old Ekman entered
the University of Chicago. There, he discovered Freud. ‘I could quote
Freud on any topic you might mention, he recalled in interview.
‘It must have been pretty obnoxious’ Convinced that psychotherapy
was a cure for the world’s ills - a popular view in the fifties - Ekman
sat in on psychotherapy sessions. ‘I was impressed that there was
more than the words going on. So I thought for my research ... I
would look at what people did with their face and body. Little did
I realize that that would then occupy my whole life, and I'd never
get back to psychotherapy:
At the beginning, Ekman steered clear of the face. It was too
complicated, and no one quite knew how to deal with it. No one,
that is, except for Silvan Tomkins, a philosopher more than a psychol-
ogist, mistrusted by his colleagues because he was a theorist at a
time when, in Ekman’s words, ‘psychology didn't like theoretical
people’? Brought together by the editor of an academic journal,
Tomkins and Ekman discussed the central problems of non-verbal
communication. Ekman had been given a grant to find out what was
universal and what was culture-specific about expression and gesture.
Everyone seemed to have an opinion, but nobody had managed to
amass any evidence. The celebrated anthropologist Margaret Mead
believed that facial expression was culture-specific. Others, including
Silvan Tomkins, disagreed; they believed facial expressions were
universal and innate, and they fielded an especially heavy hitter:
Charles Darwin.
In his speculative book, The Expressions of Emotions in Men and
Animals, published in 1872, Darwin wrote: ‘It has often struck me
as a curious fact that so many shades of expression are instantly
recognised without any conscious process of analysis on our part.
No one, I believe, can clearly describe a sullen or a sly expression;
yet many observers are unanimous that these expressions can be
recognised in the various races of man-
Darwin had plenty of circumstantial evidence. He had, for example,
noticed how his six-month-old son would respond to his nurse's
show of mock sadness by turning down the corners of his mouth.
150 A NATURAL HISTORY OF SEEING

Far too young to have acquired the concept of sorrow, the baby
already knew how to express it. (Subsequent studies have shown that
babies of four to seven months old can already tell the difference
between ‘happy; ‘sad; and ‘surprised.) Darwin, then in his early sixties,
was plagued with illness, and unable to travel to confirm his intuition.
Instead, he relied on far-flung correspondents, including John Scott,
curator of the Botanic Gardens in Calcutta, a remarkable observer
who came closest to confirming Darwin's conjecture.
Ekman was no anthropologist; to confirm whether or not Tomkins
and Darwin were right, Ekman phoned a neurologist, Carleton
Gajdusek, who had been in New Guinea studying a disease called
kuru, a human form of bovine spongiform encephalopathy (BSE),
notorious for being spread by the consumption of human flesh.
Because kuru had such a long incubation period — ten years or so
- Gajdusek had to invent novel ways of monitoring the disease’s
progress. The solution he came up with was film - nineteen miles
of it, a medical record astounding enough to win him the Nobel
prize, and which captured, along the way, the day-to-day lives of two
isolated Stone Age peoples; two contrasting ways of life that, within
a couple of years, would be gone for ever. Ekman is, to this day, the
only person ever to watch the lot. (Not even Gajdusek could bring
himself to watch it all.) It took him a year: and in all that time,
Ekman never saw an expression he hadn't seen before. There was
nothing new.
Convinced that Tomkins and Darwin were right - that facial
expressions were universal and not merely cultural products - Ekman
then had to produce the evidence. The films were not enough. He
was going to have to go out and study these people himself. It was
not a prospect he altogether savoured. Although one of the tribes
was a peace-loving bunch, the other was very violent. When Tomkins
came to see one of Ekman’s movie shows, he could tell immediately
which was the violent tribe, just by the looks on people's faces.
Ekman took along his friend and long-term collaborator Wallace
Friesen, and together they set out into the south-eastern Highlands
of New Guinea. They met the peace-loving Fore people - and stayed
put. “The violent one was too violent, and I got too scared to go in
there; Ekman admits.
“SEEING AND THINKING 151
With them, Ekman and Friesen brought stories and photographs.
They sat with amenable groups of Fore tribespeople and told them
stories. Then they showed them the pictures. The pictures were of
people the Fore had never seen, Berkeley students, expressing a range
of feelings: happiness, sadness, shock, fear, envy — the full gamut of
human emotion (or at least, as much of it as the researchers could
bear to carry with them; humans are capable of making around 7,000
emotional expressions).
Ekman and Friesen wanted to see if the Fore would be able to
recognise ‘foreign’ displays of emotion. They could, and did -
although cultural differences did make some small differences in
how emotions were interpreted. For example, the Fore had a hard
time distinguishing surprise from fear - presumably because their
culture was not particularly geared to delivering as many pleasant
surprises as Ekman’s was. None the less, Darwin’s conjecture was
conclusively proven. Every human is equipped with a 7,000-‘word’
emotional vocabulary, and a syntax that offers multiple means of
expressing the same emotion. Trivial local differences turn up now
and again, but the differences are as nothing, when compared to
the similarities.

In 1964, the Finnish psychologist Tapio Nummenmaa published

The Language of the Face, in which he asked people to say, from cut-
out pictures of eyes and mouths, what emotions were being expressed.
He found that simple emotions, like sadness or happiness, could be
read from the mouth alone, but complex emotions (for example,
surprise and frustration) required additional cues from the eyes.
The human eye is built to be noticed, and gaze direction can itself
be used to convey emotional meaning. The lateral rectus eye muscle
is labelled ‘amatoris in early anatomies, because lovers use it to direct
their flirtatious glances. A downward gaze can indicate sadness;
looking down and away suggests shame and guilt; looking away is
a reliable sign of frustration or disgust. Simply widening the eye can,
in conjunction with other facial movements, express everything from
shock to arousal to doubt.
The expressions we read on each other's faces are of two basic
types. First, there are the gestural cues which we use, half-consciously,
152 A NATURAL HISTORY OF SEEING

to clarify our spoken conversations. These movements of the head

and brows, jaw and lips, don’t necessarily express how we feel; they
simply emphasise what we say. They are a form of punctuation. The
second type of expression conveys emotion — and how. Throughout
our lives, expressions of emotion instantly reflect our ever-changing
moods, automatically, and truthfully.
The problem is, it’s hard to separate the emphasis system from the
emotive system. This happens whenever we watch Woody Allen.
Among the director's many gifts is a talent for raising the inner corners
of his brows when he talks. He does this to stress his speech, the way
you and I would frown or nod. I cannot do what he does, and chances
are you cant either. For most of us, the muscle responsible for this
inward movement of the brows is not under our conscious control.
It occurs only when we are exhibiting genuine sadness.
Using a ‘sadness’ cue to punctuate his conversation gives Woody
Allen either an endearing vulnerability, or a saccharine earnestness,
depending on your point of view. Either way, Allen cannot be held
responsible: he is simply, like the rest of us, using the tools he has
at hand to communicate.
Because the face is capable of expressing two things at once -
how we feel, and what we mean - we should perhaps be reconciled
to the false smiles, manufactured looks of interest, and all the other
deceitful gestures of which our eyes and faces are capable. Deceit is
often a necessary part of communication; we conceal what we feel
to make room for what we mean. The false smile is particularly
handy, since we use it most when we are genuinely happy. From an
evolutionary standpoint, signalling enjoyment has never been very
important; it takes a relatively large amount of sheer joy to make us
smile involuntarily.
Our eyes reveal our inner state, whether we want them to or not.
It is impossible to manipulate our rate of blinking for any length of
time, and the way our irises dilate is quite outside our control. This
is extremely handy when we're looking for arousal in a potential
mate: when aroused, the eyes blink more often, and the iris dilates.
But though the eye tells the truth, even a human eye, painted like a
target, is hard to read at any distance. Lovers may look longingly
into each other's eyes, but generally we stand too far away to judge
SEEING AND THINKING 153
the fine responses of each other’s pupils — and who in their right
mind would try to watch the rate of someone's blinking? With prac-
tice, we might be able to spot these clues, but few of us ever bother
because we are far too busy with that other great, and greatly unre-
liable, form of human communication: language. Encoding words
in a coherent fashion and decoding them again is a tough task; it
takes most of our attention, and we tend to miss much of the non-
verbal information our faces express.
It is perhaps not surprising that we have therefore evolved a sort
of truce flag - a unique, highly visible method of telling the absolute
God-given truth in a way that gets us believed: we cry. Crying is
very difficult to fake. Even actors have to generate some feeling before
they cry. Reflecting on this, Marc Hauser, an evolutionary psychologist
and a professor at Harvard University, was put in mind of an idea
that the Israeli evolutionary biologist Amotz Zahavi proposed many
years ago, that you can infer the honesty of a social signal by
measuring the cost of the expression. Hauser, applying this principle
to the eye, regards tears as the human equivalent of a dog rolling
belly-up to show submission. “Unlike all of the other emotional
expressions, tearing as an emotional expression is the only one that
leaves a long-term physical trace, he says. ‘It blurs one’s vision, there-
fore it’s costly:*°
Humans have been pack animals for most of their history. Their
survival depended upon co-operation. In such circumstances, there
is plenty of opportunity for petty deceit, but very little advantage to
blatant lying. In a tight-knit community, a truly pathological liar
stands to lose far too much if he or she gets caught out.
Yes, the eyes are the windows on the soul, but we don't count each
other's blinks, and we don't press our faces close, to watch each other's
pupils wax and wane. We don't have to. Liars get found out quite
quickly without all that. “Trusting others is not only required, but it
makes life easier to live; says Paul Ekman. ‘It is only the paranoid
who forgoes such peace of mind’
SIX

Theories of vision

1 — Theon’s arrow

At every stage, we will have to reimagine the universe .. .

Arthur Zajonc*

Up here in the study, the wind stirs the leaves in the tree outside my
window. I am writing this in long-hand and my neglected computer
screen dances to the weird, aquatic rhythm of my favourite screen saver.
These things do not distract me, nor does the sight of my pen, inching
its way across the paper. Nor do the one hundred and one other objects
in my field of view. My eyes are much better at filtering out distractions
than my ears are. (Try as I might, I cannot listen to music while I work.)
In comparison to sight, hearing, touch, taste and smell are relatively
passive senses. They are not nearly so capable of focusing on one partic-
ular input, filtering out the junk and white noise of daily living.
This intuition seems confirmed and redoubled when you consider
that we see far more than we hear, smell, or touch. We are deluged
by the visual. The other senses have it easy: we only hear what makes
THEORIES OF VISION 155
a sound, we only smell what gives off a scent; we only touch what
makes physical contact.
Perhaps that is why our vision is so good at concentrating on the
thing at hand; at filtering out the noise. The eye cannot afford passively
to drink in the scene before it. We would drown in the chaos. It must
select. It must hunt.
Theon of Alexandria, considering the shape and movements of
the eye, saw clearly the difference there must be between vision and
the other senses. The ear, for example, is shaped to receive the air
which carries sounds. Clearly, ears are collecting devices. Eyes, on
the other hand, are spherical, mobile; they are not passive gatherers,
but hunters. Theon is writing in the fourth century AD in the great
Library of Alexandria, a not-especially-gifted mathematician
preparing simplified lecture notes for his considerably less gifted
students. The heyday of the library is long past, and his death, in
405, will save him from the knowledge of the brutal assassination
of his daughter, Hypatia, fated to be Alexandria's last head librarian.
Theon’s argument is that form follows function. When we look at
the eye in action, it is clear that the eye is not merely ‘receiving airs,
like the ear or the nose. It is hunting. Though we cannot see them,
therefore, the eye must see by emitting a ‘dart’ of some sort or another.
Theon is no eccentric. His arguments are bolstering an idea of vision
which has pertained for over eight hundred years, and will continue
to shape thoughts about vision for at least another millennium.
Why should the eye emit ‘darts’? Why would anyone cook up such
a bizarre fiction, simply to explain vision? Isn't it obvious that light
enters the eye?
It isn't. The idea that the eye sees by capturing light is so counter-
intuitive that it was only accepted - and only in the broadest detail
— a mere five hundred years ago.

Arthur Zajonc, professor of physics at Amherst College, Massachu-

setts and the author of a handsome history of light, Catching the Light,
has invented an apparatus specifically to demonstrate that the subject
of his life’s work isn’t actually there. He calls it, with nice irony, his
light box.
It is, as the name implies, a box, filled with light, except that Zajonc
156 A NATURAL HISTORY OF SEEING

has been careful to ensure none of it actually reflects off the sides
of the box. Standing beside it, you can see the lamp burning merrily
away, and the hole where the light enters the box, and then, peering
into the box through a peephole, you see — nothing.
Nothing. An absolute blackness.
You can push a wand through the side of the box, and in doing
this you can see — the wand. It is brilliantly lit, but it's only the wand
you are seeing. Not the light. The box is full of light, and yet you
cannot see it.
We see objects. We see bright objects, and dark objects. The sun
is very bright. A piece of coal in the shade is dark. In the sunlight,
the coal glitters; it even shines. These observations do not of them-
selves imply any connection between vision and light. Light is a
property of things. We see bright things, not brightness.
For the early Greek philosophers of the atomist school, vision was
the ability of the eye to see objects. Visual impressions do not arrive
muddled up in a sensory soup, as sounds and smells do. Entering a
room, I see a table, a chair, a bed. I do not need to sort these objects
out. Even though the chair is behind the table, I perceive, from the little
I can see of it, that the chair is a chair. Neither do I struggle to turn
my attention from one object to another. Leucippus, and the atomist
tradition he founded — a tradition that would last, with many revisions
and controversies, for five hundred years — based his theory on what
is still the most important phenomenon of vision: its relationship to
attention. We see objects, not splodges, or waves, or impressions.
Armed with nothing but an appreciation of matter and a notion
of how things are arranged in space, ignorant of optics, ignorant of
anatomy, operating mostly without mathematics, the atomists
promulgated a daring explanation of vision: that objects release thin
films, the way a snake sheds its skin, and these films (or eidola),
enter the eye and are apprehended directly by the mind.
Calling upon nothing more than idea of matter in motion, the theory
of eidola explains the key phenomenon of vision - our ability to recog-
nise objects even as we see them, whatever the angle, without any effort
of interpretation. While we might hear a dog bark, smell a dog’s presence,
or feel a dog’s muzzle against our hand, and by those tokens infer the
presence of a dog, only when we see the dog do we apprehend it,
THEORIES OF VISION 157

directly, without inference. The dog’s eidola stream from the dog
into our eyes, so that the dog is, quite literally, making its presence felt.
The theory works on several levels. It is, at its simplest, a physical
explanation, with implications for the structure and arrangement of
matter. For the atomists, all matter was an arrangement of particles.
Depending on the density of their packing, these particles are more
or less free to move. In the era before optics, mathematics or any
theory of energy or radiation, the movement of one kind of material
through another — the passage of eidola through the air, or through
the jelly of the eye - is a significant idea.
On another level, the theory of eidola has considerable psycholog-
ical appeal, explaining at one fell swoop the whole mysterious busi-
ness of visual apprehension. That chair in the corner - it has no physical
effect that I am aware of. It does not squeak, smell, or brush up against
me. It is mute, separate from me. How do I know it exists? How do I
know - even more remarkably, how do I know at this distance - that
it is a chair? Because its eidolon is swallowed by my eye.
The theory of eidola hypothesises that vision is, after all, the conse-
quence of an actual physical contact. Everything, in atomist philos-
ophy, is an object. To exist is to exist as an object, and it is in the
nature of every object to throw off images of itself. We recognise
things, not by any act of memory or comparison or analogy, but
because the things themselves - or at least their atom-thin shells —
integrate themselves physically into consciousness.
But how could the atom-thin film of a mountain be expected to
fit through the tiny aperture of an eye? Come to think of it, how
could (say) ten thousand such films, all made of the same stuff, pass
without interference, all at one moment, from one mountain into
the eyes of ten thousand observers? Even if this were possible - why
do objects not run out of mass from which to generate their eidola?
Was the universe shrinking? Did objects grow, like snakes, to make
up the mass lost in throwing off image films? And so on — a cata-
logue of awkward questions. |
A purely material theory of sight violates what we know about
how materials behave; the atomists, in trying to get around this
problem, were doomed to invent ever more arcane and unlikely theo-
ries of matter. Throwing matter into confusion, simply to explain
158 A NATURAL HISTORY OF SEEING

vision, is just not reasonable. Any competing theory of vision would,

as its first requirement, have to treat the rest of the material universe
rather more gently than did the atomists.
This theory would, ideally, address the telling handful of cases
where vision and apprehension do not arrive hand in hand. Consider,
for example, the sorry case of the man who drops a needle on his
rug. He knows the needle is lying on his rug. He is looking at his
rug. But he cannot see the needle. Why not? According to the theory
of eidola, image-films of the needle should be streaming into his
eye. What prevents them?
While the theory explains wonderfully well why it is that we see objects,
it conspicuously fails to explain how we are able to turn our attention
from one object to another; how we can focus on one object to the
exclusion of another, and how, in the case of the needle lying on the rug,
we can miss an object that is unquestionably within our field of view.

Ironically, the body of theory that best addressed these problems

was neither innovative, nor handsomely rational; nor did it seem
from a known author. As far as we can tell, this competing line of
thought is a throwback to philosophy’s shamanic past, pre-dating
atomist theories by several hundreds of years, marking almost a
return to the beautiful, intransigent dogmas of Ra the sun god.
For the Nile-dwelling worshippers of Ra, living in the New
Kingdom of Egypt around 1500 Bc, light was Ra’s act of witness.
Light was the god’s sight, and things existed because Ra saw them.
Ra sees — and all is illumined. What happens if we humanise this
dogma - strip the divinity out and place a bare forked human being
in its place? This is precisely what some of the poetry and philosophy
of fifth-century BC Greece achieved; the result is the idea that every
eye emits some sort of ‘visual ray’ in order to see.
According to Empedocles, the Sicilian shaman and poet who lived
from around 492 to 432 BC, the eye actively observes the world, illu-
minating it with its own spirit, ‘as when a man, thinking to make an
excursion through the night, prepares a lantern?
The idea of visual rays, or ‘extramission, became a coherent theory
for the first time in the writings of Plato, a disillusioned Athenian
politician who sought the consolations of philosophy around 399 BC.
., THEORIES OF VISION 159

Plato integrated the idea of extramission into a philosophical

scheme that is a strange mixture of political and esoteric impulses.
Since, under Plato’s prescription, objects are not ‘making themselves
felt’ as eidola, Plato had to explain how we are able to recognise
things. I have not seen this particular chair before, so how do I know
itis a chair? Plato's explanation presupposed the existence of universal
Forms — metaphysical archetypes to which all objects in the universe
at least roughly correspond. The mind compares the image it has
received to the metaphysical library of Forms.
If this explanation seems ugly (and it is) we ought, anyway, to
concede that Plato’s Forms explain why two differently coloured,
differently sized and generally rather dissimilar objects can both be
apples; how a cushioned couch and a stone bench can both be seats;
how a palace and a hovel can both be houses. We group similar items
together, and we recognise familiar forms more quickly than unfa-
miliar ones. The atomists had no explanation for this; Plato did.
I have to admit that I have a partisan preference for the atomists
over Plato. If the atomists stand accused of mangling matter in
support of their theories of vision, how much more questionable are
the arguments of Plato, who invents and populates an entire realm
of ideal being to explain the same thing.
But this is the conventional prejudice of a twenty-first-century
materialist. Plato’s world was very different from ours. Matter has
always been around to be tested, manipulated and interrogated. The
mysteries of cognition are much less obliging; and if sometimes we
have a sense of there being categories ‘out there’ against which we
index the properties of material things - think of terms like ‘redness,
‘roughness, ‘goodness, ‘monarchy, ‘love’ — who are we to deny them
a place in theory? We live in a materialist age, so we tend to assume
that all theories have to be materialist. They don't, and we certainly
shouldn't expect any thinker in 400 Bc to have denied or margin-
alised their spiritual sense, or discounted the possibility of other, as
yet unbreached, unexplored realms of being.

With its main claim routed, it is inevitable that the theory of eidola
collapsed before the theory of extramission. Extramission explains
why our hapless chap can't find his needle: the visual ray emitted by
160 A NATURAL HISTORY OF SEEING

his eye has to hit the needle first. It explains how we can move our
attention from one object to another: the eye's ray is narrow, taking
in one object at a time. It explains why we clearly see just a tiny part
of the visual scene, while the rest is a blur: only that part of the visual
ray reflected directly back into the eye is strong enough to be perceived
properly. It even explains why we can see better in daylight than at
night: by inventing a visual ‘ray, the theory can presume to invent, for
that ray, any properties it wishes; chiefly, in Platos theory, that it creates
‘a single homogeneous body with daylight that makes material contact
with the object seen. A reduction in light means a reduction of the
medium with which the visual ray can coalesce, and this, in turn,
means a reduction in vision. Nocturnal animals presumably emit more
powerful visual rays. Indeed, nocturnal eyes positively shine. Can you
not see the visual fire burning in a cat's eyes at night?
So we come to the secret of extramission’s robustness; both folk
wisdom and common sense support it. First, consider the way the
eyes of wolves and owls and other nocturnal creatures shine at night:
what can this be other than the visual ray lighting their path, ‘as
when a man prepares a lantern’? Second, Theon’s argument from
form: does the eye not look as though it is active, while all the other
organs of sense are so obviously passive? Finally - and this is Theon’s
clincher — we might ask ourselves why, when we try to make some-
thing out in the distance, do we screw up our eyes? Why all that
muscular effort, if we weren't doing something? If images merely
entered, wouldn't it make more sense to widen the eye?

Extramissions chief failing is, of course, that ray: how can sight, that
most immediate of senses, rely on such an exhausting-sounding
physical effort? .
Aristotle, a former pupil at Plato’s academy, was not at all convinced
‘that the ray of vision reaches as far as the stars, or goes to a certain
point and there coalesces with the object, as some think. This is an
argument from personal incredulity: surely the eye’s rays can't reach
the stars, ‘coalesce’ with them, and return with news of their nature,
at a glance? These sorts of arguments deserve to have big health warn-
ings chained to them; the world, after all, is full of unlikely wonders.
Happily, Aristotle knew this, and thought more deeply. He observed
THEORIES OF VISION 161

that nothing is visible unless it is lit, and that all light comes from
fire. (The ‘cold fire’ contained in luminous materials was a provoking
curio for Aristotle and later Aristotelians, well into the Western
Renaissance - but it was not one that seriously challenged Aristotle's
homely observation.) Fire, therefore, makes things visible.
Not everything is visible at once. Objects get in the way and occlude
“each other. Aristotle's ruse was to suppose that this is equally true
of the air; that it too hides objects from the viewer, unless it is lit.
For Aristotle, transparent objects - air included - were a special class
of object that reacted to light. If this ‘state change’ - from opaque to
transparent — occurs in all parts of an object all at once, we no longer
have to worry about how long a ‘visual ray’ takes to leave the eye,
coalesce (somehow) with the object in view, and return (somehow)
with (some sort of) visual information about what it struck.
Aristotle’s achievement was considerable. He explained how even
distant objects are perceived instantly, and why objects are visible
by day and not by night, all without recourse to the visual ray.
In dropping the ray, however, Aristotle had to throw away our ability
to explain how we focus on objects. If there is no ray, and we simply
receive visual impulses, why don't all the objects in our field of view
cry out for our attention all at once? We are back where we started.
Around this time - 300 BC — the study of vision began to fracture.
It had proved intractable. No one, however gifted or imaginative,
was able to come up with a single explanation for all the different
phenomena of vision.

2 - Straight lines

Science is facts; just as houses are made of stones, so is

science made of facts; but a pile of stones is not a house and
a collection of facts is not necessarily science.
Henri Poincaré

Twenty-odd years after Aristotle's death, Euclid, a committed

Platonist and a prominent Alexandrian mathematician, wrote his
own theory of vision. Euclid’s Optica contains not a single line about
162 A NATURAL HISTORY OF SEEING

the mechanisms of the eye, nor about the experience — the psychology
~ of seeing. Euclid’s interest in vision was purely mathematical.
He was one of the founding fathers of geometry; in the visual ray
he spotted an elegant, real-world demonstration of the behaviour of
rays and angles. In Optica, Euclid developed a simple optics of the
sort that today any schoolchild would recognise. So it is very hard
for us to appreciate quite how revolutionary this was. For the first
time, a visual theory was being expressed in purely mathematical
terms. For the first time, vision was being treated as a purely
geometrical problem.
Ironically, for a tradition that would prosper well into the last
century, we know frustratingly little about the early days of optics
and virtually nothing about the motivations and assumptions of its
first thinkers. Was mathematics to provide a full explanation of
vision? Was it in competition with other traditions? Or did optics
start out as a sideline - a neat real-world example of how objects
relate geometrically in space? We do not know. Nothing about Euclid’s
early life is certain. We do not even know the year of his death.
A hundred and fifty-odd years after Euclid’s Optica, Claudius
Ptolemy, another Alexandrian, developed and expanded the mathe-
matics of his day in a wealth of writings, most of which have been
preserved. Sadly, among the few books that are missing are the very
ones in which Ptolemy discusses the physics, psychology and philos-
ophy of vision. Even his optics fares poorly, surviving only in an inco-
herent translation by the Byzantine scholar Admiral Eugene of Sicily.
(Eugene is not entirely at fault; the original was lost, and he was having
to work from a rather poor Arabic translation.) It is tempting to spec-
ulate that these losses contributed to the split between the mathematical
tradition that gave birth to optics, and the more psychological, philo-
sophical approach to vision, typified by Plato.
Tempting; but in the end, it was inevitable that optics would domi-
nate vision theory for the next thousand years. Unpicking the role
of light in sight was, quite simply, the single most important break-
through in vision research. While psychology played its part in the
breakthrough, the greatest gains in understanding were achieved
through the clear, cold proofs of Euclidean geometry.
* * *
/ THEORIES OF VISION 163

I am writing this book in wartime. Every other night, another casualty

is seen being rushed through the doors of the Al-Kindi General Hospital
in Baghdad. One of the more depressing aspects of this project has
been the way the news of Iraq's seemingly bottomless slide into
dereliction has thrown up the places and names of the Arab Renais-
sance — a period quite as significant as the Renaissance in Christian
“Europe - which transmitted, codified, expanded, and improved the
intellectual inheritance of over a millennium of Mediterranean culture.
Baghdad's beleaguered general hospital is named after Abu Yusuf
Yaqub ibn Ishaq al-Kindi, scion of the royal house of Kindah, and
father of modern medicine and pharmacology. Al-Kindi was born
late in the eighth century AD in al-Kufa, where his father was governor.
His adult life was spent first in Basra, then Baghdad - the twin poles
of the Islamic world.
The intellectual revolution that gripped Europe six hundred years
later was driven by artisans, supported and patronised by a leisured
class of middle-ranking nobility. The Islamic Renaissance was of a
different order, a cultural project consciously incited and bankrolled
by the more-or-less benevolent dictatorship of the Mohammedan _
Caliphate. Cosmopolitan and aspirational, the empire had to reach
beyond its own traditions and borders to achieve its cultural ambi-
tions, and had relied for the most part on a succession of Christian
Syrian doctors to fill its libraries and write up its intellectual achieve-
ments. Al-Kindi’s soubriquet ‘the Philosopher of the Arabs’ not only
saluted his exceptional grip on logic, geometry, mathematics, music
and astrology; it also acknowledged that he was the first native Arab
thinker to make much impression on the world.
Though Al-Kindi was the Caliphate’s first native great thinker, he
seems to have been immune to the temptations of chauvinism. “We
ought not to be embarrassed of appreciating the truth and of
obtaining it wherever it comes from? he wrote, ‘even if it comes
from races distant and nations different from us. Nothing should
be dearer to the seeker of truth than the truth itself, and there is
no deterioration of the truth, nor belittling either of one who speaks
it or conveys it.”
Famously avaricious, Al-Kindi spent his most productive years in
anonymous seclusion in an affluent business suburb of Baghdad,
164 A NATURAL HISTORY OF SEEING

caring more for his garden and private zoo than for his neighbours.
Inevitably, he became a target for envy.
The Banu Musa brothers - Jafar, Ahmad, and Al-Hasan — were
recruited to Baghdad’s House of Wisdom at the same time as Al-Kindi
by the Caliph Al-Mamun. Al-Mamun was building up the greatest
library since the Alexandrian; translators were in high demand,
particularly those with the expertise to unpick Greek scientific texts.
We don't know precisely what sparked the enmity between Al-
Kindi and the brothers, though there are some intriguing hints: the
degree to which Al-Kindi could not be bothered to translate his own
Greek texts, but preferred to adapt the translations of others: the scale
and significance of the private collection of manuscripts he amassed
while working for the library. Neither do we know how long the
brothers’ dislike festered before the death of Al-Mamun, followed by
that of his brother and successor Al-Muttasim, left the House of
Wisdom in political turmoil. When the despot Al-Mutawakkil
succeeded to the Caliphate in 847, the brothers showed their hand.
Ingratiating themselves with the new Caliph, they saw to it that Al-
Kindi was beaten and his extensive library confiscated.
The old cosmopolitanism was over, and it is possible, given Al-
Kindis lack of interest in religious argument, that Al-Mutawakkil’s
maltreatment was of a piece with his persecution of his non-orthodox
and non-Muslim subjects. Al-Kindi lived the rest of his life in a sadly
reduced city, its vistas marred here and there by the ruins of churches
and devastated synagogues.
Al-Kindi’s optical work is an impressive, if confused, attempt to
re-bind the physics, geometry and psychology of vision, with geom-
etry for the first time carrying by far the greater part of the argu-
ment. Al-Kindi, the good geometer, was an extramissionist. His
arguments for the existence of a visual ray were predicated on the
fact that we see most clearly in the central part of our field of view,
and rather fuzzily elsewhere. His discussion of this point is inter-
esting, in that it is the first clear instance we have of someone using
reading to elucidate the behaviour of the eye. If the eye simply
‘received impressions’ passively, Al-Kindi argued, there would be
no need for the eye to jump from word to word on a page, for the
whole page is in view all the time one is reading. The eye’s movement
’ THEORIES OF VISION 165

along a line of text suggested, to Al-Kindi, that the eye was bringing
something to bear on the page — a narrow cone of what we would
call attention, but which for Al-Kindi was a real, if invisible, ray.
Al-Kindi’s greatest single contribution to vision research was, iron-
ically enough, a mistake, a happy inconsistency between his views
on the behaviour of the visual ray and an important insight he had
into the behaviour of light. Al-Kindi remarked that light pays no
heed to the objects off which it reflects. The form of the object does
not influence light’s behaviour. Light simply rebounds off every point
of every surface, at every angle.
This is, essentially, a philosophy of radiation, in which everything
affects everything else, all the time, through the transmission of power.

Al-Kindi’s theory of light as radiation was neatly captured, six

hundred years later, in this figure by Leonardo da Vinci. If light
is scattered off an object at every point, in every direction, then
an infinite number of focused images of the object will be
available to observers regardless of their location or distance.
166 A NATURAL HISTORY OF SEEING

Gone, in Al-Kindi’s philosophy, is the transactional view of the world,

promulgated by the atomists, where objects send exquisitely scaled-
down material messengers, like picture postcards, into the individual
observer's eye.
When Al-Kindi applied this observation to vision, all went well
at first. He assumed that all radiations behave the same. If the visual
ray behaves like light, hitting every point of every available surface
in a straight line, this serves to explain why we see objects in perspec-
tive. Imagine, for example, a wheel. Seen face on, it makes a circle
in the eye; lying on the ground, however, it makes an ellipse.
Al-Kindi used the same simple analogy between visual rays and
light rays to explain why the central part of our field of view is so
much clearer than the periphery. Assuming that the whole surface
of the cornea is responsible for vision, he argued that any object
directly in front of the eye would be exposed to visual rays emitting
from every part of the cornea. An object to the side, however, would
receive only a proportion of the rays, and thus it would make a rather
poorer mental impression.
This is the problem Al-Kindi failed to spot: if visual rays behave like
light rays, they presumably behave like light rays during their return
trip to the eye, as well as on their outward trip. In which case, why
should they return to exactly their point of origin on the surface of
the cornea? What is there to prevent rays from each point of the object
from striking every point on the cornea, blurring the view hopelessly?
By introducing the idea that every point of a surface reflects light
in every direction, and by making a metaphorical link between light
and the visual ray, Al-Kindi opened up a world of trouble for himself,
and a world of possibility for subsequent thinkers. Al-Kindi had
effectively, if unintentionally, proved the superfluity of the visual ray.
Leonardo da Vinci's figure makes this point with particular
eloquence: it is light, and light alone, that makes vision possible.

By his mid-forties, Abu ‘Ali al-Hasan ibn al-Hasan ibn al-Haythem had
reached the end of his tether with his desk job. As minister for Basra,
it was his job to mediate between and contain all the religious factions
in the city - work for which he felt little enthusiasm and less aptitude,
-/ THEORIES OF VISION 167

coming eventually to the conclusion that truth resided solely in the

sciences. Science was Al-Haythem’s private passion, and his consider-
able knowledge was the product of years of private study. Encouraged
by his growing scientific reputation and hungry for adventure, Al-
Haythem (spelled ‘Alhazen’ in most European sources) made his bid
for freedom and went off, with a party of engineers, to build the Aswan
dam.
Today's Aswan dam is three and a half kilometres long, and holds
back about 165 square kilometres of water. The reservoir is about
480 kilometres long; approximately the distance from London to
Newcastle. Filling it took twelve years and displaced 90,000 people.
Al-Haythem and his party of engineers, arriving in the year 1009,
had bitten off rather more than they could chew.
The only reasonable thing to do was to go home and admit defeat.
There was, however, one obstacle —- the Caliph Al-Hakim, Al-
Haythem’s cruel and neurotic patron, a man who, in true Mikado
style, had honed his homicidal cruelty on dogs (he disliked their
barking, and had them all killed) before working his way up to
people. No one remained Al-Hakim’s enemy for long.
Al-Haythem contrived to elude Al-Hakim’s displeasure by feigning
madness. Given the Caliph would occasionally take it upon himself
to ban certain vegetables, an alibi of insanity must have seemed quite
reasonable. Al-Hakim sighed, wrote off his losses, and consigned Al-
Haythem to house arrest.
Al-Haythem did not have to try too hard to appear crazy. After all,
he blacked out his windows and spent twelve years conducting
experiments in the dark. In 1021, Al-Hakim died. Al-Haythem opened
his windows and announced to his bemused guards that he wasn't mad,
after all; hed only been pretending. Incredibly, they believed him.
Al-Haythem was interested, first and foremost, in the geometrical
behaviour of light. He developed several beautiful and influential
demonstrations of the way light moves in straight lines. One experi-
ment, involving three candles and two rooms connected by a small
hole in the common wall, would revolutionise ideas about vision and
the eye.
Al-Haythem placed three candles in an otherwise unlit chamber.
A small aperture in the partition wall connected this chamber to a
168 A NATURAL HISTORY OF SEEING

Al-Haythem’s optical experiments are the first we know

of ever to use a camera obscura.

second chamber. When the candles were lit, three clear patches of light
formed on the far wall of the second chamber: proof enough that light
travels in straight lines. Al-Haythem also noticed that arrangements
of light passing through a small aperture would necessarily appear on
the far wall in an inverted form. That is to say, light from the right-
most candle would cast an image on the left-hand side of the wall,
while light from the left-hand candle would cast an image on the
right-hand side of the wall. In further experiments, Al-Haythem
explored the implications of this observation. The demonstration illus-
trated above is one of the more playful, showing how light from a
single candle, passing through several small apertures, projects multiple
inverted images on the far wall of the second chamber.
Al-Haythem wrote the first really good studies of refraction; he wrote
about sunsets, rainbows and eclipses and was the only writer of the
period to rival the Roman anatomist Galen in his description of
the parts of the eye. He was also the first writer we know of who under-
stood the role of light in vision. Al-Haythem knew that light, when it
travels from a lighter toa denser medium, changes its direction. He
understood the implication of this for curved surfaces: that two parallel
lines, hitting a curved object of greater density, will converge. In fact,
all rays hitting a curved surface will refract, except that ray which hits
the surface at a perfect right angle. That ray alone will not be refracted.
Al-Haythem understood from his anatomical knowledge (as Al-Kindi
did not) that vision takes place inside the eye. Since rays must pass
THEORIES OF VISION 169
}

Refraction and focus. Top: light entering a denser medium veers

towards the perpendicular. Bottom: if the denser medium is lens-
shaped, the combined course-changes upon entry and exit draw
parallel rays of light to a point. Beyond this point, images captured
by the lens are inverted.

through the pupil of the eye to be seen, this effectively restricts the
number of angles at which unrefracted rays can penetrate the eye. Light
entering the eye from objects straight ahead will not be refracted, but
light entering the eye from objects to one side will be. Since the view
straight ahead is clearer than the view in the periphery, Al-Haythem
concluded that unrefracted rays are stronger than refracted rays.
This was very clever indeed. Al-Haythem had explained the char-
acter of the eye’s field of view purely in terms of light, without any
recourse to visual rays. This was a tremendous advance, combining
anatomy, mathematics and psychology to explain how we see.
But there was trouble ahead. Al-Haythem, following Galen, under-
stood that the eye is connected to the brain by the optic nerve, and
that the brain is the seat of consciousness. This left him with a serious
problem; if light enters the eye and travels up the optic nerve, what
was the lens, the ‘glacial humour; for?
It couldn't be a focusing device, because the eyeball itself is a lens.
Al-Haythem conjectured that light rays, passing through the lens, were
made parallel, so that they might travel up the optic nerve to the brain.
In the absence of good anatomical understanding of the retina, this
is a clever idea, but it is seriously flawed, for it imputes unexplained
170 A NATURAL HISTORY OF SEEING

properties to the lens. How can a lens take a cone of converging light
and make it diverge, so that it runs in a tight, parallel beam? There
are two pressing reasons why Al-Haythem, at this crucial point in his
scheme, abandoned optics in favour of what is, at best, speculation.
First, unless the light passing from the glacial humour runs in a
parallel beam, he could not explain how light reaches the brain. He
had no theory of electricity or nerves, so direct communication of
light to the brain was his only means of explaining how the brain sees.
The second problem is more subtle. Even if Al-Haythem supposed
some other, non-optical, means by which the optic nerve could pass
visual information to the brain, the image cast on the optic nerve
by a regular lens would make no sense. The image would be inverted,
just as the candle images are inverted in the drawing on page 168.
And this is an obvious nonsense. Inverted images cannot possibly
play a role in vision. How could they?
If they did, would we not see the world upside-down?

OQ. MWDSNLWG BIEN; cst?

The oldest surviving pictorial representation of the visual

system, from Al-Haythem’s Book of Optics, written around 1038.
' THEORIES OF VISION 171

3 -— The world turned upside down

We see nothing, save through reason.

Arthur Schopenhauer, On Seeing and Colours (1815)

In 1962, the anthropologist William Hudson showed a group of adults

and children living in the Zambian bush two drawings of an elephant.’
In one, the elephant was seen from above; in the other, the same
elephant was squashed, as if by a steam-roller, so that its legs and
trunk were splayed out to the sides of the body. The children preferred
the ‘squashed’ drawing, because it contained more of the elephant.
When Westerners are shown the same drawings, they prefer the
unsquashed one. Although there is less elephant in it, they consider
the picture more realistic, since it captures what it would be like to
see an elephant from that particular angle.
Both choices are sophisticated, aesthetic decisions. The pictorial art
of the veldt typically conveys ideas of value and meaning; the art of
the post-Renaissance West typically simulates the rules of optics. People
prefer some representations over others. Representations are plastic,
modified over time. They change.
Anyone with the ability, opportunity and inclination to read a
op) A NATURAL HISTORY OF SEEING

book like this is going to be living in a world of representations: a

world of signs, moving images, and visual puns. Our world is so rich
in visual representations that some philosophers ask whether we are
losing touch with the real world entirely.
Images and signs were certainly part of the world in Al-Haythem’s
time, but they were far from being dominant. The relationship
between what was seen and what was really out there seemed, for
that reason, perfectly straightforward. The eye apprehended the truth
which lay before it. Why make life more complicated?
The trouble with an eye that sees upside-down is that it requires
the brain to turn things the right way up. The brain cannot simply
apprehend what is out there — it has to manipulate the information
it receives and then re-present it. But to whom, exactly, is this right-
way-round image being re-presented? Is there a screen in the brain,
and a watcher, watching it?
Al-Haythem and his successors winced away from the idea that
eyes received inverted images, for the very good reason that for this
simple optical quirk to work, a new philosophy of mind was required
— a view of perception that would drive a wedge between the world
itself and the apprehending mind.
If everything we see is manipulated before we see it, how do we
know that anything is true? Suddenly, the world is split in two: the
world we sense, and the world represented by the senses. If every-
thing we sense is a representation of what’s really out there — well
- what's really out there?
We don't know. We can’t know, any more than I can know, from
looking at the icons on my computer screen, exactly what is going
on in the silver box under my desk. I can control the box, as much
as it will let me. I can drag files into a ‘wastebasket’ I can shuffle files
from one ‘folder’ to another. I can draw pictures. I can type words,
delete them, and get them back again at the touch of a button. But
I cant understand what's really going on inside this box of tricks,
because the icons I click and drag need not in any way resemble
what's really going on inside the box.
This vision of the world reduces philosophy and science to a user’s
manual, a description of how we interact with the world, which says
nothing whatsoever about the way the world really works. It’s a
|
“THEORIES OF VISION 173
predicament which Donald Hoffman, an expert in computer vision,
recognises and explores with unusual candour in his book Visual
Intelligence:

Neither biology nor quantum theory dictates the nature of the

_ Telational realm. Nor does any other science. Each studies certain
phenomena, and describes these by precise theories. In no case
do the phenomena or the theories dictate the nature of the
relational realm. We might hope that the theories of science
will converge to a true theory of the relational realm. This is
the hope of scientific realism. But it is a hope as yet unrealised,
and a hope that cannot be proved true.

For six hundred years, the philosophers of Europe’s Dark Age fought
shy of this demoralising, ultimately unavoidable, conclusion. The best
of them was William of Ockham. Born in Surrey just before the turn
of the fourteenth century, Ockham bequeathed the European
philosophical tradition some powerful weapons. ‘Ockham’s razor
(often spelled Occam’s razor) is the informal-sounding but powerful
observation that the simpler an explanation is, the more likely it is
to be right. Applying his own stricture to the theories of vision that
were slowly penetrating European thought, Ockham argued that the
mind apprehends things, at a distance, through the application of
the intellect. By imprisoning Plato's realm of perfect Forms within
the bounds of the individual skull, Ockham made our ability to
recognise, categorise and discriminate a matter of mind, rather than
a quasi-mystical condition of nature. This simplifies vision theory
immensely; the eye sees patterns which the mind can apprehend.
Generally speaking however, the monasteries and universities, the
great seats of learning of medieval Europe, more or less neglected
optics, regarding it as a discipline whose philosophical riches were
exhausted. The thirteenth-century polymath, Roger Bacon, attempted
to reconcile the theories of vision, but the result was indigestible.
(His list of sources was useful though, as it included Al-Haythem.)
Outside the academies, the practical applications of optics were
growing ever more numerous. The arrival, in the fifteenth century,
of flat glass mirrors was probably the trigger for the series of artistic
174 A NATURAL HISTORY OF SEEING

experiments which led to the discovery of linear perspective.

The mirror image - a three-dimensional scene, captured on a flat
rectangular surface — laid before the artist a challenge hard to resist:
the possibility of painting pictures that could, like mirror images, be
mistaken for reality.
Since a painting is a still image and cannot shift, as a mirror image
can, to accommodate the changing position of the spectator, perspec-
tive art requires that we look at it from a particular angle. This was
not the case with the Gothic art which preceded it, nor with the school
of ‘naive art in our own day; both, though ostensibly ‘representational,
arrange their objects higgledy-piggledy on the picture plane.
Much has been written on the pervasive influence of perspective
art. For our purposes its main significance is the intellectual climate
it encouraged. What more can optics say about vision? Nothing —
and this is the point. Leon Battista Alberti (1404-1472), writing up
the perspective techniques of the artist and optical illusion-maker
Filippo Brunelleschi, had this - and this alone - to say about the
how the mind apprehends what it sees: “This dispute is very diff-
cult, and is quite useless for us:
‘We; in this context, were the artisans supplying curios and orna-
ments to an unprecedentedly wealthy merchant class. Were they aware
that their optical games were revealing the secrets of vision more
quickly than half a millennium of philosophical bean-counting? No.
The writings of the great artists reveal a remarkably narrow field of
interest. Albrecht Diirer’s Painter’s Manual of 1525 says nothing about
colour. Leonardo da Vinci, who studied linear perspective and drew
over one thousand images through the camera obscura, was also one
of the finer anatomists of his day - yet he thought that the optic
nerve is connected directly to the lens and wrote of, not one, but six
methods by which upright images reach the brain - each method
more wrong-headed than the last. Even Johannes Kepler — the math-
ematician and astronomer who provided us, in 1611, with our current
model of how the eye uses light - was only incidentally interested
in vision. Understanding the eye was, for him, only a stage on the
path to understanding the movements of the planets.
* THEORIES OF VISION 175
Tycho Brahe was the greatest astronomer of Kepler’s day. He was
also the best-funded. Brahe’s uncle had died saving King Frederick
Il of Denmark from drowning. (Returning from a naval battle with
the Swedes, the king contrived to topple off his own drawbridge.)
In a gesture of gratitude, in 1576 King Frederick gave his nephew,
Tycho, an island and money for an observatory. Lots of money: a
maintenance grant of 500 daler per year, revenues and firewood from
the manor of Kullagaard, the use of eleven farms near Helsingborg,
a canonry of Roskilde cathedral, and the fiefdom of Nordfjord in
Norway - altogether about one per cent of the crown revenue.
‘My opinion of Tycho is this; Kepler wrote to Michael Maestlin,
his old mathematics teacher, ‘he is superlatively rich, but he knows
not how to make proper use of it, as is the case with most rich
people. Therefore, one must try to wrest his riches from him.’ Kepler’s
ability to turn a phrase is a godsend to biographers. Kepler’s mother
was, according to his own account, ‘small, swarthy, gossiping and
quarrelsome, of a bad disposition; while his father was a man ‘vicious,
inflexible, quarrelsome and doomed to a bad end... Saturn in VII
made him study gunnery.
He isn’t making any of this up. The Keplers were as mad as a bag of
cats. Johannes’s aunt was burned as a witch, and we still have the lawyers’
bills from the time he defended his mother from the same charge.
While he was teaching mathematics in Graz, Kepler was forced
to choose between embracing Catholicism (the Keplers were
Lutheran) or being expelled from Austria. Tycho Brahe’s timely
offer of an assistant’s job was brokered by their mutual friend, the
anatomist Johannes Jessen. For the next year, Kepler and Brahe
worked together on the problem of the orbits of the planets. Kepler
had to wait until Brahe’s death, the following year, before he was able
to obtain the great man’s jealously guarded astronomical data, stealing
it from under the noses of Brahe’s acquisitive relatives. Only then
was he able to make the astronomical breakthrough for which he is
best known, ascribing elliptical paths to the planets as they orbit the
sun. While Brahe was alive, Kepler's studies were of a humbler cast.
For instance, he was given the task of explaining why the moon
shrank suddenly during a solar eclipse.
This nonsensical observation was not a little embarrassing for Tycho
176 A NATURAL HISTORY OF SEEING

Brahe, whose reputation stood - and stands - on the unerring and

unprecedented accuracy of his-observations, the very best achieved
without telescopes. How could it be that an astronomer capable of
measuring one minute of arc — one-sixtieth of a degree of the sky -
could not get a sensible figure for the size of an object as big as the
moon? Assuming that the moon wasnt playing tricks, Brahe’s observing
equipment had to be introducing the discrepancy. But how?
Kepler’s explanation, in a letter he wrote to Maestlin, is typically
facetious, more concerned with the fact that he got his pocket picked
while he was working (‘By Hercules!’ he wrote, ‘what an expensive
eclipse!’) than with the details of his observation. Kepler studied the
eclipse of 9 September 1600 in Graz, using a camera obscura mounted
on axles - the first occasion we know of where tracking was used
in astronomy. The camera obscura allowed direct measurement of
the diameter of the moon, projected on the wall opposite the aper-
ture. Sure enough, the diameter of the moon, silhouetted against the
sun's corona, was significantly smaller than the diameter of the posi-
tive image the moon made when observed through the same equip-
ment at night.
What was going wrong? Kepler’s first recourse was to the library,
to the works of the thirteenth-century writer, John Pecham. When
he could make neither head nor tail of Pecham’s ‘arcane theorising,
Kepler turned to experiment. Using taut threads to represent rays
of light, a book suspended from the ceiling to represent a viewed
object, and a hole cut in a table to represent the aperture of the
camera obscura (the reactions of his domestic staff are not recorded),
Kepler constructed an apparatus to demonstrate, to his own satis-
faction, why the camera obscura, though useful, was unreliable.
The camera obscura is unable to bring objects into perfect focus.
Images cast on to the back wall of a camera obscura might look
crisp, but perfect focus cannot be achieved. This is because light
passes into the chamber, not through one ideal focus point, but
through an aperture. Light rays from every point of the object pass
into the box at many angles, which taken all together make the shape
of the aperture. The aperture itself introduces blur, so that light from
a full moon, blurring ever so slightly, will make a large image, while
light bleeding round the edges of the moon during an eclipse, blurring
“ THEORIES OF VISION 177
ever so slightly, will make a smaller one. The aperture can never be
small enough to eliminate all blur; and the smaller the aperture, the
dimmer the image.
This was, all in all, a rather disturbing conclusion for Kepler to
reach. The eye was a kind of camera obscura, at least in the sense
that light entered it through a narrow aperture, the pupil. If the
behaviour of light through apertures led to inconsistent results, how
trustworthy, from an astronomical point of view, was human vision?
Tycho Brahe, Kepler’s mentor, claimed to have achieved an obser-
vational accuracy of one minute of arc. What if he were wrong?
Kepler had to establish whether the human eye was reliable, and if
it was, how that reliability was achieved. Without that knowledge,
the best astronomical data of the day could not be relied on.
Kepler did not have to work in a vacuum. He could turn to any
one of the thirty classical works on optics republished in the sixteenth
century, including books by Euclid and Al-Haythem. But what finally
drove Kepler to his astonishing result was the work of Felix Platter,
a man entirely unconcerned with the nature of light.

Felix was the son of Thomas Platter, a highland shepherd and

vagabond turned boarding school teacher. He made the most of his
father’s success. A graduate of the University of Montpelier, by 1560
he was one of Basel’s more fashionable doctors, attracting aristocratic
and bourgeois clientele from as far afield as southern Alsace, Wiirt-
temberg, and the Black Forest. Fame came at a cost. On a family trip
to Grachen, his father’s rural birthplace, Felix woke up one morning
to find the girls of the region lined up outside his house with gifts
of their own urine, hoping for (and to Felix’s great credit getting) a
free check-up from the famous doctor from Basel.
Felix Platter’s medical interests were shaped by the anatomical
advances of the Renaissance. These advances were not so much to
do with the increasing acceptability of human dissection (though
this was certainly important) as with improvements in printing,
which provided anatomists with decent, accurate, contemporary
illustrations of previous findings. In 1588, Felix Platter published De
Corporis Humani Structura, a slim volume containing fifty anno-
tated anatomical plates. Felix’s anatomy is conscientious, careful, and
178 A NATURAL HISTORY OF SEEING

for the most part unexceptional — an excellent representative example

of the anatomy of his day. In one respect, however, Platter’s work
was certain to startle the informed reader.
With no knowledge of optics, Platter did not know how important
it was that the eye should not see inverted images. Indeed, it is
extremely doubtful whether he knew that images could be inverted.
He was free to observe, as neutrally and as accurately as he could,
how the eye was made. With a trained gaze and no received opinions,
Platter saw what no one had seen before. He saw the crystalline
humour of the eye, just behind the iris, towards the front of the
eyeball. He saw ciliary fibres holding the crystalline humour before
the pupil of the eye, and he saw that the whole apparatus arose
neither from the optic nerve, nor the retina, but from the purple,
blood-rich layer - strangely reminiscent of the inside of a grape —
which lies just behind the white of the eye.
Optics was not Felix’s subject, but he was not without intelligence
or imagination. It seemed clear enough to him what was going on:
the crystalline humour ‘is the looking glass of the optic nerve; and,
placed before the nerve and the pupil, it collects the species passing
into the eye as rays and, spreading them over the whole of the retiform
nerve [retina], presents them enlarged in the manner of an interior
looking glass, so that the nerve can more easily perceive them.”
This is wrong. Platter forgot that magnifying glasses only work to
magnify an image when used at close quarters. Hold a magnifying
glass above this page and the words will appear larger, because light
fans out from the page at all possible angles; the glass gathers all the
rays that hit it, bending them back towards your eye. To the eye, it
appears that light from the object is coming in from many more
angles — which is to say, that the object is larger. However, rays of
light hitting the lens from a distant object are, to all intents and
purposes, running parallel to each other. The lens still gathers up
this light and makes it converge at a focal point but because the rays
are already running parallel, they come together much closer to the
magnifying lens than would diverging rays. What the eye sees is an
image produced behind the focal point - in other words, an inverted
image.
Platter’s anatomy is, none the less, very good, and it was only sensible
“THEORIES OF VISION 179
that Kepler, encouraged by his old friend Johannes Jessen, should
put his trust in his findings. Kepler’s theory of vision applied the
rules of optics to Platter’s model of the eye. Light hitting the pupil
is refracted, making the light rays that pass into the eye run more
or less in parallel. When this parallel beam passes through the crys-
talline humour (the lens), it is once again refracted, and produces
an inverted image upon the retina. From Platter and Jessen, Kepler
was familiar with the idea that the retina is an extension of the brain,
connected to it by the optic nerve, so he had no difficulty in asserting
that it is the retina that receives images. But how such inverted images
are perceived the right way up is a matter, not of optics, but of inter-
pretation: a mapless and mysterious territory into which Kepler,
wisely, did not venture, saying ‘the armament of opticians does not
take them beyond this first opaque wall encountered within the eye’
So Kepler's innovation was, ironically enough, to apply already
known facts to an already mapped structure - less of a discovery,
more a brave insistence upon the inevitable. Still, we shouldn't under-
estimate the intellectual chutzpah involved. Kepler took consider-
able pains to put the retinal image the right way up again, but
eventually abandoned the attempt in disgust. (‘And there was no end
of this useless labour, he complains.)®
We do not know to what extent Kepler was taxed with the impli-
cations of his own work. His focus was directed outward, at the
motions of the planets. He drew no diagram to illustrate his theory
of how we see, and the image traditionally used to illustrate Kepler's
work is from the Enlightenment philosopher René Descartes’ La
dioptrique, of 1637.
It is a fitting and provoking choice. The optics are Kepler’s and in
all essentials they match our current understanding of how light
behaves within the eye. But there is another detail, a flourish added
by Descartes. At the bottom of the figure is an observer, studying the
back of the eye, which has been scraped clear to make visible the
inverted image within. On one level, this figure simply stands in for
the anatomist. (Descartes himself once stripped the sclera and choroid
from an ox’s eye and used the eye to project an image on to a piece of
paper.) At the same time, the figure also suggests a contemplative
intelligence, standing somewhere behind the eye. This personifica-
180 A NATURAL HISTORY OF SEEING

tion of the conscious mind cannot but remind us of the dreadful

conundrum faced by all theorists of vision - including Descartes -
who came after Kepler. If everything we see is a representation of
the world, then seeing becomes just representation. If I cannot appre-
hend the world directly, but can only percieve representations of it,
then where exactly am I, and where, exactly, is the world?
SEVEN

Nervous matter, visually endowed

Bodies are not designed machines. They are palimpsests of half a

billion years of gradual change. After this long, no anatomical struc-
ture performs just one function.
When the Roman anatomist Galen noted the mesh of major blood
vessels lying across the retina, he concluded that the retina existed
to feed the eye. For Galen, the retina was the mesh of blood vessels.
That beneath these vessels lay another, nervous, layer only became
apparent when Kepler's optical description of the eye insisted that
the retina must be, in anatomist William Bowman’s delightful phrase,
‘a sheet of nervous matter visually endowed.
The story of the discovery of the retina is vexed and broken. To
tell it is to set many plates spinning at once. I hope that they continue
to spin in your memory by the time I gather them up at the end of
the chapter.
We begin with the story of ‘a splendid toy, the ophthalmoscope,
which did so much for our understanding of the retina, but did not
let us see it. The second part describes how the elusive tissue of the
retina finally gave itself up to close study, but told us virtually
nothing. The extraordinarily beautiful drawings of the Spanish
histologist Santiago Ramon y Cajal were required before anyone
182 A NATURAL HISTORY OF SEEING

could begin to really understand how nerve tissues ticked. Meanwhile

the infant science of physiology set about explaining the retina’s
working, and racked up some impressive successes. ‘The final part of
this chapter describes how the physiological observations of Ernst
Mach, in 1867, were happily married to the anatomical work of
Haldan Hartline, sixty-five years later. This marriage ushered in our
current model of the retina.

1 — A splendid toy

Everything in the vertebrate eye means something.

Gordon Lynn Walls

Immediately behind the retina, cupping the pigment-stuffed tips of

the photoreceptive rod and cone cells, lies a single layer of cells, the
pigment epithelium. Its heavy black pigment, like the backing plate
of a camera, stops light scattering off the back of the eye.
Nocturnal vertebrates do not have pigment epitheliums. They have
mirrors. By bouncing light back through the eye this mirror, the
tapetum, effectively doubles the amount of light the eye can see.
Strap a light to the barrel of your gun, and wait for night. Stand
behind a likely tree, and aim your light-cum-rifle into the darkness.
Wait for two bright discs to light up in the dark, and fire. The discs
are the eyes of your prey: you cannot miss. This method of hunting,
called pit-lamping, is an old prospector’s trick. In the days before
electric lamps, prospectors lit their way using lamps powered by
acetylene and water. The lamp doubled as a handy hunting device,
hence ‘pit-lamping’ Strictly speaking, you don't need a lamp, or even
a rifle, to go pit-lamping. All you need is light. When the native
Americans of the Shuswap River in British Columbia go night-fishing
for salmon, they light fires in their boats, using wood that doesn’t
crackle. Salmon eyes are good reflectors, and the fishermen spot and
spear their prey by firelight, with very little trouble.
In the interests of a quiet life, I should point out to my Canadian
readers that if you go out pit-lamping, you will be arrested. There
are a number of reasons for the ban. There's the safety issue: a lamp’s
NERVOUS MATTER, VISUALLY ENDOWED 183

beam can pick out eyes at a great distance, but it isn’t nearly powerful
enough to reveal the whole animal. The eyes you're aiming at might
be the eyes of a deer; but they might just as easily be the eyes of
your neighbour's cow. But I suspect the real driving force behind
the ban is the fact that pit-lamping is incredibly easy; game stocks
would plummet if every fool with a lamp and a gun could bag his
or her limit every season.
For most of human history, it was assumed that bright eyes were
generating their own light. As late as the middle of the seventeenth
century, Thomas Willis, the co-founder of the Royal Society, met an
old man who claimed to read books by the light coming out of his
eyes.
The problem was not that mirrors in the eye were unthinkable;
animal mirrors - in the form of fish scales — are familiar enough,
and anyone curious and strong-stomached enough to dissect, say, a
horse’ eye, will find its back wall covered with a shimmery coating.
The problem - long before photoreceptors were discovered - was
to understand how mirrors could possibly benefit vision.
What were seventeenth-century ‘mechanical’ philosophers and
anatomists to make of mirrors in the eye? If the eye existed to gather
light, why should it reflect all that carefully gathered light back into
the environment? Even more puzzling, why should nocturnal animals,
starved of light, waste this carefully gathered light so recklessly?
The difficulty of interpretation is evident in the Anatomia Humani
Corporis (1685) by Professor Govert Bidloo. Bidloo (1649-1713), a
playwright turned anatomist, served as personal physician to William
Ill, first in Holland and later in England, where he was elected a
Fellow of the Royal Society. His Anatomia is generally regarded as
the finest anatomic work of the seventeenth century. In his discussion
on the eye, Bidloo observes that the ‘glowing’ nocturnal eye is really
emitting no more light than it receives. The careful wording is impor-
tant: nowhere does he say that the eye is merely reflecting light.
Bidloo kept his options open. }
The Academy of Sciences, Paris's rival to the Royal Society, was
founded in 1666 by France's finance minister, Jean-Baptiste Colbert.
The academy immediately distinguished itself by the quality of its
anatomical work. Anatomists like Claude Perrault and Jean Mery
184 A NATURAL HISTORY OF SEEING

were at the forefront of efforts to advance anatomy as a compara-

tive science — efforts which were greatly aided by the regular dissec-
tion of rare animals from King Louis XIV’s private zoo.
In 1703, Mery noted that a cat’s eyes shine much more brightly
if you hold the cat under water. This - if you can get past the image
of the experiment itself - suggests that eye shine has at least
something to do with optics. But even here, at the birth of modern
anatomy, Mery - a hard-nosed Mechanical if ever there was one,
a man who undertook clandestine dissections whenever the oppor-
tunity arose - was not prepared to explain eye shine purely in
terms of mirrors and optics, and regarded his findings as ‘myste-
rious. (The mystery is easily punctured. The submerged eye is not
actually any brighter. It is simply that an eye in water no longer
refracts the light to the same degree, so instead of being emitted
in a beam, light sprays out of the cat’s eye, making its shininess
easier to spot.)
Karl Asmund Rudolphi (1771-1832), a Swedish-born naturalist,
is best known for his Grundriss der Physiologie. This work, which
argues that the human genus should be divided into species, not
races, is notorious for lending a racist impulse to German and Scan-
dinavian intellectual life, with dire and well-known consequences.
In the year of its publication, 1821, well over a century after Mery
dunked his cat, Rudolphi turned his attention to the directionality
of the shining eye. He was able to show that the reflecting eye will
emit light along exactly the same line as the direction of the in-
going rays. No chemical or biological process is taking place - a
point he demonstrated by the simple expedient of shining lights into
the eyes of a decapitated cat.
Rudolphis macabre experiment brings the mystery of the reflecting
eye to a head, as it were. Why should a light-starved eye reflect the
light? Rudolphi made the matter no less mysterious: he did, however,
suggest a way forward for research. The eye reflected light back along
the line of illumination. If there were a way of capturing and studying
this reflection, then it would be possible to study the back of the
living eye.
Jan Purkinje — who just happened to be Rudolphi’s son-in-law —
attempted this in 1823:
NERVOUS MATTER, VISUALLY ENDOWED 185

When I observed the eye of a little dog from a certain direction,

that light seemed to be thrown back, until I discovered that
the light is reflected from the hollow surface of the lens into
the eye and then returned. When the experiment was imme-
diately repeated with human beings, the same phenomenon
. occurred: indeed, the whole pupil lit up in a beautiful orange
colour.

What Purkinje is describing in such fascinated detail has become,

in flash photography, a familiar and rather annoying problem: ‘red-
eye’ — the ability of even the human eye to give off a bright orange-
red glow if exposed to powerful light.
When he came to study the inside of the eye Purkinje, like many
a researcher after him, kept getting in the way of his own light. How
was an observer to place his line of sight along the same line as the
line of illumination?
The physicist James Clerk Maxwell (1831-1879) seems to have
solved the problem by 1854, but quite how is a mystery. In a letter
to his aunt, Jane Cay, he wrote: ‘People find no inconvenience in
being examined, and I have got dogs to sit quite still and keep their
eyes steady. Dogs’ eyes are very beautiful behind, a copper-coloured
ground, with glorious bright patches and networks of blue, yellow,
and green, with blood-vessels great and small.”
Maxwell did not pursue his enquiry, aware, perhaps, that several
competing mechanisms were doing the rounds of the learned
societies, all claiming to solve the illumination problem and thereby
make it possible to study the inside of the living eye. Ironically, the
best of these early ‘ophthalmoscopes’ pre-dated Maxwell's doodling
by about seven years, but no one knew about it.
Charles Babbage (1791-1871), computer pioneer and inventor of
gigantic mechanical calculating engines, had a talent for poverty.
Having invented everything worth inventing - from a speedometer,
to a heliograph, to a cowcatcher - he expected at least a knighthood
for his labours, but at the age of seventy-nine, impoverished,
maddened in his last hours by the din of Italian organ-grinders,
Babbage died, alone and uncelebrated. Among the inventions that
ought to have made his fortune was a solution to the illumination
186 A NATURAL HISTORY OF SEEING

problem: in 1847, curious about the double vision he experienced

in one eye, Babbage invented the ophthalmoscope.
Babbage took a piece of mirror and scraped off two or three small
spots of the silvering. He mounted the mirror in a brass tube, at an
angle of forty-five degrees. An opening in the tube let light in. Light
hitting the mirror was reflected down the tube into the patient's eye.
Light reflected and focused by the eye travelled back up the tube to
the centre of the mirror, passed through the clear spots in the glass,
and arrived at the observer's eye. That was the theory, but to establish
the practical usefulness of his invention, Babbage needed a professional
champion.
William Mackenzie, the Glasgow doctor whose Practical Treatise of
the Diseases of the Eye, of 1830, had become a standard work, exchanged
letters with Babbage about his double vision, and recommended the
eminent ophthalmologist Thomas Wharton Jones. Wharton Jones (who
had earlier got himself unwittingly entangled in the Burke and Hare
body-snatching scandal) was one of London's leading ‘eye men. Edward
Jenner called him one of the greatest Englishmen who ever lived. There
was only one problem: he was short-sighted.
When he looked down the barrel of Babbage’s invention and saw
nothing of what he had been promised - no network of blood vessels,
no radial structure, no sign of a nerve head —- Wharton Jones advised
Babbage that his invention was useless. For this, Wharton Jones has
come in for not a little retrospective stick. This is not entirely fair.
Babbage had invented an ophthalmoscope; it just wasn't a very good
one. The beam of light issuing from the eye is slightly convergent;
had Babbage thought to add a weak diverging lens just behind the
mirror, the images gathered by his instrument would have been a
great deal easier to see: even Wharton Jones might have been
impressed. As it was, the disappointed ophthalmologist saw nothing
but a red splodge.?
He did, however, keep detailed notes of the demonstration. In
2003, these were followed, and a copy of Babbage’s instrument was
made for the British Optical Association Museum in London. It is
clear from this that the first commercially successful ophthalmo-
scopes resemble Babbage’s prototype far more closely than did the
instrument generally held to be the first ophthalmoscope - the one
NERVOUS MATTER, VISUALLY ENDOWED 187

cobbled together four years later from cardboard, glue and glass
slides from a microscope set by Hermann von Helmholtz.
Helmholtz (1821-1894), a physiologist with ambitions for a career
in physics, arrived to run Kénigsberg’s physiology department in
1848. Jan Purkinje’s work on illuminating the eye had, by this time,
largely been forgotten, but a student at Konigsberg, William
Cummings, independently stumbled upon many of the same findings.
Helmholtz was intrigued by Cummings’ work, and it was while he
was trying to work out how to demonstrate it to his students that
he threw together his ‘splendid toy.
Splendid it was: in its more finished incarnation, the ophthalmo-
scope revealed the back of the eye in exquisite detail, assisting greatly
in the diagnosis and treatment of eye diseases. Its popularity was
slow to build, however. Helmholtz enjoyed playing the showman
among his acquaintances. (“The ophthalmoscope is, perhaps, the most
popular of my scientific performances; he once remarked.)* But
Helmholtz, a physicist and physiologist, took little interest in the
modifications and improvements necessary to make the instrument
a commercial success. In the year following its manufacture, only
eighteen ophthalmoscopes were sold.
What needed the most work was the light source. Having one’s
eyes examined by an ophthalmoscope powered by oil, gas, and even
petrol must have daunted many a client. It says much for the consti-
tution of the nineteenth-century public that the device eventually
found favour as a parlour game.

2 - An invisible quarry

If you are going to study how the brain works, it is a good

idea to start with the simplest part.
Jeremy Nathans

In the living eye, the retina is invisible.

Not only are its neural layers thin (wavelengths of light are thicker),
they are transparent. More than that, they have optical properties
virtually identical to those of the vitreous humour which fills the eyeball.
188 A NATURAL HISTORY OF SEEING

The match is so close, light passing from the humour into the nervous
tissue does not glint or shimmer. (If it did, the consequences for
vision would be disastrous.) However good the magnification, the
study of the living retina reveals barely more than was visible to
Galen eighteen hundred years ago.
The ophthalmoscope’s contribution to understanding was consid-
erable but, as we shall see in a moment, it was indirect. You cannot
see the retina through an ophthalmoscope.
What can you see? You can see the larger branches of the blood
vessels forming a mesh across the back of the eye - a spider's web
centred on the head of the optic nerve. Because this is the retina’
only other visible feature, it is very tempting to suppose that light
is brought into focus here. Leonardo thought so, with good reason.
If light is focused on the nerve head, then this can be the ‘seeing’
part of the eye, and the rest of the retina, with its mesh of blood
vessels, can go on being a simple blood delivery system.
In 1619, the Jesuit astronomer Christoph Scheiner spotted the
problem with this common-sense explanation: the optic nerve head
can't be the ‘site of foveation because it is in the wrong place. In his
drawings, Scheiner put the optic nerve head several millimetres
towards the nose, away from where light is brought to a focus against
the retina. It is the retina, and not the nerve head, that is ‘seeing’

Prince Charles Stuart, the second son of Charles I of England and

Henrietta Maria of France, arrived in London to claim the throne
on his thirtieth birthday, 29 May 1660. He received a warm welcome.
England's experiment with republicanism had collapsed. But the
clock could not be turned back. Charles II found his powers and
privileges severely limited by Parliament, leaving him in the humil-
iating position of having to rely, for day-to-day funds, on a pension
from Louis XIV of France.
Charles, made canny by his many years in exile, concealed his
absolutist ambitions well. Only now and again did he behead his
ladies-in-waiting. He learned the trick from his acquaintances at the
Royal Society, to which he gave his royal seal of approval in 1661.
They, in their turn, learned it from the French priest and physicist
Edmé Mariotte.
NERVOUS MATTER, VISUALLY ENDOWED 189

THESE
WORDS WILL
VANISH

STARE AT THE CROSS

The blind spot. Cover your right eye and focus on the cross. Move
the book towards you, and the message to the left will disappear.

Mariotte, at first believing that the nerve head must be the seat
of vision, discovered that - visually at least - the nerve head did
precisely nothing. It was blind: each eye had a small blind area in
its field of view, corresponding to the location of the nerve head.
Mariotte had discovered the ‘blind spot.
Cover your right eye, focus on the cross above and move the book
slowly towards you. At a certain point (and with a little practice) the
words to the left of the cross wink out of existence, revealing the white
paper beneath. Move the book closer, and the words reappear.
Why - other than in experiments like these - don't we notice the
blind spot? As David Brewster once wrote, We should expect, whether
we use one or both eyes, to see a black or dark spot upon every
landscape within fifteen degrees of the point which most particularly
attracts our notice. The Divine Artificer, however, has not left his
work thus imperfect ...the spot, in place of being black, has always
the same colour as the ground’>
Repeat the experiment with a red piece of paper, and the words
will disappear to reveal red paper. Repeat the experiment with wall-
paper, and the pattern of the wallpaper will be revealed. Your visual
system is ‘filling in the blind spot with the surrounding texture. A
miracle? Not really. Peripheral vision is very bad at detail, and because
the blind spot is in the periphery of vision, textures can be mapped
over the blind spot very crudely, with no loss of fidelity.
Charles II liked winking at girls. He had a reputation for that sort
of thing. But if he winked at you, there was always the possibility
that he was thinking how you would look without your head.
* * *
190 A NATURAL HISTORY OF SEEING

William Bowman’ drawing of the optic nerve head. Bowman,

made a baronet by Queen Victoria for his services to medicine,
went on to found the Ophthalmological Society, which later
became the Royal College of Ophthalmologists.

The nerve head is blind: at this busy junction there is no room for
the apparatus of seeing.
So where exactly is this apparatus?
The Dutch microscopist Antoni van Leuenhoek hinted at the presence
of nerve fibres elsewhere in the retina, but the Italian naturalist and
physiologist Felice Fontana (1730-1805) is generally given the credit
for spotting them first. Fontana is best remembered for conducting
the first human brain stimulation experiments. By applying electricity
to specific regions of the brains of corpses, Fontana had been able
to make their faces spasm in interesting ways. When a law was passed
forbidding his work with the dead, Fontana advertised for — and got
- live volunteers. If Fontana had not existed, it would have been
necessary for Hammer Films to invent him: Fontana’s love of anatomy
was such that he used to bring the bits he was studying to the dinner
table - a habit not without its risks, since among his discoveries were
the poison sacs of the viper.
As he was examining a rabbit’s eye with a microscope, Fontana
spotted whitish bands running across the back of the eye, terminating
in a fibrous tangle at the nerve head. This was the first good anatom-
ical evidence for the retina’s visual function. Unfortunately, close
examination of other vertebrate eyes did not always throw up similar
NERVOUS MATTER, VISUALLY ENDOWED 191

findings. The human eye was particularly unenlightening: following

Scheiner’s meticulous drawings, we find that the image in a human
eye is brought to focus not at the optic nerve, but at an area called
the macula. The macula is a smooth depression in the retina,
eminently suited to the role of seeing — except that there is no tell-
tale white streak. At the centre of the macula, and barely the size of
a pinhead, lies the fovea. But why should vision be concentrated
upon such an absurdly tiny point, when the rest of the retina is
clearly equipped for seeing?
It was the German naturalist, Gottfried Reinhold Treviranus
(1776-1837) who, looking through a microscope in 1834, spotted a
layer of densely packed cylindrical structures in the retina. Assuming
(correctly) that these must be visual cells - photoreceptors, in today’s
terminology - he then had to explain how they functioned. The
trouble was, in the heat of observation, he believed he had got his
sample back to front. This seemed the only possible explanation for
what he had seen: an array of highly specialised visual cells, covered
with a mat of blood vessels. This arrangement was obviously nonsen-
sical; Treviranus duly reported that his newly observed photore-
ceptor cells must lie above the blood vessels, pointing up into the
vitreous humour.
Treviranus had been right the first time. Three subsequent studies,
by G.G. Valentin (1810-1883), J. Henle (1809-1885), and E. Briicke
(1819-1892), showed that his cylindrical structures really did lie
beneath a mat of blood vessels. These three then drew a conclusion
as irresistible as it was wrong: since their view was obscured by blood
vessels, these structures could not be visual cells, after all.°
Well into the nineteenth century, the best anatomists (and Treviranus
was certainly one) were doomed to chase each other's tails, as they
tried to unpick a structure too fine for their instruments to handle.
Strange then, that the experiment that finally settled the matter
should be the ophthalmoscope - a machine through which one could
not see the retina at all.

If by any chance you have an eye test looming, you may want to pay
particular attention when your optometrist lights up the ophthal-
moscope and brings it to bear on your eye. As light from the
192 A NATURAL HISTORY OF SEEING

instrument enters your pupil at an oblique angle, the view of the

darkened surgery will momentarily vanish in an orange-red haze,
crazed by the tangled shadows of blood vessels. This is “Purkinje’s
tree, an indirect glimpse of the eye’s blood supply, which Jan Purk-
inje first conjured in 1823.
A glimpse of Purkinje’s tree is as eloquent a demonstration of the
vertebrate eye’s oddity as you could wish for - what are blood vessels
doing in front of the seeing parts of the eye?
A further mystery looms when the ophthalmoscope is moved
about. As the light source moves, the shadows of the blood vessels
shift with the change in the angle of the light. This means that the
blood vessels lie some distance away from the seeing part of the eye.
What fills this apparently empty space? One obvious candidate
presents itself: the retinas elusive ‘nervous matter, so easy to spot in
the white ‘visual streaks’ of rabbits, but so oddly absent in the human
eye. If the nervous matter connecting up the retina was excessively
thin and optically neutral, then this would explain why it could not
be spotted with microscopy, but only inferred from the waving
branches of Purkinje’s tree.
Purkinje’s tree demonstrates that the seeing parts of the retina lie
underneath two layers of tissue: a layer of blood vessels, and a layer
of invisible nerves. Strange as the idea seemed, it was more or less
accepted by the time the German anatomist Heinrich Miiller found
a way to use fixatives to harden the retina. Studying the retina in
cross-section, Miiller found never-before-seen nerve layers between
the retina’ blood vessels and its photoreceptors. Unfortunately, only
Miiller could see them. Even his star pupil, Hermann von Helmholtz,
could not be induced to see what Miiller saw down the same micro-
scope. For a while, these nerve layers were as controversial and
contested as Percival Lowell’s Martian canals, and for the same reason:
because they lay on the very edge of visibility; fancy and expecta-
tion played as large a part in their mapping as actual observation.
The retinas true complexity was revealed only in 1870, when Max
Schultze (1825-1874), the German biologist and director of the
Anatomical Institute at Bonn, made exquisitely detailed cross-
sectional drawings. Schultze’s retina, though literally paper-thin,
boasted no fewer than ten neural layers.
4

NERVOUS MATTER, VISUALLY ENDOWED 193

Blood vessels lie over layers of
nerve tissue, and layers of nerve
tissue lie over the visual cells -
this arrangement is common to
all vertebrate eyes. The problem
now was to explain how such a
Heath Robinson arrangement
could work.
Enter (yet again) Jan Purkinje.
An early riser, Purkinje
would often wake before dawn
and take a walk before break-
fast. Everyone seems to have
their own story about when and
how he noticed the optical
phenomenon known as the
‘Purkinje shift. Some say that,
as dawn approached, he walked
past his favourite geraniums
and noticed with pleasure how
the dark red blooms waved
against a ground of pale green
leaves. On his return he walked
past the same flowers lit by the
SCHULTZE RETINA morning sun; now, pale red
Max Schultze’s 1872 drawing of blooms stood out brightly
the layers of the retina. against dark green foliage.
Other accounts have it that the
patterns of his bedroom carpet seemed altered at different times
of day. The chances are that all these stories are true: Purkinje
was a superb, eclectic observer.
Purkinje found that changes in light level affect how colours register
on the eye. At dusk, blues, greens and yellows are brighter, while colours
at the opposite end of the spectrum (reds and oranges) are dim. In
bright daylight, the difference in brightness is levelled out and, in some
cases, reversed, so that blues seem dim in comparison to reds. Purk-
inje drew the inescapable conclusion, published in 1825, that the eye
194 A NATURAL HISTORY OF SEEING

contains two sets of visual equipment, one for daylight vision and one
for dusk and dawn.
Where Purkinje’s new-fangled experimental psychology led,
anatomy eventually followed. In 1866, Max Schultze distinguished
between two distinct types of visual cell in the retinas of birds.
Because rod-shaped cells were more abundant in the eyes of
nocturnal birds, Schultze proposed that rods were adapted for seeing
in poor light. Birds that operated during the day had very few rods,
but many more cones. So cones were the cells that saw in good light.
Considering the human retina, which contains healthy populations
of both rods and cones, Schultze was able to improve upon Purkinje’s
explanation of his ‘shift. Rods enable vision at night; they are insen-
sitive to colour, but respond in some degree to all visible wavelengths
of light, giving us monochrome vision. Cones give us colour vision
during the day, because each type of cone is sensitive to a different
wavelength. If we average their sensitivities, we find that cones, as a
class, react to longer wavelengths more strongly than rods do. During
daylight, a red object appears lighter than a blue object emitting the
same amount of light. At dusk, as we swap visual attention from
cones to rods (a neat piece of visual orchestration and one that is
still largely mysterious), the red object dims and the blue object
appears brighter.
Because red is not a vivid colour in poor light, it is quite a bad
choice of livery for emergency vehicles. Vehicles with green and
yellow livery suffer fewer accidents. By the same token, red is an

100 rod vision cone vision

& 60
&
& 40
c
& 20

400 500 600 700

Blue Green Yellow Red

Wavelength (nm)
NERVOUS MATTER, VISUALLY ENDOWED 195

excellent colour for the instrument displays of ships and aircraft. In

the dark, it is important that pilots can read accurately without losing
their night vision. Since their rods are effectively blind to red light,
pilots can read red-light instruments with their cones without
blinding their night-adapted eyes.
From his work with birds’ eyes, Schultze knew that cones and
rods, in whatever proportion they happened to be present, were not
mixed together willy-nilly. Cones tended to gather at the focus point
of the retina, pushing the rods to the sides. (Some accounts say that
there are no rods in the fovea. This isn't strictly true: only in the
very centre of vision do cones push rods away entirely.)
If you are star-gazing, it is best to look very slightly to one side
of the star you are studying. Your rods are much more sensitive than
your cones, and will be able to pick up the object quite easily. Look
at the star directly, and its image will fall on the dense nest of cone
cells packing the fovea, so the chances are you wont see a thing.
Once we know that day- and night-time vision use different
regions of the retina, a number of ‘disadvantages’ to the back-to-
front retina fall away.
Yes, the rods are overlaid with blood vessels, and yes, the vessels
certainly get in the way, but (saving ophthalmoscopic tricks) we can
no more see our blood vessels than a camera, pressed up against a
chainlink fence, registers the chainlink. To set against the slight blur
that they add to an already blurred image, the blood vessels give the
vertebrate eye the considerable advantage of an extra food supply.
Schultze noted that the optics of primate eyes perfectly focus just
one degree of space. At this point, the retina is bent into a fovea -
4 shallow dish crammed with cones. Blood vessels, which overlie the
retina everywhere else, do not intrude on the fovea. The area's relative
thinness and lack of complexity — it is little more than a single layer
of naked photoreceptors - means that it can be more than adequately
fed from behind, by the choroid layer.
But the fovea’s very simplicity is a source of mystery. Where are
its neural layers? Why is the seat of focused vision also the thinnest,
simplest part of the retina?
196 A NATURAL HISTORY OF SEEING

The Scots physicist James Clerk Maxwell probably contributed more

to science than any other Victorian philosopher. Legends abound: it
is said his work on the optics of index-graded lenses in fish was inspired
by a breakfast of kippers. Occasionally scientific curiosity got ahead
of him, as this passage, from a letter to his wife, Katherine, reveals:

There is a tradition in Trinity that when I was here I discovered

a method of throwing a cat so as not to light on its feet, and
that I used to throw cats out of windows. I had to explain that
the proper object of research was to find how quick the cat
would turn round, and that the proper method was to let the
cat drop on a table or bed from about two inches, and that
even then the cat lights on her feet.’

If his list of key contributions omitted cats, it nevertheless covered

fields as diverse as optics and liquid dynamics. He also made major
contributions to our understanding of the fovea.
Maxwell was particularly interested in an unusual observation
made ppthe inventor Samuel Soemimenne in 1795. Soemmering
noticed that the fovea
and its immediate
surrounds (an area
called the macula)
turned bright yellow
when removed from the
eye. At first, it was
assumed that this
startling colour change
was caused by contact
with the air. Maxwell
wondered whether
there was not another
explanation. It could be
that the structures and
tissues that normally
overlaid the macula
James Clerk Maxwell. protected it from a
NERVOUS. MATTER, VISUALLY ENDOWED 197

particular wavelength of light. Once they were removed, this wave-

length triggered the maculas change of colour.
Maxwell identified this wavelength by looking through a prism at
a long vertical slit. As he tilted the prism, an elongated dark spot ran
up and down the blue part of the spectrum, but could not be persuaded
to pass through the other colours. The position of the dark spot
indicated which wavelength of light was being blocked by the eye.
Maxwell’s interest was further aroused by a report written by
Wilhelm Karl von Haidinger (1795-1871), director of the Imperial
Institute of Vienna. Haidinger described how, while studying a
mineral sample through a polarising glass, he noticed a dark propeller
shape hovering above the paper laid across his desk. When he rotated
the filter above the paper, the propeller turned, too.
The figure set natural philosophers an irresistible challenge:
another Scots physicist and inventor, Sir David Brewster and the
mathematician George Stokes both essayed an explanation of
‘Haidinger’s brushes’ as the propeller figure came to be called. But
Maxwell had an advantage: he already knew that there was a structure
in the eye that filtered out violet light.
To explain a puzzling visual phenomenon, Maxwell could usually
be counted on to come up with a piece of apparatus, preferring demon-
stration to argument. At a meeting of the British Association in 1850,
he produced ‘a small piece of apparatus which he had made by
cementing together . . . sectors of sheet gutta-percha, these being so
cut out of the sheet and put together that its fibrous structure .. .
should radiate, at least approximately, from a central point. This
arrangement ... reproduced, or rather simulated, Haidinger’s phenom-
enon.® His accompanying explanation went something like this:
Haidinger had been studying his sample through a polarising filter.
That meant the light entering Haidinger’s eye was polarised; all the
light waves vibrated along the same plane. The dark brushes
Haidinger noticed suggested that there was a second polarising filter,
radial in structure, crossing his vision, and this could only be inside
his eye, lying over the visual cells. When Haidinger rotated his polar-
ising filter, he changed the angle at which the light in his eye was
polarised, and the filter in his eye blocked the light along a different
diameter; so it, too, appeared to rotate.
198 A NATURAL HISTORY OF SEEING

The subtlety of Haidinger’s phenomenon held a further, final clue

as to the nature of the filter in the eye. Haidinger’s brushes did not
block all the light. The brushes were not black; they were the faintest
of shadows. So, the filter did not block all the light, just a portion.
Maxwell, calling to mind his prism experiments, suggested, first, that
the filter blocked only violet light, and second, that the active ingredient
in the filter was macular pigment — the same pigment that Soemmering
saw turn bright yellow when the macula was exposed to daylight.
Over the photoreceptors in the macula lay a single layer of
incredibly fine fibres, radially arranged like the spokes of a wheel
(or the gutta-percha fibres in his home-made apparatus). If these
fibres contained macular pigment, then they presumably offered the
seeing cells some modest protection against violet light.
So much for solving the mystery of Haidinger’s brushes. The pres-
ence of radial fibres lying over the fovea was more suggestive still: it
indicated how the fovea, so naked and unadorned, was attached to the
rest of the retina. What if those very fine fibres were nerve fibres?
If they were, the fovea made considerably more sense. It was indeed
attached to the rest of the retina, but as much overlying material as
possible had been swept to the sides, so the cones could enjoy an
unimpeded view. The fine, virtually invisible, fibres connecting the
fovea to the rest of the retina also doubled as a filter, protecting the
photoreceptors from ultraviolet light.

There is a discipline to good observation — one that fine artists spend

their lives acquiring. Had the astronomer Percival Lowell been an artist,
it is doubtful he would ever have persuaded himself into seeing canals.
As it was, in 1895, his book Mars was published. ‘Certainly, Lowell wrote,
‘what we see hints at the existence of beings who are in advance of, not
behind us, in the journey of life? His reputation never recovered.
In the same year a Spanish anatomist, Ramon y Cajal, on becoming
a member of the Royal Academy of Sciences in Madrid, made an
announcement that, for his own field, was just as shocking: the
nervous system, he said, was made up of individual cells. Cajal (1852-
1934) was born in the foothills of the Spanish Pyrenees. A vandal,
hell-raiser and passionate artist, he would vanish into the hills for
days with stolen brushes, paints and paper. With Cajal around, no
NERVOUS MATTER, VISUALLY ENDOWED 199
whitewashed wall was safe. His father Justo, a physician, was deter-
mined to stop his son frittering his life away in Bohemian poverty
and sent him to a school famous for its beatings and short rations.
To keep him out of trouble outside school hours, he was appren-
ticed first to a barber and then to several shoemakers.
Cajal spent some years as an army physician. In Cuba, he contracted
malaria, then tuberculosis, and returned to Spain in poor health. His
father, by then professor of dissection at the University of Zaragosza,
got him work as a demonstrator of anatomy. But Cajal had never ceased
to be an artist. Now, at last, he found his proper subject matter: the
structures and tissues of the body. In 1877, ‘using every peseta saved
from the service in Cuba, Cajal bought an old-fashioned microscope.
Why he chose the workings of individual tissues and cells for his
life's work is simple to explain. His health was poor, his finances meagre.
On these small structures, he could work cheaply, and from home.
Which, by all accounts, was no hardship. His marriage to Silveria Fahanas
Garcia was a happy one: as well as his seven children, five of whom
survived into adulthood, we have the testimony of a friend who declared
‘half of Cajal is his wife? We can only regret that, despite several engaging
autobiographical writings - including a charming late memoir subti-
tled ‘Reflections of an arteriosclerotic — Cajal maintained the conven-
tional decencies of the period and said not a word about his family.
Cajal’s interest in the nervous system was first provoked in 1887.
The psychiatrist Simarro Lacabra, knowing that Cajal was preparing
a book on laboratory technique, thought it worth showing him some
nerve tissue stained by a new method invented by the Italian neuro-
anatomist Camillo Golgi. Golgi’s stain employed gold and silver
compounds that hardened within the tissue to be studied, leaving a
visible tracery. Golgi however, never quite mastered his own method:
his staining agent might stain one type of tissue in one sample, and
quite another type in another. Frustrated, Golgi nearly gave up the
use of his own invention.
Cajal turned the stain’s fickleness to his advantage. So what if the
stuff stained one tissue, then another, then another? Was it not
obvious that, each time this happened, different kinds of tissue were
being stained? For Cajal, ‘a look was enough . . ideas boiled up and
jostled each other in my mind. A fever for publication devoured me.
200 A NATURAL HISTORY OF SEEING

Before Cajal, most observers, faced with an undifferentiated mesh

of uniform nerve fibres, had taken what they saw at face value and
reported that the nervous system was simply a giant web or ‘reticulum.
Cajal teased out the inter-relationships between nerve cells and traced
the interweavings of neural communities. There is no question that
he was vitally aided by Camillo Golgi’s stain, but it was the quality
of Cajal’s observation itself - his artist's eye - that enabled him to
lay out the nervous system, revealing that it was made up of many
independent, but interlinked, cells.
The identification of many different kinds of nerve cell was
provoking enough, but Cajal went further, showing how the nerve
cells receive electrical impulses through incoming fibres, dendrites,
and conduct signals to distant locations through outgoing fibres,
axons. This meant that the nervous system was to be understood
not just through the behaviours of cells, but through the relationships
between cells of different kinds.
Golgi, a lifelong adherent to the idea of the reticulum, found Cajal’s
work outlandish. The two men met, for the only time, at Stockholm

Cajal’s drawing of a dog’s retina shows rods (a) and cones

(b) connecting to several different kinds of bipolar cell. Draw-
ings like this put paid to the idea that the nervous system was a
mesh of undifferentiated fibres.
NERVOUS. MATTER, VISUALLY ENDOWED 201

in 1906, when they shared the Nobel Prize for Physiology or Medicine.
This was the first time a Nobel prize had been shared, and the
Caroline Institute’s decision to split it between the two men in this
way was almost scandalous.
‘I thought that [Cajal] had deserved receiving a full, and undivided
Nobel Prize, Gustaf Retzius, a former member of the institute,
remembered in his autobiography. ‘[A]sked about this by the Nobel
Council of the staff of professors at the Caroline Institute, I expressed
this opinion of mine decidedly; he writes, resorting to angry italics
even after an interval of forty-two years.°
It is enlightening to compare the speeches of the two men. Golgi
is clearly a product of the nineteenth century, whereas Cajal’s speech
might have been written yesterday. As the Swedish histologist Emil
Holmgren reported to the Nobel Committee, ‘Cajal has not served
science by singular corrections of observations by others, or by
adding here and there an important observation to our stock of
knowledge, but it is he who has built almost the whole framework
of our structure of thinking’
Though Cajal was always meticulously polite about Golgi and
generous about his achievements, this was the civility we expect
victorious youth to show to redundant old age. Golgi knew it, and
hated it. When they received their award, Golgi spoke first, and used
the occasion to defend the reticulum while at the same time stealing
as much of Cajal’s glory as possible. He convinced no one.

3 — Contrast

Every light is a shade, compared to the higher lights, till you

come to the sun; and every shade is a light, compared to the
deeper shades, till you come to the night.
John Ruskin

Ernst Mach (1838-1916) called himself a physicist. In 1864 he took

a job as professor of mathematics in Graz, and in 1866 he was also
appointed professor of physics. He is best remembered for working
out what happens when the speed of a travelling object exceeds the
202 A NATURAL HISTORY OF SEEING

speed of sound, and the speeds of supersonic aircraft today are

commonly given in multiples of the speed of sound: “Mach numbers.
Mach’s early years in Vienna in the 1850s were dogged by lack of
funds. He was hardly poor, but he could not afford to fund purely
physical researches. Less well explored - and for that reason, more
amenable to cheap and cheerful experiment - were those still-unla-
belled areas where physics, physiology, and the psychology of sensa-
tions overlapped. In 1867, a professor and newly married, Mach
made a series of simple observations, uncovered an underlying mech-
anism of vision, and transformed our understanding of vision,
perception and the nervous system.
His breakthrough was to explain an optical illusion generated by
a series of grey bands, each one a little lighter than its left-hand
neighbour. (Mach used black-and-white spinning tops to create very
precise shades of grey, but his findings are now more usually
presented in printed form.) Each “Mach band’ is evenly printed,
without grain or shading, and yet each band appears fluted. It is as
though the bands have been lit from the side; the edges lying against
darker neighbours appear lighter, while edges lying against lighter
neighbours appear darker, as though in shadow. The fluting is an
illusion — but why should the eye manufacture dark where there is
no dark, and light where there is no light?
Mach realised that the eye is interested in boundaries. By
exaggerating the contrast between neighbouring bars, the eye reveals
7

NERVOUS MATTER, VISUALLY ENDOWED 203

the line along which they join. Mach worked out a detailed
mathematical model of this process, suggesting how neighbouring
points on the retina would interact with each other to create the
Mach band illusion.
Of course, it was only a model; there was no guarantee that real
structures in the retina behaved that way.

The American physiologist Haldan Keffer Hartline (1903-1983)

received his medical degree in 1927, on the solemn understanding
that he should never attempt to treat patients.
This was Hartline’s version, anyway, and it is typical of the man’s
dry humour and eccentric reputation that he was widely believed.
His laboratory was once described by his long-term collaborator,
Floyd Ratliff, as a ‘slightly disorganized but extremely fertile chaos,”*
and if he wasn't to be found there, then he was most likely engaged
in outdoor adventures of one sort or another: though slightly built,
Hartline was an athlete, with several first ascents in the Wyoming
Rockies to his name. He also enjoyed sailing, and flew his own open-
cockpit plane.
Though he once shared lab space with George Wald, the man who
discovered how the retina used vitamin A, and sailed with Roger
Granit, who found out much about the retina’ electrical behaviour,
the three men never collaborated. It took the Royal Caroline Institute
in Stockholm to bring them together, when it awarded them the Nobel
Prize in 1967.
In the 1920s, when Hartline began his studies, the discipline of
anatomy was at an impasse. On the one hand, new tools were being
developed to explore the workings of the nervous system - the last
great anatomical unknown. ‘Listening in’ to the electrical activity of
the nervous system was, according to the British medical pioneer
Edgar Douglas Adrian (Lord Adrian of Cambridge), ‘a new, very
powerful microscope to work with. On the other hand, if the things
under study are inadequately prepared, nothing useful will be
revealed, however good the microscope. In 1927, Adrian and his
co-worker Rachel Matthews had managed to record the electrical
activity of the optic nerve of a conger eel. This was an astonishing
technical achievement, but the recording was meaningless, a white
204 A NATURAL HISTORY OF SEEING

noise made up of the chatter of hundreds of thousands of individual

fibres.
It had been known since 1865 that the retina reacted electrically
to light, but to really unpick how the retina behaved would take
recordings of single nerve fibres — and how could one possibly isolate
a single fibre?
Hartline and his collaborator Clarence Graham claimed that it
was by pure good fortune that they chose the horseshoe crab for
their research. If true, this was definitely a case of chance favouring
the prepared mind: it would not have escaped their notice that the
horseshoe crab was ancient, a ‘living fossil’ that had been swimming,
crawling and spawning in the Earth’s tidal shallows for at least 250
million years. The closest living relative to the crystal-eyed trilobites,
horseshoe crabs boast coarsely faceted compound eyes, whose
photoreceptors are a hundred times the size of human rods and
cones (the biggest of any known animal), and optic nerves that can
be frayed into thin bundles which are easy to split down to a single
active fibre.
In 1932, Hartline and Graham managed to record the activity
of individual fibres. They found that every nerve signal was iden-
tical. This meant that any information the nerve contained was
being conveyed purely by its level of activity; the shape or amplitude
of each signal was irrelevant. However, other, more anomalous
results, suggested that something peculiar was happening to infor-
mation at the level of the retina.
When Hartline began work on the horseshoe crab’s compound
eye, he assumed that each individual ommatidium would work
independently. But he soon noticed that extraneous lights in the
laboratory, rather than increasing the rate of fire of a receptor, often
caused it to fall silent. Why should light stop a receptor from firing?
Hartline realised that the ommatidia were not independent. Each
acted in concert with its neighbours. If one ommatidium was brightly
lit while its neighbour was dimly lit, the dimmer signal became even
weaker than normal. The result was a greater difference between the
two signals - in other words, an enhancement of contrast.
Contrast seemed to be the only fact about the illuminant which
found itself translated into electrical signals. Turn the lights up, and
NERVOUS MATTER, VISUALLY ENDOWED 205

edge line bar

Mutual inhibition is evident in these sketched responses of

fields of ganglion cells, confronted with different kinds of edge.

the nerve fibres grew excited - for a moment - before settling back
to their rest state. Turn the lights down and the same thing happened.
At no point did the nerve fibres seem to care about the actual level
of illumination. Hartline was impressed: this ‘contrast only’ form of
vision meant that the eye could detect small local variations in light
intensity, even though ambient light varies a million-fold between
sunlight and starlight.
Mach’ conjecture had turned out to be correct: contrast is accen-
tuated in the eye by a process Hartline dubbed ‘mutual inhibition.
Imagine a group of six photoreceptors, all attached to a single
nerve cell. It is dark. The photoreceptors are silent and so is the
nerve. The light goes on. All six receptors fire. For a moment, there
is a lot of activity, until the nerve cell issues a six-fold inhibiting
signal to its pool of receptors, and everything falls silent again.
Now, imagine that light catches just one receptor. The receptor
fires, but the other five, still in darkness, stay silent. The nerve cell
responds by sending a single inhibiting signal, but this one signal is
shared between all six receptors. This means that the lit photoreceptor
will remain fairly active while its quiet neighbours are effectively
damped. An illumined receptor whose neighbour is in the dark will
give a stronger signal than a receptor surrounded by other illuminated
receptors, and a poorly illuminated receptor will have its already
small signal suppressed to virtually nothing if it lies next to a well-
illumined cell.
Now imagine a million nerve cells, each one overlapping its neigh-
bour, each performing the same trick, making the dark side of an
206 A NATURAL HISTORY OF SEEING

edge darker and the light side of the an edge lighter - and you will
start to get some idea of how cells that inhibit each other actually
make vision possible.
Computer modelling can give us a good visual idea of what this
system achieves, as David Marr found during his ground-breaking
(a)

How the retina sees: David Marr filtered these images

using responses typical of nerve cells in the retina. Areas
of detail are revealed, while patches of constant illumination
are rendered an even grey.
NERVOUS MATTER, VISUALLY ENDOWED 207

work at the Massachussetts Institute of Technology. These images,

from Vision, his celebrated account of that research, shows what
happens to a black-and-white photograph when it is filtered using
the kinds of responses typical of a particular kind of retinal nerve
cell, called a ganglion cell. Large expanses of light and dark in the
original photographs (areas which tend to reflect absolute light levels,
but little else) are effectively evened out to a medium grey while
areas of high contrast are massively exaggerated, even to there being
flanking strips of white to the black wires of the chainlink fence.
Rendering a nice photograph down to something that looks suspi-
ciously like a poor photocopy does not look like much of a trick.
We should remember, however, that the eye has evolved to report
salient features, not pretty pictures. By reporting only the lines of
contrast, the retina avoids having to prepare endless, uninteresting,
and massively redundant reports about plain surfaces.
We receive only local contrast information from the eye, and infer
shading at our leisure. This point was demonstrated quite spectac-
ularly in 2003, by Dr Steven Dakin of University College, London,
when he filtered the iconic image of Cuban revolutionary Che
Guevara to remove everything but the contrast information.’? Look
very closely at this image, and you will see that the lit parts of Che's
face are exactly the same shade of grey as his beard and facial
shadows. The lines alone reveal which side of each edge is supposed
208 A NATURAL HISTORY OF SEEING

to be bright, and which is supposed to be dark. Whether we look at

the original black and white image, or the filtered image, we are
using contrast cues to recover patterns of light and shade. In the
discrimination of a scene, contrast is everything.
Enthused by his findings, Hartline turned from the eyes of horse-
shoe crabs to the eyes of vertebrates. The vertebrate eye is much
more complex and compact, and the technology needed to tease out
and listen to a single fibre from a vertebrate optic nerve did not yet
exist. But Hartline realised that this was not a problem. The optic
nerve was a conduit; all the really interesting activity took place at
the retina, where the optic nerve was already spread out in a thin
layer. The retina was a ready-made dissection of the optic nerve.
Through a combination of canniness and relentless practice, Hart-
line eventually teased out single nerve fibres from a frog’s retina. He
had expected the nerve cells to mutually inhibit, much like the nerve
cells in the retina of a horseshoe crab. The reality was much more
exciting: the frog’s nerve cells reacted in many different ways to the
light. Along with cells that reacted to contrast, there were cells that
responded to changes in ambient light intensity, regardless of whether
the light grew brighter or darker. Others reacted only when the light
was dimmed; still others responded to the movement of dark spots
or shadows. As the list grew longer, it became apparent that all these
cells were, at heart, operating like the cells in the horseshoe crab
retina. The difference was that the horseshoe crab processed its visual
information once, across a single layer of nerve cells. The frog, on
the other hand, boasted several layers of nerve cells, and visual infor-
mation, analysed several times by successive layers, was processed
to reveal more and more about the world: where objects were, which
way they were moving, maybe even what colour they were. Individual
nerve cells never act independently, Hartline explained, during one
lecture; ‘it is the integrated action of all the units of the visual system
that gives rise to vision?
Before Hartline, it was usually assumed that the optic nerve carried
just a single sort of visual information — a picture, if you will. Now
we realise that even before visual information leaves the retina, it is
being integrated and analysed for significant features. Different kinds
of information are processed by different kinds of cell, and carried
NERVOUS MATTER, VISUALLY ENDOWED 209

down the optic nerve along distinct pathways to distinct areas of

the brain.
World War II interrupted Hartline’s work, and its value was not
truly recognised until the early fifties. Since then huge strides have
been made: cells in the frog retina were discovered which were
capable of ever more complex responses; other animals were studied,
and people began to appreciate how specialised each animal's retina
was; cells which appeared identical under the microscope served
completely different functions in different retinas, depending on how
they were strung together.
Why does visual information need to be manipulated so soon -
almost the moment light touches the eye? Why can’t it simply be
piped down the optic nerve to the brain? A quick look at the numbers
reveals the answer. There are 126 million photoreceptors in the
human retina, but only a million or so fibres in the optic nerve. Were
every photoreceptor connected to the brain by a nerve fibre, the
optic nerve would be as thick as the eyeball. This would be an horren-
dous logistical problem for any species, and particularly ruinous for
humans, since it would make it impossible for us to move our eyes.
The more salient the information coming from the retina, the less
of it we need. One million well-edited signals are better than 126
million chaotic ones. How the editing is done depends on the time
of day. The retina contains two separate mechanisms of vision (one
might almost say that it is two retinas in one). One is for night
(where sensitivity to any and all light is essential) and one is for day
(where light of different wavelengths can be processed to generate
colour vision).
At night, visual information is gathered from the rod receptors,
which lie across the whole field of vision. Although there are one
hundred and twenty million rods in the eye - rods are twenty times
more numerous than cones - they do not generate twenty times the
traffic down the optic nerve. To efficiently detect very faint levels of
illumination, the rods are connected by nerve fibres into fewer, larger
units. By pooling signals, the rods effectively become fewer in number
and larger in extent. With so many rods merging their information
before piping it down a single fibre of the optic nerve, there is a
good likelihood that even small amounts of light will be gathered
210 A NATURAL HISTORY OF SEEING

and registered. The price paid for this sensitivity is fuzziness: fields
of rods are much larger than single photoreceptors and will inevitably
produce an image with a coarse grain.
This balance between sensitivity and acuity is finely managed.
Closer to the fovea, rod fields are smaller, as they attempt to catch
fine detail; towards the periphery, the fields are larger, to detect the
very faintest objects and movements. What is more, the arrangement
is dynamic; as light levels fluctuate, the fields change size: at the
threshold of human vision, pools of up to a thousand rods enable
us to spot sources so dim and distant, it is said we can see the light
of a single candle shining seventeen miles away."
Daylight vision, on the other hand, prioritises information from
the centre of vision, at the expense of peripheral vision. Rather than
condensing information from the fovea, nerve cells may carry infor-
mation from single cone cells straight to the optic nerve. (The ‘grain
of the field of view is at its finest here.) It is usual for a cone to be
connected to more than one set of nerve cells; information from the
cones is duplicated, and the fovea, which occupies a mere half of
one per cent of the retina’s surface, produces information so impor-
tant that it takes up forty per cent of that part of the brain which
first receives visual information.
Hartline and his successors produced an intricate picture of how
vision works but it was a monochrome picture, a picture suited to
night-time, and deep water, and eyes that struggle to distinguish
form within the murk. Eyes adapted to take advantage of bright light
boast complexities Hartline could not possibly have hoped to unpick.
Avoid vertebrates because they are too complicated} Hartline
advised his students, ‘avoid colour vision because it is much too
complicated, and avoid the combination because it is impossible?
EIGHT

Seeing colours

Whenever I mention that I'm writing a book about eyesight, I am

invariably treated to the same story: how one day, asa child, it dawned
on the person I am speaking to that their experience of colour might
be entirely their own. Who is to say that my experience of green is
the same as yours?
The near-infinity of the colour space we perceive makes colour
a mysterious and personal sensation. We don't all see the same
colours, any more than we all have identical toothaches, but just as
we can agree on what a toothache is before comparing dentist stories,
so we can all agree on what ‘orange’ and ‘purple’ and ‘green are before
we go on to argue about the precise hue of the plaid our friend is
wearing. We never agree entirely, but we always agree enough.
We can all match objects of the same colour and some are better
at this than others. Many chemists can discriminate between colours
as finely as wine connoisseurs discriminate between bottles of
Pauillac. For now, let us agree that there is sufficient orderliness to
colour for us to group objects together in consistent ways, and agree
with each other’s choices most of the time. Where we disagree, the
disagreements are not crazy or puzzling. A colour-blind man (for
reasons of heredity, it is usually a man) may put green and red objects
22: A NATURAL HISTORY OF SEEING

together, but that’s about as wild as it gets. No one has ever come
up with unique multi-coloured categories. We map colour space in
slightly different ways, but we are all mapping the same colour space.
Where does this orderly colour space reside? Are objects coloured?
Does light come in different colours? Do we make up colours in our
heads? Is colour a mere linguistic convention?
The first part of this chapter attempts to answer these and similar
questions. The rest of the chapter is an historical account of colour
vision. There is certainly a conventional way of telling that history,
but that may be no bad thing according to John Mollon, professor
of visual neuroscience at Cambridge:

Each newcomer to the mysteries of colour science must pass

through a series of conceptual insights. In this, he or she reca-
pitulates the history of the subject. For the history of colour
science is as much the history of misconception and insight as
it is of experimental refinement.'

My potted version of that history ends in 1973, with Edwin Land’s

‘retinex’ theory of colour vision; then, in the final section, we leap-
frog back to the workshops and studios of London and Paris in the
1880s, for it was there, among the artists of the Divisionist or pointilliste
school, that experiments in colour perception laid the groundwork
for the most fashionable field of current vision research — the study
of visual attention.

1 — Colours and words

I wish you ... could eradicate the insane trick of reasoning

about colours as identified by their names. People seem to
think that blue is blue, and one blue as good as another.
C.J. Monro to James Clerk Maxwell, 3 March 1871

What colour are the walls in your kitchen? Are they nol? Or are they
wor? To help you, nol is the word the Burinmo hunter-gatherer people
of Papua New Guinea give to a sort of greenish-bluish-purple and
SEEING COLOURS 213

wor is yellow. Well, yellowish-orange, with hints of brown and green.

Which, funnily enough, is as good a description of the walls in our
kitchen as any in the paint catalogue. (We got rather over-excited
with our colour choices when we moved in.)
Are our kitchen walls wor? I'm not sure. Even if you showed me
something certifiably wor-coloured, I might not be able to pick out
the colour again. English lacks a name for this colour, and with no
easy label, it's devilishly difficult for me identify it. This was the finding
of Jules Davidoff, a psychologist at Goldsmiths College, London, who
is studying how language affects our perception of colour.
The Burinmo language distinguishes between nol and wor, but
makes no distinction between blue and green. Shown a green and
asked to memorise it, the hunter-gatherers found it difficult, after an
interval of only a few seconds, to select it from a tray of other blues
and greens. It’s not that we don't see the colours we have no names
for; it's simply that the colours we can name are lodged in our memory
in a way that others are not. Nameable colours are the beacons by
which we navigate colour space. The more names we learn, the more
ably we navigate. This was brought home to me, with some force, as
I researched this chapter. I have never, until recently, been able to
identify indigo. Although I’m not alone in this — about half the people
I spoke to about it share my incapacity - this failing has always
niggled me. There are, according to no less an authority than Isaac
Newton, seven colours in the rainbow, and I can only see six. At the
blue-green end of the spectrum, there is supposed to be a band of
indigo between the blue and the violet - so why has it been so hard
for me to spot?
Indigo is an example of a colour word that has fallen out of fashion.
It’s in the dictionary, but hardly ever crops up in conversation. Had
you pointed at something indigo, until last week I'd have seen a dark
blue. Questioned further, I would have happily conceded that there
was an element of red in the blue, but didn’t that make indigo just
a fancy name for purple? Not any longer. Research for this chapter
has piled my desk with any number of examples of different colours,
arranged in any number of colour systems. I’m surrounded by a veri-
table swatch table of indigos. And because they’re all helpfully labelled
‘indigo, the colour has begun to lodge in my memory. Last week I
214 A NATURAL HISTORY OF SEEING

walked into a Sainsbury's supermarket, and what did I see but Sains-
bury’s store livery: orange on-~ miracle! - indigo.
So how many colours are there in a rainbow? It depends to whom
you speak: Newton, it must be said, had a vested alchemical interest
in there being seven distinct colours. Though I recognise indigo, I
can't honestly say that it makes a distinct band between blue and
violet. Similarly, the orange and red bands seem to me to be so
hopelessly merged that there is no separating them.
The Greek poet Homer praises:

Jove’s wondrous bow, of three celestial dyes,

Placed as a sign to man amid the skies

The philosopher Xenophanes (c.570-465 BC) mentions three colours:

purple, yellow and crimson. I don’t know how many colours the
Burinmo people see in a rainbow; not many, I'd wager. They only
have five colour words and two of them are nol and wor.
It seems that, the more insulated the society, the fewer colour
words it will employ. In 1969, the anthropologists Brent Berlin and
Paul Kay wrote a description of how colour terms arise. The most
isolated societies do little more than distinguish light from dark. As
societies emerge into the wider world, red is the first ‘true’ colour
they identify, followed by a greenish-yellow. Green and yellow are
invariably distinguished before blue is distinguished from black. The
rest follow in no particular order. We don't have to go anywhere very
exotic to see elements of this scheme in operation, for most modern
languages contain traces of the evolution of colour. For example, up
until the Industrial Revolution, the Welsh ‘glas’ meant ‘mountain-
lake-coloured, and did for both green and blue. Over time, the influx
of new dyes and textiles, English influence, and growing urbanisa-
tion, made it expedient for the Welsh to distinguish between green
and blue. In modern Welsh, ‘glas’ simply means blue; ‘gwyrdd’ means
green. The native Japanese language has just four native colour terms;
only late in its history did it start borrowing others from Chinese
and English. The trace of these borrowings is evident, since the loan
words obey different grammatical rules.
If Berlin and Kay’s scheme has fallen out of favour with today’s
SEEING COLOURS 2S

anthropologists, it is not because it is ‘wrong’ exactly; it is rather that

it has proved insufficient. Many languages don’t distinguish colours
from other visual properties: texture, reflectance and brightness are
often described using terms that double as the names of colours.
The translator's job is made even harder by the various vicissitudes
and accidents to which words are prone. Often, colour terms derive
from the names of dyes and dyestuffs. (Just a couple of hundred
years ago, ‘pink’ was a noun.) And since a dyestuff often generates
more than one colour, this may be why the English word ‘blue derives
from ‘flavius, a Latin word meaning yellow.
Have special pity for anyone translating an ancient Greek text.
So-called Greek ‘colour words’ have no direct English equivalents.
Worse, they don't refer to colours, relating more to texture, consis-
tency and quality, with colour a small, often irrelevant, part of the
whole meaning. The sea is the colour of wine, but so are sheep.
Honey, sap and blood are all chloros which, as far as we can tell, is
a sort of yellow-green.
But what green recruit has not felt blue at times? The English
language is not short of colour metaphors, and neither was ancient
Greek. If we look for a wider meaning for ‘chloros’ we find that it
can also mean ‘fresh, ‘fluid; and ‘living. Studying these metaphorical
meanings, we begin to see what the devil the poet is on about.
The odd thing about Greek poetry (and Homer’s Iliad in particular,
the best, earliest and most substantial example) is that the metaphor-
ical meanings are the only meanings employed. Why does Homer
never speak directly and simply about colour? William Ewart Glad-
stone (1809-1898), four times British Prime Minister under Queen
Victoria, and a great classicist, was unequivocal in his criticism of
Homer's colour palette: ‘Although this writer has used light in various
forms for his purposes with perhaps greater splendour and effect
than any other poet, yet the colour adjectives and colour descriptions
of the poems are not only imperfect but highly ambiguous and
confused ... we find that his sense of colour was not only narrow,
but also vague, and wanting in description.”
From this, Gladstone leapt to some rather sweeping conclusions.
Consider his book, Homer and Homeric Age (1858), the first sally in
a lifelong attempt to reconcile the literature of heroic Greece with
216 A NATURAL HISTORY OF SEEING

the literature of the Bible. Not content merely to remark upon the
green sheep and wine-dark-seas of the Iliad, Gladstone - with
youthful enthusiasm - leapt from the particular to the universal with
two extraordinary claims, one far exceeding the other for sheer
chutzpah. First, he said that the ancient Greeks were colour-blind.
This is by no means a silly idea. Colour-blindness is a possible expla-
nation of Homer’s choice of colour terms, for reasons explored in
the next chapter. Gladstone’s second claim, however, far outstrips the
first: that among the Greeks‘ .. the organ of colour was but partially
developed? He believed he had discovered prima facie historical
evidence of the way human vision had evolved. According to Glad-
stone, what we think of as normal colour vision arose after the
composition of the Iliad.
The enormity of the idea was irresistible: here, at last, was cultural
evidence of human evolution. The German philologist Lazarus
Geiger was inspired by Gladstone's claim to investigate Greek liter-
ature, the Vedic hymns and the Zend- Avesta, amongst other writings,
to see whether the development of colour vision could be traced in
other ancient literatures. Needless to say, he found what he was
looking for:

All, at first, was vague in color ... but gradually a difference

was perceived, and men were compelled to find some term to
express this newly observed appearance . . . green was for a
long time regarded as yellow . .. Not only was the sky not called
blue, but nothing was called blue, and it was impossible to call
anything blue... the men of that time did not and could not
call anything blue.

As the years passed, scepticism grew. Advances in archaeology left

Gladstone's attempts to reconcile mythology with the Bible looking
rather risible. (In her novel Middlemarch (1871-1872), George Eliot
neatly captures the contemporary mood with her portrayal of the
pedant Casaubon, who has devoted his life - and, by the bye, his
young wife's — to “The Key to All Mythologies.)
And aside from the historical niceties, there is always the possi-
bility that Homer didn’t want to talk about colour. The idea of pure
SEEING COLOURS 217

colour could hardly have existed in the ancient Greek world. Only
a handful of pigments were available to them: white, blue-black, red
and yellow-green. These, according to Empedocles, were the primary
colours: those from which all others are made. (If you play with the
precise pigments to which he refers, you really can generate a full
spectrum of colour, albeit a rather muted one.) The Greek world was
not saturated, as ours is by artificial colour. The Iliad is a rich
compendium of surfaces, reflections, mists and tricks of the light.
Why should the ancient Greeks have separated out colour for special
emphasis, rather than texture or lustre?
Perhaps the Greeks - or Greek writers — did not consider colour
very important. Seven hundred years after the composition of the
Iliad, in the third century AD, Heliodoros managed to write a sixty-
thousand word romance, the Aethiopica, without once using the
words red, green or blue. This same lack of interest has been
encountered recently; in 1971 a team of Danish anthropologists went
to Polynesia to study colour perception among the islanders. But in
one village, they were told, “We don’t talk much about colour here.‘

In 1996, I drove north from Helsinki towards the Arctic Circle. The
Finnish countryside is a curious, fractal landscape of lakes and forests,
repeating endlessly, with minimal variation. Either the tedium drives
you mad or you achieve a kind of satori. It was in this Zen mood that
I reached Rovaniemi and the borders of the Arctic. I parked the car,
wandered around Santa’s house for a little while (why come all this
way and not say hello?), bought an outrageous hat, and walked back
to my car. I had still to find an hotel before I began the final, surreal,
reindeer-dodging part of my journey through the tundra to the rock-
strewn coast around Lake Inari. It was evening. I stepped out from
behind the shadow of Santa’s grotto and stopped, pole-axed.
The sun was setting. It was October, when the days are more or
less the same length they are in Britain. It was the wrong time of
year for the midnight sun, the Northern Lights, or the other heavenly
exotica for which these latitudes are rightly famed. What brought
me to such an abrupt halt was an ordinary evening and an average
sunset — except the colours were all wrong. The sky was no longer
the sky I recognised. It was neon. The pink edging of the clouds was
218 A NATURAL HISTORY OF SEEING

an artificial hot pink. Where I expected orange, there were flushes

of peach and cherry-red. The sky itself had a slight violet cast. It was
a sky painted by a painter who, to my eyes, had never seen a sunset
and was relying on written accounts. It was a sky painted, perhaps,
by Akseli Gallen-Kallela, the master of Finnish landscape painting,
whose early, bucolic works had left me cold and unconvinced when
I had viewed them, days before, in Helsinki’s Ateneum Art Museum
- and now I knew why. Gallen-Kallelas strange, forced, artificial
choices of colour were the right colours for the Finnish landscape.
This far north, things are lit differently.
The Finnish sky is more blue than the southern English sky. The
more atmosphere sunlight passes through, the more its shorter, blue
wavelengths are scattered by air molecules. (If this didn’t happen,
the daylit sky would be as black as the interior of Arthur Zajonc’s
light box.) The sun's rays pass obliquely through the atmosphere to
Finland, making its blue sky particularly intense; to reach southern
England, sunlight follows a more direct path. The sunlight’s shortest
route through the atmosphere is at the equator, where the skies are
more white than blue. (The noon sky - when the sun is directly
overhead — is less blue than in the morning or late afternoon.)
When we look at the sun at dawn or sunset, it has a distinct
reddish-orange cast, because all the short, bluish wavelengths have
been scattered, and only the longer wavelengths are reaching us along
a direct path. The contrast between the reddish sun and the blue sky
makes for pleasing sunsets in England, rather dull fare at the equator
(the sky is at its least blue here and the sun at its least red), and
firestorms of cherry-reds and hot pinks in Finland.
How many colours are there? In Vincente Minnelli’s 1956 biopic
of Vincent van Gogh, Lust for Life, the tortured artist chases across
Europe after an elusive ‘light, for the very good reason that this is
exactly what artists of that time did. J.M.W. Turner’s obsessive pursuit
of different lights was so well-known it became the subject of news-
paper parody. Nineteenth-century Europe saw an explosion of novel
colours as synthetic pigments immensely extended the artist’s range
of possible effects. Industry was colouring the world in inadvertent
ways, too. I live in Sydenham, a humble London suburb immortalised
in 1871 by the Impressionist painter, Camille Pissarro. Pissarro, like
f SEEING COLOURS 219

Claude Monet, came to London for the pollution. The smoke belching
from the capital’s chimneys was an irresistible draw, generating
colours and light effects that were not just odd, dramatic, and atmos-
pheric, but new.
That there might be physiological limits to the number of possible
colours is a fact not at all apparent to the eye, which seems capable
of discriminating between an infinite range of hues. Is there, some-
where in the world, a never-before-seen colour lurking on the wings
of a tropical bird, glinting in the scales of a reef-dwelling fish, or
buried deep in the lustre of a rare gem fresh from a Mexican mine?
How can we be sure that there are no more new colours to find?

2 — Primary colours

Seventeenth-century philosophers were interested in the behaviour

of light, and the coloured lights visible through prisms were a partic-
ular source of fascination. But if science was going to say anything
useful about light, it had to decide what it was competent to study.
How, in the seventeenth century, was one to hunt down the mysteries
of colour, let alone measure them?
Newton's Dutch contemporary, Christiaan Huygens (1629-1695),
clearly delimited his philosophy’s powerful but narrow abilities: ‘In
true philosophy, he wrote, ‘one conceives the causes of all natural
effects in terms of mechanical motions’ The trick, for seventeenth-
century optics, was to separate what was quantifiable —- the mechanical
behaviour of light - from the ineffable experience of colour in the
eye and heart of man. This was not a denial of the subjective realm,
but a recognition of what the science of the day could and could
not do.
In 1666, in a room of his birthplace, the manor house of Wools-
thorpe, Lincolnshire, Isaac Newton (1643-1727) began experi-
menting with the ‘celebrated phenomenon of colours. Newton knew
full well that he would not, by the light of the prevailing philosophy,
be able to say much that was useful about the phenomenon of colour.
‘Rays, to speak properly, have no Colour, he wrote. ‘In them, there
is nothing else than a certain power and disposition to stir up a
220 A NATURAL HISTORY OF SEEING

sensation of this Colour or that? This intuition - that colour is not

a property of things but is generated by the eye itself - is spot on;
and the slightly desperate hand-waving (‘a certain power and dispo-
sition’) marks the point at which the niceties of objective measure-
ment leave off and the woolly posturings of medievalism take over:
posturings he was determined to avoid.
Newton's experiments with prisms - a common toy and fairground
gewgaw — were directed not at the colours, slippery beasts that they
were, but at the shape a beam of light made as the prism fanned it
out. ‘It was at first a pleasing divertissement to view the vivid and
intense colours; Newton wrote to Henry Oldenburg, Secretary of
the Royal Society, ‘but after a while applying myself to consider them
more circumspectly, I became surprised to see them in an oblong
form; which, according to the laws of refraction, I expected should
have been circular ..? A neat, round beam of light enters a darkened
chamber through an aperture. It passes through a prism, and makes
an oblong pattern on the screen beyond. ‘Comparing the length of
this coloured spectrum with its breadth, I found it about five times
greater, a disproportion so extravagant that it excited me to a more
than ordinary curiosity of examining from whence it might proceed.’
To better understand what the prism does to the light, Newton
drilled a narrow aperture through the screen, and allowed a narrow
portion of the oblong to pass into a second darkened chamber, and
through a second prism. Was the light once more dispersed? No.
Then - and only then - did Newton admit colour into his account.
He noted that the thin beam of light retained not only its shape, but
also its hue. A blue ray remained blue, and a red ray remained red,
no matter whether he reflected it off coloured surfaces, passed it
through coloured media, or added extra prisms. Nothing disrupted
the beam’s spatial or chromatic integrity.
Whatever property first caused the light to fan out could not,
therefore, be a property of prisms. This left only one possibility: he
had uncovered a property of light.
Newton tentatively assigned physical values to his different
coloured lights. Red light was only slightly deviated from its path
by a prism; violet light was turned from its course much more sharply.
Oddly, there was no white light in the spectrum. Where had it gone?
SEEING COLOURS 22

When Newton brought the coloured lights to shine on the same spot
of wall, white light reappeared. White light was not one thing: it was
a mixture of different kinds of light. So what is colour? Newton had,
in a sense, found a way to measure colour; he could record how
weakly or strongly coloured lights were refracted. But he could not
say that blue light refracts by such-and-such amount because it is
“blue. This left him with the job of somehow separating the physical
properties of colour from their phenomenological properties.
He failed. Worse, he knew that he had failed. His alibi — he claimed
his pet dog Diamond knocked over a taper and set light to papers
containing his best thoughts — is so desperate one is almost inclined
to believe it. More telling, perhaps, is the fact that he delayed publi-
cation of his Opticks by twelve years, by which time his fiercest conti-
nental critics were safely dead.

Bring the lights of the spectrum together, to shine on the same spot
of wall, and they reconstitute white light. Newton was inclined to
take this observation at face value, but not every light source has as
rich a spectrum as sunlight, and two years later, Huygens pointed
out that as few as two colours (blue and yellow were his example)
were sometimes sufficient to reconstruct white light. It was a trick
Newton himself never mastered: ‘I could never yet by mixing only
two primary colours produce a perfect white, he lamented.° However,
he confirmed that white is produced when just three colours, spread
reasonably evenly across the spectrum, are brought to a point.
What was the nature of light, if just two or three colours could
produce a sensation of whiteness in the eye? Were there really no
more than two or three primary colours, from which all other colours
derived?
Newton thought of light as a stream of particles of different sizes,
each generating a different spectral colour in the eye. Newton's great
rival, Robert Hooke, studying the gaudy multi-coloured shimmer of
soap bubbles, advanced a very different theory of light and colours,
arguing that light was a mixture of rays which came in a number -
perhaps an uncountable number - of different wavelengths.
Homely evidence against corpuscular theories of light had already
appeared by 1665, in a posthumous collection of optical observations
22:2 A NATURAL HISTORY OF SEEING

by the Jesuit physicist, Francesco Maria Grimaldi. Grimaldis close

observation of light revealed that light blurred as it emerged from a
narrow aperture, just as coastal waves bend when they enter the narrow
aperture of a harbour. (He coined the word - diffraction - for this
phenomenon.) He lit an object with a single source of light, passed it
through a narrow aperture, and noted that when a second, similarly
narrow source of light was brought to bear on the same point, this
would sometimes cause the point to grow dimmer. We would call this
the ‘interference effect: Grimaldi’s observations hardly amounted to a
wave theory but they were incredibly damaging to any assertion that
light was a kind of matter.
How long Newton would have persevered with his corpuscular
theory in the face of mounting evidence is anyone’s guess. We know
that he toyed with a wave theory of light, and it is tempting to
suppose that it was chiefly his animosity towards his arch-rival Robert
Hooke that persuaded him otherwise.
When, on 12 November 1801, the Royal Society received conclu-
sive proof that their hero Isaac Newton had been in error — that
light was after all a wave — their response was unedifying. The hostile
reception given to the bearer of these bad tidings, a twenty-eight-
year-old doctor, Thomas Young, was sufficient to stall what up till
then had been a most promising career.
Born in 1773 to Quaker parents, the eldest of ten children, Thomas
‘Phenomena Young (so nicknamed by his Cambridge contempo-
raries) could read fluently by the age of two, and by sixteen was
proficient in Latin and Greek and acquainted with eight other
languages including Hebrew, Arabic and Persian. Young had been
ruffling the feathers of intellectual London since 1793 when, as a
twenty-year-old medical student, he showed members of the Royal
Society how the human eye accommodates to objects at different
distances by using the ciliary muscles to alter the curvature of the
lens. In 1797 an uncle left him £10,000 and a London house, into
which he moved in 1800. As a man of independent means, Young
devoted his spare time to investigations into sound and light, with
contributions so frequent and numerous that he sometimes had to
invent pseudonyms to avoid the charge of neglecting his medical
work.
SEEING COLOURS 223
By 1801 Young was professor of natural philosophy at the Royal
Institution and one of its first lecturers. The RI’s public talks became
famous for their hazardous stagecraft: early audiences regularly
contended with toxic fumes, safety lamps plunged into explosive gases,
powerful electromagnets dangled above their heads, and model volca-
noes.” Young's demonstrations were, by comparison, relatively staid.
However, what he lacked in showmanship, he more than made up
for in diversity. In 1801, he lectured on acoustics, air pumps, animal
life, astigmatism, and astronomy (to cover only the first letter of the
alphabet). This was also the year he showed that light was a wave.
“The experiments I am about to relate; Young assured his audience
(at a later and fuller demonstration, this time in 1803), ‘may be
repeated with great ease, whenever the sun shines, and without any
other apparatus than is at hand to every one.® He split a narrow
beam of sunlight with a slip of card, about one-thirtieth of an inch
thick. The card, held edgewise to the tiny beam of light whose diam-
eter was only slightly greater than the thickness of the card, split it
into two slivers, one on each side. These famous demonstrations -
a refined version of Grimaldi’s work of some 150 years before -
proved that light is a wave: the light bent around around the edges
of the card and spread out to produce an interference pattern, just
as ripples on a pond interfere.
The Royal Society’s old guard - keepers of the Newtonian flame
— shrugged and shook their heads. One of the younger members
went a step further, and sharpened his pencil to a wicked point. Five
years Young’s junior, Lord Henry Peter Brougham (1778-1868) also
had something of the prodigy about him; he was destined to become
Lord Chancellor and a powerful opponent of the slave trade. At the
time of Young’s demonstration, he was a versatile and prolific writer
for the Edinburgh Review (founded a year earlier, in 1802). He
contributed eighty articles to the first twenty issues, ranging over
science, politics, literature, surgery, mathematics and the fine arts.
Brougham had already presented a modest paper concerning light
and colour to the Royal Society in 1795 and an (even more modest)
paper on prisms in 1798; achievements which, in his own mind,
entitled him to launch a savage, satirical and anonymous attack on
Young in the pages of the Review.
224 A NATURAL HISTORY OF SEEING

Young had no appetite for a scrap. He abandoned his researches

into light and turned his polymathic mind to other, less controversial
subjects; he came up with the first good scientific definition of energy
and helped translate the Rosetta Stone.

If violet light had a wavelength of 0.0000016 inches (400 nanometres)

and red light had a wavelength not quite double that (about 700
nanometres) what would be the nature of the radiation that lay outside
those wavelengths? Might there be such a thing as invisible light?
The British astronomer Sir William Herschel (1738-1822) was
interested in the relative energies (measured as heat) of differently
coloured light. He placed a sensitive thermometer just beyond the
red part of the spectrum - and discovered infra-red radiation.
Learning about William Herschel’s discovery, the German chemist
Johann Ritter (1776-1810) set about detecting ‘invisible light beyond
the violet end of the spectrum. From his own experiments, as well
as from the new lore of photography, Ritter was familiar with the
way silver chloride turns black when exposed to light. He also knew
that blue light caused a speedier reaction than red light. Using a
prism, Ritter exposed silver chloride to an area just beyond the violet
end of the spectrum and noted an even more intense reaction. Silver
chloride reacted somewhat to visible light, but what really got it
going was ultraviolet - a light nobody could see.
By the turn of the nineteenth century, it was clear that what-
ever correlation there might be between the physics of light and
the perception of colour was mediated through biology. Although
Thomas Young is still remembered, this has little to do with his
disregarded lectures of 1802 and 1803. Auguste Fresnel achieved
a larger, better set of observations and proofs only a few years
later — testament to his gift for experiment and the intellectual
advantages of a quiet home life. (A Napoleonic court, rightly
suspecting that he harboured Royalist sympathies, had sentenced
him to live with his mother.) The power of his deductive reasoning
led Young to the theory by which he is best known today. As James
Clerk Maxwell was later to say, in a lecture to the Royal Institu-
tion: ‘So far as I know, Thomas Young was the first who, starting
from the well-known fact that there are three primary colours,
SEEING COLOURS 225

sought for the answer to this fact, not in the nature of light, but
in the constitution of man?’
The spectrum contains only a few of the colours we see around
us. There is no brown band in the rainbow, although the natural world
throws up any number of different browns. Light of different wave-
lengths appears differently coloured to us, but it cannot be the light
itself that is coloured. Colour must be the means by which the brain
tells us what mixture of wavelengths is being reflected by an object.
How many wavelengths must we see to perceive, in their mixture,
a virtual infinity of different colours? Young knew the number could
not be large, because a room lit by red lamps, for example, is not very
much dimmer than a room lit by white. If the eye had to have many
different kinds of colour-receptive abilities, for many different kinds
of light, there wouldn't be enough red-sensing ability to see clearly
in a red-lit room. So ‘it becomes necessary, Young wrote in 1801, ‘to
suppose the number limited, for instance, to the principal colours,
red, yellow and blue: With no anatomical evidence, Young had inferred
the existence of three types of colour receptor in the human eye.
Young plucked the colours red, yellow, and blue out of the air.
From the writings of Newton and Huygens, Young had noted that
a mix of any three coloured lights will generate white light, provided
their colours are reasonably spaced out across the spectrum. (He
seems to have found no use for Huygens’ observation that just two
well-chosen spectral colours can perform the same trick.) But Young
was also intrigued by two other observations: that green and red
light, shone on the same spot of wall, will generate pure yellow light
and violet and green light will generate blue.
Young eventually revised the triumvirate of colours perceived
‘directly’ by the eye. In his article ‘Chromatics’ for the Encyclopaedia
Britannica, he wrote that the eye detects mixtures of red, green, and
violet light, manufacturing from these colours all the other colours
we can see, even such apparently simple and ‘primary’ colours as
blue and yellow.

Flooded with light, cones become chemically exhausted and need time
to recharge. Deprived of light, cones become hypersensitive. Just as
staring at a black dot for thirty seconds or so can leave a bright dot
226 A NATURAL HISTORY OF SEEING

persisting in your vision for more than a minute, so staring at strong

colours can upset the eye’s colour balance.
Consider the image of the Union Jack in the colour section. The
expected colours have been swapped for their complementaries. A
complementary colour is the colour which, when added to the
original colour, generates white light. A colour and its complemen-
tary will do this because under their mixture all cones are being
stimulated equally.
In the figure, the red parts of the Union Jack have been swapped
for cyan (the blue that arises when green and blue light overlaps).
The dark blue parts (very close to violet in the spectrum) have been
swapped for yellow, the colour made by overlapping red and green
lights; and the white has been swapped for black. Concentrate on
the dot at the centre of the image for thirty seconds, then switch
your gaze to the wall. For a fleeting moment, you will see the after-
image of a Union Jack, in its true colours. Why? First, the parts of
your retina exposed to black have been rested and recharged; once
your gaze turns to the wall, their greater readiness to excitation
generates an illusion of bright whiteness. By exhausting the cones
receptive to shorter wavelengths of light, the cyan bars of the Union
Jack have primed those portions of the retina to perceive red. The
yellow, by contrast, drains both the ‘red’ and ‘green’ cones, so that
when the parts of the retina exposed to yellow view the wall, a dark
blue, almost violet, after-image is generated.

The eye is not a painter’s palette. The eye gathers and interprets a
mix of wavelengths. Add a wavelength and you give the eye an extra
morsel of information to work with. The more wavelengths there
are, the richer the view. This is why, when shopping for clothes, it’s
best to take them to the front of the shop, where you can see them
in daylight. Daylight reveals more colour information than artificial
light, and some artificial lights are better than others. (Fluorescent
lights are hopelessly drab.)
On the painter's palette, something different happens. The more
pigments are added, the less colour information the mix contains.
Pigments swallow most wavelengths of light; it is the wavelengths
they do not absorb which bounce into our eyes. A jar of turmeric
j SEEING COLOURS 227

Blue cone Rod Green cone

100
Red cone

(%)
Sensitivity

ty) |__ ! 1 ! J
400 450 500 550 600 650 (nm)
violet blue green orange red
yellow

absorbs all light but yellow; a jar of paprika absorbs all light but red.
Mix the two and the light reflected will be orange - a mixture of
red and yellow light. It won't be a particularly vivid orange, because
the yellow turmeric and the red paprika each absorb a little orange
light. The more colours we mix in, the duller the result: eventually
all wavelengths of light will be absorbed by at least some of the
mixture. Strictures against the mixing of pigments are a commonplace
of painting manuals.
These two perfectly correct and completely incompatible ways of
looking at colour - subtractively like a painter, or additively like a
projectionist — can trip up the keenest. David Brewster, researching
colour vision in the 1840s, made the mistake of treating coloured
lights as paints on a palette. Instead of shining lights through different
coloured filters on to a single point, thereby adding monochromatic
lights together, Brewster shone light through filters placed on top of
each other. Add red light to green light, and you get a vivid yellow.
Place a green filter under a red filter, however, and the little light
that penetrates both is a darkish red.
Brewster managed to convince himself that each part of the spec-
trum was made up of three primary colours: yellow, red, and blue.
He persisted for years in this eccentric opinion, which seems odd,
given the precision of his observations. Thinking he was measuring
the relative power of his three ‘primary lights, Brewster inadvertently
measured the relative impact different colours have on the eye, and
so produced what, twenty-five years later, Helmholtz realised were
228 A NATURAL HISTORY OF SEEING

the first really good drawings of how different cones react to different
wavelengths of light.
Brewster's diagram clearly revealed the wavelength sensitivities
of our three types of cone: violet on the far right (which Brewster
called blue), green in the middle (which Brewster, for some reason,
reckoned was yellow) and on the left - orange. This is easily the
diagram’s biggest surprise and one Brewster never spotted: there is
no cone attuned to red light. The best our so-called ‘red’ cones can
capture, according to Brewster's diagram, is a yellowish orange. Later
assays, like the one on the previous page, produced readings in
which the long-wave pigment did not even peak at orange, but at
yellow-green!
If we are surprised to learn that blue and yellow are perceived as
mixtures, how much more startling it is to discover that red — most
vibrant and unsettling of colours; the colour of danger, blushes, poison,
blood, and sexual arousal — is not directly perceived, but constructed
whenever yellow cones are excited and blue cones fall silent. Indeed
the sensation of red can be generated by removing all bluish wave-
lengths and enriching the yellowish-green part of the spectrum. The
fact that red is patently more than this formula is striking evidence
that colour is neither an objective property of objects nor of the light
they reflect: it is a construction of mind.
This is not the only surprise the diagram contains. Brewster had
every reason to expect that the curves on his diagram would exhibit
a tidy symmetry. He found, however, that there were modest but signifi-
cant variations in the relative strength of his colours as he moved
across the visual spectrum. Accordingly, Brewster nudges his violet
curve to the right of the diagram, while his red and yellow curves
run more closely together. This first-class observation is but a hint of
what would be revealed by later researchers, using more sophisticated
equipment. Today's benchmark measurements create a picture even
less symmetrical than Brewster's. The ‘green’ and ‘orange’ curves run
even closer together, so that they are practically overlapping, while
the ‘violet’ curve sits out on a limb, detecting wavelengths of much
smaller wavelengths. This far-from-symmetrical arrangement suggests
that while blue cones detect wavelengths far into the shortwave band
of visible light, red and green cones both respond to very similar long
; SEEING COLOURS 229
wavelengths. Couple this with the fact that the three sensitivity curves
overlap, so that even the simplest-seeming colour is detected by at
least two types of cone (light between violet and green will stimulate
all three types!) and one begins to wonder whether it makes any sense
at all to talk about cones detecting this or that ‘colour’ of light. It would
be a shame to eschew talk of ‘red’ ‘green’ and ‘blue’ cones: there is a
certain prettiness and poetry to the habit. But there is no anatomical
justification for it: red, green, blue, nol, wor, wheatmeal, and every
other colour we see (except for extreme violet, detectable by blue
cones only) is detected by at least two types of cone.

So far, we have talked about lights. No one has mentioned shadows.

We have talked about the colours of the spectrum. No one has
mentioned earth colours — the browns and olive greens that light
alone cannot create.
Otto von Guericke is one of those people footnotes were invented
for. His genius is hard to pin down. He invented a machine for the
production of static electricity but he did not realise what it was. He
invented a barometer and a vacuum pump, and he directed the
rebuilding of the city of Magdeburg, razed during the Thirty Years
War in 1631. Most importantly for us, in 1672 he wrote the first ever
methodical account of the colour of shadows.
Light an object with a red light: it appears to cast a green shadow.
Light the object with a green light, and its shadow appears red. Shine
a yellow light, and a blue shadow is cast. A blue light casts - well,
one wants to say a ‘yellow’ shadow, but oddly, there is no such thing;
instead the shadow appears brown. It is one of the foibles of the eye
that it perceives dim yellow as a separate colour. (‘There is also, no
doubt, a linguistic element to the way we now define yellow so
narrowly: in the eighteenth and nineteenth centuries, ‘Indian yellow’
paint, which used a pigment made from the dung of cattle fed on
poisonous mango leaves, lent European interiors a hue that to modern
eyes seems more brown than yellow — not to say rather diarrhoeic.)
Dazzled by brightly lit patches of one particular colour, the brain
is somehow being tricked into filling the shadows with a comple-
mentary colour, just as, a couple of pages back, it was tricked into
constructing a true-colour Union Jack on a white wall. The difference
230 A NATURAL HISTORY OF SEEING

— the crucial difference - is that we can explain the appearance of

the true-colour Union Jack by supposing that our colour receptors
are exhausted by prolonged exposure to lights of certain wavelengths.
However, shadows cannot be explained in this way. A strong colour
in one part of the scene should not, by Young’s theory, upset the
colour balance of receptors directed at quite another part. And our
experience of coloured shadows is quite different from our experi-
ence of after-images. After-images are blurry and float about in our
vision: coloured shadows are vivid, precise, and fully integrated with
the scene before us. This observation reveals a whole new order of
colour; a system of antagonistic pairs, in which the absence of one
colour brings forth the sensation of another, so that blue ‘opposes’
yellow, and red ‘opposes’ green.
Ewald Hering, a German physiologist and psychologist who taught
at the University of Leipzig, came up with this formula in 1872.
Hering wondered, with devastating insouciance, why certain colours
should be unimaginable. While it is perfectly easy to imagine a
greenish yellow, or a bluish green, or a yellowish red, it is considerably
harder to think of a “bluish yellow, or ‘reddish green. Hering mapped
these ‘impossible colours’ over the spectrum and deduced hitherto
unknown relationships between the colours. According to his scheme,
the eye was sensitive to four wavelengths, not three, and these sensi-
tivities were arranged in opponent pairs: blue and yellow, and red
and green. Part of Hering’s supporting evidence was the observation
that people are never colour-blind for a single colour. Invariably, they
are missing one or other of his opponent pairs: those who are colour-
blind to red are likewise blind to green, and the few who fail to see
blue also fail to see yellow. White and black arguably made a third
pair, bringing the number of Hering’s primary colours to six.
There was genius in including white and black among the
primaries. By introducing light and shade into his discussions, Hering
was able to explain how we see earth tones. The eye does not reveal
absolute values for how bright or how dim things are in a particular
scene. Absolute light values are almost meaningless. Place a cube of
coal on a table and shine a spotlight on it. Beside it, in the deep
shadow beyond the circle of light, place an identical cube of chalk.
Both coal and chalk shine with the same vigour. Chalk readily reflects
’ SEEING COLOURS 231

what little light reaches it; coal reflects the brightest light only with
reluctance. To see these objects for what they are, what matters is
that the coal is bathed in light, while the chalk lies in shadow. The
eye ignores the fact that the coal and the chalk glow with the same
intensity: it points up the local differences in illumination; the white
glare surrounding the coal, and the deep darkness around the chalk.
‘No one has ever confused coal with chalk.
Only if we rob an object of all context can the eye be fooled.
A full moon at night, lit by a sun we cannot see, and suspended
in the non-reflecting vacuum of space appears white - yet moon-
dust is black.
Hering realised that the eye renders light and dark by comparing
the light levels in neighbouring regions of space. This awareness
enabled him to describe how browns and olive greens arise, even
though no mixture of lights in a darkened lab can generate them.
When yellows and greens are surrounded by areas of greater
illumination, browns and olives appear; they are seen as mixtures
of black with either yellow or green. To interpret these colours
correctly, we need a context. Look at a brown or olive surface
through a long, non-reflective black tube and you will see either
orange, yellow, or green. The dun, ‘earthy’ quality of the colour is
quite stripped away.
Given that Hering had managed to explain the perception of
common, yet hitherto mysterious colours, it seems strange that his
work should have encountered resistance. The problem was that
Hering’s idea of opponent pairs sat uneasily with Young's explana-
tion of colour blindness. Young maintained that some people were
missing one out of three colour-perceiving ‘particles’ (or, as we
would say now, one out of three types of cone cell). This might
not have mattered — three-quarters of a century had passed since
Young’s paper, and few remembered it — but his theories had recently
acquired a new champion in Hermann von Helmholtz, the most
influential vision scientist of the nineteenth century.
Helmholtz’s most lasting achievement in the study of vision is his
Handbuch der Physiologischen Optik - a work which required
frequent updating in his own lifetime, not least because of the sheer
number of arguments and objections Ewald Hering put up against
232 A NATURAL HISTORY OF SEEING

it. The two men could agree

on nothing. Their argu-
ments ranged across vision
research, from _ spatial
perception to eye move-
ments, stereoscopic vision,
and, finally - with
Helmholtz retired from
research and reportedly
sick of Hering’s bickering —
to colour vision.
Helmholtz was immov-
ably wedded to the theory
of trichromacy. He came up
with the idea indepen-
dently, only later - by
Hermann von Helmholtz. happy accident - discov-
ering Young’s neglected
paper. Young’s arguments, re-aired, confirmed the rightness of
Helmholtz’s trichromacy idea to everyone's satisfaction — everyone's
but Hering’s. Hering’s theory of opponent pairs flatly contradicted
both Young and Helmholtz.
Although Hering is often cast as the iconoclast in accounts of the
ensuing argument, it was he who came closest to resolving the dispute.
It struck Hering that he and Helmholtz might both be right; that
they had described different stages in the perception of colour.
Perhaps, as Helmholtz said, there were indeed just three kinds of
colour-sensitive receptor in the eye. In that case, Hering’s ‘opponent
pairs’ were not structures but descriptions of the activity of the three
receptors. When the receptors were excited, red, yellow, or white were
perceived. When the receptors grew less excited, one perceived their
opponent colours: green, or blue, or black.
Hering’s attempt at a rapprochement got nowhere. His theory and
the Young-Helmholtz theory were at loggerheads for nearly a century.
Each had its champions, and each side was convinced theirs was the
‘correct’ explanation of colour vision. The debate was heated, endless,
and in some cases, downright unpleasant. Even today, vision theorists
SEEING COLOURS 233

fall into two broad camps: those who reckon Hering was a visionary
and those who reckon he was a lucky eccentric.
What was so extraordinarily difficult about Hering’s ideas? First,
he proposed that a reduction in nervous activity might lead to a
sensation of certain colours — green, blue, and black — just as readily
as an increase in activity would lead to the sensation of red, yellow,
and white. But how could reduction in nervous activity give rise to
sensation? He did not know. Next, he proposed that to perceive the
full gamut of colours — including olives and earth tones — the eye
integrated information from neighbouring parts of space. How? No
one knew.
What Hering needed was a good model of the nervous system
and there wasn't one. There wouldn't be until well into the following
century. Hering died in 1918, when Haldan Hartline - who unpicked
the mechanisms of contrast — was a teenager, and Edwin Land -
who realised Hering’s ideas in mathematics and technology - was
just nine years old.

3 — Old wine in new bottles?

Ed is some star in his studies and we are sure that he will

make a name for himself and Alma Mater in college.
Caption to a college year-book photograph of Edwin H. Land

After Thomas Edison, Edwin H. Land was the most prolific inventor
ever to do business with the US Patent and Trademark Office. By
1982, when he retired from Polaroid, the corporation he had founded
in 1937, Land had over five hundred patents and a personal fortune,
and left a legacy in vision research that remains controversial. Did
Land’s ‘retinex’ theory of colour vision really advance our under-
standing - or was it simply a repackaging of old ideas, some of them
going back as far as the eighteenth century?
Edwin Land was born in Connecticut in 1909, the only son of a
prosperous scrap-metal dealer. Optics fascinated him from boyhood;
he was only nineteen when he invented the sheet polariser which
launched his extraordinary career. His energy and determination
234 A NATURAL HISTORY OF SEEING

were legendary. Making his first polarising filter required the use of
a gigantic electromagnet, which was kept under lock and key in
Columbia University’s physics laboratory. Refused a key, Land went
to the sixth floor, climbed on to a ledge, and edged his way along
the side of the building to gain entrance to the lab through an open
window. After a year of this sort of thing, Land decided to continue
his education at the New York Public Library.
His career was not without its checks and disappointments. For
decades, he argued that the most important application for his polar-
ising filter would be as an antiglare device for car headlamps. The
car manufacturers did not listen, and late in the day they came up
with a cheaper solution: dipping headlamps. Land, who had expected
to save lives, instead benefited humanity in more indirect ways: he
made sunglasses, 3-D films, innumerable laboratory and industrial
instruments, and, one sunny day in 1944, while wandering through
Santa Fe zoo with his daughter, he dreamt up polaroid photography.
As a part of his research into instant colour film, Land repeated
some of James Clerk Maxwell’s experiments.’ He was particularly
interested in Maxwell’s attempt, at the Royal Institution on 17 May
1861 to produce colour images using three-colour projection.
Maxwell had taken three black-and-white transparencies of a tartan
ribbon: one through a red filter, one through a green filter, and one
through a blue filter. He used magic lanterns to superimpose his
transparencies on a screen, filtering the red image with a red filter,
the green image with a green and the blue with a blue. The experiment
was a success, producing ‘a coloured image .. . which, if the red and
green images had been as fully photographed as the blue, would
have been a truly-coloured image of the ribbon’ (Maxwell knew that
the photographic materials of his day reacted rather unevenly to
different wavelengths of light.)
One evening, at the end of a long series of experiments with three
projectors, Land and his assistants shut off their blue projector and
took the green filter out of the green projector. Then, one of Land’s
assistants, Meroe Morse, called their attention to the screen. The red
projector was still running, projecting the red record on the screen
in red light, and the unfiltered green projector was projecting the
green record with white light. That combination of red and white
SEEING COLOURS 235
lights should, in Morse’s mind, produce something pinkish. But there
was the original image, its every colour still identifiable. How could
red and white lights throw blues and greens on the screen?
Land was not particularly impressed. The complementary colours
thrown by shadows had been known to von Guericke four hundred
years earlier, and colour opponency - the greenish tinge given to
objects adjacent to red-lit surfaces, the bluish cast of shadows cast by
yellow lights, and so on — was one of the mysteries of colour perception
with which Hering had twitted Helmholtz. Certainly, explaining the
mechanism of colour opponency had been a problem for Hering, but
Haldan Hartline had demonstrated how nerve cells in the retina detect
contrast by mutual inhibition. So it was generally assumed that oppo-
nency must involve inhibitory structures like the ones Hartline had
discovered.
‘Oh yes, Land explained to Morse, ‘that’s colour adaptation - and
they went home.
At two oclock the following morning, Land sat up in bed. Colour
adaptation? What colour adaptation?
If colour adaptation were a simple matter of opponency, the most
they could have expected of that red-and-white image was a few
greenish edges here and there. But the image cast on the wall had
been better than that - much better — it contained a full spectrum.
Somehow their eyes had deduced, from the little evidence on the
screen, the complete spectrum of the original image. How was a
mystery, but Land, considering what information the eye had to work
with, was able to sketch the broad outlines.
First, the eye had information about the reds in the image. This not
only told it what objects were predominantly red; it also revealed what
proportion of red was contained in the colours of all the other objects.
Second, the eye had information about the greens in the image. The
trouble was, there was no way the eye could know this. The green
filter had been removed from the projector and the green record
was projected in ‘noisy, multiple-wavelength white light. The eye was
receiving information, not about ‘green’ light so much as information
about light that was ‘not very red. None the less, the contrast between
red wavelengths and white illumination was enough for the eye to
manufacture an entire spectrum.
236 A NATURAL HISTORY OF SEEING

Most colour-vision experiments up till then had used single objects

or swatches of a single colour. These experiments had been uninfor-
mative, and now Land knew why: the experimental set-up had reduced
the eye to a mere wavelength detector. Place a swatch of isolated colour
under a green lamp and it appeared green. Placed under a red lamp,
it glowed red. Big deal. It was only when objects were put into context
that the eye could deduce colours, the way it had that night in Land's
lab.
Land’s ‘retinex’ theory of colour vision arose from a series of
dramatic experiments, designed to see how far colour perception
could be stretched. Land’s team put colours in context: they used
boards of intersecting shapes and colours, called ‘mondrians’ after
the artist whose work they resembled. People studying the mondrians
reported colours accurately, even under the most extreme changes
in lighting. For instance, by adjusting the lighting, a swatch might
give off a greenish wavelength one moment and a reddish wavelength
the next. If the swatch was seen in isolation, this was what was
observed: the swatch changed from green to red. When viewed as
part of a mondrian, the object’s colour remained stable.
Context is everything. The eye has no interest in absolute levels
of illumination, nor in absolute colour values. Every colour is
perceived in relation to every other, just as every patch of light is
perceived in relation to every patch of shade. In the real world, this
makes colours remarkably stable. Reduce the visual context, and we
become susceptible to illusions of contrast like the squares in the
colour section. The three squares in this figure are particularly inter-
esting, since they demonstrate three different contrast effects at work,
affecting our perception of lightness (how bright a colour is), hue
(what colour it is), and saturation (how easily it can be distinguished
from a grey of the same lightness).
Until then, researchers had rather assumed that colour progressed
in a paint-by-numbers fashion, ‘filling in’ a black-and-white sketch
point by point. Now it was clear that colour perception used all three
aspects of colour — its lightness, its hue, and its saturation — to detect
boundaries and sketch forms. Colour vision was not so peculiar; it
was simply a multi-dimensional version of monochrome vision.
In the colour section, Mach’s bands are adapted to reveal how we
SEEING COLOURS 237

perceive hue and saturation. The ‘fluting’ effects, though milder than
those generated by the original, demonstrate that hue and saturation
are, like lightness, processed by a contrast-enhancing mechanism.
Land’s own demonstrations were spectacular. Using two yellow
lights with wavelengths only 20 nanometres apart - the difference
is barely detectable to the eye - he was able to create full-colour
images.
Ironically, Maxwell's original 1861 experiment had done much the
same but without his knowing it. Though Maxwell appreciated that
the photographic processes of his day registered wavelengths
unevenly, neither he nor anyone else realised quite how insensitive
they were. In truth, Maxwell's ‘red record’ and ‘green record’ contained
no unique information, as his film stock was only capable of regis-
tering blue light! Not until 1873 was there a photographic emulsion
sensitive to any light but blue. Good responses to the whole visible
spectrum were not achieved until 1882. The history of science is full
of these curious footnotes: the truly ‘original’ experiment is rare
indeed, and every discovery seems, with hindsight, to be but the final
piece in a puzzle that has been put together over many years by
many people.
Edwin Land had so many precursors —- Maxwell and Hering among
them - that it seems only reasonable to ask, what did he really
achieve? His retinex theory stated that colour perception depends
strictly on the neural structure of the human visual system: colour
is subjective. Isn't this exactly what Isaac Newton said? Retinex, inci-
dently, was coined from the words ‘retina and ‘cortex, because Land
did not know - any more than anyone else — whether the retina or
cortex had the key role in colour perception. But wasn't this the very
question explored by Hartline and his successors?
To put it bluntly, what was original about Land’s theory? The math-
ematics he drew up to explain it were tremendously useful, making
retinex a true theory of colour vision, whereas everything that had
gone before was, strictly speaking, only a description. None the less,
many academics, nettled by Land’s showmanship and annoyed by
his lack of reference to earlier work, considered him a mountebank
who was pouring old wine into new bottles. This reaction, though
understandable, was not really fair. If journalists wanted to cast Land
238 A NATURAL HISTORY OF SEEING

as an iconoclast, overturning ‘establishment’ views on colour, it was

hardly in his power to correct them. Land was an historian of vision
theory, and the lack of cited sources in his papers probably reflects
a lack of patience for academic niceties more than an attempt to
hoodwink his audience.
His practical work, his algorithms, and even his basic kit — the
mondrians and projectors — have become the staple of laboratories
dedicated to the development of better cameras and even artificial
vision. Edwin Land died, aged 82, on 1 March 1991.

4 — Fusion and attention

Some say they see poetry in my paintings; I see only science.

Georges Seurat

Draw the following figure on a piece of card, stick a pencil through

the centre, and spin it anticlockwise, not too fast.
The effect is modest, but enchanting: before your very eyes, the
card takes on all the colours of the rainbow. The outermost lines are
red, the next ones are green, the next pale blue, and the innermost
lines are dark violet. Flicker colours were discovered in 1826 by a
French monk, Bénédict Prevost, but proved to be one of those puzzles
SEEING COLOURS 239

that no one quite knew what to do with. The phenomenon has been
forgotten and rediscovered at least a dozen times, most famously by
the toymaker C.E. Benham, whose odd little top sold well and immor-
talised his name into the bargain. Nobody knows how Benham’s top
works, or how significant it is. It could be the key to unlock a radical
_new idea about colour perception - or it could simply be a case of
the overloaded brain getting its wires crossed.
The brain quite often treats separate streams of information as a
single, anomalous input. Here, courtesy of Hermann von Helmholtz,
giant of German natural philosophy, is another toy to confound the
philosophically minded. Using a circle of card and a pencil, make a
spinning top. Mix blue and yellow paint to make green, and paint
the mixture over the centre of the disc. Paint half the outer ring blue,
half yellow. When you spin the disc, the centre looks - well, green,
obviously — but the circumference looks lighter, and greyish.
In an age where coloured lights were awkward to project and
disco had yet to be invented, spinning a coloured top like this was
a good, cheap way of demonstrating the difference between coloured
lights and coloured pigments. Different lights mixed together make
white, while a mix of different pigments makes a dirty mess. But
Helmholtz’s top demonstrates something else: when a point in space
changes colour rapidly enough, the colours combine in the eye. Blue
and yellow, spun together, make white (or at least grey: the light
reflected isn't bright enough to make a true white). Where has this
white come from?
An object flickering yellow and blue will excite different colour
receptors, and signals from these receptors will be sent off at
different times, linked to the rate of flicker. But there comes a point,
as the rate of flicker increases, where the signals arrive so close
together that the brain treats them as arriving simultaneously.
Photoreceptors respond quite slowly to changes in light intensity:
locust photoreceptors, for example, need 25 milliseconds to report
a change in light - and insect receptors are typically faster than
human ones. A spinning top presents the human eye with a rate
of flicker it is simply not equipped to handle, so blue light and
yellow light blend to white.
Helmholtz’s top was no more original than Benhams. It was a
240 A NATURAL HISTORY OF SEEING

demonstration, not a discovery. The principles of visual fusion were

already established, at least among artists, by 1731, when the
Amsterdam miniaturist turned entrepreneur, Jacques Christophe Le
Blon, embarked upon yet another of his ruinous manufacturing
schemes. Le Blon, constantly on the look-out for ways to churn out
cheap pictures for the burgeoning middle-class art market, had
already lost his shirt on a system of three-colour printing. Unde-
terred, he turned to the textile industry. Weavers were certainly up
to the task of producing complex designs: what made woven pictures
so prohibitively expensive was the number of different coloured
threads used in their manufacture.
So why not use three threads, each dyed a primary colour? Woven
finely together, might they not conjure up a full spectrum? Le Blon's
experiments were encouraging. Seen from a distance, the colours of
adjacent threads fused, as the colours on a spinning top merge to
produce novel colours. In both temporal and spatial fusion, the visual
system is unable to distinguish between several inputs and treats
them as one.
Le Blon was canny enough to realise that this sort of colour mixing
was quite unpainterly. Rather than mixing colours, reducing the
amount of colour information in the mix, he was laying differently
coloured threads side by side. Like differently coloured beams of
light directed at a single point, the different-coloured threads, fused
in the eye, provided ever-richer colour information. In theory, red,
yellow and blue threads, woven tightly enough together, would be
perceived as white. The problem was illumination. The primary-
colour threads could never conjure up a perfect white, because they
were simply not reflective enough; the best he could achieve with
them was a ‘light cinnamon.
Le Blons list of threads increased to four, because he needed black
thread to generate different shades of a single colour. With the neces-
sary but philosophically untidy addition of white, the list came up to
five. And it kept growing. Dr Cromwell Mortimer, reporting to the
Royal Society explained, ‘though he found he was able to imitate any
picture with these five colours, yet for cheapness and expedition, and
to add a brightness where it was required, he found it more convenient
to make use of several intermediate degrees of colours?° So much
SEEING COLOURS 241

for slashing the cost of manufacture: the illumination problem put

paid to any hope of mass-producing pictures in textile form. Le Blon
had won himself a permanent place in the history of vision; but the
business failed.
Le Blons story ends happily: Louis XV, impressed with his tenacity,
gave him a monopoly to develop colour printing in France. Finally,
Le Blon came up trumps: the four-colour method of printing we
use today was his invention. He died in 1741, aged seventy-four, four
busy and successful years into the project.
Le Blon, a professional painter, knew the value of harnessing illu-
mination to generate colour. Apart from the commercial possibilities
of mass production, a painting system that juxtaposed a handful of
colours to generate a full spectrum would, by its very nature, generate
a uniquely luminous art.
In the mid-1880s, a young painter and print-maker, Paul Signac,
ran into Georges Seurat at the newly established Salon des Indépendants,
an open exhibition which Signac had helped set up in reaction to the
art establishment of the day. Seurat was exhibiting Bathers at Asniéres,
which had been rejected by the official Impressionist salon. Bathers
was composed entirely of dabs of colour. The colour contrast between
neighbouring dabs was often extreme; however, the dabs were so small
that, seen from a distance, the colours appeared smoothly blended.
Another arresting quality was the canvas’ comparative evenness of
illumination. This was a painting in which shadows seemed to glow
as brightly as the lighted surfaces. No wonder the official salon found
the work defective; even Renoir — no stranger to the pastel end of the
paint chart - found that Seurat’s output set his teeth on edge: ‘Imagine
Veronese’s Marriage at Cana done in petit point? he complained.
In composing his paintings along the same lines as Le Blon’s
tapestries, Seurat ran into many of the same problems: the technique
does not allow for the fact that, just sometimes, black is black and
white is white. For Seurat - an experimental artist in charge of his
choice of subject matter - these limits were part of his signature. He
might not be able to emulate Veronese with this technique, but then
Veronese, master of chiaroscuro, could never have emulated him.
Signac was impressed enough to begin to explore the possibilities
of making paintings out of swatches. He also wrote about the technique,
242 A NATURAL HISTORY OF SEEING

and gave it a name: Divisionism. Though it trips off the tongue less
readily than pointillisme (which, strictly speaking, applies only to
Seurat’s style of painting), Divisionism is the proper term for any
picture that relies for its effect upon the way colour information from
many small areas will fuse in the eye. When you next sit down in front
of the television, with its matrix of red, blue, and green dots, you will
be sitting in front of a Divisionist picture. Of course, the dots on a
television screen are so small that you would have to press your face
up to the screen before the picture broke up into individual patches
of colour. The canvases of Seurat and Signac were not so subtle — nor
were they meant to be.
Bathers is presently in the National Gallery in London.At a distance
of about a metre you can focus your attention on any part of the
canvas and examine the individual dots. If you want to appreciate
how these dots fuse, refocus - not your eyes but your attention —
and the colours will blend to reveal the painting as a whole. You do
not need to ‘zoom’; you do not need to track in towards your subject
and out again as though you were the camera in a particularly tricksy
Alfred Hitchcock dolly shot. All you have to do is will the change
from one zone of attention to another.

There are more than forty different types of nerve cell in the retina,
each with their own ‘field of view. There are cells which edit the
responses of the photoreceptors and bipolar cells that compress that
information by factors of between twelve and one thousand,
depending on the light level; no one knows how this adjustment is
made. The bipolar cells pass their information on to the six layers
of ganglion cells, each layer connected with the others by at least
thirty different types of amacrine cell. There, the information is once
again savagely edited, until only a million signals remain to be carried
along the optic nerve to the brain.
Some researchers believe that somewhere in that bewildering mix
lurks a mechanism of attention. Perhaps the receptive fields of the
ganglion cells — which, although they vary wildly across the retina,
always come in overlapping large and small sizes - are part of that
mechanism: as we study Seurat’s Bathers, at one moment we attend
to our large-field ganglion cells, and see a pair of purplish swimming
SEEING COLOURS 243

trunks; at another, using small-field ganglion cells, we study the

painter's individual dabs of blue and red paint. How easy that would
be! However, other experiments suggest that we have not just two
but up to half a dozen ‘zones of attention™
Colour, which seemed such an intractable phenomenon in the
1950s, now offers us - through the work of artists, printers, and
inventors, as much as through the work of academics - a window
on to the next great unsolved problem of vision: attention.
No one knows how we focus our attention. No one even knows
how to ask the right questions about attention. If the eye is an outpost
of the brain, it is - by the same logic - an incursion of the light.
When our eyes are drawn to a change in the scene, what draws them:
our desire to see, or the change in the scene? Is one a cause, the
other an effect? Or - in Goethe’s memorable phrase - do ‘the two
together constitute the indissoluble phenomenon?
We can no longer talk about the visual organs as any kind of
mechanism. That kind of thinking will lead us nowhere, as the
German mathematician and philosopher Gottfried Leibnitz pointed
out in his Monadology of 1714:

Suppose that there be a machine, the structure of which

produces thinking, feeling, and perceiving; imagine this machine
enlarged but preserving the same proportions, so you could
enter it as if it were a mill. This being supposed, you might visit
inside; but what would you observe there? Nothing but parts
which push and move each other, and never anything that could
explain perception.

In the absence of theories — and lacking even the language in which

to cast our theories — we have only descriptions. Of all our descrip-
tions of visual attention, Seurat’s paintings are, rightly, the most
famous. If anyone is so crass as to ask what Seurat’s paintings are
‘about, the following reply is more revealing than most: they are
about the nature of visual attention.
NINE

Unseen colours

From Manchester, in February 1794, a Quaker schoolmaster, John

Dalton, wrote to his friend and mentor Elihu Robinson,

I am at present engaged in a very curious investigation: I

discovered last summer with certainty, that colours appear
different to me to what they do to others: The flowers of most
of the Cranesbills appear to me in the day, almost exactly sky
blue, whilst others call them deep pink; but happening once
to look at one in the night by candlelight I found it of a colour
as different as possible from daylight; it seemed then very
near yellow, but with a tincture of red; whilst no body else
said it differed from the daylight appearance, my brother
excepted, who seems to see asI do...

So, a modest pink flower served as the subject of Dalton’s first scien-
tific paper, and launched him on a quite extraordinary career. Dalton
(1766-1844), son of a weaver, went on to develop modern atomic
theory, formulated several laws on the behaviour of gases, and
attained an unlooked-for celebrity for his services to science, phil-
anthropy, and education. Across Europe, even those unfamiliar with
UNSEEN COLOURS 245
his reputation are familiar with his name: Daltonism is a common
term, in many languages, for colour-blindness.
In the nineteenth century, physiological descriptions of colour-
blindness provided clues to human heredity; contemporary genetics
has since repaid the favour, by providing us with a rich and complex
history of human colour perception. Their fertile conversation is the
subject of this chapter.

1 — Dalton’s eyeball
Extraordinary Facts relating to the Vision of Colours - Dalton’s first
scientific paper, published in 1798 — caused something of a stir. Not
only was industry making the world more colourful; it was making
ever-greater use of colour for its own safe and efficient running.
Signals and signs were proliferating, smoothing the relationship
between workers and their machines. What could be clearer, after
all, than a system of colour signals; a red light for Stop and a green
light for Go? Dalton’s paper — the first reasonably rigorous description
of the nature and prevalence of colour-blindness — suggested that
colour was not the universal language it appeared to be.
A century on, in the early hours of 15 November 1875, two express
trains collided head on at Lagerlunda, Sweden. The authorities were
mystified. How could the driver of the late-running northbound
express not have seen the red light waved by the station-master at
Bankeberg station? The puzzle inspired the physiologist Alarik
Frithjof Holmgren to test the vision of the 266 employees of the
Uppsala-Gavle railway. Thirteen, including a station-master and an
engineer, proved to have difficulties telling red from green. By the
end of the year, colour tests had been prescribed for railway and
shipping personnel in Sweden, and four year later, in 1879, the Amer-
ican government followed suit, commissioning Dr William Thomson
to devise a colour-blindness test for railway and shipping employees.
Given that the Lagerlunda crash had nothing to do with colour-
blindness, this seems a lot of work and fuss. (A reevaluation of the
original records showed that it had actually been caused by a failure
to follow the rules of the railway.”) It is not at all clear what tests
246 A NATURAL HISTORY OF SEEING

for colour-blindness have achieved. Those of us who are colour-

blind are more than capable of compensating for our disability;
colour-blind drivers have no trouble remembering that at traffic
lights, the red light is at the top, the green light is at the bottom.
Anyway, since almost every colour-blind person has at least some
sense of colour, practice and experience easily overcome common
errors.
The futility of colour testing was shown up most sharply by Mr
Trattles, a British seaman denied his First Mate’s certificate after
failing a colour vision test. Trattles took his case to Winston Churchill,
then President of the Board of Trade. His plight was discussed in
both Houses of Parliament, and a certificate was finally issued to
him after he took a steamer trip, in front of witnesses, down the
Thames, and correctly identified every navigation light he passed.
But bureaucracy turns slowly. Although the Australian government
will let you fly a commercial jet if you're colour-blind (and they are
the first country to do so), many major airlines still won't employ
you. In Britain, restrictions in some branches of the services, the
police and civil service are often strongly enforced. And don't even
think about driving a train unless you live in Russia. A training
course there, using pigment tables, has.greatly improved colour
recognition among colour-blind railway employees (rather, I suppose,
as studying swatch tables has lodged the previously ‘invisible’ colour
indigo in my memory).
Holmgren’s curiosity and industry were laudable; but the rapid
government action he inspired reflects less a mature measure of risk
and more the anxieties of an age accustoming itself to a new form of
human calamity: the industrial accident. Today, colour-blindness is no
more life-threatening than it ever was. It is, though, a significant hand-
icap to anyone whose job involves toiling through the colour-coded
arcades of the World Wide Web; too few Web designers remember to
make their natty designs friendly to colour-deficient eyes.

To the end, Dalton maintained the modest habits of his Quaker

upbringing. He never married, and few men could claim him as their
friend. He died, a recluse, on 27 July 1844. Nevertheless, when his
brethren turned up for the funeral, they found themselves contending
UNSEEN COLOURS 247
with crowds of fellow mourners. A staggering forty thousand people
lined up outside Manchester Town Hall for the chance to file past
the great man’s coffin. The funeral procession was a mile long. Shops
and business closed as a mark of respect. Dalton, an active and distin-
guished public figure, was internationally recognised as the founder
of modern chemistry.
The following day Dalton’s doctor, Joseph Ransome, began one of
medical history’s more peculiar post-mortems. Dalton’s eyes had not
gone the way of the rest of him but lay in a dish on Ransome’ table.
Ransome poured the humours of one eye into shallow glass dishes.
He noted that neither the aqueous humour in front of the lens, nor
the vitreous humour behind it, contained any trace of colour. The lens
was yellow, as you would expect from so elderly an eye. To double-
check his result, Ransome carefully removed layers from the back of
the second eye and looked through it. Examined through Dalton’s
dissected eye, red and green objects retained their natural hues.
Dalton, who had single-handedly developed modern atomic
theory and transformed our understanding of matter had, thanks to
Ransome, reached beyond the grave to conduct his final, posthumous,
experiment, a test of the theory he had nursed for fifty years to
account for his colour-blindness.
The result was conclusive: he was wrong,

The peculiarities of his impoverished colour vision niggled at Dalton

throughout his life. His letter to Elihu Robinson sets out the problem
vigorously, describing how the cranesbill in his garden changed
colour under different lights. This was peculiar, because others did
not see the change. To them, the Pelargonium zonale was the same
colour, whether swaying in the morning garden or studied by candle-
light. Significantly, that colour was pink - a colour Dalton could not
see. (He once shocked his co-religionists by turning up to prayer in
red stockings — he thought they were grey.)
What caused Dalton to see two colours in a flower when every-
body else saw only one? And how was this related to his inability
to distinguish green from red?
Thomas Young's theory of trichromacy provided a possible
solution to Dalton’s problem. In the normal trichromat eye, ail three
248 A NATURAL HISTORY OF SEEING

types of cone need to be stimulated equally to see white. Two-colour

vision is much more easily ‘bleached out’ because light has only to
stimulate two cones equally to be perceived as white. One by-product
of seeing white more often is that objects appear to glisten. (The
degree to which Homer emphasises the lustre of objects over their
hue lends circumstantial weight to Gladstone's suggestion that the
ancient Greeks were colour-blind.)
If Dalton was red-blind, Young argued, then a pink flower would
have no apparent colour: it would simply reflect ambient light. In
daylight it would reflect the blueness of daylight; indoors, it would
reflect the light from a yellow candle.* Young voiced his opinion -
and Dalton rejected it. What set him so firmly against Young's idea?
His biography holds a clue.
Dalton was a self-educated man barred, because of his religion,
from studying at Oxford or Cambridge. He acquired much of his
early knowledge from magazines, and his skills as a teacher were
honed by running a school in Kendal, in the Lake District, while he
was still only twelve years old. (He constantly had to evade threats
of violence from recalcitrant pupils older and bigger than he was.)
As a consequence, Dalton never really learned the trick of listening
to other people’s opinions. This proved, in many cases, to be a key
to success: ‘Having been in my progress so often misled by taking
for granted the results of others, he wrote, ‘I have determined to
write as little as possible but what I can attest by my own experience.
Dalton’s intellectual self-reliance lent his work an honesty and a
novelty that only rarely let him down.
Alas, it also prevented him from ever solving the colour-blind-
ness problem. This talented and original chemist, biologist, and
astronomer came up with no better theory than that the humour
filling his eyeballs must be blue. The day after his death, Dr Ransome
found no evidence to support him, and today, Young’s explanation
of Dalton’s colour-blindness is the one we are most comfortable with.

The patterns of heredity Dalton sketched out in his first paper

remained mysterious during his lifetime, but provided valuable clues
to succeeding generations, so that colour-blindness became the first
human genetic trait to be linked to a specific chromosome.
UNSEEN COLOURS 249
The ophthalmologist Johann Friedrich Horner was born in Zurich
on 27 March 1831, and died there fifty-five years later. A leading
campaigner for hygiene and measures against cholera, Horner
combined teaching with a crammed research schedule and a busy
private practice; over the course of his career he treated a staggering
100,000 patients.
His medical studies introduced him to the ophthalmoscope, at the
time only a few years old, and inspired him to specialise in ailments
of the eye. In 1876, he applied his medical knowledge to the mysteries
of “‘Daltonism. In place of Dalton’s anecdotal findings, Horner
presented the first scientific account of the hereditary transmission
of Daltonism: colour-blind fathers have colour-normal daughters
who are, in turn, mothers of colour-blind sons.
Horner noticed that this pattern had also been described in
connection with haemophilia. These ailments were being passed on
from generation to generation in the same way, via the mechanism
which dictated the sex of a human child - and in fact were firm
evidence of the existence of a sex-determining mechanism.
Edmund Beecher Wilson, America’s first cell biologist, working at
the turn of the twentieth century, was in no doubt that the whole
of an animal's development ‘is capable of a mechanical or physico-
chemical explanation, and Horner’s work provided firm evidence in
support of this conviction.’ Wilson asked ‘whether the embryo exists
preformed or predelineated in the egg from the beginning or whether
it is formed anew, step by step in each generation’ Certain devel-
opmental decisions — like the sex of the animal - seemed fixed and
were relatively impervious to changes in the environment. Those
decisions must somehow be stored in the single fertilised egg cell,
from which the animal grew.
It had been suspected, for some time, that the ‘tiny threads’ in cell
nuclei, spotted between 1879 and 1882 by the German anatomist
Walther Flemming, might be storing instructions for cell reproduc-
tion and development. Armed with the latest refinements in optical
microscopy, staining, and tissue sectioning, Wilson set about isolating
the evidence, and eventually published a paper showing that males
possessed one sex chromosome whereas females possessed two. This,
Wilson concluded, must be the difference that decides the sex of the
250 A NATURAL HISTORY OF SEEING

organism. Even as the paper went to press, Nettie Stevens — the fore-
most female biologist of her day - submitted a paper which flatly
contradicted him. Stevens had discovered, quite independently, that
males have two different chromosomes, ‘X’ and ‘Y, while women
have two ‘X’ chromosomes.
Stevens was right but so, oddly enough, was Wilson, for he had
stumbled upon a rare case. Some males are born with only one sex
chromosome. Much more frequently, though, they heed Nettie
Stevens, and boast two. Stevens had saved Wilson no end of tail-
chasing. By 1911 Wilson and his colleague, Thomas Morgan (Stevens's
old professor: a fitting coincidence), were able to show that colour-
blindness must somehow piggy-back on the X-chromosome.®
This is why girls only rarely suffer from colour-blindness. The only
way a girl can be colour-blind is if both her parents carry colour-
blind X-chromosomes. If (as is much more likely) a girl inherits a
colour-blind X-chromosome from just one parent, she will still be
able to see the full complement of colours, because, as an embryo,
only some of her cells will follow the instructions of her colour-
blind X-chromosome; others will listen to her second, colour-normal
X-chromosome. Although the initial ‘choice’ of which X-chromosome
to listen to is random, tissues instructed by the healthier X-chromo-
some tend to dominate the development process (the phenomenon
of “X-chromosome inactivation’). So, a woman carrying one ‘colour-
blind’ X-chromosome will still see the normal complement of colours.
Her sons will not be so lucky. A son inherits only one X-chromo-
some, and it must come from his mother. (An X-chromosome from
the father would give the child two X-chromosomes, making it a
girl rather than a boy.) Since one of the mother’s X-chromosomes
carries a gene for colour-blindness, there is a fifty per cent chance
that he will be born colour-blind.
This is where most accounts of colour-blindness end. But there
is more to tell, and in the telling, we get a much richer idea of how
human colour vision evolved.

Air-dried by Ransome and donated to the Manchester Literary and

Philosophical Society, Dalton’s eyes survived the German bombing
which destroyed so many of his valuable papers. In 1990, a team of
UNSEEN COLOURS 251

vision researchers obtained permission to re-examine the shrivelled

remains.’ This was, as one of the team, John Mollon, described it,
an exercise in ‘molecular biography’: a sincere (if peculiar) act of
homage to establish, once and for all, which of the several common
forms of colour-blindness afflicted Dalton.
Dalton, it turned out, was not ‘red-blind’ at all. He was green-
blind. Mollon’s team of ‘molecular biographers’ showed, by genetic
analysis, that Dalton’s green cones were missing. Where you would
expect there to be green cones, there were simply more red cones.
- This condition is ‘deuteranopia; and it is fairly uncommon. It is much
more usual for the green cones to be recognisably ‘green’ but dysfunc-
tional in some way. This gives rise to the milder condition of deuter-
anomaly, an umbrella heading, which covers a number of possible
problems, including pigments that react poorly to light, and pigments
that react to unexpected wavelengths. There are equivalent conditions
that affect the pigments for long-wavelength and short-wavelength
light: protanopia, protanomaly, tritanopia, tritanomaly . .. With so
many varieties of ‘colour-blind’ vision, it seems extraordinary that
anyone ever acquires perfect colour-normal vision. In fact, few do.
Jeremy Nathans, professor at Johns Hopkins University School of
Medicine, has a Golden Brain. This is an award from the Minerva
Foundation, a Berkeley-based charity with a special interest in brain
and vision research. Received in 1989, the award recognised Nathans’s
work in unpicking the genetics of colour vision — work which threw
a bright light on the many puzzles of colour-blindness. Like Walter
Gehring, Nathans started his career studying the genetics of
Drosophila. It was not long, however, before he fell in love with
rhodopsin — that ancient, ubiquitous pigment whose slow mutations,
read in the right way, are an historical record of vertebrate evolution.
Having isolated the gene that codes for rhodopsin in human rods,
Nathans went on to isolate the genes that code for human colour
receptors. He found that the gene for ‘blue’ receptors lies in an
extremely stable position on chromosome 7, a gene shared equally
between men and women. This is why an inability to discriminate
blues and yellows is extremely rare, and does not follow the inher-
itance patterns of the more usual types of colour-blindness. Nathans
was not surprised to find the genes for red and green receptors on
252 A NATURAL HISTORY OF SEEING

the X-chromosome, as it was long established that red-green colour

blindness was sex-linked.
With further study, he was able to show why colour-blindness is
so common. The genes for red and green receptors are virtually iden-
tical (their DNA sequences differ by only two per cent) and lie next
to each other on the X-chromosome. When eggs are formed in a
female embryo, the X-chromosomes from the maternal grandmother
and grandfather exchange sections randomly, to form the egg’s brand-
new X-chromosome. During this process, the ‘red pigment’ and ‘green
pigment’ genes can easily fall out of sequence. If you imagine the
confusion that could reign if two people with the same name move
in next door to each other, you begin to appreciate why accidents
are common during chromosome assembly. Misaligned and redun-
dant ‘red’ and ‘green’ genes account for ninety-five per cent of all
variations in human colour vision.
Nathans, who enjoys normal colour vision, was startled to discover
evidence of confusion on his own X-chromosome. He had expected
to find genes for two receptors: red and green. It turned out he had
three. Slips and duplications of this sort are common. Far from there
being an easy one-to-one relationship between gene and pigment, it
is quite usual for up to nine genes on the X-chromosome to cluster
together in an attempt to code for ‘red’ and ‘green’ This is why percep-
tion of middle- and long-wavelength light seems to vary even among
individuals who see colour normally.
Nathans’s work threw up an intriguing possibility. ‘Red’ genes
seem particularly liable to mutation; when more than one ‘red’ gene
is present on the X-chromosome, it often turns out that all the ‘red’
genes are defective, coding for receptors which register light towards
the greenish part of the spectrum. The result, for men, is some
degree of colour-blindness. What about women? A woman might
have one X-chromosome containing a healthy red gene, while the
other contains mutant ‘greenish-red’ genes. As we've seen, which
X-chromosome to ‘listen to’ is a decision embryonic cells take at
random. If there is a developmental disadvantage in following the
instructions of one X-chromosome, then cells which obey the
healthy chromosome will tend to dominate.
But where is the developmental disadvantage in having an extra
UNSEEN COLOURS 253

colour receptor? Might some women not enjoy four-colour vision?

The first known human tetrachromat is ‘Mrs M; an English social
worker. She was tracked down by Dr Gabriele Jordan, then at
Cambridge University and now at the University of Newcastle. Jordan
found her by testing the colour vision of women who had sons with
.a specific type of colour-blindness. Equipped with four receptors
instead of the usual three, Mrs M sees rare subtleties of colour.
Looking at a rainbow, she can segment it into ten different colours.*
The rest of us can see only seven — if that many.
Before Mrs M, we thought that the human brain was ‘wired’ for
three-colour vision. We were wrong. The brain can handle however
many channels of information the eye cares to throw at it. If our
colour perception is so plastic, the idea that the human colour sense
has changed markedly during recorded history no longer seems so
fanciful.

2 - A history of accidents

We are the products of editing, rather than of authorship.

George Wald

Gordon Lynn Walls - a renowned anatomist and expert on animal

eyes — was by most accounts not an easy man to get along with. He
was arrogant and dictatorial towards his students; of his colleagues,
he was demanding. Hed gone in for zoology because, he said, the
mathematics in his engineering course became too difficult. Much
of what he said about vision was not original, and some of it was
wrong. Walls died of a heart attack in August 1962, aged fifty-seven.
He is greatly missed.
He was one of vision research’s great communicators, as able to
write for children as to referee and edit original research. He was
also, as one obituary put it, ‘kind and warm, but ... at great pains
to conceal it. For every student he brow-beat in an argument there
was one — often the same one - whose casual question he answered
with an hour-long lecture of spell-binding clarity. If he was a martinet,
he was also a showman, demonstrating night vision experiments by
254 A NATURAL HISTORY OF SEEING

Bee
PSPS
BY ps
Oe;Les
6!eat

Gordon Lynn Walls's comparison of the visual cells of a leopard frog

and a tiger salamander. The four leopard frog cells are on the left.

Visual cell types in lizards.

‘UNSEEN COLOURS 255

A selection of rod cells; the right-most rod is human.

256 A NATURAL HISTORY OF SEEING

moonlight. If he was opinionated, he had a lot to be opinionated about:

he could hold an audience for hours on any subject for which he felt
enthusiasm. He was once spotted, in a rare leisure hour, conducting
impromptu, unofficial tours of the Palomar Observatory.’
Walls’s stories about eyes drew together the evolution of materials
and structures, the rules of optics, and animal behaviour in new and
exhilarating ways. He was as fascinated by how eyes moved as by
how they handled light; in how they looked, as much as how they
saw. He was especially intrigued by the sheer variety of rods and
cones to be found in vertebrate eyes. He traced the family traits of
different kinds of photoreceptor and was the first to realise that rods
are a specialised cell that must have evolved from a primordial cone.
Visual systems adapt to different niches over time; their story is
traceable through their genes and - more graphically and entertain-
ingly — through their anatomy. There is the type of frog which, having
returned from a nocturnal environment to a daylight one, boasts
two kinds of rod - and so has acquired a simple, ad hoc, form of
colour vision. Then there are the geckos. Geckos are descended from
lizards, animals which have spent so long in the sun that they have
lost their rods. When geckos adopted nocturnal habits, they had only
cones, so geckos are unique in being able to see at night using cones.*°
If only one line of Walls’s writings survives into posterity, it will be
the title to an essay, a line which handsomely sums up the credo of
vision science since Descartes: ‘Everything in the vertebrate eye
means something’
Wallss chief written contribution to vision research is the 785-page
epic The Vertebrate Eye and Its Adaptive Radiation, to which he also
contributed many illustrations. There is room here for only a tiny
sample of Walls’s meticulous drawings of the photoreceptors of
different species. The further one delves into Walls’s catalogue of verte-
brate wonders, the more striking nature's variety becomes. There are
cones that are tipped with oil. The oil droplet is coloured, restricting
the wavelength of light the cone can detect. By tweaking detectable
wavelengths this way, many vertebrates enjoy superbly accurate, less
‘noisy’ colour vision. Many vertebrates — but not mammals.
Many vertebrates have double or twin cones; variations in response
between the cones provides instant information about the wavelength
al
j
UNSEEN COLOURS 257

of light. Mammals, on the other hand, have only single photorecep-

tors.
Many vertebrate eyes adapt to different lighting conditions by
sliding their rods in and out of the pigment epithelium. Mammals
have no such talent.
_ Most vertebrates enjoy several dimensions of colour vision, and
tetrachromats are common. But not among mammals. Among
mammals, even three-colour vision is a rarity. Most are - from a
human point of view - colour-blind.
Since so many different kinds of vertebrates boast the same visual
equipment - oil droplets, retractable rods and so on — we know this
equipment must have evolved by the time mammals arose, around
240 million years ago. So why did mammals ‘throw away’ so much
good visual gear?

Every ecological niche is filled by a living thing of a characteristic

build, size, and shape. By studying the fossils of extinct animals, we
can often infer what kind of niche they filled. The earliest mammals
arose around the same time as the dinosaurs. These small, rodent-
like creatures evolved into two groups: marsupials, who give birth
early and nourish the developing embryo in a pouch; and placentals,
who give birth to more developed, capable young. Kangaroos are
marsupials. Humans are placentals. Marsupials evolved to fill daylit
niches, and they have enjoyed stable three-colour vision for over 240
million years,*t whereas placentals evolved to occupy nocturnal
niches. Humans are the descendants of night-dwellers, and our excel-
lent night vision (and it is excellent, however much people in indus-
trialised countries never get to appreciate it) is one consequence of
our spell in the dark. Another consequence is our relatively gimcrack
and faulty colour vision.

The cobbled-together quality of human colour vision particularly

shows in the anatomy of the retina. You may recall that in Chapter
Four I mentioned the problem of chromatic aberration: the fact that
lights of different colours come into focus at different depths inside
the vertebrate eye. In humans, green and yellow lights come into focus
at the very centre of the fovea, but blue light reaches the fovea and
258 A NATURAL HISTORY OF SEEING

surrounding retina in a blurred wash. The human eye handles chro-

matic aberration by arranging its cones in concentric circles. Blue
cones are not found in the fovea but are scattered among the rods
filling the rest of the retina.
There is more to this story: first, blue cones are rare, representing
only two per cent of all cones. Another thing: look again at the figure
on page 227; as orthodox a diagram as you could wish for, which
shows the sensitivity of human photoreceptors to light of different
wavelengths. The red cone is most excited by greenish-yellowish light;
the green cone is most excited by green light and the blue cone is
most excited by blue light. This diagram has been copied so often,
you may be forgiven for thinking that it is not controversial. But there
is a problem.
The vertical axis of the graph measures the percentage of cones
stimulated by a particular wavelength of light, or if you prefer, the
percentage chance that a particular cone will react to a particular
wavelength. It does not indicate how powerfully the cones respond.
The aggregate responses of blue cones, which represent only two per
cent of the total cone population, are nothing like as powerful as
the responses of red or green cones. The diagram does not reflect
this disparity in signal strength, something which is often overlooked.
Blue-cone signals are maddeningly few. George Wald never
managed to detect any, and this was a man whose Nobel prize recog-
nised his work in identifying cones by spectroscopy. Their weakness
has driven at least one contemporary writer on vision — Gerald Huth,
formerly of the University of Southern California — to dismiss their
existence entirely, and construct, in their place, an entirely original
(and often persuasive) theory of colour perception.'? A more conser-
vative interpretation is that our brains must be paying especial atten-
tion to ‘blue’ signals.
The idea that blue-cone information was being handled as a ‘special
case’ was given a boost in 1998, when a paper in Nature reported that
the blue-cone signals in Macaque monkeys are sluggish - running
about 30 milliseconds behind signals from red and green cones.
Further studies uncovered evidence that there are indeed two separate
channels for colour information - a red-green and a blue channel -
which are anatomically separate. In the foetus, they develop in different
UNSEEN COLOURS 259
ways and at different rates; their tissues don't even share the same
immune responses.’* The obvious implication — that they developed
independently — is irresistible. But can we prove it?

In 1986, Jeremy Nathans found firm genetic evidence of human

colour visions split nature.’* The gene for blue cones is ancient: it
“pre-dates the entire mammalian line, and is common to many different
vertebrates. Establishing the age of red and green cones was more
problematic. A survey in 1981 found that all vertebrates have a
‘reddish-greenish receptor of some sort, and the gene that codes for
this may be, in many cases, as ancient as the gene that codes for the
blue cone. But this gene has mutated many times; mammals, in their
reacquisition of colour vision, may have acquired, lost, and reacquired
dimensions of colour vision throughout their evolutionary history.
The history of our own species is a case in point: one of our ances-
tors managed to duplicate its ‘greenish-reddish’ cone so as to acquire
separate classes of ‘red’ and ‘green’ cone. We can date this event — the
birth of primate trichromacy — by comparing the anatomies of living
species on different continents.*® The primates of South America,
which broke away from Africa around sixty-three million years ago,
have only two genes to code for cones. Their blue cone is analogous
to our own; the other is a ‘greenish-reddish’ cone. So, the appear-
ance of green and red cone types must have taken place more recently
than 63Ma. More abstruse genetic studies suggest that primate three-
colour vision is younger still; just thirty to forty million years old.
In evolutionary terms, that is a last-minute job. No wonder it goes
wrong so often.
Human colour perception is not the single, seamless sense it appears
to be, but the detritus left behind by an untold number of genetic acci-
dents. Our plastic and sense-hungry nervous system wires these mis-
matched parts together into a single, often imperfect, machine.
The 607 islands of Micronesia are spread across one and a half
thousand square miles of the Western Central Pacific Ocean. There
are no mountains, or none worth the name; their highest point is
the volcano Agrihan which, at under a thousand metres, is somewhat
lower than Snowdon, in Wales. Most of these very beautiful islands
are little more than scraps of coral, barely breaking the waves that
260 A NATURAL HISTORY OF SEEING

surround them. In a big enough storm, some of them fail to do even

that. In 1775, a thousand people were living on the island of Pingelap
when Typhoon Lengkieki engulfed the island. Nine out of ten
islanders were killed outright; almost everyone else starved to death.
Only twenty survived. Little by little, the island and its people recov-
ered, and although intermarriage between close relatives was, of
necessity, common, four generations lived and died on the island
without suffering any obvious medical problems. So, it was a much-
recovered community that suffered the final and oddest consequence
of the typhoon. Children began to be born without a sense of colour.
Their world was monochrome; a fuzzy, easily dazzled matrix of greys.
With time, the trickle of cases became a flood, with each succeeding
generation producing more and more totally colour-blind (achro-
matopsic) children. Today, between five and ten per cent of the island's
population of 3,000 are unable to see any colours at all.
By tracing the islanders’ genealogies, we can see what went wrong.
One of the original survivors of the typhoon, the island’s young chief,
replenished the island with children, but he passed on a rare genetic
defect. In the normal run of things such defects are almost never
expressed normally, and certainly the young chief would not have been
aware of any problem. However, in Pingelap, intermarriage between
his descendants spread the defect throughout the population so that,
with each new generation, the chances of developing the defect
astronomically increased. Yet the Pingelap islanders, afflicted with
total colour-blindness and photophobia, are far from devastated by
their condition and continue to flourish. They have repopulated their
island, digging their disability ever deeper into the gene pool with
each passing generation. Today, one in three islanders is, like their
illustrious forefather, carrier of the condition.
For humans, the evolutionary pressure to maintain colour vision
is decidedly slack. Colour-blindness affects many of us - but we do
not seem greatly inconvenienced. Eight per cent of males are born
colour-blind, which suggests that colour vision is something that
our species can learn to live without. The history of human colour
vision, as it emerges from researches into our genome, seems to have
more to do with the vagaries of genetic mutation and mistake, than
with any grim narrative of do-or-die adaptation.
UNSEEN COLOURS 261

3 — Inside Leibnitz’s mill

‘There are two, virtually independent, pathways for colour informa-

tion, and they both terminate at the brain’s first major hub for visual
information, the lateral geniculate nucleus (LGN). The LGN treats
the information it receives much as the horseshoe crab’s retina treats
light. For the LGN, contrast is everything. It compares the signals
from red cones to the signals from green cones, and enhances the
difference. Given that red cones and green cones respond to similar
wavelengths of light, this is just as well: if the differences in signal
were not enhanced, we would never be able to distinguish all the
colours that lie between red and green. The LGN processes blue-
cone signals differently, comparing blue-cone responses to the
combined responses of red and green cones. What is not blue is red-
and-green; in other words, yellow.
If this scheme sounds oddly familiar, look no further than Ewald
Hering’s theory of opponent colours. Proposed in 1872, this theory
holds that two colour pairs — red-and-green and blue-and-yellow -
provide the axes for our two dimensions of colour vision. So, was
Hering right?
In a sense, yes: there is such a thing as colour opponency and
there are two channels for colour information. Hering’s physiological
insight found its foundation in anatomy. His achievement was real.
The trouble is, as experiments grow in sophistication — so that we
experiment with very many precisely graduated colours and not just
the bold ‘primaries - the muddier our results become. Most worry-
ingly of all, our reported experiences of colour do not tally with the
activities of our colour channels as measured by instruments.
This may sound like an anomaly apparent only to a handful of
technicians but, given the right conditions, the matching problem
expands to threaten the very foundations of colour science. In his
essay “The Case of the Colourblind Painter; the neurologist Oliver
Sacks described the sad case of an artist, poisoned by carbon monoxide
and involved, hours later, in a car crash. Returning to his studio, he
262 A NATURAL HISTORY OF SEEING

found his brightly coloured canvases had become murky grey daubs.
The paintings were meaningless and unfamiliar, for there was no colour
in them anywhere. This man, with perfectly functioning eyes, through
subtle but devastating damage to his brain, could no longer see colour.’”
The London-based neurologist Semir Zeki has considered similar
cases among people and monkeys and has found that, while damage
of this sort robs an individual of the experience of colour, the mech-
anisms of colour vision continue to function.’* Asked to match up
coloured counters, people with no experience of colour are still able
to match up colours perfectly. They just don't see them.

There is no reason why there should ever be an easy, one-to-one rela-

tionship between the structures of our brains and the phenomena we
experience. When, as in the case of opponent colours, this almost
happens, the confusion generated is immense; John Mollon, for one,
believes colour science disappeared into a cul de sac for years because
of the general, and mistaken, assumption that anatomy and genetics
had somehow ‘discovered’ the hiding-place of Hering’s opponent
colours.’?
Because they have no anchor in our experience, our mechanical
models of how the mind works are astonishingly varied, changeable,
and temporary. Inevitably, an analogy is drawn between the mind
and whatever happens to be the most complicated information-
handling system of the time. Up to three hundred years ago, the soul
was the fluid part of a complicated plumbing problem, then steam
power provided a better idea of how power and information might
be generated and handled in the body. By 1861, electricity was being
harnessed to shuffle information through a telegraph system, and it
was generally thought that consciousness was conducted through —
surprise, surprise — a network of electrical switches.
The computer provided the next big, and relatively futile, change
of analogy. In 1966, at the Massachusetts Institute of Technology,
the robotics researcher Marvin Minsky handed one of his graduate
students what he thought would be a relatively easy summer project:
to build a system of machine vision. This raises a complacent smile
on the faces of today’s students: how naive of Minsky, to imagine
that the human brain is like a computer!
UNSEEN COLOURS 263

The parade of mechanical analogies troubled Ernst Mach, who

was to become one of the modern era's great philosophers of science.
Mach understood that we needed a new vocabulary to talk about
sensations. ‘Bodies do not produce sensations; he wrote, ‘but
complexes of elements (complexes of sensations) make up bodies:”°
Our bodies are not the same today as they were yesterday. Thought
by thought, cell by cell, calorie by calorie, the body plays its part in
the flux of things. We talk as if bodies are real and abiding, while
sensations are fleeting, but sensations have their origins in the physics
which contains us. Colours, sounds, spaces, and times are the ultimate
elements, and we their temporary nexus. ‘By the recognition of this
fact; Mach assures us, ‘many points of physiology and physics assume
more distinct and more economical forms, and many spurious prob-
lems are disposed of?
Mach’s curious, disembodied way of discussing sensation is
catching on, but slowly. It is still depressingly easy to find ‘new
accounts of consciousness that talk about the brain as though it were
a fleshy computer; and a digital computer at that.
In the next, and final, chapter we shall see where this kind of
thinking is leading us - for both good and ill.
TEN

Making eyes to see

—————

A camera is no more a copy of an eye than the wing of a

bird is a copy of that of an insect. Each is the product of an
independent evolution; and if this has brought the camera
and the eye together, it is not because one has mimicked the
other, but because both have had to meet the same problems,
and have frequently done so in the same way.
George Wald

Around 1790, three men came to Bahrain from what is now the
United Arab Emirates and sought treatment from the American
mission doctor there. He could do nothing for them. Their eyes had
been burned out with a hot iron.
There was, at that time, a story circulating of a European woman
who, while visiting the ladies of the Persian court, came across a
small boy walking about blindfolded. She asked him what the game
was. He replied that he was practising; when he grew up, he was
bound to have his eyes put out sooner or later.
Around 1828, as the Ottoman Empire began its slow decline, it
seemed the practice of blinding one’s political enemies fell out of
fashion. But blinding the enemy enjoyed a revival during the First
MAKING EYES TO SEE 265

World War, especially on the Western Front, where mustard gas

attacks blinded thousands. Though many recovered some vision, their
eyes were irreparably damaged.
Now it is the turn of the New World.
One night in 1994, a year into the United States’ involvement in
a UN humanitarian mission to Somalia, Emil Bedard and a company
- of 175 marines found themselves in front of a mob of about 10,000
demonstrators. They were completely defenceless. They had guns -
lots of guns — but that was the problem. If they used them, they
would kill unarmed civilians, and then the mob would surely kill
them.
‘Several marines took serious injuries ranging from lacerations to
broken jaws, Bedard recalls. ‘As I stood with them, I thought, “How
can we deal with this more effectively?” A year later, he got his answer.
In February 1995, US Marine Lt Gen Anthony Zinni led the 1st Marine
Expeditionary Force on a mission to bring out 2,400 UN troops from
Mogadishu. For years, he had been lobbying for weapons that would
allow his men alternatives to killing people. Of the new weapons he
took to Somalia, two hit the headlines. One was a ‘disabling’ foam so
innocuous that evening news broadcasts ran footage of its deployment
alongside a clip from the film Ghostbusters where the actor Bill Murray
is felled by slime.* The other was a laser powerful enough to blind
people.
‘The rescue operation was a complete success. The problems the
US marines had with their non-lethal weaponry cannot be held
against them: they were unfamiliar weapons, deployed for the first
time in a dangerous environment. They were used moderately.
Commanders on the ground were sufficiently concerned about the
risks of permanently blinding civilians they ordered that the lasers
be ‘de-tuned’ turning them into little more than powerful torches.
On 1 March 1995, commandoes of the US Navy SEAL Team 5
were positioned at the south end of Mogadishu airport when they
saw a Somali man aim at them with a rocket-propelled grenade.
They used their laser to target him, and a seat sniper killed him.
The team - trained in non-lethal combat and looking forward to
fighting a new, bloodless kind of battle - were not at all pleased:
‘We were not allowed to disable these guys because that was
266 A NATURAL HISTORY OF SEEING

considered inhumane, one complained. ‘Putting a bullet in their head

is somehow more humane?’ .
All weapons cause horrible injuries, intentionally or not. The
problem with dazzling lasers is that the specific horrifying injury
they inflict — blindness - cannot be treated. False legs let you walk,
after a fashion; prosthetic arms give you some limited dexterity; but
there are no prosthetic eyes. In September 1995, the US Department
of Defense banned the development of lasers specifically designed
to blind people. In 1996 the ban became global, under the Geneva
Convention.
None the less, attempts to produce an ‘eye-safe’ dazzling laser continue.
The Veiling Glare Laser, under development at the Joint Non-Lethal
Weapons Directorate in Quantico, Virginia, uses near-ultraviolet light
to make the lens of the eye fluoresce; it is painlessly, but effectively,
dazzling. The long-term risks to vision have yet to be assessed.* Mean-
while, cruder devices find favour in the field. In May 2006, the US
military - wary of shooting civilians who fail to heed their warnings
at checkpoints in Iraq — fitted some of their M4 rifles with a laser that
is like shining a big light in your eyes; according to Lt Col Barry Venable,
a Pentagon spokesman. ‘Calling them weapons would be a misnomer}
he adds, with an eye to international accords and trigger-happy jour-
nalists. ‘It’s not the “death ray” *
But the use of strong lights to control civilians raises an important
wider point — well-made by Emil Bedard, though he is now a vocal
and active convert to non-lethal weaponry: “ . . not everyone will
accept these capabilities solely because they are intended to be non-
lethal. Some will argue that these capabilities, if misused or in the
wrong hands, may have catastrophic results’

Throughout human history, we have sought to take control of vision.

The earliest recorded attack on enemy vision dates from the Pelo-
ponnesian War, in the fifth century Bc, when Spartan forces lit a
mixture of wood, pitch, and sulphur under the walls of a city to
incapacitate the defenders. For most of this chapter, Iwant to consider
the happier consequences that follow from that desire: our attempts
to restore sight to the blind and, more recently, our attempts to build
eyes so that machines can be given the gift of sight.
MAKING EYES TO SEE 267

The first part considers the early history of eye medicine, and the
difficulties confronting surgeons who restore sight to a person who
has had no previous experience of vision. The chief, and besetting,
problem with such an operation is that vision is learned through
movement. A person with restored sight may be aware of light but
they will probably not be able to see. In the second part of this chapter,
we will see how the relationship between movement and vision
inspired some important physiological studies, and informed the most
productive of today’s assays at machine vision. Finally, we look at
how machine vision and eye medicine are married in the latest devel-
opments in prosthetic vision. The devices promise renewed sight to
the blind, an enriched visual experience for us all, and a possible
future more nightmarish than any a non-lethal weapon could inspire.

1 — ‘Suppose a man born blind ..?

Even his fiercest critic admitted Henry Blackbourne ‘could couch

cataract well.° Operating in the early seventeenth century, Black-
bourne was an itinerant oculist. Oculists were self-appointed
surgeons, whose range of operations typically included cures for
deafness and the remedial treatment of hare lip and cleft palate. Their
signature operation - the one that had them setting up scaffolds on
the village green and ringing bells to gather an audience - was
‘couching® A lens afflicted by cataract blocks the view: if the lens is
moved out the way of the pupil, the afflicted eye gains unimpeded
(if blurry) vision. Since extraction of the lens was not possible at
this time without a great deal of trauma, couching needles were used
to push the cataracted lens down inside the cushion of the iris.
The operation blinded as many people as it cured, but that was
neither here nor there, since cataract, left untreated, is a one-way
ticket to total blindness. Blackbourne’s critic, Richard Banister,
writing in 1622, was more concerned with the way Blackbourne
‘would cozen and deceive men of great sums of money by taking
incurable diseases in hand’ Blackbourne’s greatest weakness was
women: ‘He was amorously given to several ... so that his cozening
made him fearfully to flee from place to place:
268 A NATURAL HISTORY OF SEEING

Banister wanted to get rid of itinerants like Blackbourne. Even

more, he wanted to get rid of the ‘boundless boldness of many
women, who for lack of learning cannot be acquainted with the theo-
retic part, and yet dare venture on the practical. I believe, scarce
three of thirteen hundred can define or describe the names and
natures of the hundred and thirteen diseases of the eye...
He had a steep hill to climb. Just as much as Blackbourne, he was
working from folklore and guesswork. To give us some measure of
how little was known about the eye at this period, we need only look
at the records of Walter Bayley. Bayley was Queen Elizabeth I's physi-
cian, a learned man who wrote on the preservation of eyesight. Bathing
sore eyes in urine is excellent advice - urine is sterile and mildly anti-
biotic — but in an age that regularly used urine as a facial cleanser, it
was hardly ground-breaking. His other big recommendation — drinking
ale to strengthen the sight - is even harder to justify.
The problem is more than a mere lack of knowledge. There was
a palpable reluctance on the part of the medical establishment to
treat the eye with any degree of seriousness. As late as 1666, nearly
half a century after Banister’s sterling efforts at gentrification, Robert
Turner (author of The Compleat Bone Setter) gave as treatment for
squint: *.. . the blood of a turtle or head of a bat, powdered of course.
Or for weak sight, the eyes of a cow hung around the neck’
In Britain, major advances in eye surgery had to wait until the estab-
lishment of Moorfields Hospital, the world’s first dedicated eye hospital,
in 1805. But there were some glimmers of hope. In 1728, William
Cheselden, of St Thomas's Hospital in London, removed cataracted
lenses from the eyes of a boy of thirteen. This was not the first such
operation to restore sight to a blind person; the earliest we know about
gave vision to a thirty-year-old man, in 1020. It is, however, the first
full account we have of the operation and its aftermath, and it began
a modest but growing fashion for making the blind to see.
Studying the case, the French physician Julien Offray de La Mettrie
(1709-1851) — author of Man a Machine and a notorious materialist
- noted the difficulty the boy had in seeing with his newly clear
eyes. He had vision but he lacked education in the ways of seeing.
The idea that learning has a role to play in sensation reaches forward
to our own day, to the work of Gerald Westheimer, Michael Land,
MAKING EYES TO SEE 269

and others, who are studying how the eye learns to sort salient infor-
mation from the world. It also reaches back to the work of William
Molyneux and the empiricist philosopher John Locke, whose dialogue
on sensory learning provided Locke’s Essay Concerning Human
Understanding of 1690 with one of its most celebrated passages.
Molyneux wrote to Locke:

Suppose a man born blind and now adult, and taught by his
touch to distinguish between a cube and a sphere of the same
metal. Suppose then the cube and sphere were placed on a table,
and the blind man made to see: query, whether by his sight,
before he touched them, could he distinguish and tell which
was the globe and which the cube? .. . The acute and judicious
proposer answers: not. For though he has obtained the expe-
rience of how the globe, how the cube, affects his touch, yet he
has not yet attained the experience that what affects his touch
so or so, must affect his sight, so or so . . .7

Cheselden’s operation (and subsequent, improved attempts) seemed

to bear this out: curing a person of blindness was one thing; enabling
them to see was quite another.
In 1777, Franz Anton Mesmer (1734-1815) ran into dreadful
trouble when he attempted to restore the sight of Maria, daughter
of Empress Maria Theresa’s Court Councillor, Joseph Anton von
Paradis. How Maria von Paradis lost her sight - it faded steadily
between two and five years old — is not known. It is generally assumed
that her blindness was ‘psychogenic’; that there was nothing wrong
with her eyes, her optic nerves, or the visual centres of her brain,
and that her blindness was the product of a dissociative disorder. It
was (to put it unkindly) all in her mind. This would certainly serve
to explain Mesmer’s success. Today, his treatments, which involved
the judicious application of what he called ‘animal magnetism, carry
an unmistakable whiff of snake oil. But the evidence suggests that
Mesmer was sincere. Mesmer’s life’s work was to put the phenomenon
of faith healing on a scientific basis. If he failed, it was an interesting
failure, revealing, to later researchers, the power of medical suggestion
(the ‘placebo effect’).
270 A NATURAL HISTORY OF SEEING

When Mesmer met her, Maria was eighteen years old, a musical
prodigy (Mesmer’s friend Mozart wrote his piano concerto in B flat
for her) and a favourite of the Empress. In exerting (at the very least)
the power of suggestion, Mesmer was able to unlock Maria's
psychogenic blindness, and for a while, she appeared to be recovering
her sight. There are many versions of what went wrong and each is
more purple than the last. Because Maria's family received a pension
because of her blindness, one version has it that her mother forcibly
removed Maria from Mesmer’s care, slapping her into submission
and a return of her blindness. If her mother were truly vicious, this
might, for the prurient, suggest a cause: psychogenic blindness is
often triggered by traumatic and unspeakable events. Another has
it that Mesmer seduced his eighteen-year-old client. I wouldnt put
it past him: Mesmer’s biography, a bewildering mix of genuine intel-
lectual labour and blatant charlatanism, inspires wry admiration
more than moral confidence. However, this story is no more substan-
tiated than the first; we had better dismiss them both.
Much more, revealing, I think, is the hostility Mesmer faced from
the doctors who had been treating Maria unsuccessfully for the
previous ten years. Their complaints, however much they were moti-
vated by jealousy, were public documents, soberly expressed, and
devastating in their simplicity: if Maria could see, as Mesmer claimed,
why could she not be taught to identify anything? Why could she
not call anything by name? Mesmer had given Maria back her sight,
or the suggestion of sight, but Maria did not seem to know what to
do with it. Eventually, amid accusations and counter-accusations, the
Empress had a quiet word: ‘all this nonsense’ had to stop. Mesmer,
disappointed, left Vienna for Paris, and Maria reacquired her
blindness.
It served her well. Able to focus again on the sound world she
had inhabited since she was five, Maria became a celebrated composer
and musician. She wrote at least five operas, as well as cantatas, lieder,
and many piano works; her longest European tour lasted three years.
I find her loss of vision sad, but I labour under a sighted person’s
prejudices. I doubt any blind readers will find much to mourn in
the life of one of the most accomplished, independent, well-traveled,
and successful women of her age.
MAKING EYES TO SEE aiygl

Mesmer’s notoriety, the jaundiced hindsight we apply to his works,

and all the frills and furbelows of court gossip and newspaper prattle
have obscured a valuable, and utterly typical, case. Maria, restored
to the miraculous world of vision, reckoned she was better off blind.
As his treatments took effect, Mesmer reports Maria saying: ‘How
comes it that I now find myself less happy than before? Everything
that I see causes me a disagreeable emotion. Oh, I was much more
at ease in my blindness. Richard Gregory and Jean Wallace's moving
report from 1963, ‘Recovery from early blindness’, contains strong
parallels with Maria’ story.
The two Cambridge researchers tell the story of $.B. a man born
in 1906, blind since he was ten months old. His sight was restored,
by corneal grafts, when he was fifty-two. Gregory and Wallace docu-
mented his recovering vision, his attempts to develop a visual
memory, and the painful steps he took to relate the world he knew
to the confusing play of colours and forms in his newly opened eyes.
They also record his depressions, and how he changed from an ebul-
lient and assertive blind man to an under-confident and morose
sighted one. ‘His story is in some ways tragic; they wrote. ‘He suffered
one of the greatest handicaps, and yet he lived with energy and
enthusiasm. When his handicap was apparently swept away, as by a
miracle, he lost his peace and his self-respect:*
S.B. had been trained as a cobbler. Early in their study, Gregory
and Wallace thought it would be a pleasant to take him to the Science
Museum in London. There, in a glass case, stood a screw-cutting
lathe; a machine S.B. had been familiar with throughout his life, and
always wished that he could use.

He was quite unable to say anything about it, except that he

thought the nearest part was a handle .. . He complained that
he could not see the cutting edge, or the metal being worked, or
anything else about it, and appeared rather agitated. We then
asked a Museum Attendant for the case to be opened, and S.B.
was allowed to touch the lathe. The result was startling; he ran
his hands deftly over the machine, touching first the transverse
feed handle and confidently naming it as a handle, and then on
to the saddle, the bed and the head-stock of the lathe. He ran
272 A NATURAL HISTORY OF SEEING

his hands eagerly over the lathe, with his eyes tight shut. Then
he stood back a little and opened his eyes and said: “Now that
I’ve felt it I can see:

We are physical beings. As infants, we map objects by touch, and by

relating touch to visual experience, we refine our personal rules about
how the visual world is to be interpreted. After years of disuse, we
can hardly expect a blind person’s brain to be able to disentangle a
world full of colour, shade, depth, and movement, nor should we
assume that laggard sight can ever ‘catch up’ enough to be fully inte-
grated with other, mature, experienced senses.
One experiment - disturbing, though necessary if we ever want
to treat blindness properly - deprived kittens of visual stimulation
at critical moments in their development. The animals never learned
to see properly. What is true of that handful of unfortunate cats is
just as true of human beings. To date, no person, blind during child-
hood and restored to sight, has managed to enjoy all the pleasures
of the visual world. All of them retain and rely upon at least some
‘blind’ skills.

2 — How bumping into things teaches us to see

A popular Victorian visual novelty was a handsomely furnished but

doll-sized room, nailed over a peephole in a door. One can imagine
the consternation of observers, opening the door to this sumptuous
interior to be confronted by a different room altogether.
In 1952, Adelbert Ames Jr (1880-1955) took the novelty room to
new levels of philosophical sophistication in a series of witty exper-
iments conducted at the Dartmouth Eye Institute, of which he was
a founder. Ames had been a painter, until his studies of optics got
the better of him. While becoming the world’s leading expert on
disordered binocular vision, he took considerable delight in fash-
ioning optical illusions to demonstrate the level of guesswork
involved in seeing. His most famous construction is the Ames Room,
a distorted cube which, looked at through a window, appears regular.
However, when two people stand either side of the room, one appears
MAKING EYES TO SEE 273
dwarfed, while the other appears to have swelled to giant propor-
tions; their head practically touches the ceiling.
Since sudden changes in size never happen except in Wonderland,
you would be forgiven for thinking that the illusion must break down
at this point. But a photograph taken through the window of an
_Ames Room (see last page of colour section) shows how persistent
the illusion can be. The eye is faced with two irreconcilable absur-
dities: either the people have shrunk and expanded, or everything
else is lined up to fool the eye. As far as the eye is concerned, it is
easier to believe that something has gone wrong with the people
than that something has gone wrong with everything else.
This is interesting, because it reveals how devoted our eyes are to
the laws of optics. Even more intriguing is what happens if you throw
tennis balls through the window, or poke at the interior of the room
with a long stick. Physically testing the environment reveals its true
distorted shape; after a little exploration of this kind, the eye ceases
to be fooled. Gradually, it swaps one absurdity for the other, and sees
the room for the strange, topsy-turvy space that it is.
Experimental psychologists possess a bottomless appetite for these
sorts of games.? Hermann von Helmholtz, a keen hiker, once wrote
about the effect of bending over and looking through his legs at the
surrounding mountains. (Sudden inversion of the view accentuates
one’s depth perception wonderfully. I tried this minutes before tack-
ling some desperate, broken ascent on the Pic du Midi in the Pyrenees,
and given the effect on my morale, I rather wish I hadn't.) In the
1890s, George Stratton of the University of California went one better,
and built volunteers several lens and mirror systems to view things
from unfamiliar angles. For one experiment, Stratton wore inverting
goggles, turning his world upside-down; he found that things righted
themselves after a couple of days. Once he took the goggles off, it
took about the same amount of time for his vision to right itself
again.
Did Stratton ‘get used’ to his inverted vision, or did the world
‘spring upright’ for him? Are these simply two ways of describing
the same experience? This question - which steers perilously close
to ‘is my green the same as your green territory — inspired Ivo Kohler,
at the University of Innsbruck, to repeat Stratton’s work. As far as I
274 A NATURAL HISTORY OF SEEING

can tell, Kohler holds the world record for the longest continuous
psychological experiment. For 124 days, Kohler wore glasses fitted
with binocular prisms, reversing his field of view left to right. There
can be no doubt that he adapted to his strange glasses: he wore them
even while riding around Innsbruck on his motorbike.”? In another
experiment, using spectacles that turned the view upside-down,
Kohler noted that visual adjustment is not a steady process. In the
early stages, when the world still appears upside-down, touching
objects tends to make them look suddenly normal. Even touching
things at one remove - for example with a long stick - is enough
to turn the world the right way up. In addition, objects right
themselves if they seem logically absurd while inverted. A candle
might appear inverted until it is lit, when it flips up the right way.
“The different sense-perceptions, Stratton wrote, ‘are organized into
one harmonious spatial system. The harmony is found to consist in
having our experience meet our expectations. We do not learn to
see through our eyes alone, but by the movements of our bodies
through our environment. Learning to see is as much about moving
and touching as it is about focusing and converging the eyes. As
long as our environment remains reasonably stable, so that objects
remain solid, retain their colouring, and don't deliberately line up to
appear joined when they're not, we rarely run into difficulty.

Trust the world to teach your senses: this lesson has not been lost
on the robotics community. At the Massachusetts Institute of Tech-
nology, Rodney Brooks has spent twenty-five years directing an arti-
ficial intelligence (AI) laboratory unlike any other. For much of the
1980s, it resembled a grown-up version of the family living room
when he was a child of twelve, building robots that could respond
to light, wander around the floor, and negotiate obstacles. (His father
was a technician in a defence lab and used to bring home the bits.)
Brooks’ eighties robots looked like insects, but they are a lot simpler.
They have no memory to speak of and little scope for action. What
they do have, in spades, is input. They are equipped to sense every
bump, knock, and scrape. Brooks's team did not teach their robots
anything. They did not program them, in the usual sense. They did
not give them any a priori information about the environment. They
“ MAKING EYES TO SEE 275
trusted that the world itself, perceived through primitive senses,
would shape their robots’ behaviour. The lab turned out the most
life-like robots of their time, robots that developed simple responses
to simple hazards. Brooks's approach — dubbed ‘embodied or‘situated’
AI — revolutionised artificial intelligence.
_ In 1997, Brooks featured in a bizarre documentary by Errol Morris,
Fast, Cheap and Out of Control, about people whose work drew its
inspiration from the animal world. Its title very neatly sums up the
kind of world Brooks's robots are ushering in. Why build one heavy,
slow, extremely expensive and complicated machine to do a task, when
you can share that task out among a swarm of simpler, smaller, dumber
machines? Why build a single, impossibly expensive robot to map
Mars, when you could more easily cloud the Martian skies with tens
or hundreds of thousands of tiny, expendable, mechanical ‘bees’?
Fast, cheap, out-of-control robots. They have no complicated
instructions. They have no sense of mission. What they have is senses;
you could say that they are little more than ‘senses on legs’ As, year
by year, their senses are made subtler, so they are able to develop a
more refined response to their environment.
Robots built with Mars in mind are the first recipients of a new
generation of mechanical eyes; organs which mimic the navigational
abilities of compound eyes." Like dragonflies they detect the line of
the horizon, using ocelli made up of a handful of ommatidia. Like
bees, they measure their airspeed and their nearness to obstacles by
optic flow, noting the rate at which textures pass neighbouring
ommatidia. Like ants, they navigate using polarised light. The latest
really good prototype (it can hover over a target using its own senses,
and needs no instructions) is a miniature helicopter built by the
Australian researchers, Javaan Chahl and Mandyam Srinivasan. With
a length of 1.5 metres and a weight of seven kilogrammes, it is much
too big to go to Mars — not so much a bee as an albatross — but it
represents a huge logistical advance over poor Sojourner Rover. This,
the Mars Pathfinder’s mobile unit, crept off its ramp on 5 July 1997.
It had no senses and had to be directed centimetre by centimetre,
across the Martian surface, from Earth, 120 million miles away. In
thirty days, it covered fifty-two metres.
* * *
276 A NATURAL HISTORY OF SEEING

Brooks has not been slow to capitalise on the practical applications

of his work. (His cv as an entrepreneur is as long as his list of acad-
emic achievements.) We should not, however, lose sight of why robots
like these are made: they are built as experiments, to answer questions
about the nature of intelligence. The MIT lab’s headline achievement
of the 1990s was Cog, a humanoid robot with moving hands, arms
and eyes. It had a sense of touch, and a sense of vision. It could reach
out and touch objects, and it could both see and feel what it touched.
Cog has gone into honourable retirement now. With its experience
to guide them, what will Cog’s successors make of their world? Will
they go a step further, and take an interest in things? Will they learn
to point? Will they learn to follow each other's gaze, or even —- wonder
of wonders — acquire theory of mind? If being human is something
we learn through copying the visual behaviours of our parents and
our peers — as the work discussed in Chapter Five suggests — any
serious attempt to build a thinking machine will have seriously to
consider not just its physical senses, but also its social environment.
This idea dates back surprisingly far, to the work of the Russian
Lev Vygotsky, who put forward this view in the 1920s. Unfortunately,
his work was suppressed by the Soviet authorities, and was not avail-
able in the West until the 1960s.’? “The mechanism of social behavior
and the mechanism of consciousness are the same, Vygotsky wrote.
“We are aware of ourselves in that we are aware of others; and in an
analogous manner, we are aware of others because in our relationship
to ourselves we are the same as others in their relationship to us.
One Cog may never be enough. We may have to build a whole
society of Cogs; a task so patently Herculean that some critics wonder
whether there is much future in robotic approaches to AI.*4

3 — New eyes for old

We do not know his name - anonymity has been preserved for his
family’s sake — and his case has never been discussed in any formal
scientific papers. We know he was blind and that, towards the end of
the 1990s, he volunteered for a trial of an innovative system of prosthetic
vision, using brain implants. We know he found the implants a nuisance
MAKING EYES TO SEE 277
because one day, not really thinking about it, he took hold of the wires
connecting his brain to the machinery, and tugged.’ His death, following
a massive infection, brought to a crashing halt vision science’s most
extraordinary adventure to date: the attempt to bypass the eyes entirely,
and bring sight to people who are profoundly blind. But the project is
far from abandoned, and how such a technology is applied will become
an important ethical debate for the next generation.
People blind from birth or early childhood have difficulty learning
to see. However, in the developed world, the vast majority of blind
people lose their sight in adulthood. They know how to operate in
the visual world, and would certainly benefit from having their sight
restored. Because there are so many different ways to lose sight — so
many things that can go wrong in the complex machinery of vision
- the pressure is on to develop a system that can address all causes
of blindness at once: a final solution to the problem of sightlessness.
Work on projecting pictures directly into the brain began in
Cambridge, England, in the early 1960s, under the leadership of
Giles Brindley. Although there was no realistic prospect that these
experiments would restore her sight, a volunteer - a fifty-two-
year-old woman who had been blinded by glaucoma - agreed to
the installation of eighty electrodes over her visual cortex and,
under her scalp, an array of radio receivers. The idea was that a
transmission, picked up by a particular receiver, would stimulate
a corresponding electrode, and the tiny charge emitted by the elec-
trode would stimulate a portion of the woman’s visual cortex. But
what would she see? Brindley hoped to exploit a finding by the
pioneering Swedish neurologist Salomon Henschen. In 1893,
Henschen had been studying patients who had sustained injuries
to their visual cortex. Signals which start their journey in the
retina arrive at the area of the visual cortex which today we call
‘V1. Henschen tried matching the location of wounds to V1 to
the damaged areas of his patient’s visual field, and met with a
success that was positively daunting: he found that the scene
captured by the eyes is literally mapped over the surface of V1.
It is, admittedly, a funny sort of map: upside down, yes, and also
back-to-front, inside-out, sides-to-middle, convoluted and every
other form of twisted you could name. None the less, his discovery
278 A NATURAL HISTORY OF SEEING

- that there really is a screen (of sorts) in the brain —- was confirmed
during the Russo-Japanese War and during World War I in studies
of wounded soldiers.
By installing eighty electrodes over his volunteer's V1 region,
Brindley hoped that he had taken the first step in transmitting intel-
ligible maps on to a blind person's visual cortex.
In theory, Brindley’s system might one day be used to provide his
subject with a kind of crude, eighty-pixel system of prosthetic vision.
But neither Brindley nor his volunteer expected that day to come
any time soon.
It didn't: predictably, less than half the woman's electrodes gener-
ated any kind of visual image. The others caused her to see stars.
And while some electrode responses caused her to see a single star
in a single location, others made her see several stars at once. There
was another problem: matching the location of the electrodes to
the location of the stars proved problematic. According to
Henschen, the correspondence should have been exact. But trying
to replicate the rippled surface of V1 with a rigid array of electrodes
and receivers proved heinously difficult; further complicated by
the fact that, whenever the volunteer moved her eyes, the stars
appeared in different parts of her visual field.
None the less, Brindley and his collaborator, the neurosurgeon
Walpole Lewin, had every reason to be optimistic. In 1968 they wrote:

Our findings strongly suggest that it will be possible, by

improving our prototype, to make a prosthesis that will permit
blind patients not only to avoid obstacles when walking, but to
read print or handwriting, perhaps at speeds comparable with
those habitual among sighted people.**

Another major player in the field was the American prosthetist,

William Dobelle (1941-2004). Dobelle’s lifelong interest in what he
called ‘the spare parts business’ began at home in Massachusetts; he
was just thirteen when he applied for his first patent for an improved
artificial hip. The senses were his particular fascination; lectures
delivered in the mid-seventies predicted just about every major
advance and approach in sensory prosthesis for the next thirty years..7
MAKING EYES TO SEE 279
Like Brindley in the UK, Dobelle drove forward primary research
with a series of ground-breaking operations. In 1978, he and his
colleagues implanted an array of sixty-four electrodes in the brain
of a genial, thirty-five-year-old Brooklynite, Jerry, whose optic nerves
had been severed by a gunshot wound. Though their system was in
essence the same as Brindley’s, Dobelle’s team got better results, not
least because the devices used in such Operations were growing
smaller, year after year. Dobelle’s study was reassuring: V1 did ‘map’
the visual scene, as Henschen had claimed. But there were still prob-
lems: like Brindley’s, Dobelle’s team found that many electrodes
produced multiple stars. More seriously, they found that their equip-
ment, if used continuously, dangerously overheated the brain.
Both Brindley and Dobelle reined in their expectations. With luck,
miniaturisation would eventually solve the heating issue. Meanwhile,
their systems of induced ‘dot vision seemed best suited, not to vision,
but to the reading of Braille. Getting blind people to see a blind
script sounds like a waste of time, but the results from both teams
were impressive. By 1979, Dobelle'’s patients could read Braille by
‘sight’ five times more quickly than they could by using their fingers.
The wait went on. And on: for nearly twenty years, no major
human experiment was conducted. What would be the point? Until
the stimulating electrodes were reduced to the size of individual
neurons, the prospects for this sort of prosthetic vision were stalled.
In 1996, a team led by E.M. Schmidt implanted the first such micro-
electrodes into the cortex of their volunteer, a forty-two-year-old
woman, blind since the age of twenty. The results were excellent, and
other human trials quickly followed. In hindsight, perhaps too little
thought was given to the greater risks of using this new equipment.
The microelectrodes were so fine, they could be implanted directly
into the brain, but their low output meant that they had to be phys-
ically connected to the system's machinery. This meant wires coming
out from the brain, vastly increasing the risk of infection and accident.
Ultimately, for one volunteer at the US National Institute of Health,
it meant his death, and an immediate moratorium for the project.
For both Brindley and Dobelle, who had waited so long to see
their visions realised, the death marked a sea-change in their careers.
Brindley began his professional career in neuroscience (and had
280 A NATURAL HISTORY OF SEEING

supervised through his doctorate the young David Marr, who went
on to be a hugely influential figure in the study of visual perception).
By the eighties, his interest in prosthetics had led him into many
diverse areas, from spinal injury to urology. He became world-famous
in 1983, by dropping his trousers at an American Urological Associ-
ation meeting in Las Vegas. It was a performance worthy of a Mel
Brooks film: ‘he walked down the aisle and let us touch it, one attendee
remembered; ‘People couldn't believe it wasn’t an implant: Brindley
was demonstrating how a powerful muscle relaxant, injected directly
into the penis, can generate an erection. It is generally reckoned that
Brindley’s show-stopping lecture ushered in the Viagra age.
Dobelle, on the other hand, who had once enjoyed ‘spare parts’ almost
for their own sake — he made significant contributions to hearing
implants and artificial hearts - had been deeply bitten by the vision
bug. Frustrated by what he saw as the needless new restrictions on his
research, he moved to Portugal and continued his work with human
volunteers there. It was a decision that led him away from the scientific
mainstream; he no longer published in the scientific journals, preferring
to talk to the press. He had plenty to talk about: by 2000, he had given
restricted vision back to Jerry, his volunteer of 1978.
The “‘Dobelle eye’ Jerry demonstrated to astounded journalists
consists of an ultrasound scanner, mounted on spectacles, which
feeds spatial information to a 10lb computer worn at the waist.
(Advances in computers were crucial to the system's success: only a
year before, Jerry's apparatus had weighed two hundred times that
and was as big as a bookshelf.) Once the information has been simpli-
fied, to provide information only about the edges of objects, it is
broadcast to the sixty-four platinum electrodes implanted on the
surface of Jerry's visual cortex. Jerry sees enough of this information
to be able to read two-inch tall letters at a distance of five feet, and
navigate hitherto unexplored sectors of the New York subway system.
Dobelle implanted arrays in seven new volunteers, and
announced the development of a 512-electrode system which, in
mass-production, would cost no more than the training bill for
an average guide dog. It was a project he was still working on at
his death, aged sixty-two, in 2004.
+ * *
J
MAKING EYES TO SEE 281

The future of systems like the Dobelle eye is in doubt. They began
life as spin-offs from studies into how the brain handles electrical
stimulation. Now that the basic research is done, stimulating the
brain to manufacture eyeless vision seems risky, long-winded, and
even unnecessary."
A major selling point for such systems has been that a retina stricken
“with blindness quickly deteriorates. A system of prosthetic vision has
to stimulate the brain, because following the loss of sight, all parts of
the retina lose the ability to see. However, this is not the case. Although
common blinding conditions like macular degeneration destroy the
architecture of the retina, its nerve cells remain remarkably healthy
and functional. Even when the photoreceptors are destroyed, as
happens in retinitis pigmentosa, the ganglion cells to which they
connect stay in a workable condition for decades.'? If there were a
way of installing a plate of photosensitive ‘cells in the eye, could we
get it to talk to the undamaged neural layers of the retina?
Until very recently, artificial retinas were a completely unreal-
isable fantasy; only since about 2004 have we been able to contem-
plate building something so tiny and so finely worked. Now that
they have become a real possibility, vast sums are being invested
in their development.
But where should they go? The back-to-front arrangement of the
vertebrate eye makes it difficult to know where to put a prosthesis.
Should it sit on top of the retina, covering the ganglion cells but far
away from the site of the old, dysfunctional photoreceptors? Or
should it go under the retina - with all the risks and dangers such
a procedure implies — so that it can take over the job of the dead
photoreceptors? Sod’s Law dictates that whichever approach I pick
as the most promising, it will have been abandoned by the time you
read these words. Still, in spite of all the risks, and some recent,
rather worrying data on how quickly they corrode, putting a pros-
thesis under the retina seems the approach most likely to succeed.”
A prosthesis of this sort only needs to detect light; visual processing
can be left to the surviving layers of the retina. A prosthesis that sits
on top of the retina and talks directly to ganglion cells has to do all
this processing on its own - a daunting task.
Two teams — one from America and one from Germany - lead
282 A NATURAL HISTORY OF SEEING

the race to develop a working sub-retinal prosthesis. Their contrasting

methods and approaches reflect the cultural differences in the two
scientific communities. The German project is funded by the German
Federal Research Ministry, as part of a multi-million-euro project
to realise an ‘intelligent retina. It has yet to proceed to human trials.
The American team is a company called Optobionics, founded by
Dr Alan Chow, a paediatric ophthalmologist from Illinois, and his
brother Vincent, an electrical engineer. After eight years of animal
trials, the Chows announced that their patented prosthetic retinas
worked well enough that the animals’ brains were responding to
their signals. They subsequently found that they were wrong; they
had not taken into account the possibility that the animals’ visual
centres were responding to infra-red radiation.
Nonetheless, the US Food and Drug Administration approved
their application to proceed with clinical trials. To date, ten volun-
teers have been implanted with the devices, and they all report
improvements in their vision. However, all of them had at least a
little residual vision, which makes assessing small improvements
very difficult. What is more, they all reported improvements across
the whole of their visual field - a neat trick, given that the Chows
arrays are implanted only in the periphery of their vision. The
improvements cannot be because the devices work the way the
Chows thought they would. At best, the prosthesis is ‘encouraging’
the rest of the retina to rewire itself. If true, this would be truly
wonderful but it seems more likely that the volunteers are simply
reporting the kind of small visual changes that you would expect
after major ocular surgery.

Assume it happens. There seems no very good reason why it won't.

Artificial retinas will not only renew, but may even eventually extend,
the capacities of our eyes.
The eye is tuned, by evolution, to our prehistoric needs, but we
have since utterly transformed ourselves and our world. By building
our own retinas, we can custom-build a different, broader visual
experience for ourselves, and today’s remedial surgery for blindness
will very soon be extended to elective surgery for better, broader
sight. Artificial retinas can easily be tuned to see other parts of the
MAKING EYES TO SEE 283

electromagnetic spectrum. If a blind person can see with a couple

of retinal chips, why not adapt one chip for daylight use and one
for night? And if blind people can be equipped for night vision,
why shouldn't sighted people get in on the act?
Come to think of it, why stop at light? An artificial retina might
conceivably be sensitive to a whole host of novel sensations. Anyone
for radar? Spectroscopic analysis, anybody?
The Information Revolution is looking a little tired. The Perception
Revolution, meanwhile, is cracking its way out of the egg.

4 —- What you see is what you get

Flowers, crystals, busts, vases, instruments of various kinds,

&c., might thus be represented so as not to be distinguished
by sight from the real objects themselves.
Charles Wheatstone”

Imagine you are a surgeon, conducting keyhole surgery.

Not the easiest job in the world: first, where are you supposed to
look? At your hands? At the point where your instruments disappear
into the tiny incision? At the ultrasound monitor, mounted above
the table, which shows what youre up to inside your patient? Frankly,
youre all over the place. Your attention is scattered and diffused. No
wonder the trustees want to replace you with a robot.
But wait — put this pair of goggles on. The goggles have ordinary
clear lenses, and a casual glance will not reveal their secrets. But you
know that inside these goggles there is a microwave receiver, which
tracks the motions of your head as you work. That information is
sent, via a cable or a radio link, to a computer, which projects images
on to the partly silvered surface of your goggles. It’s a kind of head-
up display, of a very special kind. Because the computer knows where
you are looking, it can paste images on to your goggles in such a
way that they blend seamlessly with your surroundings.
Now, when you operate, it appears that youre actually looking
inside the patient’s body. Hey presto: you have X-ray eyes.
* * *
284 A NATURAL HISTORY OF SEEING

Funny how some stories sound futuristic, even when they are long
out of date. I wrote that story for Wired magazine around 1996.
Since then, visual prosthesis has advanced almost beyond recog-
nition and with it has come a huge improvement in our ability to
sow our individual visual worlds with virtual, pre-processed infor-
mation (a process dubbed ‘augmented reality’ or AR’). News stories
about simple head-up displays and augmented-reality goggles share
column inches with tales of lasers that map images across the retina
as though it were a television screen, induction loops that excite
the optic nerve, and trans-cranial magnetic stimulators that conjure
visual fancies in the brain.
Some of these stories are more fantastic than others, but even the
most speculative eventually acquire a grain of truth, as the technology
seems hell-bent on fulfilling every tech journalist's moistest wet
dream. Each year, as half a million children go blind for want of a
bit of butter, a telephone directory’s-worth of articles appear, and
they all begin the same way: ‘Imagine you are a fighter pilot’ Imagine
you are an astronaut. ‘Imagine you are a deep-sea diver ..?

Here's an idea: give everyone a pair of goggles. With them, and access
to a gargantuan shared image database (are the good people at Google
listening?) the everyday might become a canvas for us all. We might
paste Gormenghast over Pall Mall, or glimpse Dickens’s London down
Fleet Street. Driving to Edinburgh for work, we might choose to give
it the skyline Stevenson knew and loved. Goggles on, and every day
becomes an adventure in new territory. We might go shopping in the
drowned London of Wyndham’s The Kraken Wakes, catch a plane to
Blade Runner’s LA, make a Narnia of Birmingham, an Oz of the New
Forest, or erect Middle Earth on Salisbury Plain ...
Map fantasy over reality! Free the world of physical constraints!
Build Lang’s Metropolis on Earth!
Those of us old enough to remember a world without the Internet
will recall that we have heard this kind of talk before. Indeed, in the
early twentieth century, the self-same millennial promises were made
about another innovative medium: cinema. Will augmenting our
individual visual worlds free us of the constraints of space, time, and
the body?
MAKING EYES TO SEE 285

I wonder.
One of the more frightening things I stumbled across while
researching this book was the illustration accompanying an article
about augmented reality in the April 2002 edition of Scientific
American. The illustration took the author Steven Feiner’s lead in
suggesting a future in which virtual images are mapped contextu-
ally over the outdoor environment. The images were captions, laid
across an observer's view of a city street:

Near a bus stop: ‘Local #23 NEXT BUS 30 Seconds:

Over an office building: ‘OFFICE SPACE AVAILABLE Contact
Megalopolis Realty [email protected]
Outside a cinema: “FEATURES Jaws of Terror 2:00, 4:00, 6:00°

Is this the best we can do? Are our eyes to be stuffed with nothing
better than the usual commercial and political blandishments? Will
augmented reality revolutionise our way of being, or will it prove,
like so much of the Internet, to be just a new way of doing business
as usual?
If we are going to supplement the eye, it is high time we considered
what we are going to supplement it with. What do we want our new
sensory world to look like? What do we want from our eyes?

In 2000, William Dobelle gave Jerry back his sight; enough that he
can (just about) navigate a subway system. But Jerry's vision is, by
any measure, very poor. The visual world is very complex. The images
captured by the 2000-vintage Dobelle eye were excessively simple.
This is why Dobelle and his team went to such pains to program
the Dobelle eye's computer to simplify the scenes it captures, reducing
them to a few bright lines. In the summer of 2000, Dobelle was
working with Jerry at his house when Jerry’s eight-year-old son,
Marty, piped up: ‘You guys are out-of-date. Why don't you take digital
signals straight from the TV or computer?’
Now Jerry wants to deal stocks and shares over the Internet.
He wants to plug his eyes into the New York Stock Exchange.
‘He’s hot for that, Dobelle said.”
EPILOGUE

The invisible gorilla

The story of the eye is filled with material wonders: mirror eyes,
fibre-optic eyes, and eyes crammed full of exquisite lenses; shells
that see and spiders that steer by the sun; eyes less than one-tenth
of a millimetre across, and eyes one third of a metre across. Birds
can focus on two things at once. Tube worms see with their feeding
tentacles, chitons with eyes scattered like pimples over their shells.
The fish Bathylychnops exilis has four eyes: one set to look up, and
one set to look down; but at what? Nobody knows. The surface-
feeding fish Anableps anableps can see clearly both in and out of
the water. Mantis shrimps use polarised light to swap messages
that no other animal can see. Even a kilometre below the sea’s
surface there is, incredibly, still plenty of material my story could
have drawn on, if only there were room. Bioluminescent
photophores tattoo the sides of lanternfish, fang-toothed fish dangle
lures baited with light-making bacteria, and one deep-water fish,
Aristostomias, communicates with others by generating pulses of
red light - a colour no other neighbouring species can see.
The story of the eye is epic: this book is but the shortest précis
of its wonders.
* THE INVISIBLE GORILLA 287
Imagine a world without sight.
Colour does not exist. There is no plumage, no camouflage, no
courtship display, no phosphorescence in the deep ocean trenches.
No flowers.
What a dull world, and what a simple one. Animals develop behav-
_lour to respond to information. Intricate behaviours can never evolve
in the absence of enough information. And since behaviours are the
bedrock of consciousness, in a world without sight, consciousness
will never arise.
Because humans enjoy the luxury of consciousness, we can appre-
ciate seeing for its own sake. For us, seeing isn't necessarily tied to
any other useful activity. We tend, naturally enough, to concentrate
less on the act of seeing, and more on the view; and assume that the
eye is simply a window on the world.
But humans too have their very specific, and idiosyncratic, ways
of seeing.

Humans have an exceptionally narrow, acute focus of concentration,

coupled with a taste for studying social and transactional behaviour.
Of course, what we see reflects the real world, but we should never
assume our eyes are telling us the whole story. Welcome to the world
of John Grimes: a world where birds and swimming-suits change
colour, people swap hats and heads, and nobody seems to notice.
Grimes, of the University of Illinois, presented his first work on
‘change blindness’ in 1992, at a conference on perception in
Vancouver, British Columbia. Sitting in the audience was the Amer-
ican philosopher Daniel Dennett. Only the year before - in a book
with the deliberately provoking title Consciousness Explained -
Dennett had attacked the idea that we carry around pictures of the
world inside our brains, saying it was much more likely that we make
assumptions and use our senses to confirm or update these assump-
tions, as and when necessary. So, our assumptions never feel like
assumptions, and our illusion of seeing a wide field of view in detail
never feels like an illusion.
Even Dennett was startled by Grimes’s findings. Grimes’s exper-
iments exploited the fact that when our eyes saccade, we are func-
tionally blind. (If this weren't the case, vision would resemble
288 A NATURAL HISTORY OF SEEING

hand-held camera-work; all nauseating sweeps and irritating judder.)

Grimes set his volunteers to study photographs on a computer screen.
When their eyes performed a saccade, the computer quickly swapped
one photograph for another. In Grimes’ first experiments, the differ-
ence in the images was subtle. Hats worn by people in one photo-
graph might be swapped round in the next.
Nobody noticed.
So Grimes tried swapping heads.
Nobody noticed.
Grimes made objects change colour.
Nobody noticed.
‘I wish in retrospect that I'd been more daring, since the effects
are stronger than I claimed; said Dennett, following Grimes’s original
presentation.’ By 1996, Grimes had pushed his study even further,
swapping a red parrot that filled a good quarter of his computer
screen, for a green one. One in five of his volunteers failed to notice.
Rather than remember all the incidental details of the world
around us, we rely on our ability to spot changes. If change happens
when our eyes are saccading, we are blind to the change, and it passes
us by. Happily, very few things in the world change in the time it
takes for our eyes to saccade. But what about when we blink? And
what if something comes between us and what were looking at,
blocking our view?
A sunny day on the campus of Cornell University, in New York
State, 1998. The passer-by looks fifty-something. His magnificent
whiter-than-white hair is the first thing you notice. Perhaps he is a
professor. In any event, he seems to know his way around; when a
young man stops him to ask for directions, he happily obliges. A
moment later, two men barge between the student and the professor.
They're carrying a door. For a brief moment, the professor and the
young man are separated.
Now the video gets interesting. The camera is looking down on
the scene from a height; from here, we can see what the professor
cant: as the door passes between them, the boy asking for directions
swaps places with one of the men carrying the door.
The interruption passes. The professor carries on giving directions;
the new man listens and nods attentively. He is wearing a different
“THE INVISIBLE GORILLA 289

shirt, a different jacket, and we can safely assume that his voice is
different, but the professor doesn’t notice. Though the man has
changed, the professor did not witness the change, and so he assumes
the man is the same as the one he was talking to a moment ago.
Why would he think otherwise? You might think his memory would
wake him up to what is going on; on the contrary, fully half the
“unwitting subjects of this devious and very funny experiment -
dreamt up by the Harvard psychologist Daniel Simons and his
colleague Daniel Levin of Kent State University, Ohio — never spotted
the swap.
Memory is expensive. It is far cheaper for us to rely on the present
moment for our information. If the world were full of trickster
psychologists, we may have had to evolve differently. Happily, it’s not,
and were really rather easily tricked. We don't need to be blind to
changes to miss them; because we only see two degrees of the world
with any clarity, we tend to miss things that happen outside our
focus of attention, and the more we attend to something, the more
extreme our ‘attention blindness’ becomes.
This is the thesis of Simons’ most celebrated experiment, in collab-
oration with Harvard’s Christopher Chabris.
You are about to watch a short video of two teams of students
playing basketball in a corridor. One team wears white, one wears
black. You have to decide which team passes the ball most often.
Concentrate: the answer's important.
Run tape.
Have you got your answer? Good. Now were going to watch the
video again. This time, don’t worry about the number of passes. Just
sit back and watch the game. Do you see anything you hadn't spotted
before?
That’s right.
A gorilla.
A two-metre-tall pantomime gorilla just wandered through the
middle of the basketball game. If you were able to concentrate, and
put out of your mind the fairly obvious fact that Simons must have
something screwy up his sleeve, there's a fifty per cent chance you
missed it - even when it stopped and waved at you.
There’s no camera trickery, no movement of images during a
290 A NATURAL HISTORY OF SEEING

saccade, no sneaky undergraduates barging past you with a door. All

Simons did was ask you to follow a game. For fully half the people
who see this video, the ball-counting task is absorbing enough to
make them miss an entire gorilla.
In 1909, the zoologist Jacob von Uexkiill proposed that the signifi-
cant world of an animal was the sum of things to which it responds.
The rest goes by almost unnoticed. And is it not ironic, that in 538
million years of natural selection, eyesight should evolve from a
simple light-detecting cell, pass through numerous variations and
generate countless different ways of seeing, and come at last to serve
as the dominant sense of the planet’s dominant species — an animal
who sees only what it wants to see?
FURTHER READING

General
Gregory, R.L. 1998. Eye and Brain: The Psychology of Seeing. Fifth edition.
Oxford, Toyko: Oxford University Press. [Uses optical illusions to reveal the
mechanisms of perception.]
Helmholtz, H. v. 1856-67. Trans. J.P.C. Southall 1962. Handbuch der Physio-
logischen Optic. New York: Dover. [The foundation stone of modern vision
science, still relevant and readable.|
Marr, D. 1982. Vision. New York: WH Freeman & Co. [Striking, still-current
model of vision, steeped in computer metaphors. ]
Morgan, M. 2003. The Space Between Our Ears: How the Brain Represents Visual
Space. London: Weidenfeld & Nicolson. [A good recent attempt to shed
computer-based metaphors when discussing vision and mind.]
Oyster, C.W. 1999. The Human Eye: Structure and Function. Sunderland, MA:
Sinauer Assoc. [Sumptuous 800-page anatomy of the eye, with many
fascinating digressions into the history, biography and theory of vision.]
Wade, N. 2000. A Natural History of Vision. Cambridge, MA: MIT Press. [Vision
theorists — often in their own words — from the Greeks to the Victorians. ]
Zajonc, A. 1993. Catching the Light: the Entwined History of Light and Mind.
Oxford: OUP. [Thoughtful, often surprising history of light and colours;
292 FURTHER READING

readers scandalised by my omission of Goethe from these pages had better

go here to decompress.]

Prologue — Youth and age

Daw, N.W. 1995. Visual Development. New York: Plenum.
Hamer, R.D. & Schneck, M.E. 1984. ‘Spatial summation in dark-adapted human
infants’ Vision Research 24/1, pp. 77-85.
Weale, R.A. 1992. The Senescence of Human Vision. Oxford: OUP.
Wiesel, T.N. 1999. ‘Early explorations of the development and plasticity of the
visual cortex: a personal view’ Journal of Neurobiology 41, pp. 7-9.

One - The commonwealth of the senses

Catania, K.C. 1999. ‘A nose that looks like a hand and acts like an eye: the
unusual mechanosensory system of the star-nosed mole’ Journal of Compar-
ative Physiology A: Sensory, Neural, and Behavioral Physiology 185, pp. 367-
372.
Findlay, J.M. & Gilchrist, 1.D. 2003. Active Vision. New York, Oxford: OUP.
[A technical analysis of the relationship between eye movements and
perception. ]
Frisby, J. 2004. ‘Bela Julesz 1928-2003: a personal tribute’ Perception 33, pp.
633-637.
Gilchrist, I.D., Brown, V., Findlay, J.M. 1997. ‘Saccades without eye movements:
Nature 390, pp. 130-131.
Julesz, B. 1995. Dialogues on Perception. Cambridge, MA: MIT Press.
Wade, N.J. 1994. ‘A selective history of the study of visual motion after-effects’
Perception 23, pp. 1111-1134.
Wade, N. & Tatler, B. 2005. The Moving Tablet of the Eye: the Origins of Modern
Eye Movement Research. Oxford: OUP.
Wade, N.J. 2005. “The original spin doctors — the meeting of perception and insanity’
Perception 34/3, pp. 253-260. [Describes Purkinje’s researches into vertigo.]
Walls, G.L. 1962. “The evolutionary history of eye movements. Vision Research
2, pp. 69-80.
Weiskrantz, L. 1986. Blindsight: A Case Study and Implications. Oxford: Clarendon
Press.
Wheatstone, C. 1838. ‘Contributions to the physiology of vision 1: on some
remarkable, and hitherto unobserved, phenomena of binocular vision? Philo-
sophical Transactions of the Royal Society of London 128, pp. 371-394.
FURTHER READING 293

Yarbus, A.L. 1967. ‘Eye movements during perception of complex objects?

In Riggs, L.A. (ed). Eye Movements and Vision. New York: Plenum Press,
pp. 171-196.

Two - The chemistry of vision

Boll, F. 1877. ‘Zur anatomie und physiologie der retina’ Arch. Anat. Physiol. (Physiol.
Abt 4/35). Trans. R. Hubbard 2003 in Vision Research 17, pp. 1249-1265.
“Campion-Vincent, V. 1999. “The tell-tale eye? Folklore 110, pp. 13-24.
Carpenter, K.J. 2003.‘A short history of nutritional science: part 1 (1785-1885).
Journal of Nutrition 133, pp. 638-645.
Evans, A.B. 1993. ‘Optograms and fiction: photo in a dead man’s eye. Science-
Fiction Studies 20/3, pp. 341-361.
Goulet, A. 2005. ‘South Sea daggers and the dead man’s eye: foreign invasion
in fin-de-siécle optogram fiction? Cahiers Victoriens et Edouardiens 61.
Lehman, S. 2004. ‘George Wald: a scientist who studied vision and saw a world
in crisis: Optics and Photonics News 15/2, p. 18.
McLaren, D.S. The Control of Xerophthalmia: a Century of Contributions and
Lessons. Basel: Sight and Life. Available at: www.sightandlife.org/booksAll/
BooksHTML/allBooks.html
Sommer, A., & West Jr., K.P. 1996. Vitamin A Deficiency: Health, Survival, and
Vision. New York, Oxford: Oxford University Press.
Wolf, G. 2001. “The discovery of the visual function of vitamin A? Journal of
Nutrition 131, pp. 1647-1650.
Wolf, G. 2002. “The experimental induction of vitamin A deficiency in humans?
Journal of Nutrition 132, pp. 1805-1811. ;
Wolken, J.J.1995. Light Detectors, Photoreceptors, and Imaging Systems in Nature.
New York, Oxford: Oxford University Press.

Three —- How are eyes possible?

Axelrod, J. 1988.‘An unexpected life in research’ Annual Review ofPharmacology
and Toxicology 28, pp. 1-23.
Conway-Morris, S. 1998. The Crucible of Creation: the Burgess Shale and the
Rise of Animals. Oxford: Oxford University Press.
Conway-Morris, S. 2003. Lifes Solution: Inevitable Humans in a Lonely Universe.
Cambridge: Cambridge University Press, p. 166.
Darwin, C. [1859] 1982. The Origin of Species. London: Penguin Books.
Gehring, W.J. 1998. The Homeobox Story. New Haven, CT: Yale University Press,
p. 207.
294 FURTHER READING
GehringWJ., 2002. ‘The genetic control of eye development and its implications
for the evolution of the various eye-types’ The International Journal of
Developmental Biology 46, pp. 65-73.
Gehring, WJ. 2002. “The journey of a biologist: Balzan Prize winner 2002 for
Developmental Biology? Available at: www.balzan.com/en/preistraeger/
gehring.cfm
Gehring, WJ. 2004. ‘Historical perspective on the development and evolution
of eyes and photoreceptors? The International Journal of Developmental
Biology 48, pp. 707-717.
Parker, A. 2003. In the Blink of an Eye. London: Perseus Books. [Lively and
original account of the role of vision in the Cambrian, though Simon
Conway-Morris’s review, ‘On the first day, God said . . : (American Scientist
91/4) bursts Parker’s bubble somewhat. ]
Wald, G. 1968. “The molecular basis of visual excitation’? In Granit, R. (ed). Les
Prix Nobel en 1967. Stockholm: Nobel Foundation.
Walter, N. 2002. ‘From transdetermination to the homeodomain at atomic
resolution — an interview with Walter J. Gehring? The International Journal
of Developmental Biology 46, pp. 29-38. [Arguably the worst title ever given
to a good interview. ]

Four - The adaptable eye

Dawkins, R. 1996. Climbing Mount Improbable. London: Penguin Books.
[Contains a spectacular account of eye evolution. The figure by Michael
Land - a whimsical-but-serious map of eye variety - is the cherry on a
very rich cake.]
Exner, S. 1891. Die Physiologie der facettierten Augen von Krebsen und
Insekten. Vienna. Trans. R.C. Hardie 1989 as The Physiology of the
Compound Eyes of Insects and Crustaceans: a Study. Berlin, New York:
Springer Verlag.
Fortey, R. 2001. Trilobite: Eyewitness to Evolution. London: HarperCollins.
Gislén, A., Dacke, M., Kroger, R.H.H., Nilsson, D.-E., Warrant, E.J. 2003. ‘Supe-
rior underwater vision in a human population of sea gypsies: Current Biology
13, Pp. 833-836.
Land, M.F. 1997. ‘Visual acuity in insects? Annual Review of Entomology 42,
PP. 147-177.
Land, M.F & Nilsson, D.-E. 2002. Animal Eyes. New York, Oxford: Oxford
University Press. [The best single-volume work on eye variety.]
FURTHER READING 295
Nilsson, D.-E. 1989. ‘Optics and evolution of the compound eye: Facets of
Vision. In Stavenga, D.G. & Hardie, R.C. (eds). Berlin-Heidelberg: Springer-
Verlag, p. 3075.
Wald, G. 1953. ‘Eye and camera’ Scientific American Reader. New York: Simon
& Schuster, pp. 555-568.
‘Walls, G.L. 1942. The Vertebrate Eye and Its Adaptive Radiation. Bloomfield
Hills, MI: Cranbrook Institute of Science.
Zucker, C.S. 1994. ‘On the evolution of eyes: would you like it simple or
compound?’ Science 265, pp. 742-743.

Five — Seeing and thinking

Byrne, R., & Whiten, A. (eds). 1988. Machiavellian Intelligence. New York,
Oxford: Oxford University Press, pp. 34-49.
Darwin, C. 1872. The Expressions of Emotions in Men and Animals. London:
John Murray.
Ekman, P. 2001. Telling Lies: Clues to Deceit in the Marketplace, Politics and
Marriage. New York: W.W. Norton & Co.
Emery, N.J. 2000. “The eyes have it: the neuroethology, function and evolution
of social gaze? Neuroscience and Biobehavioral Reviews 24, pp. 581-604.
Hoffman, D.D. 1998. Visual Intelligence. New York, London: W. W. Norton &
Co. [Visual discrimination of objects reduced to a set of (numbered!) rules.
An off-putting approach, but it pays off.]
Kreisler, H. 2004. ‘Face to face: the science of reading faces? Conversation with
Paul Ekman on 14 January at the Institute of International Studies, UC
Berkeley. Transcript available at: www.globetrotter.berkeley.edu/people4/
Ekman/ekman-cono.html
Miller, G.F. 1997. ‘Protean primates: the evolution of adaptive unpredictability
in competition and courtship? In Whiten, A., & Byrne, R.W. (eds).
Machiavellian Intelligence II: Extensions and Evaluations. Cambridge:
Cambridge University Press, pp. 312-340.
O’Connell, S. 1997. Mindreading: How We Learn to Love and Lie. Oxford: Heine-
mann.
Tomasello, M., & Call, J. 1997. Primate Cognition. Oxford: OUP.

Six - Theories of vision

Al-Haythem, Ibn (Alhazen) Trans. I.A. Sabra 1989. The Optics of Ibn AI-
Haytham. London: The Warburgh Institute (University of London).
296 FURTHER READING

Al-Kindi, Abu. In Hozien, M. (ed). 1993. Islamic philosophy on-line. Available

from: www.muslimphilosophy.com/kindi/default.htm
El-Ehwany, A.E.‘Al-Kindi? In Sharif, M.M. (ed). [1963] 1983.4 History ofMuslim
Philosophy 1. Karachi: Pakistan Philosophical Congress. p. 424.
Hoorn, W.v. 1972. As Images Unwind. Amsterdam: University Press Amsterdam.
[Heavy-going but rewarding analysis of visual theory from the Greeks to
Newton and Goethe.]
Ladurie, E. L.-R. Trans. A. Goldhammer 1997. The Beggar and the Professor: a
Sixteenth-century Family Saga. Chicago: University of Chicago Press. [A
cheerful and fascinating history of the Platter family.]
Lindberg, D.C. 1976. Theories of Vision: from Al-Kindi to Kepler. Chicago:
University of Chicago Press.
Sabra, A.I. 1981. Theories of Light: from Descartes to Newton. Cambridge:
Cambridge University Press.
Wertheim, M. 1999. The Pearly Gates of Cyberspace: a History of Space from
Dante to the Internet. New York, London: W.W. Norton & Co.

Seven — Nervous matter, visually endowed

Bentivoglio, M. 1998. ‘Life and discoveries of Santiago Ramon y Cajal’ Stock-
holm: The Nobel Foundation. Available from: www.nobelprize.org/
medicine/articles/cajal/index.html
Bruce, V., Green, P.R., Georgeson, M.A. 1996. Visual Perception: Physiology,
Psychology and ;Ecology (3rd edition). Hove, E. Sussex: Psychology Press.
[Vision science’s very readable postgraduate-level bible.]
Cajal, S.R. y. 1906. “The structure and connexions of neurons’ (Nobel lecture).
.In (Elsevier) 1967. Nobel Lectures, Physiology or Medicine 1901-1921.
Amsterdam: Elsevier Publishing Company.
Hartline, H.K. 1967. “Visual receptors and retinal interaction’ (Nobel lecture).
In (Elsevier) 1972. Nobel Lectures, Physiology or Medicine 1963-
1970.Amsterdam: Elsevier Publishing Company. Available from: www.nobel
prize.org/medicine/laureates/1967/hartline-lecture.html
Hartline, H.K., Wagner, H.G., Ratliff, F. 1956. Inhibition in the eye of Limulus:
The Journal of General Physiology 39, pp. 651-673.
Holmgren, E. In Grant, G. 1999. ‘How Golgi shared the 1906 Nobel Prize in
Physiology or Medicine with Cajal’ Available from: www.nobelprize.org/
medicine/articles/grant/index.html
Sharpe, L.T., Whittle, P., Nordby K. 1993. ‘Spatial integration and sensitivity
FURTHER READING 297
changes in the human rod visual system? Journal of Physiology 461, pp.
235-246.
The John Moran Eye Center at the University of Utah maintains ‘Webvision:
the organization of the retina and visual system, a thorough and colourful
guide to the subject by Helga Kolb, Edurado Fernandez and Raph Nelson.
Visit www.webvision.med.utah.edu

Eight — Seeing colours

Ball, P. 2002. Bright Earth. London: Penguin. [An erudite history of colour.]
Campbell, EW. 1994. ‘Dr. Edwin H. Land (1909-1991)’ Biographical Memoirs
of Fellows of the Royal Society 40, pp. 195-219.
McElheny, V. 1998. Insisting on the Impossible: the Life of Edwin Land. Reading,
MA: Perseus Books.
Mollon, J.D. 2003. “The origins of modern color science? In Shevell, S. (ed).
Color Science. Washington: Optical Society of America.
Newton, I. [1730/1952]. Opticks, or a Treatise of the Reflections, Refractions,
Inflections and Colours of Light. New York: Dover.
Bruce MacEvoy’s guide to watercolour painting includes a very good account
of colour at www.handprint.com/hp/wel/wcolor.html
The Virtual Colour Museum, a cultural history of colour written by Narciso
Silvestrini and Ernst Peter Fischer, is available at www.colorsystem.com

Nine — Unseen colours

Crescitelli, EF. 1963. Obituary: Gordon Lynn Walls (1905-1962). Vision Research
3> Pp- 1-7.
Davenport, D.A. 1984. ‘John Dalton’s first paper and last experiments. Chem-
Matters, April 1984, p. 14.
Hunt, D.M., Dulai, K.S., Bowmaker, J.K., Mollon, J.D. 1995. “The chemistry of
John Dalton’s color-blindness’ Science 267, pp. 984-988.
Hunt, D.M., Dulai, K.S., Cowing, J.A., Julliot, C., Mollon, J.D., Bowmaker, J.K.,
Li, W.H., Hewett-Emmett, D. 1998. “Molecular evolution of trichromacy in
primates: Vision Research 38, pp. 3299-3306.
Ramachandran, V.S., & Blakeslee, S. 1998. Phantoms in the Brain. London:
Fourth Estate. [The humble blind spot harnessed to a dazzling account of
consciousness and perception. ]
Sacks, O.1995.An Anthropologist from Mars. London: Pan Macmillan. [Includes
“The Case of the Colourblind Painter’]
298 FURTHER READING

Sacks, O. 1996. The Island of the Colourblind. London: Pan Macmillan.

Zeki, S. 1980. “The representation of colours in the cerebral cortex? Nature 284,
pp. 412-418.

Ten — Making eyes to see

Brindley, G.S., & Rushton, D.N. 1974.‘Implanted stimulators of the visual cortex
as visual prosthetic devices? Transactions of the American Academy of
Ophthalmology & Otolaryngology 78, pp. 741-745.
Brooks, R.A. 2002. Robot: the Future of Flesh and Machines. London: Allen Lane.
[A personal, sometimes sardonic take on the world of artificial intelligence.
Among other accomplishments, Brooks has edited the International Journal
of Computer Vision.]
Chua, C.N. 2000. ‘Ophthalmology in the British Isles’ Available from:
www.mrcophth.com/Historyofophthalmology/ophthalmologyinuk.htm
Dobelle, W.H. 2000. Artificial vision for the blind by connecting a television
camera to the visual cortex? Transactions - American Society for Artificial
Internal Organs 46/1, pp. 3-9.
Dobelle, W.H., Quest, D.O., Antunes, J.L., Roberts, T.S., Girvin, J.P. 1979. ‘Artificial
vision for the blind by electrical stimulation of the visual cortex’ Neurosurgery
5, pp. 521-527.
Gregory, R.L. & Wallace, J.G. 1963. Recovery from early blindness — a case study.
(Experimental Psychology Society Monograph No. 2) Available from:
www.richardgregory.org/papers/index.htm
Lindblom, J., & Ziemke, T. 2003. ‘Social situatedness of natural and artificial
intelligence: Vygotsky and beyond. Adaptive Behavior 11/2.
Rizzo, J.F, et al. ‘Retinal prosthesis: an encouraging first decade with major
challenges ahead? Ophthalmology 108/1, pp. 13-14.

Epilogue - The invisible gorilla

Noé, A., Pessoa, L., & Thompson, E. 2000. ‘Beyond the grand illusion: what change
blindness really teaches us about vision? Visual Cognition 7, p. 93.
Simons, D.J., & Levin, D.T. 1998. ‘Failure to detect changes to people during
real-world interaction: Psychonomic Bulletin and Review 4, p. 644.
Spinney, L. 2000. “How much of the world do we really see?’ New Scientist 2265.
The Visual Cognition Lab at Illinois University maintains footage of attention
blindness experiments at www.viscog.beckman.uiuc.edu/djs_lab/demos.html
NOTES

Prologue — Youth and age

= Gordon, M. 1997. ‘Unravelling the real story behind the Cyclops? New
Scientist, 153/2068, p. 16.
Gould, G.M. & Pyle, W.L. [1896] 1997. Anomalies and Curiosities of Medi-
cine. Seattle, WA: The World Wide School. Available at: www.gutenberg.org/
etext/747
Maurice, D.M. 1998.'An ophthalmological explanation of REM sleep: Exper-
imental Eye Research 66, pp. 139-145.
Siegel, J.M. 2005.‘Functional implications of sleep development? PLoS Biology
3/5, p. 178.
Tarusovy, B.N., Polivoda, A.I., Zhuravlev, A.I. 1961. ‘Study of the faint spon-
taneous luminescence of animal cells? Biophysics 6, pp. 83-85.
Cepko, C. Interview with Norman Swan on The Health Report: the Retina
and the Brain, ABC Radio National, Australian Broadcasting Commission,
Monday 13 December 1999. Transcript available at: www.abc.net.au/rn/
talks/8.30/helthrpt/stories/s73272.htm
Livesey FJ., Cepko C.L. 2001. “Vertebrate neural cell-fate determination:
lessons from the retina’ Nature Reviews: Neuroscience 2/2, pp. 109-118.
The distinction between rods and cones is apparent in the jawless fish of
the Devonian period, 417-354Ma. See Bowmaker, J.K. 1991. ‘Evolution of
photoreceptors and visual pigments. In Cronly-Dillon, J.R. & Gregory, R.L.
(eds). Evolution of the Eye and Visual Pigments. Boca Raton, FL: CRC Press,
pp. 63-81.
Schefrin, B.E., Bieber, M.L., McLean, R., Werner, J.S. 1998. “The area of
complete scotopic spatial summation enlarges with age: Journal of the Optical
Society of America A (Optics, Image Science and Vision) 15/2, pp. 340-348.
But see Schefrin, B.E., Hauser, M., Werner, J.S. 2004. “Evidence against age-
related enlargements of ganglion cell receptive field centers under scotopic
conditions. Vision Research 44/4, pp. 423-428.

One — The commonwealth of the senses

Keller, H. [1905] 2002. The Story of My Life. New York: Penguin Books.
bp
F Yarbus, A.L. 1967. ‘Eye movements during perception of complex objects.
In Riggs, L.A. (ed). Eye Movements and Vision. New York: Plenum Press,
pp. 171-196. ‘
iss) Dennett, D. 1991. Consciousness Explained. New York: Little Brown &
Company, pp. 339-342.
Catania, K.C. 1999. ‘A nose that looks like a hand and acts like an eye: the
unusual mechanosensory system of the star-nosed mole. Journal of
Comparative Physiology, A: Sensory, Neural, and Behavioral Physiology 185,
Pp. 367-372.
Wolken, J.J. & Florida, R.G. 1969. “The eye structure and optical system of
the crustacean copepod Copilia’. Journal of Cell Biology 40/1, pp. 279-285.
O’Shea, S. 2003. Letter to the Octopus News Magazine online message board,
24 May. Available at: www.tonmo.com/forums
Milius, S. 2003. ‘Moonlighting: beetles navigate by lunar polarity: Science
News 164/1, p. 4.
Shashar, N. & Cronin, T.W. 1996. ‘Polarization contrast in octopus? The
Journal of Experimental Biology 199, pp. 999-1004:
Shashar, N., Rutledge, P.S., & Cronin, T.W. 1996. ‘Polarization vision in cuttle-
fish — a concealed communication channel?’ The Journal of Experimental
Biology 199, pp. 2077-2084.
Adapted from Vilhjalmsson, T. 1997. “Time and travel in Old Norse society:
Disputatio 2/89, p. 114.
10 Dowling, J.E. 2000. ‘George Wald, November 18, 1906 - April 12, 1997-
Biographical Memoirs 78. National Academy of Sciences Press. Available at:
www.nap.edu/html/biomems/gwald.html
NOTES 301
11 Chiao, C.-C., Cronin, T.W. & Osorio, D. 2000. ‘Color signals in natural
scenes: characteristics of reflectance spectra and effects of natural illumi-
nants. Journal of the Optical Society of America A 17, pp. 218-224,
12 Morgan, M.J., Adam, A., Mollon, J.D. 1992. ‘Dichromats detect colour-
camouflaged objects that are not detected by trichromats. Proc. R. Soc. Lond.
B Biol. Sci 248/1323, pp. 291-295.
13 Marshall, N.J. 2000.‘Communication and camouflage with the same “bright”
colours in reef fishes’ Philosophical Transactions of the Royal Society of
London B 355, pp. 1243-1248. Available at: www.dx.doi.org/10.1098/
rstb.2000.0676
Milius, S. 2004. “Hide and see: conflicting views of reef-fish colors? Science
News 166/19, p. 296. Available at: www.sciencenews.org/articles/ 20041106/
bob8.asp
14 ‘Bullet Shrimp’ Mirror, Wide 10 April 1998, p. 11.
15 Chiao, C.-C., Cronin, T.W., Marshall, N.J. 2000. “Eye design and color
signaling in a stomatopod crustacean Gonodactylus smithii: Brain, Behavior
and Evolution 56, pp. 107-122.
16 Leonardo da Vinci. In Wheatstone, C. 1838.
i7 Wheatstone, C. 1838. ‘Contributions to the physiology of vision 1: on some
remarkable, and hitherto unobserved, phenomena of binocular vision’ Philo-
sophical Transactions of the Royal Society of London 128, pp. 371-394.
18 Julesz, B. 1995. Dialogues on Perception. Cambridge, MA: MIT Press.
19 Of course, producing a pattern of truly random dots on a piece of paper
— without the aid of a computer - would be the truly heroic part of such
an enterprise. Bela Julesz used an IBM-704 computer to generate his patterns
on a tape which could then be converted into a television signal. See Mollon,
J.D. 1997. ‘“... On the basis of velocity clues alone”: some perceptual themes
1946-1996, The Quarterly Journal of Experimental Psychology 50A/4, pp.
859-878. 9
20 Hartline, H.K. 1972. ‘Visual receptors and retinal interaction’ In Nobel
Lectures, Physiology or Medicine 1963-1970. Amsterdam: Elsevier Publishing
Company, p. 281. Available at: www.nobelprize.org/medicine/laureates/
1967/hartline-lec
21 The quotations here are taken from Wade, N.J. 2005. “The original spin doctors
— the meeting of perception and insanity. Perception 34, pp. 253-260.
22 Hubel, D.H. 1988. Eye, Brain and Vision. New York: Scientific American
Library, p. 81.
302 NOTES
23 Oyster, C.W. 1999, p. 178.
24 Gilchrist, I.D., Brown, V., Findlay, J.M. 1997. ‘Saccades without eye move-
ments: Nature 390, pp. 130-131.
25 Weiskrantz, L. 1986. Blindsight: a Case Study and Implications. Oxford:
Clarendon Press.
26 MacKenzie, D. 1999. ‘An eye for danger: New Scientist 2188, p. 21.

Two - The chemistry of vision

1 Oomen, H.A. 1958. ‘Clinical experience on hypovitaminosis A’ In Kinney,
T.D. & Follis Jr, RH. 1958. Nutritional disease. Federation Proceedings 17,
Suppl 2, No 3, Pt 2, pp. 111-128.
2 Campion-Vincent, V. 1999. “The tell-tale eye’ Folklore 110, pp. 13-24.
3 de l'Isle-Adam, V. Trans. A. Symons 1925. Claire Lenoir. New York: Albert
& Charles Boni. ;
4 Goulet, A. 2005. ‘South Sea daggers and the dead man’s eye: foreign
invasion in fin-de-siécle optogram fiction? Cahiers Victoriens et Edouar-
diens 61.
5 Boll, F 1877. ‘Zur anatomie und physiologie der retina? Arch. Anat. Physiol.
(Physiol. Abt) 4-35.
6 Wald, G. 1953. ‘Eye and camera’ In Scientific American Reader. New York:
Simon & Schuster, pp. 555-568.
7 Vernois, M. 1870. ‘Etude photographique sur la retine des sujets assassinés’
Revue photographique des hopitaux de Paris 2, pp. 73-82. In Campion-
Vincent, 1999.
8 Dew, W. 1938. I caught Crippen. London: Blackie & Son.
9 Magendie, F. In Grmek, M.D. 1974. ‘Francois Magendie’ Dictionary of scien-
tific biography 9, pp. 6-11.
10 Adapted from Boyle, R. In Hunter, M. 2004. The Robert Boyle Project. Avail-
able at: www.bbk.ac.uk/boyle
11 Wolf, G. 1978. ‘A historical note on the mode of administration of vitamin
A for the cure of night blindness? American Journal of Clinical Nutrition
31, pp. 290-292.
12 Celsus, A.C. In McLaren, D.S. [1] The Control of Xerophthalmia:
a Century of Contributions and Lessons. Basel: Sight and Life. Available
at: www.sightandlife.org/booksAll/BooksHTML/allBooks.html
13 Elliot, R.H. 1920. Tropical Ophthalmology. London: Oxford Medical Publi-
cations, p. 78.
NOTES 303

14 Wald, G. 1968. “The molecular basis of visual excitation? In Granit, R. (ed).

Les Prix Nobel en 1967. Stockholm: Nobel Foundation.

Three — How are eyes possible?

1 Axelrod, J. 1988. ‘An unexpected life in research? Annual Review of Phar-
macology and Toxicology 28, pp. 1-23.
“2 Masland, R.H. 2001. “The fundamental plan of the retina? Nature Neuro-
science 4/9, pp. 877-886.
3 Okano, T., Kojima, D., Fukada, Y., Shichida, Y. 1992. ‘Primary structures of
chicken cone visual pigments: vertebrate rhodopsins have evolved out of
cone visual pigments. Proceedings of the National Academy of Sciences of
the United States of America 89, pp. 5932-5936.
Okano, T. 2002. ‘Pinopsin is a chicken pineal photoreceptive molecule
(letter). Nature 372, pp. 94-97.
Su, C-Y. 2006. ‘Parietal-eye phototransduction components and their poten-
tial evolutionary implications? Science 311/5767, pp. 1617-1621.
4 Paley, W. [1802] 2006. Natural Theology. New York, Oxford: Oxford Univer-
sity Press.
5 Gehring, WJ. 1998. The Homeobox Story. New Haven, CT: Yale University
Press, p. 207.
6 Oyster, C.W. 1999, p. 71.
7 Angier, N. 1995. ‘With new fly, science outdoes Hollywood’ New York Times
24 March: A1 (col. 2), A15 (col. 1).
8 Halder, G., Callaerts, P., Gehring, W.J. 1995. ‘Induction of ectopic eyes by
targeted expression of the eyeless gene in Drosophila’ Science 267,
pp. 1788-1792.
9 Dickinson, WJ. & Seger, J. (response, Gehring, WJ.) 1996. ‘Letters: eye evolu-
tion? Science 272, pp. 467-471.
10 Travis, J. 1997. ‘Eye-opening gene. Science News 151/19.
11 Goethe, J.W. In Seamon, D. & Zajonc, A. 1998. Goethe’ Way of Science. New
York: State University of New York Press, p. 199.
12 See Conway-Morris, S. & Gould, S.J. ‘Showdown on the Burgess Shale:
Natural History 107/10, pp. 48-55.
13 Parker, A. 2003. In the Blink of an Eye. London: Perseus Books.
14 Conway-Morris, S. 2003. ‘On the first day, God said ..? American Scientist
91/4.
15 (quotation) Allen, G. In Mollon, J.D. 2003.
304 NOTES

Four — The adaptable eye

Nilsson, D.-E. 1989.‘Optics and evolution of the compound eye: Facets of Vision.
Stavenga, D.G. & Hardie, R.C. (eds). Berlin-Heidelberg: Springer-Verlag, p. 3075.
Mollon, J.D. 1997. ‘“... On the basis of velocity clues alone”: some perceptual
themes 1946-1996. The Quarterly Journal of Experimental Psychology 50A/4,
pp. 859-878.
See, for example, Thakoor, S., Morookian, J.M., Chahl, J., Hine, B., Zornetzer,
S. 2004. ‘BEES: Exploring Mars with Bioinspired Technologies.’ Computer,
37/9,
pp. 38-47
Schénhammer, R. 2000. ‘Flying (human) bodies in the fine arts - dreams
and daydreams of flying’ Paper presented at the 17th Annual International
Conference of the Association for the Studies of Dreams, 4-8 July, 2000,
Washington, DC. Available from: www.psydok.sulb.uni-saarland.de/ voll-
texte/2005/550/
Kelber, A., Balkenius, A., Warrant, E.J. 2002. ‘Scotopic colour vision in
nocturnal hawkmoths’ (letter). Nature 419, pp. 922-925.
Gould, G.M. & Pyle, W.L. [1896] 1997.
Boyce, N. 2000. ‘Eye wish? New Scientist 167/2253, p. 33.

Five — Seeing and thinking

= Lorenz, K.1992.‘Analogy as a source of knowledge? Nobel Lectures, Physiology
or Medicine 1971-1980. Lindsten, J. (ed). Singapore: World Scientific
Publishing Co.
Helmholtz, H. v. 1860. Treatise on Physiological Optics, vol. 3. In Watson, R.I.
(ed). 1979. Basic Writings in the History of Psychology. New York: Oxford
University Press.
Millard, R. 2002. The Tastemakers. London: Scribner.
Chance, M.R.A.1957.‘The role of convulsions in behavior? Behavioral Science
2, pp. 30-45.
Tiger, L. ‘Real-life survivors rely on teamwork? Wall Street Journal, 25 August
2000, p. A14.
Miller, H. 1991. ‘Eye contact’ (letter). New Scientist 1781, psy
O’Connell, S. 1997. Mindreading: how we learn to love and lie. Oxford:
Heinemann, p. 60.
Thompson, P. 1980. ‘Margaret Thatcher: a new illusion? Perception 9, pp.
483-484.
Kreisler, H. 2004. “Face to face: the science of reading faces’ Conversation
°
-

; NOTES 305

with Paul Ekman on 14 January at the Institute of International Studies,

UC Berkeley. Transcript available at: www globetrotter.berkeley.edu/people4/
Ekman/ekman-cono.html
10 Brockman, J. 1999. ‘Animal minds: a talk with Marc D. Hauser’ Edge 54.
Available at: www.edge.org/documents/archive/edge54.html

Six - Theories of vision

Zajonc, A. 1993. Catching the Light. Oxford: Oxford University Press.
Al-Kindi. In Hozien, M. (ed). 1993. Islamic philosophy on-line. Available at:
www.muslimphilosophy.com/kindi/default.htm
Deregowski, J.B. 1972. ‘Pictorial perception and culture’ Scientific American
227, pp. 82-88.
See, for example, Borgmann, A. 1999. Holding on to Reality: the Nature of
Information at the Turn of the Millennium. Chicago: University of Chicago
Press.
Hoffman, 1998, p. 199.
wl
nn Kepler, J. In Koestler, A. 1989. The Sleepwalkers. London: Penguin Books, p.
280.
N Platter, F. 1583. De Corporis Humani Structura et Usu. In Lindberg, 1976, p.
176.
Kepler, J. 1604. Ad Vitellionem paralipomena. Translated in Donahue, W.H.
2000. Optics: Paralipomena to Witelo and the optical part of astronomy. Santa
Fe, NM: Green Lion Press, pp. 151-152.

Seven — Nervous matter, visually endowed

= Purkinje, J. In Wyman, M. 2001. “The history of veterinary ophthalmology
with particular emphasis for Ohio’ 25th annual Waltham/OSU Symposium
on Small Animal Ophthalmology. Available at: www.vin.com/VINDBPub/
SearchPB/Proceedings/PRo5000/PRo0525.htm
Campbell, L. & Garnett, W. 1882. The Life of James Clerk Maxwell. London:
Macmillan & Co. Digital preservation © 1997, 1999 Sonnet Software, Inc.
Available at: www.sonnetusa.com/bio/maxwell.asp
Keeler, C.R. 2004.‘Babbage the unfortunate’ British Journal of Ophthalmology
88, pp. 730-732.
Helmholtz, H. v. Trans. E. Atkinson 1999. Popular Lectures on Scientific Subjects,
Second Series. Bristol: Thoemmes Press, p. 278.
Brewster, D. 1832. In Ramachandran, V.S., & Gregory, R.L. 1991.
306 NOTES

6 Schickore, J. 1998. ‘The Microscopic Anatomy of the Retina, 183 5-1855.

Paper presented at the History of Science Society Annual Meeting 23
October. Missouri: Kansas City.
7 James Clerk Maxwell’s letter to his wife Katherine, 3 January, 1870. In Camp-
bell, L. & Garnett, W. 1882. p. 241.
8. Letter from William Swan, 2 April 1882. In Campbell, L., & Garnett, W.
1882. pp. 236-237.
9 Retzius, G. Trans. G. Grant 1948. Biografiska Anteckningar och Minnen, vol.
2. Uppsala: Almqvist & Wiksell, p. 246.
10 Holmgren, E. In Grant, G. 1999. ‘How Golgi Shared the 1906 Nobel Prize
in Physiology or Medicine with Cajal’ Available from: www.nobelprize.org/
medicine/articles/grant/index.html
11 Bernhard, C.G. 1967. Presentation speech, the Nobel Prize in Physiology
or Medicine. In Granit, R. (ed). 1968. Les Prix Nobel en 1967. Stockholm:
Nobel Foundation.
12 Dakin, S.C. & Bex, PJ. 2003. ‘Natural image statistics mediate brightness
“filling in”? Proc R Soc Lond B Biol Sci 270, pp. 2341-2348. Dakin and
colleagues at University College, London, are now conducting brain scans
to try and isolate the ‘filling in’ mechanism triggered by his illusion.
13 Sharpe, L.T., Whittle, P., Nordby, K. 1993. ‘Spatial integration and sensitivity
changes in the human rod visual system?’ Journal of Physiology 461, pp.
235-246.

Eight — Seeing colours

1 Mollon, J.D. 2003. “The origins of modern color science? In Shevell, S. (ed).
Color Science. Washington: Optical Society of America.
2 Gladstone, W. 1877. “The colour sense? Nature 19thC.2, pp. 366-388.
3 Geiger, L. In Bellmer, E.H. 1999. “The statesman and the ophthalmologist:
Gladstone and Magnus on the evolution of human colour vision, one small
episode of the nineteenth-century Darwinian debate’ Annals of Science 56/1,
Pp. 25-45.
4 Gage, J. 1993. In Ball, P. 2002. Bright Earth. London: Penguin, p. 49.
5 Newton, I. 1671/72. ‘A Theory Concerning Light and Colours’
Philosophical Transactions of the Royal Society of London 80. Available at:
www.newtonproject.ic.ac.uk/texts/cul3970_n.html
6 Newton, I. [1730] 1952. Opticks, or a Treatise of the Reflections, Refractions,
Inflections & Colours of Light. 4th edition. New York: Dover.
NOTES 307

7 Knight, D. 2002.‘Scientific lectures: a history of performance’ Interdisciplinary

Science Reviews 27/3, pp. 217-224.
8 Young, T. 1804. ‘Experimental demonstration of the general law of the
interference of light? Philosophical Transactions of the Royal Society of
London 94.
. 9 Campbell, EW. 1994.‘Dr Edwin H. Land (1909-1991): Biographical Memoirs
of Fellows of the Royal Society 40, pp. 195-219.
10 Mortimer, C. 1731. ‘An account of Mr Christopher Le Blon’s principles of
printing, in imitation of painting and of weaving tapestry, in the same
manner as Brocades: Philosophical Transactions of the Royal Society of
London 37, pp. 101-107. In Mollon, J.D. 2003.
11 Julesz, B. 1995. Dialogues on Perception. Cambridge, MA: MIT Press, p.
27.

Nine — Unseen colours

- Dalton, J. In Davenport, D.A. 1984. John Dalton’s first paper and last exper-
iments. ChemMatters, April 1984, p. 14.
Nv Frey, EG. 1975.‘A railway accident a hundred years ago as reason for system-
atic testing of colour vision? Klinische Monatsblatter fiir Augenheilkunde
167/1, pp. 125-127.
ies) Zorpette, G. 2000. ‘Looking for Madam Tetrachromat’ Red Herring,
1 November.
4 Hubel, 1988. p. 165.
5 Wilson, E.B. In Al-Awgati, Q. ‘Edmund Beecher Wilson, America’s first cell
biologist? Columbia magazine, Fall 2002. Available at: www.columbia.edu/
cu/alumni/Magazine/Fall2002/Wilson.html
6 Jaeger, W. 1992. ‘Horner’s law. The first step in the history of the
understanding of X-linked disorders? Ophthalmic paediatrics and genetics
13/2, pp. 49-56.
7 Hunt, D.M., Dulai, K.S., Bowmaker, J.K., Mollon, J.D. 1995. “The chemistry
of John Dalton’s color blindness’ Science 267, pp. 984-988.
8 Jameson, K.A., Highnote, S.M., Wasserman, L.M. 2001. ‘Richer color
experience in observers with multiple photopigment opsin genes: Psycho-
nomic Bulletin & Review 8/2, pp. 244-261.
9 Flom, M.C., Stern, C., White, H.E. 1963. University of California: in memoriam
Gordon Lynn Walls, Optometry; Physiology: Berkeley. Available at:
www.crsltd.com/research-topics/walls/obituary.html
308 NOTES
10 Roth, L. & Kelber, A. 2004. ‘Colour vision in nocturnal and diurnal geckos:
The Lund Vision Group, Lund “University. Available at: www.biol.lu.se/
funkmorf/vision/almut/gecko/html]
11 Arrese, C., Hart, N., Thomas, N., Beazley, L., Shand, J. “Trichromacy in
Australian Marsupials’ Current Biology 12/8, pp. 657-660.
12 Huth, G.C. ‘A new explanation for light interaction with the retina of the eye
based on nanostructural geometry: rethinking the vision process: Available at:
www.ghuth.com/vision/?p=5 5
13 Cottaris, N.P. & de Valois, R.L. 1998. “Temporal dynamics of chromatic
tuning in macaque primary visual cortex: Nature 395, pp. 896-900.
14 Smithson, H.E. & Mollon, J.D. 2004.‘Isthe S-opponent chromatic subsystem
sluggish?’ Vision Research 44, pp. 2919-2929.
15 Nathans, J., Thomas, D., Hogness, D.S. 1986. ‘Molecular genetics of human
color vision: the genes encoding blue, green, and red pigments. Science 232,
pp. 193-202.
16 Mollon, J.D. 1999.‘Color vision: opsins and options: Proceedings of the National
Academy of Sciences of the United States of America 96, pp. 4743-4745.
17 Sacks, O. 1995. An Anthropologist from Mars: Seven Paradoxical Tales.
London: Picador.
18 Julesz, B. 1995, Dialogues on Perception. Cambridge, MA: MIT Press, p. 47.
19 Mollon, J.D. 1997.‘*...On the basis of velocity clues alone”: some perceptual
themes 1946-1996. The Quarterly Journal of Experimental Psychology 50A/4,
p. 867.
20 Mach, E. [1886] 1987. Contributions to the Analysis of the Sensations. Peru,
IL: Open Court.

Ten — Making eyes to see

a Bedard, E.R. 2002.‘Non-lethal capabilities: realizing the opportunities? Defense
Horizons 9.
2 Burgess, L. Troops train to use non-lethal weapons to control crowds, reduce
civilian deaths’ Stripes Sunday magazine, 22 December 2002.
3 Rappert, B. 2003. Non-lethal Weapons as Legitimising Forces? Technology,
Politics and the Management of Conflict. London: Frank Cass Publishers.
4 Hambling, D. 2002. ‘“Safe” laser weapon comes under fire? New Scientist
2359, p. 5.
5 Dunham, W. 2006. ‘Pentagon uses laser device at Iraqi checkpoints’ Reuters
18 May, 17:13:42 GMT.
NOTES 309

6 Adapted from Chau, C.N. 2000. ‘Ophthalmology in the British Isles’ Avail-
able from: www.mrcophth.com/Historyofophthalmology/ophthalmology
inuk.htm
7 Locke, J. [1690] 2004. Essay Concerning Human Understanding 2/9.8. London:
Penguin Books.
-8 Gregory, R.L. & Wallace, J.G. 1963. Recovery from early blindness - a case
study. (Experimental Psychology Society Monograph No. 2) Available from:
www.richardgregory.org/papers/index.htm
9 See, for example, Snyder, F.W. & Pronko, N.H. 1952. Vision with spatial inver-
sion. Wichita, KS: McCormick-Armstrong. (A detailed and entertaining
account.)
10 Julesz, B. 1995. Dialogues on Perception. Cambridge, MA: MIT Press, p. 3.
11 ‘Development of biomorphic flyers’? NASA tech briefs November 2004,
p. 54. Summary available from: www.nasatech.com/Briefs/Novo4/NPO30554
-html
12 Lindblom, J. & Ziemke, T. 2003. ‘Social situatedness of natural and artificial
intelligence: Vygotsky and beyond: Adaptive Behavior 11/2.
13 Vygotsky, L.S. 1979. ‘Consciousness as a problem in the psychology of
behaviour’ Soviet Psychology 17/4, pp. 3-35.
14 See, for example, Trevarthen, C. ‘Can a robot hear music? Can a robot
dance? Can a robot tell what it knows or intends to do? Can it feel pride
or shame in company? Questions to the nature of human vitality: In Prince,
C.G., Demiris, Y., Marom, Y., Kozima, H., Balkenius, C. (eds). 2002. Proceed-
ings of the Second International Workshop on Epigenetic Robotics: modeling
cognitive development in robotic systems (Lund University Cognitive Studies
94). Lund: Lund University, pp. 79-86.
15 Wagenaar, D.A. 2004. ‘Cortical stimulation for the evocation of visual
perception? Term paper for CNS247: Cerebral cortex, R. Andersen, Caltech.
Available from: www..its.caltech.edu//pinelab/wagenaar/papers/cns247.pdf
16 Brindley, G.S. & Lewin, W.S. 1968. ‘The sensations produced by elec-
trical stimulation of the visual cortex’ The Journal of Physiology 196, pp.
479-493.
17 See also Dobell, W.H. & Mladejovsky, M.G. 1 974. The directions for future
research on sensory prostheses? Transactions - American Society for Artifi-
cial Internal Organs 20, pp. 425-429.
18 Heiduschka, P. & Thanos, S. 1998. ‘Implantable bioelectric interfaces for
lost nerve functions’ Progress in neurobiology 55, pp. 433-461.
310 NOTES

19 Humayun, MLS., Sato, Y., Propst, R., de Juan Jr., E. 1995. ‘Can potentials
from the visual cortex be elicitéd-electrically despite severe retinal degen-
eration and a markedly reduced electroretinogram?’ German Journal of
Ophthalmology 4, pp. 57-64.
20 Ananthaswamy, A. 2002. ‘Eye strain’ New Scientist 2329, p. 14.
21 Wheatstone, C. 1838. ‘Contributions to the physiology of vision 1: on some
remarkable, and hitherto unobserved, phenomena of binocular vision’ Philo-
sophical Transactions of the Royal Society of London 128, pp. 371-394.
22 ‘Computer Helps Blind Man “See” Reuters 09:20 AM Jan. 17, 2000 PT.
Available from: www.wired.com/news/technology/o,1282,33691,00.html

Epilogue — The invisible gorilla

1 Spinney, L. 2000. ‘Blind to change? New Scientist 2265, pp. 28-32.
PICTURE CREDITS

The developing eye’s temporary blood supply (Walls, G.L., 1942) 5

Record of eye movements while studying a face (Riggs, L.A., 1967) 18
A retinal rod expands when exposed to light (Wolken, J.J., 1995) 20
The ‘muffin par illusion (after Hoffman, 1998) 32
The problem with depth perception (after Julesz, 1995) 35
The world’s first random dot stereogram (in Julesz, B., 1960) 36
Jan Purkinje’s moving self-portrait (Psotnickova, J., Jan Evangelista Purkyne,
Orbis, Prague, 1955) 40
Wilhelm Kiihne (Hubbard, R., 1977: ‘Preface to the English translation of
Boll's On the Anatomy and Physiology of the Retina; and of Kihne's
‘Chemical Processes in the Retina, Vision Research 17, p. 1247) 57
Kiihne’s rabbit optogram: capturing a dying vision (Kiihne, W. [1878.]
1882, ‘Uber den Sehpurpur, Untersuchungen aus dem physiologischen
institute der universitat Heidelberg, 4, pp. 169-249) 59
G.L. Walls’s impression of an idealised rod cell (Walls, G.L., 1942) 73
Walter Gehring as a graduate student in 1964 (Walter, N., 2002) 82
Evolution of a computer-generated eye (Nilsson, D.-E., and Pelger, S., 1994) 98
If humans had compound eyes (Kirschfeld, K., 1976) 108
Optic flow (after Grindley) 113
The compound eye of the fly (after Land and Nilsson, 2002) 118
The visual models of birds (after Hoffman, 1998) 131
The Thatcher illusion (Peter Thompson) 148
Objects are visible from every angle (adapted from Richter, J.P., The
Literary Works of Leonardo da Vinci, Phaidon, London, 1939) 165
Our first picture of the visual system (Ibn Al-Haythem, in Polyak, S.L.,
The Retina, University of Chicago Press, Chicago, IL., 1941) 170

A choice of elephants (redrawn from Hudson, 1962) 171

Descartes’ optics of the eyeball (in R. Dioptr., AT-VI-116) 180

The optic nerve head (Bowman, W., 1849) 190
312 PICTURE CREDITS

An early cross-section through the retina (Polyak, S.L., 1941) 193

James Clerk Maxwell (© Royal Society) 196
Different cell types in the retina of a ‘dog (Cajal, S.R. y, ‘La rétine
des vertébrés, La Cellule 9, pp. 121-133, 1892) 200
Mutual inhibition at work in the retina (after Marr, D., 1982) 205
Mutual inhibition reveals the edges of things (in Marr, D., 1982) 206
Using edge information to ‘fill in’ an image (Steven Dakin) 207
Different human photoreceptors see different wavelengths (after
Lythgoe, 1979) 227
The variety of rods and cones (Walls, G.L., 1942) 254, 255

PLATES
The compound eye of the Antarctic krill Euphausia superba (Gerd Alberti
and Uwe Kils)
The carapace-eye of a brittlestar (in Fitzgerald, R., 2001 © American
Institute of Physics)

The eye of a decapod shrimp (Michael Land)

Close-up of the surface of a crayfish eye (Michael Land)

The mantis shrimp (Jeffrey Jeffords)

David Brewster's study of the visible spectrum (Brewster, D., 1831)

Some flowers boast ultraviolet patterns (Thomas Eisner, Cornell University)

The banded butterflyfish Chaetodon striatus (Tom Doeppner)

Slits, horseshoes and pinholes: pupil shapes are astonishingly various

(Dan-Eric Nilsson)

The compound eye of Phacops (Riccardo Levi-Setti)

Erbenochile (Phil Crabb, courtesy of the Natural History Museum, London)

A star-nosed mole’s nose (Kenneth Catania)

A naked mole rat’s teeth (Kenneth Catania)

The nerve cells of the retina, drawn by Santiago Ramén y Cajal (Instituto Cajal)
PICTURE CREDITS 313
Camillo Golgi’s staining method reveals the architecture of the visual cortex
(Hubel, D., 1988)
David Hosack’s drawing of the musculature of the eyeball (© The Royal Society)

Two views through an ophthalmoscope (in Wilmer, W.H., Atlas fundus oculi,
Henry Kimpton, London, 1934)

A corrosion cast of the human eye (Jane Olver)

Suspended in space: the eye's lens (in Streeton, B.W., ‘Anatomy of the
zonular apparatus Chapter 14 in Tasman, W., and Jaeger, E.A., eds.,
Duane’ Foundations of Clinical Ophalmology 1 (Lippincott-Raven,
Philadelphia. Reproduced with permission)
The pineal body of the desert spiny lizard Sceloporus magister (Jan J. and E.
Carol Roth, Fig. 9.6 in Wolken, J.J., 1995)

Ectopic eyes on the legs of a fruit fly (Gehring, W.J., 2002)

Ectopic fly eyes triggered by the activation of a mouse gene (Gehring, W,.,
2002)

Portrait of Thomas Young by Henry P. Briggs (© Royal Society)
Portrait of Sir Charles Wheatstone by Charles Martin (© Royal Society)
A family photograph of Santiago Ramon y Cajal (Instituto Cajal)

Edwin Land prepares an experimental ‘mondrian (J.J. Scarpetti)

Interior and exterior views of Adelbert Ames’s distorted room (in Ittelson,
WH., The Ames Demonstrations in Perception: A Guide to Their
Construction and Use, Princeton University Press, Princeton, NJ, 19 52)
INDEX

acuity, visual, archetypes, visual, 131-3, 159

at night, 209-10, 257 arcus senilis, 13
of compound eyes, 107, 108 (fig.) area centralis, 125
of human eyes, 107, 126, 134 Aristostomias, 286
underwater, 118-19 Aristotle, 160-1
Adrian, Edgar Douglas (Lord Adrian artificial intelligence, 275-6
of Cambridge), 203 assumptions, visual, 131-3
after-images, 226, 230 astigmatism, 127
ageing, 12-13 atomism, 156-8
Alberti, Leon Battista, 174 attention, 154-5, 158, 160, 161, 212,
Al-Haythem, Abu (Ibn al-Haythem, 242-3, 287-9
Alhazen), 166-70, 168 (fig.), 170 augmented reality (AR), 283-5
(fig.) autism, 147
Al-Kindi, Abu, 163-6 Axelrod, Julius, 76-7, 77 (fig.)
Allen, Grant, 98
Allen, Woody, 152 Babbage, Charles, 185-6
amatoris (muscle), 151 Bach-y-Rita, Paul, 19-20
Ames Jr, Adelbert, 272-3 Bacon, Roger, 173-4
Anableps anableps, 126, 286 Ballard, Dana, 136
aniridia, 84 Banister, Richard, 267-8
Anomalocaris, 93, 95 Banu Musa brothers, 164
apposition, 107 (fig.) Baron-Cohen, Simon, 147
aqueous humour, 7 Bateson, William, 82
INDEX 315
Bathylychnops exilis, 286 calcite, 27, 104-5
Bayley, Walter, 268 Callaerts, Patrick, 85
Bedard, Emil, 265, 266 Cambrian period, 88-99, 101
bee, 26, 76, 107-8, 111-12 camera obscura, 167-8, 168 (fig.),
behaviourism, 142 176-7
belladonna, 42 camouflage, 30, 75, 141
* Benham, C.E., 239 among reef fish, 30
Benham’s top, 238-9, 238 (fig.) broken by dichromacy, 30
Berlin, Brent, 214 Canadia spinosa, 97
Bidloo, Govard, 183 Catania, Kenneth, 22
bifocals, 8 cataract, 13, 29-30, 267-8
Bisazza, Angelo, 50 Celsus, Aulus Cornelius, 64
Blackbourne, Henry, 267-8 cephalopod eyes, 21 (fig.)
blind spot, 189 (fig.) Chabris, Christopher, 289
blinding (social practice), 264-6 Chahl, Javaan, 275
blindness, 13, 127, 145 Chance, Michael, 141-3
psychogenic, 269-70 Charles II of England, 188-9
see also blinding, prosthetic Che illusion, 207 (fig.)
vision, restored sight Cheselden, William, 268, 269
blinking, 152-3 Chiao, Chuan-Chin, 30
Bloch, Carl E., 66 Chlamydomonas, 73
blue cones (S cones), 251, 258-9 choroid, 6, 195
blue morpho (butterfly), 96 Chow, Alan and Vincent, 282
Boll, Franz Christian, 57 chromatic aberration, 121 (fig.),
Bowman, William, 181, 190 (fig.) 121-2, 257
box jellyfish (Chironex fleckeri), 50 chromosome 7, 251
Boyle, Robert, 61-2 ciliary muscle, 8, 222
Brahe, Tycho, 174-7 Claretie, Jules, 56
Brewster, David, 189, 197, 227-8 Cog (robot), 276
Brindley, Giles, 277-80 collosal squid (Mesonychoteuthis
brittlestar, 22 hamiltoni), 23-4
Brooks, Rodney, 274-6 colour-blindness, 211-12, 216, 230,
Brougham, Henry, 223 231, 244-53, 260, 261-2
Briicke, Ernst, 115, 191 genetics of, 248-50, 251-2, 259-60
Brunelleschi, Filippo, 174 prevalent among men, 250
Buckland, William, 89-90 see also dichromacy, monochro-
Burgess Shale, 96 macy
Burinmo, 212-13, 214 colour perception, 193-5, 219, 224-
Burroughs, Edgar Rice, 102 33, 235-7, 259-60
butterflyfish, 141 evolution of, 216, 248, 253, 256-
7, 259, 260
Cajal, Santiago Ramon y, 181, 198- role of language in, 213-15
201, 200 (fig.) colour words, 212-17
Calamactis praelongus, 75 colours, 30-31, 194-5, 211-21, 245
316 INDEX

additive mixing, 226, 239, 240 Descartes, René, 179, 180 (fig.)
complementary, 226, 229 Dew, Walter, 60
Earth, 229, 230, 231 Diadema setosum, 75
flicker, 238-9 dichromacy, 30, 248, 257
impossible, 230 Dickinson, W. Joe, 86
opponent, see Hering, Ewald diet and vision, 52-3, 63-7
primary, 221, 225, 227-8, 230, diffraction, 28, 108-9, 120, 222,
240 223,
structural, 96-7 Divisionism, 212, 242
subtractive mixing, 227, 239 Dixon Jr, Thomas, 56
compound eyes, 102-18, 103 (fig.), Dobelle, William, 278-80, 285
107 (fig.) dung beetle, 25
Condylura cristata, 22 Direr, Albrecht, 174
cones, 8, 125, 194-5, 210, 225-6,
228, 229, 231, 248, 251-3, 254-5 Ebers papyrus, 63
(fig.), 256, 258-9; see also red, Ediacaran fauna, 90
green, blue cones eidola, 156-8
contact lenses, 126-7 Ekman, Paul, 148-51, 153
contrast (visual), 146, 202-3, 204-8, Eldredge, Niles, 90
231, 236-7 Elliot, Robert Henry, 64
Convoluta roscoffensis, 75 empathy, 146-7
Copilia, 22 Empedocles, 158
cordierite, 27 Engelmann, Theodor, 72
cornea, 5 Erbenochile, 104
vitamin A deficiency and, 64-6 Erythropsis, 73
corrective surgery, 127 Euclid, 161-2
couching, 267 Eugene of Sicily, 162
Cox, Joseph, 41 Euglena gracilis, 72-3, 94
crystallins, 6, 119 evolution,
Cummings, William, 187 convergent, 20
Cyclops, 1-2 theories of, 78-9, 88-91, 98
Exner, Sigmund, 43, 76, 115-17
da Vinci, see Leonardo expressions (facial), 147-53
Dakin, Steven, 207 eye’s role in, 151-2
Dalton, John, 244-5, 246-8, 250-51 extramission, 158-60, 164-6
Darwin, Charles Robert, 56, 78-9, eye
88-90, 149-50 biggest, 23-4, 123
Davidoff, Jules, 213 carries meaning, 102, 140-41,
Dawkins, Richard, 82, 138 145-7, 151-3
deceit, 143, 152-3 ectopic, 82, 84-5
Deilephila elpenor, 117 evolved many times, 79, 85-6
Dennett, Daniel, 19, 287-8 protozoan, 72-4
depth perception, 31-8, 114-15, 128 shines in the dark, 160
dermal light sense, 75-6 smallest, 73
INDEX B17,

eye movement, 41-8, 138-9, 146 Gladstone, William Ewart, 215-16,

necessary to vision, 45 248
stabilises vision, 41 glaucoma, 13
eye muscles, 45-6 Goethe, Johann Wolfgang von, 42,
absent, 47-8 243
eyeball, 3 (fig.) Gogh, Vincent van, 218
eyeless, 83-6 Golgi, Camillo, 199-201
” eyelids, 124 Gonnelli, Jean, 17
Goodall, Jane, 144
face, see expressions Gould, Stephen Jay, 90-91, 98
Feiner, Steven, 285 Graham, Clarence, 204
fibre-optics, in nature, 110 Granit, Roger, 203
filling in, 189 green cones (M cones), 251-2, 258-9
film, see moving pictures Gregory, Richard, 271-2
Findlay, John, 46-7 Grimaldi, Francesco Maria, 222, 223
flavins, 73 Grimes, John, 287-8
Flemming, Walther, 249 Grindley, G.C., 112
flight, eyes adapted to, 111-12, 275 Guericke, Otto von, 229, 235
foetal development, 1-3, 6-7
Fontana, Felice, 190 Haeckel, Ernst, 80
Fore, 151 Haidinger, Wilhelm Karl von, 197-8
Fortey, Richard, 104 Halder, Georg, 85
Fossey, Diane, 144 handedness, 50
fovea, 8, 122, 125-6, 191, 195, 196, Hartline, Haldan Keffer, 203-5, 208-
198, 210, 257 ®)y Dil, BAA
Franklin, Benjamin, 8 Hauser, Marc, 153
Fresnel, Auguste, 224 Heliodorus of Emesa, 217
Friesen, Wallace, 150-51 Helmholtz, Hermann von, 58, 115,
Frisch, Karl von, 75 133, 187, 192, 228, 231—2, 232
fusion (fig.),
235, 239, 273
spatial, 240-3 Henle, J., 191
temporal, 202, 239; see also Henschen, Salomon, 277
persistence of vision Hering, Ewald, 230-33, 235, 261
Herschel, William, 224
Gajdusek, Carleton, 150 Heterocephalus glaber, 22
Galen, 4, 168, 181, 188 Hoffman, Donald, 129, 172-3
gaze detection, 140, 145-6 hog-nosed snake, 140
Gehring, Walter, 79, 81-6, 82 (fig.) Holmgren, Alarik Frithjof, 245, 246
Geiger, Lazarus, 216 Holmgren, Emil, 201
genes, 83-6 homeosis, 82
giant squid (Architeuthis dux), 23-4 Homer, 214, 215-16
Gibson, James Jerome, 129 Hooke, Robert, 103 (fig.), 221, 222
Gilchrist, Iain, 46-7 Horner, Johann Friedrich, 249
Gislén, Anna, 118-19 horseshoe crab, 75, 109, 204
318 INDEX

Hubbard, Ruth, 70 laser weapons, 265-6

Hudson, William, 171 lateral geniculate nucleus (LGN),
Huth, Gerald, 258 261
Huxley, Julian, 131 Le Blon, Jacques Christophe, 240-41
Huygens, Christiaan, 219, 221, 225 learning, visual, 137-9, 272-4
hyaloid, 5 (fig.) Leibnitz, Gottfried, 243
lemur, 143
ichthyosaur, 122-4 lens, 100, 120-22, 169-70, 178
images graded index, 121
inverted at the retina, 48, 170, 179 human, 6, 8, 74, 119, 178; see also
sewn together at the primary cataract
visual cortex (V1), 37, 49 lens cylinder, 109, 117
imprinting, 130 Leonardo da Vinci, 33, 165 (fig.),
induction, 2, 86 174, 188
infant development, 128-9, 146-7, Lerner, Aaron, 76
a7 Leucippus, 156
infra-red, 28, 224 Leuenhoek, Antoni van, 190
plants sensitive to, 74 Levin, Daniel, 289
snakes sensitive to, 28 Levi-Setti, Riccardo, 114
intelligence, 143-4 Lewin, Walpole, 278
invertebrate eyes, see compound light, 219-23
eyes in the womb, 7
ion channels, 74 polarised, 24-7
iris, 5, 6 regulates metabolism, 74-8
variations in intensity, 24
Javal, Louis Emile, 44 wave theories of, 221-3, 224
Jessen, Johannes, 175, 179 Locke, John, 269
Jolly, Alison, 143-4 Lorenz, Konrad, 130-31, 133
Julesz, Bela, 35-8
Macaque (monkey), 258
Karrer, Paul, 68 Mach bands, 202 (fig.), 202-3, 236-7
Kay, Emma, 135 Mach, Ernst, 45, 182, 201-3, 263
Kay, Paul, 214 machine vision, 115, 275—6; see also
Keller, Helen Adams, 14 prosthetic vision
Kepler, Johannes, 174-7, 179-80 macula, 191, 196, 198
Kirschfeld, Kuno, 107, 108 (fig.), 117 macular degeneration, 13, 281
Kohler, Ivo, 273-4 macular pigment, 29
Kihne, Wilhelm, 57 (fig.), 58 Magendie, Francois, 64-5
Mallock, Henry, 107
l’Isle-Adam, Villiers de, 54 mantis shrimp, 31, 286
La Mettrie, Julien Offray de, 268 Marella splendens, 95, 97
lamp shell (Brachiopoda), 97 Mariotte, Edmé, 188-9
Land, Edwin, 212, 233-8 marmosets, 30
Land, Michael, 118, 138-9, 268 Marr, David, 206, 280
INDEX 319

marsupials, 257 nerve cells, 7, 200, 242

materialism, 159, 183-4, 219, 268-9 nervous system, 198-200, 233
Matthews, Rachel, 203 neural superposition, 117, 118 (fig.)
Matthiessen, Heinrich, 120-21 Newton, Isaac, 219-22, 225
Maxwell, James Clerk, 185, 195, 196 nictitating membrane, 124-5
_ (fig.), 224, 234, 237 night blindness, see vitamin A
‘Mayr, Ernst, 85, 86 night vision, 24, 62-3, 117, 193-5,
Mead, Allan, 143 209-10, 257
Mead, Margaret, 149 Nilsson, Dan-Eric, 99-101, 106, 117,
melatonin, 76-7 138
memory, 136-7 non-lethal weaponry, 265-6
Mery, Jean, 183-4 Notonecta, 26
Mesmer, Franz Anton, 269-71 Nummenmaa, Tapio, 151
metabolism, regulated by light, 74-8
Meyerhof, Otto, 68 objects, perception of, 131-3, 158-9
mind, analogies for, 262-3 Occam’s razor, 173
minnow, 76 ocelli, 275
Minsky, Marvin, 262 Ockham, see William of Ockham
mirrors, 74, 118, 174, 182 ommatidium, 26, 105-14, 204
Moken, 118-19 Omodeo, Pietro, 73
Mollon, John, 212, 251, 262 Ooman, H.A., 52
Molyneux, William, 31, 269 Ophiocoma wendtii, 22
Monet, Claude, 219 Ophthalmosaurus, 123-4
monochromacy, 260, 261-2 ophthalmoscope, 185-8, 191-2
Moorfields Hospital, 268 in fiction, 55-6
Morgan, Thomas, 250 opsin, 69-70
Mori, Masamichi, 65 optic chiasm, 29
Morse, Meroe, 234 optic flow, 112, 113 (fig.), 275
mosquito fish (Gambusia holbrooki), optic nerve, 13, 29, 169-70, 174, 179,
50 188, 203, 204, 208-10, 242
motion blur, minimising, 44, 111 optic nerve head, 190 (fig.)
motion perception, 42-3, 49-50 optogram, 58-61, 59 (fig.)
moving pictures, 38-9, 43 optokinetic reflex (OKR), 44
muffin pan illusion, 32 (fig.), 33 O’Shea, Steve, 23-4
Miiller, Heinrich, 57, 192
mustard gas, 265 Paley, William, 78-9
mutual inhibition, 205 (fig.), 206 Paradis, Maria von, 269-71
(fig.), 208, 235 parallax, 38
myopia, 9, 127 Parker, Andrew, 96-8
Pax6, 84
naked mole rat, 22 Pelger, Susanne, 99-101
Nathans, Jeremy, 251-2, 259 peripheral vision, 43, 50, 189, 195
nautilus, 100 persistence of vision, 43
Necker cube, 132 (fig.) perspective, 166, 173-4
320 INDEX

Phacops, 113-15 _ Quiring, Rebecca, 83

phi phenomenon, 43
photography, 87, 237 rainbow, 214
inspires vision research, 55—60 random dot stereogram, 36 (fig.)
photoreceptors, 74, 191, 194, 209, 239 Ransome, Joseph, 247, 248
evolution in vertebrates 8, 78, 98 Ratliff, Floyd, 203
(fig.), 99-101, 256-7, 259 Raudulf’s Saga, 26
in the pineal, 78 reading, 138
sensitivities of, 194, 227 (fig.), red cones (L cones), 251-2, 258-9
227-9, 258 refraction, 109, 121, 168, 169 (fig.)
variety of, 254-5 (fig.), 256-7 restored sight, 268-72
photosensitivity, non-visual, 74-6 reticulum, 200, 201
photosynthesis, YER SSIES retina, 6-8, 122, 125-6, 178-9,
Phronima, 109-10 181-2, 187-8, 190-95, 193
phytochromes, 74 (fig.), 200 (fig.), 203, 208-10,
Piccinni, Ester, 73 242-3, 257-8, 281
pineal body, 75-8 artificial, 281-3
Pingelap, 260 forensic potential, 56, 58-61
pipe rock, 95 retinal, 69
Pissarro, Camille, 218 retinal pigment epithelium (RPE), 6,
pit-lamping, 182-3 135, 182257
Plato, 158-60 retinex theory, 233, 236-8
Platter, Felix, 177-9 retinitis pigmentosa, 281
pointillisme, 212, 242 Retzius, Gustaf, 201
pointing, 146-7 rhabdom, 106
polarisation, 24-7 rhodopsin, 25, 56-9, 68-72, 251
polarised light, Rittenhouse, David, 32
communication by, 26, 286 Ritter, Johann, 224
navigation by, 25, 26, 275 Robinson, David A., 46
polarising filters, 197, 233-4 rods, 7, 13, 20 (fig.), 57-8, 73 (fig.),
in the eye, 197-8 74, 125, 194-5, 209-10, 251;
Pontella, 50 254-5 (fig.), 256, 257
presbyopia, 8 Royal Institution, 223
Prevost, Bénédict, 238 Royal Society, 183, 188, 222-3
primates, 143-7, 259 Rudolphi, Karl, 184
printing, 240-1
prosthetic vision, 19-20, 276-85 saccade, 17, 18 (fig.), 44-5, 47-8
pseudoscope, 34 with immoveable eyes, 47-8
Ptolemy, Claudius, 162 Sacks, Oliver, 261
pupil, 3, 24, 119 Scheiner, Christoph, 188, 191
Purkinje, Jan Evangelista, 40 (fig.), Schmidt, E.M., 279
40-42, 184-5 Schultze, Max, 192-3, 193 (fig.), 194
Purkinje shift, 193, 194 (fig.) sclera, 4, 141
Purkinje’s tree, 192 scorpions, 113
INDEX 321
Scott, John, 150 Tinbergen, Nikolaas, 130-1
seed shrimp, 96-7 toad, 15-16, 43
Seger, Jon, 86 Tomkins, Silvan, 149
serotonin, 77 tonic immobility, 140
Seurat, Georges, 241-2, 243 touch, 16-20, 22
sex chromosomes, 249-50 harnessed for prosthetic vision,
” shadows, coloured, 229-30 19-20
short-sightedness, see myopia role in machine vision, 274-5
Signac, Paul, 241-2 role in visual interpretation, 271-4
Simons, Daniel, 289-90 Towe, Kenneth, 114
sky, colour of, 217-18 Trattles case, 246
smalleye, 84 Treviranus, Gottfried Reinhold, 191
smooth pursuit, 44, 48 trichromacy, 225, 231-2, 247-8, 257
social behaviour, 143-5 trilobite, 93, 95, 97, 103-5, 204
social signals, 145-7; see also expres- Turner, J.M.W., 218
sions Turner, Robert, 268
Soemmering, Samuel, 29, 196
Sojourner Rover, 275 Uexkiill, Jacob von, 290
spectacles, 126 ultraviolet, 28-30, 224
natural, 124 patterns, 28
spectrum, 27-8, 220-1, 224-5, 227- protection from, 29, 198
2) uvea, 5, 178
Spemann, Hans, 82
spherical aberration, 120 (fig.) Valentin, G.G., 191
spiders, 113 Veiling Glare Laser, 266
Srinivasan, Mandyam, 275 Verne, Jules, 56
star-nosed mole, 22 Vernois, Maxime, 60
stereopsis, 33-8, 35 (fig.) vertebrate eyes, 21 (fig.), 21-2, 98
stereoscope, 34 (fig.), 99-100, 105, 119-26, 190-1
Stevens, Nettie, 250 vertigo, 41-2
stimulus, visual, 131 (fig.), 131-3 vestibulo-ocular reflex (VOR), 41
Stratton, George, 273 vision aids, 126-7; see also pros-
sunstones, 27 thetic vision
superposition, 116-17 visual cortex, 37, 277-9
neural, see neural superposition visual cycle, 69
visual streak, 125, 190
tapetum, see mirrors vitamin A
tears, 153 deficiency (xerophthalmia), 52-3,
Temnodontosaurus platyodon, 123 63-7
tetrachromacy, 253, 257 dietary sources of, 52, 65, 67
Thatcher illusion, 148 (fig.) role in vision, 68-9
Theon, 155, 160 vitreous humour, 13
Thompson, Peter, 148 Vogt, Klaus, 118
Thomson, William, 245 Vygotsky, Lev, 276
322 INDEX

Wald, George, 28, 59, 67-72, 68 X-chromosome, see sex chromo-

(fig.), 91, 203, 258 somes
Wallace, Jean, 271-2 X-chromosome inactivation, 250,
Walls, Gordon Lynn, 253, 256 252
Warburg, Otto, 68 Xenophanes, 214
wasps 14-15, 111
water, Yarbus, Alfred, 17, 146
blueing effect of, 30-31, 71 Y-chromosome, see sex chromo-
human visual acuity in, 118-19 somes
water boatman, 26 Young, Thomas, 222-5, 231-2, 248
Weiskrantz, Lawrence, 49 Young-Helmholtz Theory, 225, 230,
Westheimer, Gerald, 268 231, 247
Wharton Jones, Thomas, 186 Yudkin, Arthur Meyer, 68
Wheatstone, Charles, 33-4
William of Ockham, 173 Zahavi, Amotz, 153
Wilson, Edmund Beecher, 249-50 Zajonc, Arthur, 155-6
Wiwaxia corrugata, 97 Zeki, Semir, 262
Wolf, George, 63 Zinni, Anthony, 265
A NOTE ON THE AUTHOR

Simon Ings is a novelist and science writer. A Natural History

of Seeing was written in between the birth of his daughter and
film-making expeditions to Ladakh, Arabia’s Empty Quarter,
and Arctic Norway. His science features and interviews have
featured on national radio and in magazines as diverse as New
Scientist, Wired, and Dazed and Confused. He lives in London.
par lo Oh) Be im ; ey:

ai acsrruA ee ee
qe }

q heey ‘oh, 2% bine

US snayait vies s2plan amma my
hactin spe ciscow tie’ AagacneT a Bond ep
.- yee ore i Canis ee senibaes goa
casepeid shel iow bee zeigt ae
aia
ale ues» eet tg saetibngnd ptt has Iptpoiits
f “0 4 in > <

; po: Saat Sit| ee sh ixvaands 4 ud

OFF Sc
pi tess ny ees,
et, eeoe Ci- Tee oar Tr’ + <r ee

ir Os ae ae Pala Apay, 24fm0

Pie ay TREE,
aw as Z =
ey ala SS

cee

eae
* — x
ee

aa v + neta =’. Ca

® ae

om i
xo Lonces
o F PPP
pRore ®
“Far-ranging and wonderfully eclectic...
popular science at its best.”

IMON InGs’s most recent novel is The Weight of

Numbers. His science features and interviews have

appeared in magazines ‘as diverse as New Scientist,

Wired, and Dazed and Confused. Ings lives in London.

Visit his Web site at www.simonings.net.

CH
JACKET DESIGN BY ROBERTO DE 'VICQ DE CuUMPTI

JACKET PHOTOGRAPHS BY PETER MASON / Getry IMAGES

AutHor PHOTOGRAPH BY ANNA Davis

A
PRINTED IN THE UNITED STATES OF AMERIC
Early praise from the UK for
A NATURAL HISTORY
OF SEEING

“Keats worried that science unweaves

the rainbow. Had he read Simon Ings,
he would have realized that actually -
science gives us more reasons every
day to celebrate the world. With a
science writer’s precision and a novel-
ist’s talent for vivid analogy, Ings
weaves together science, history, the
arts, and personal observation. What to
call this elegant hybrid of research and
wit and style? I call it a work of art.”
—MICHAEL SIMS, author of Adam’s~
Navel: A Natural and Cultural History ~
of the Human Form
“A narrative as arresting and
remarkable as any fiction, accessible ©
but complete, with the reader
assumed to be an intelligent adult
on the lookout for something
substantial. An excellent guide to
one of the world’s true wonders.”
—THE TELEGRAPH

“A rich and eclectic survey,

with an intriguing nugget on
almost every page.”
—THE SUNDAY TELEGRAPH

“Surprisingly appealing and

readable. ... Read it and you will neVer
see things in the same way.”
—THE OBSERVER

“Elegant, entertaining and

/
up-to-date ... utterly compelling.”
—THE INDEPENDENT

SKS SNS

Dey
| || Ml iii
780393"067194 i
W. W. NORTON
sea 978-0-393-06719-4
NEW YORK + LONDON
WWW.WWNORTON.COM

Our Senses - Sight: The Eye
No ratings yet
Our Senses - Sight: The Eye
12 pages
2022-2023 JS3 TM1 Note
No ratings yet
2022-2023 JS3 TM1 Note
36 pages
Understanding Eye Function and Communication
No ratings yet
Understanding Eye Function and Communication
6 pages
Sense Organs
No ratings yet
Sense Organs
8 pages
0853237557
No ratings yet
0853237557
225 pages
The Human Eye
No ratings yet
The Human Eye
38 pages
Vision How It Works and What Can Go Wrong Full Version Download
No ratings yet
Vision How It Works and What Can Go Wrong Full Version Download
16 pages
Eye Facts for Curious Kids
No ratings yet
Eye Facts for Curious Kids
1 page
BIO-Project Investigatory
No ratings yet
BIO-Project Investigatory
83 pages
BlakeSekuler5e Chapter2
No ratings yet
BlakeSekuler5e Chapter2
42 pages
Year 9 Chapter 8 Sense and Control Sample Answers
No ratings yet
Year 9 Chapter 8 Sense and Control Sample Answers
80 pages
Baasic Science Note JS3
No ratings yet
Baasic Science Note JS3
26 pages
Human Eye
100% (2)
Human Eye
72 pages
100 Eyes
No ratings yet
100 Eyes
18 pages
The Eye
80% (10)
The Eye
252 pages
Protecting Young Eyes: A Guide
100% (1)
Protecting Young Eyes: A Guide
40 pages
Eye & Ear
No ratings yet
Eye & Ear
27 pages
Anatomy of the Human Eye Explained
No ratings yet
Anatomy of the Human Eye Explained
34 pages
Anatomy of The Eye
No ratings yet
Anatomy of The Eye
34 pages
Eye My
No ratings yet
Eye My
38 pages
Anatomy and Function of the Human Eye
No ratings yet
Anatomy and Function of the Human Eye
10 pages
Sight
No ratings yet
Sight
26 pages
Eyes
100% (3)
Eyes
18 pages
14 1. Sense Organs S.4
No ratings yet
14 1. Sense Organs S.4
30 pages
Children's Activity Book
100% (6)
Children's Activity Book
48 pages
Science Booklet Year 1 202
No ratings yet
Science Booklet Year 1 202
95 pages
WellBeing Special Report - Improving Your Eye Health
No ratings yet
WellBeing Special Report - Improving Your Eye Health
16 pages
How Do Our Eyes See
No ratings yet
How Do Our Eyes See
36 pages
Plan Lector Ingles
No ratings yet
Plan Lector Ingles
17 pages
Nithesh Bio IV
No ratings yet
Nithesh Bio IV
25 pages
Eye Lecture Guide
0% (1)
Eye Lecture Guide
46 pages
Human Eye and
No ratings yet
Human Eye and
8 pages
Human Eye - X
No ratings yet
Human Eye - X
17 pages
Preface
No ratings yet
Preface
8 pages
Lesson 13 - Slides
No ratings yet
Lesson 13 - Slides
31 pages
Understanding the Human Eye and Vision
No ratings yet
Understanding the Human Eye and Vision
26 pages
Sensing and Perceiving 4.2
No ratings yet
Sensing and Perceiving 4.2
18 pages
Part 11 Introduction To Psychology by Kalat 9th Edition Summarized Version by Me
No ratings yet
Part 11 Introduction To Psychology by Kalat 9th Edition Summarized Version by Me
19 pages
Human Eye and The Colorful World PDF
No ratings yet
Human Eye and The Colorful World PDF
12 pages
O'Regan
No ratings yet
O'Regan
230 pages
To Understand How The Eyes See. To Label The Parts of The Eye and Describe Their Function
No ratings yet
To Understand How The Eyes See. To Label The Parts of The Eye and Describe Their Function
14 pages
Anatomy and Functions of the Eye
No ratings yet
Anatomy and Functions of the Eye
6 pages
Tugas Bung Majalah
No ratings yet
Tugas Bung Majalah
10 pages
Form 2 East Text Book Scan
No ratings yet
Form 2 East Text Book Scan
25 pages
Sensory Receptors Explained
No ratings yet
Sensory Receptors Explained
59 pages
Human Eye and Vision Defects
No ratings yet
Human Eye and Vision Defects
7 pages
Light & Eyes: - Vision Begins When Light Comes Into The Eye - Light
No ratings yet
Light & Eyes: - Vision Begins When Light Comes Into The Eye - Light
31 pages
Anatomy of An Eye
No ratings yet
Anatomy of An Eye
66 pages
Adobe Scan Oct 19, 2022
No ratings yet
Adobe Scan Oct 19, 2022
9 pages
UNIT 6 Student Handout2023-I
No ratings yet
UNIT 6 Student Handout2023-I
12 pages
Esa80005 1586 1592
No ratings yet
Esa80005 1586 1592
7 pages
DPP PDF Light (Part - 3)
No ratings yet
DPP PDF Light (Part - 3)
8 pages
B.SC Term 2
No ratings yet
B.SC Term 2
33 pages
Acknowledgement 3
No ratings yet
Acknowledgement 3
12 pages
Special Senses Project
No ratings yet
Special Senses Project
15 pages
Sense Organs Eye
No ratings yet
Sense Organs Eye
10 pages
Onisha Biology Project 14 10 15
No ratings yet
Onisha Biology Project 14 10 15
14 pages
DK Animals
100% (5)
DK Animals
31 pages
Topic 5 Compensation & Benefits
No ratings yet
Topic 5 Compensation & Benefits
22 pages
Mpce 011 PDF
No ratings yet
Mpce 011 PDF
9 pages
Ms. Sunita Sunita (Adult, Female)
No ratings yet
Ms. Sunita Sunita (Adult, Female)
2 pages
Covid 19 Prophecy
No ratings yet
Covid 19 Prophecy
9 pages
Report Form For SBAR
No ratings yet
Report Form For SBAR
1 page
Anita Singh, Et Al
No ratings yet
Anita Singh, Et Al
10 pages
Identification and Management of Adult Patients Requiring Mealtime Assistance
No ratings yet
Identification and Management of Adult Patients Requiring Mealtime Assistance
19 pages
Brosur - CSSD Basic Course 2021 - (8th - 12th November 2021)
No ratings yet
Brosur - CSSD Basic Course 2021 - (8th - 12th November 2021)
2 pages
Special-4 Instrumentation For Package Units
No ratings yet
Special-4 Instrumentation For Package Units
1 page
Not in The Script An If Only Novel Amy Finnegan Download
100% (3)
Not in The Script An If Only Novel Amy Finnegan Download
38 pages
Suraksha Pharma Contract Manufacturing Performance
No ratings yet
Suraksha Pharma Contract Manufacturing Performance
73 pages
The Children Song Answers - VIII
No ratings yet
The Children Song Answers - VIII
6 pages
"Indian Summer: A Poetic Reflection"
No ratings yet
"Indian Summer: A Poetic Reflection"
3 pages
Evaluacion Del Programa Familias Fuertes Amor y Li
No ratings yet
Evaluacion Del Programa Familias Fuertes Amor y Li
7 pages
ELECTRICAL ENG'G SAFETY WRITTEN REPORT Final
No ratings yet
ELECTRICAL ENG'G SAFETY WRITTEN REPORT Final
3 pages
Annexure C HSE Questionaire
No ratings yet
Annexure C HSE Questionaire
8 pages
2021 PRE UNEB S.4 AGRICULTURE Revision Paper 1 EXAMINATION
No ratings yet
2021 PRE UNEB S.4 AGRICULTURE Revision Paper 1 EXAMINATION
5 pages
Writing - Opinion Essay - Staying Up Late - 1
No ratings yet
Writing - Opinion Essay - Staying Up Late - 1
3 pages
High-Efficiency Polycrystalline Solar Modules
No ratings yet
High-Efficiency Polycrystalline Solar Modules
4 pages
Chapter 2 Material Balance Fazliedited
No ratings yet
Chapter 2 Material Balance Fazliedited
36 pages
Ratificación EN ISO 26845:2008
No ratings yet
Ratificación EN ISO 26845:2008
22 pages
Eddy Current Exam Questions and Answers
75% (8)
Eddy Current Exam Questions and Answers
99 pages
DRRR Quarter 2 Module 7 10
No ratings yet
DRRR Quarter 2 Module 7 10
6 pages
Astm D6161 - 1998
No ratings yet
Astm D6161 - 1998
10 pages
Cafe Operations Checklists
No ratings yet
Cafe Operations Checklists
5 pages
GE-50300 Operator Manual
No ratings yet
GE-50300 Operator Manual
152 pages
Flame-Proofing Agent APYROL BKW Guide
100% (1)
Flame-Proofing Agent APYROL BKW Guide
2 pages
Uday Paper Article 1
No ratings yet
Uday Paper Article 1
13 pages
792 Compact AC Overhead Lines
100% (1)
792 Compact AC Overhead Lines
163 pages
Parle G
100% (1)
Parle G
26 pages

A Natural History of Seeing - The Art and Science of Vision - Simon Ings - 2008 - W. W. Norton & Company - 9780393067194 - Anna's Archive

Uploaded by

A Natural History of Seeing - The Art and Science of Vision - Simon Ings - 2008 - W. W. Norton & Company - 9780393067194 - Anna's Archive

Uploaded by

4 ¢ |

Only one percent of what we see

The blind can be taught to see

n graceful, accessible prose, novelist and sci-

and other mysteries of seeing, taking us

ability to raise the inner corners of his eyebrows.

evolution of sight and the evolution of our under-

The Weight of Numbers

All rights reserved

For information about permission to reproduce selections from this book,

Manufacturing by Courier Westford

Library of Congress Cataloging-in-Publication Data

W. W. Norton & Company, Inc.

W. W. Norton & Company Ltd.

— new eyes for old

rw iar ite _ 6 Rd fel

' i eae aed eee ht = 7 Pia

es owe 5 fee pan ON

chew tne F LF D date

3 How are eyes possible? ye

4 The adaptable eye 102

) Seeing and thinking 128

6 Theories of vision 154

2 Unseen colours 244

EPILOGUE - The invisible gorilla 286

Further Reading 291

a sitet a wie0 wt gitfeo inetparihala tear se ft

i)eae CeAte ell

Youth and age

shaped. Instead of a nose, it had ‘a fleshy growth 3/4 inch long by

ine blood vessels

The human eye.

work of the Greek physician Herophilus (335-280 sc). Herophilus is

The developing eye’s temporary blood supply (the hyaloid),

the brightness of day, as acquire a new way of seeing. In her retinas

This first year of my daughter’s life is visually simple. A few familiar

Five, I trace the evolution of that commonwealth of the senses, and

an arcus senilis made of cholesterol - nothing to worry about, not

The commonwealth of the senses

Helen Adams Keller, daughter of a former captain in the Confed-

This chapter is entitled “The commonwealth of the senses’ to convey

1 — Seeing and touching

Alfred Yarbus’s record of eye movements while

Certainly. Keller cannot see colours - I can. This is a real and

your brain a couple of hours to start seeing through your back. If

If you swim in the sea, sooner

THE COMMONWEALTH OF THE SENSES oo

cornea iris lens visual cells

Convergent evolution: though unrelated, the vertebrate eye (top)

an eye can take, so we should not be surprised when quite unrelated

THE COMMONWEALTH OF THE SENSES mae}

spondent: ‘I can assure you that the eyes of Mesonychoteuthis are

THE COMMONWEALTH OF THE SENSES 25

a plane parallel to the water's surface; some water bugs, including

THE COMMONWEALTH OF THE SENSES 27

The sun emits energy at different wavelengths, and visible light is

THE COMMONWEALTH OF THE SENSES a0)

have up to twice as many different colour receptors in the eye as do

information from the eyeballs, as the eyes converge slightly to focus

THE COMMONWEALTH OF THE SENSES 33

. a painting, though conducted with the greatest art and

This explanation is not quite sufficient. It assumes that every object

student at the Hungarian Academy of Science, embarked on a thesis

easy to judge, thanks to another technique for depth perception,

3 — The moving eye

Vision itself is a dynamic process. There is little in the world

Our eyes provide us with a picture of the world. It is an odd, magical

For the staid and sedentary professors of the University of Prague,

Jan Purkinje in a spin.

THE* COMMONWEALTH OF THE SENSES 41

interested in a new method of pacifying the insane which had come

gravity had changed direction, pulling us, not downward, but to

THE’ COMMONWEALTH OF THE SENSES 43

Take a moment to scan your surroundings: turn your head slowly

4 - The ‘problem’ of consciousness

D.B-s head was a bomb, ready to explode. The malformed blood