6G studies in

comparative
semantics ry,

paul canart
 Oo a A Cl

CC

¢
o
G 7 o 1 a
- s L C 4 4
= 5 t .

“
4

-¢
a “i

oO
\ 3 a)

%
a)
& 2 2 : - = oI P

J
4
‘VU
de
Co

:
5 “e ~ - c L

. - "1 G: t

, o-a>= Si ~ > v 3
 Studies in
Comparative
Semantics
 Studies in
Comparative
Semantics
Paul Canart

Edited by Wendy Cobcroft

With a Foreword by A.R. Chisholm

St. Martin’s Press

NEW YORK
© University of Queensland Press, St. Lucia, Queensland, 1979

Allrights reserved. For information, write: St. Martin’s Press, Inc.

175 Fifth Avenue, New York, NY 10010

Printed in Hong Kong

First published in the United States in 1979

ISBN 0—312—77087-1

Library of Congress Cataloging in Publication Data

Canart, Paul
Studies in comparative semantics.

Includes bibliographical references and index.

1. Semantics, Comparative. I. Title.
P325.5.C6C3 1979 412 78 —24328
ISBN 0—312—77087-1
To A. R. CHISHOLM

Paul Canart held Professor Chisholm in the highest esteem and affection,
and their association, stretching back over forty years, made it especially
fitting that Professor Chisholm should write a Foreword to this book.
In so doing, he has fulfilled a wish expressed by Paul Canart from the
time of the book’s inception.
It is therefore with a profound sense of gratitude that I take this
opportunity of sincerely thanking Professor Chisholm for his sensitive
Foreword, and of acknowledging that the publication of this work has
been made possible largely as a result of the very deep personal interest
that Professor Chisholm has taken in it.

W.E.C.
Contents

Abbreviations
Foreword
Preface
Introduction
Chapter I: Semantic Developments from “At Once”’
Chapter II: Semantic Groups inMalay-Indonesian
Chapter III: The Bow
Index
 Abbreviations

Ar. Arabic Mal. Malay

Cat. Catalan OEng. Old English
Ch: Chinese OFr. Old French
Dan. Danish Prov. Provencal
Du. Dutch Pt. Portuguese
Eng. English Rum. Rumanian
Finn. Finnish Rus. Russian
Pr: French Sino-Jp. Sino-Japanese
Ger. German Sk. Sanskrit
Gk Greek Sp. Spanish
Te: Icelandic Sw. Swedish
Id. Indonesian Turk. Turkish
It. Italian : cognate with
Jp. Japanese a hypothetical form
Lat. Latin lit. literally
Foreword

When, many years ago, Paul Canart was enrolled as an honours

student in French, I realized at once that his linguistic ability
was outstanding.
Born in Belgium, he completed his pre-university education
in Australia, but although his English was impeccable, French
remained his native tongue, the one in which I practically always
conversed with him. Very soon it became evident that he was a
fine Latinist also, and our discussions moved steadily into the
area of comparative linguistics.
He had some very distressing setbacks. At an early stage he
broke a leg, and after a slow recovery was beset by incipient
tuberculosis. But despite his many miseries he remained extra-
ordinarily active, and I was very glad, when the opportunity
came, to offer him a post on the French staff—an appointment
that paved the way for highly specialized philological teaching
in the Department.
Ill health was his persistent enemy. Though he faced it
manfully, even learning to adapt himself to the loss of an eye a
few years ago, it undermined him relentlessly, and he died in
1974 just before reaching the statutory retiring age.
After his appointment to a post in the Department of French,
his interest 'in comparative linguistics grew apace, and his
eloquent enthusiasm attracted many students. On several
occasions we even arranged public talks for him on the subject
of linguistic derivations, and they became widely known. I
was not surprised, therefore, when, in the late 1950s, his
“intervention” in the public discussion of a doctoral thesis at
the Sorbonne made such an impression on the jury that he was
invited to give a few talks on comparative linguistics. He had
xX Foreword

demonstrated, in his intervention, that both the candidate and

the jury had kept too narrowly to the restricted field of French
philology.
It needs to be stressed that a linguist, in the sense in which
the term applies to Paul Canart, is not a mere polyglot. He
studies many languages, but what interests him above all is
the relationship that can be discovered among them. And in
his search for such interconnections Canart was guided by a
native intuition that at times seemed almost visionary (though
it was, of course, unfailingly controlled by scholarly discipline).
Few, if any, have gone so far in the discovery and charting
of linguistic areas which, although they are remote from each
other, have common features, pointing towards some universally
valid psychological instincts in humanity.
His investigations, in the latter part of his career, led him
more and more towards the study of semantics. What he dis-
cerned in this field was not a principle of derivation and
borrowing, but a natural tendency, common to all human
groups, to seek an adequate equipment for the expression of
thought. And Paul Canart’s theory is that this equipment has
universal characteristics, whatever the language group may be
in which it is evolved.
To investigate this vast semantic area, Paul Canart imposed
upon himself the task of acquiring an accurate insight into an
astonishing number of languages—European, Semitic, Indian,
Far Eastern.
A striking and indeed strangely dramatic example of semantic
principles common to all languages is provided, in the present
book, by the study of the far-reaching influence of ‘“‘the bow”
in the evolution of words and constructions relating to tension.
The bow provided, all over the world, an abundance of both
physical and metaphorical images. They may sometimes look
like borrowings, but often this is ruled out by the distance in
time and space that often separates the language groups in
which they occur. They simply come forth from the depths of a
psychological heritage shared by all human beings.
A particularly ingenious application of Paul Canart’s research
methods is seen in his tracking down in Chinese (and Japanese)
writing of an urge indicated by the pointer in a given character,
 Foreword XI

which makes its generic purpose definite. The work done by

Canart in this respect shows how a given concept produces a
certain linguistic form and then conceals itself, as it were, but
is still subtly visible. In this, as in many other ways, the written
form of a language is usually historical rather than topical.
Hence the enormous discrepancies in the spelling of such
languages as English and Gaelic—discrepancies that manifest
themselves to various degrees in French, with its innumerable
silent consonants; in Arabic, with its silent -t ending for feminine
nouns and adjectives; in German, with its occasionally puzzling
combination of s and z, and so on.
Incidentally, Canart’s conclusion about the Chinese pointers
is curiously confirmed, it seems to me, by the evolution, in
Egyptian hieroglyphic writing, of the generic determinative,
which indicated, when necessary, the semantic background of a
hieroglyph. This demonstrates in a vivid fashion the existence
of a psycho-linguistic universality that bridges both time and
space.
Among the many fascinating conclusions and suggestions
in this brilliant book, I have to limit any further introductory
remarks to the theory that bright and dark vowels tend, in
most languages, to be used in response to a natural urge. This
is only a tendency, of course, and it would be dangerous to twist
it into a governing principle. It is certainly recognized by poets,
and is one of the expedients used for producing an incantatory
effect. There is a memorable example in Mallarmé’s famous
sonnet, “Le Vierge, le vivace et le bel aujourd’hui”’, in which all
the rhymes are built on the sound of the letter 7, compelling the
reader to feel the bite of winter, to be dazzled by the pitiless
whiteness of the landscape.

The foregoing sketch of Paul Canart’s career and research

activities is woefully inadequate. It is really little more than a
deeply sympathetic epitaph. But I venture to hope that this
epitaph is, like the pointers in Chinese writing and the Egyptian
determinatives, both revealing and commemorative, presenting
the picture of a remarkable scholar at work, and helping to record
the vastness of his linguistic vision.

A. R. Chisholm
 ile ae ‘
4 oe ~

7 dPsate Wat| Y
ial
we 6a BDae Seal
tna ae ep ay .
ccna a
ter
SO
ery a et ig
ai”, : e

oS cont ha aan
ae al” See eee a

al = u an aa

Se ate
avery ae Side
eave ene
ewe eRalae
st-@e'
7 as uae tans Soar 4 a, 5
2a qqereae.- ¢
eae ; ee eeeo) ee

awe ae Pie

’ AW eed ae
Co led >a a ed ath
at;
me ae
ery
oa. Tha
= ira
meds

coe
¢eise®.y hs ; eae al ered Ba
,ict aa R'

Mp sa a’ Aba,
— ei

at ey oe |
— ae
“aE ial
YON re
scien Mees
oot is oe ar id

non
Preface

The present work was left incomplete by the death of Paul

Canart in June 1974. Although he had put it aside during the
previous two years, owing mainly to the loss of an eye in August
1972, he had nevertheless intended to resume work on it in 1975,
following a trip to Europe. These plans were tragically cut short.
Paul Canart never revised the manuscript, and consequently
there have had to be a number of changes made to the original
text. Alterations of a stylistic nature have not been pointed out,
since these must be regarded as a necessary part of the editing of
any unrevised work; what is most important is that they do not
affect the essential content. Changes of order have been made
in the interests of a greater sense of overall unity. This, I would
suggest, requires no special justification, for it is quite normal
to try various possibilities before deciding on a final order,
and Paul Canart would certainly not have hesitated to make
changes of this kind.
The most important alteration, and the only one that needs
to be commented on specifically, concerns the organization of
Chapters II and III. “The Bow”, which originally constituted
Chapter II, has been made the subject of the final section, since
it undoubtedly provides the most suitable ending for the book.
As a result of this reorganization, the discussion of “Semantic
Groups in Malay-Indonesian” (which incidentally contains much
fascinating European material by way of illustration and com-
parison) now forms Chapter II. It must also be pointed out that
the Malay-Indonesian sections dealing with the vowel u and
the endings -ut and -ung are largely independent of the other
groups examined, apart from the obvious question of their
symbolism, and moreover they are clearly inseparable from the
XiV Preface

semantic field surrounding “the bow’. Consequently they have

been placed immediately following the exploration of “hair’’
and ‘“‘grasses”’ on pp. 108—13. These changes have made it possible
to give the chapter on the bow the prominence the author
intended it to have, without adversely affecting the discussion
relating to Indonesian.
Had Paul Canart been able to complete the work himself,
the final presentation may well have been substantially different
from the one now adopted. Moreover the additions that I
have been required to make in order to achieve some degree of
completeness would undoubtedly have been dealt with quite
differently. Precisely because of Paul Canart’s very unorthodox
approach to language, I must stress that I do not know just
how he would have treated those points for which he had jotted
down certain key words and which I have had to take up. Neither
can I be sure that he would have been completely in agreement
with my discussion of them. Thus, while my additions are based
on the author’s own material, points of interpretation included
in the writing up are my own, and it goes without saying that
any errors in these sections are entirely my responsibility. My
additions to the text are as follows:
1. The introductory and concluding paragraphs to Chapter I.
2. Pp. 93-100: the various terms were: discussed briefly in class
in 1969. Paul Canart had insisted that I write up the first part
of “The Bow” early in 1970 and he read through that draft,
which has since been somewhat modified. He took over the
writing after the initial discussion of “‘buxom”’:
3. Pp. 147—S0 and pp. 156-60.
4. Footnotes giving additional details on certain points: these
are marked Ed.
I have grouped together in what is now the Introduction
a number of general considerations, mainly concerning the
importance of as broad a geographical spread as possible, and
the need to combine a diachronic and a synchronic approach.
However, it. will be noted that there is no detailed statement of
method. Here it must be said that had Paul Canart lived to see
his book through, he would undoubtedly have incorporated
further material on this subject. But it also needs to be pointed
out that he never discussed questions of method in his lectures.
 Preface XV

He preferred to explore language in all its multiple facets,

without in any way trammelling its (and incidentally his own)
natural exuberance, for he was concerned above all else to show
the absorbing interest of language study. The five sets of lectures
given under the general title of Comparative Semantics between
1967 and 1973 and which drew on material from a wide range of
languages, both Indo-European and Oriental, had precisely this
informal, unpredictable quality, and no student who attended
these lectures could have failed to be captivated by Paul Canart’s
enthusiasm when dealing with questions of meaning. Actually
a considerable amount of the material that is discussed in
the following pages was presented in class during the period
just mentioned, the notable exception being the Malay-
Indonesian content of Chapters II and III.

In order to appreciate this book, it is essential to be aware of

the great importance Paul Canart attached to the languages of
Asia, especially Japanese and Chinese. In addition to a profound
admiration for Oriental civilization and culture, a further decisive
factor was his belief that Australia should have much stronger
links with the countries of Asia. For, despite our geographical
proximity, relations have never been in any sense close, except
perhaps from an economic point of view; culturally, Australia
has never really interested itself in Asta, and has never encouraged
to any significant degree exchanges along these lines. It was with
a view to going some way towards remedying this highly unsatis-
factory state of affairs that Paul Canart wished to foster an
understanding and appreciation of Asian languages, and in
particular Chinese and Japanese, whose use of ideographs, he
felt, made them of great value for linguistic study, and for which
he thus hada special affection.
Although he had introduced references to Chinese into his
lectures as far back as the early 1950s, and did in fact give an
address on its importance as an aid to the study of Western
philology at the 1955 Congress of the Australian Universities’
Modern Language Association held in Brisbane, it was only in
his courses on Comparative Semantics—especially from 1969
onwards—that Paul Canart allowed himself to give much greater
weight to Chinese and above all to Japanese. To what extent he
XVi Preface

was justified in including such material within the framework of

a French Department is a point that is open to question, since
these lectures, while providing a natural vehicle for the fostering
of his ideals, nevertheless had obvious practical limitations.
Be that as it may, it is beyond doubt that those who were
privileged to have Paul Canart as a teacher have benefited signi-
ficantly from their contact with his most original mind, so much
in evidence in everything he did. A few of us have been inspired
to undertake the study of Chinese or Japanese as a result of his
infectious enthusiasm and genuine interest in Asia, while those
who have not gone so far have nevertheless, I believe, gained
some awareness of the fascination of Oriental languages. The
bridging of East and West by means of lectures in Comparative
Semantics began as a bold experiment, and it proved to be a
most exciting experience for those of us who had the good fortune
to share in it.
This work aims to give pride of place to Asia, but at the same
time it must be stressed that it does not in any way attempt to
play down the value of European languages; rather does it set
out to draw attention to the fact that the time is long overdue
for the languages of Asia (and by implication the various peoples
and their respective cultures) to be given serious attention here
in Australia. After all, we are ideally situated to forge important
ties with the countries of the Asian region and to enrich our own
lives accordingly. We should be able to offer an example to other
Western nations in actively seeking to promote a genuine under-
standing and appreciation of our northern neighbours. Surely
one of the best means of doing this is to make an effort to further
encourage the study of Chinese and Japanese, whose relative
neglect up until recently has undoubtedly been due to political
factors, as well as to the inevitable psychological barrier created
by their use of ideographs. It is time for a radical change of
attitude towards this whole question, and it is my earnest hope
that this book will pave the way for just such a change.

I must finally mention that while working on this book I have

been. constantly reminded of the circumstances surrounding the
preparation of the Cours of the great Swiss linguist Ferdinand
de Saussure, some sixty years ago. Paul Canart remarked to me
 Preface XVii

on a number of occasions over the years that I would one day

make his work known, just as Saussure’s students had undertaken
a similar task. I used to insist that he would be well able to attend
to his own publications, at least for some years. But his insistence
to the contrary has turned out to be prophetic; it is tragic indeed
that he did not live to complete this book, much less to see it
published.
Where I have been rather more fortunate than those who
prepared the work of Saussure is that the material at my disposal
was largely written up, as has already been noted. Moreover I
have the author’s lecture notes, which should enable other aspects
of his research to be published in the not too distant future.
There is one further point of comparison with Saussure which,
I trust, will not be considered out of place here. Saussure’s Cours
presented an exposition of certain principles of linguistics that
revolutionized thinking in this science; the present work deals
with semantics—precisely that branch of language which
Saussure was prevented from doing more than touch upon. Of
course, no one would deny that considerable attention has been
focussed on semantics since Saussure, but part of the originality
of Paul Canart’s book lies in the fact that it illustrates aspects of
a truly comparative semantics, on a far wider scale than anything
that has hitherto been attempted.
Yet despite the underlying breadth of vision, it presents only
a fragment of the immense linguistic knowledge that Paul Canart
possessed; on the other hand, it does offer a refreshingly new
approach, especially in terms of the unique emphasis it places
on Oriental languages and their interest for semantic exploration.
How this work will be judged remains to be.seen, but I believe
that in spite of the fact that so much more could have been
written had circumstances permitted, it shows sufficient evidence
of a highly original and brilliant mind to be given serious atten-
tion, and it is hoped that it will provide a starting-point for a
re-examination of the study of semantics.

Wendy Cobcroft
 en ’

So henna!
+i, eee
plates ducaipattintn ssh
ean Seae a ; ” i

a,Siena sid tatunions

ekeadont eens eee gas
an sna eaten ee
Duapeg ita
etlnletadeitigel, Shite 4 ay Pte
» ed fey signer :
WesS-BshPoe aes = enol:
bak eye EP we eee =

eitenigits's - el is Rete tent ee ' otra

5 ate Pp Te coe
ty nr ; or
ee ee :
SE ee
et’ tc re ey 7 a
eos ep ~ :
La me here

oe hoe +" ee Se kedsvy

re. OS Kasi -_
aS, : \ eho Oe Re ae
a coe ieedayiaived be
oth ee a a

aa nem oe -

ae see aha ~~
Soe AON te ee
Introduction

Of course light comes from the East. Man has always known
this and never more so than in primitive times when, turned
towards the east, he anxiously awaited the coming of dawn,
which would at once dispel all the horrors and dangers of the
night. There is unmistakable evidence of this in the Sanskrit
terms designating the main cardinal points: the east is in front,
and thus the west is behind and the south on the right side. For
example, Sk. daksina, ‘‘south’”, is obviously cognate with Lat.
dexter. Nor is it an accident that the Yemen at the far south of
the Arabian peninsula is so called, for this is the usual Arabic
term for the right side, and in Hebrew it refers to the south as
well.
However, the light referred to here is not the physical light
of the sun, nor is the East that of India or Arabia, but rather
what is usually called the Far East. The light meant is that of
the human spirit, as it is evidenced in the way human beings
think and speak, and nowhere is this more clearly seen than in
their languages, gradually moulded as they have been through
endless stretches of time and commonly assumed by Westerners
to be radically different from their own mode of thought and
speech. But the time will ultimately come when the history of
European thought and civilization will have to take Eastern
thinking into account, and the famous Chinese inscrutability
will then be seen with new eyes and discovered to be a rather
quaint myth belonging to the dark ages.
To my mind, the study of Chinese and Japanese is undoubtedly
one of the most rewarding and exhilarating linguistic experiences
that can be undertaken, and for years I have been urging the
importance of a working knowledge of one or other of these
2 Comparative Semantics

languages for Western linguists. Moreover I have always stressed

that our problems need to be re-examined in the light of such
Eastern studies, just as scholars in the East would benefit from
a similar study of Western languages. However, it has been
suggested to me on a number of occasions that our Western
linguistics undoubtedly has much more to offer than to gain.
The present work is an attempt to show that in many ways Far
Eastern languages can throw light on what appear to be specifi-
cally Western phenomena.
Now, if our approach is to be at all fruitful, it is obvious that
language must be considered as a means of communication, that
is, it must be meaningful. To some extent, then, this work re-
presents a protest against the sterilizing tendency to keep all
meaning out of linguistic research—the tendency that has finally
led to the ultimate absurdity of linguistics without language. The
practice of some present-day “‘linguists” to concoct utterly arti-
ficial examples as a basis for analysis is hardly any less ridiculous.
Our attention will be drawn especially to range and change of
meaning; hence this will be essentially a semantic study, and
more precisely an exploration in comparative semantics.
There is a valuable lesson to be learned from this method of
exploration about the actual nature of a given object. In a certain
language and culture an object is seen in a particular way, and
this manner of seeing it tends to be taken for granted as being
the only conceivable one of viewing and considering the object.
As soon as you learn another language, say French or German
or Russian, you discover a new way of looking at those objects,
which are now seen from a new angle, and unexpected aspects
emerge gradually. However, English, Dutch and German on the
one hand, and English and French, or again French and Italian
or Spanish on the other, are so similar and have developed in
such close association with one another that they are far more
alike than they are different, and it is only in a fairly small number
of cases that the old objects are discovered to be entirely new.
At this point a warning must be sounded, for much of our
teaching of foreign languages has been based on a fallacy. The
closer languages are, the more numerous are likely to be the
so-called faux amis, that is, words which look or sound alike
but which in practice are very different in sense; and much is
 Introduction 3

made of these, since they must on no account be confused. But

where our teaching is utterly wrong is that those faux amis are
kept strictly apart, that we merely take care not to confuse them
and leave it at that, whereas it would be far more profitable to
try to discover how those ‘“‘faux amis” came to have parted
company; for it often turns out that originally they were good
friends, and that they have been separated by the accidents of
phonetic or semantic developments.
To take one particular example: it is essential, of course, not
to mix up Fr. /uxe, “‘luxury’’, and Fr. luxure, “‘lust’’, but it is
far more illuminating to discover that they were once one and
the same and that fundamentally they turn out to be specific
outcrops
of the field of “hair” and “grasses” that we shall be
exploring in Chapter HI. Only by such a process can the study
of languages as close as English/French, Dutch/German, Italian,
Spanish/Portuguese and so on be really valuable. For differences
are the things that attract our attention, and in general it may
be assumed that the more unlike two languages are, the more
valuable their study will prove to be.
Especially if we want to discover the real shape of a given
object—and here I use “‘object”’ not merely in its usual physical
sense, but also in its derived metaphorical and figurative develop-
ments which often turn out to be most illuminating—that object
must be viewed and examined and explored from as many angles
as possible; the only way to do this is through as many eyes and
new experiences, that is through as many languages, as possible.
Therefore we certainly should not confine our interest to
Romance, or Germanic, or Slavonic, or even Indo-European,
however bewilderingly vast any of these particular language
families may be; we must go far beyond these limits in our search
for the true nature of the object. And to this immense synchronic
exploration must be added simultaneous probing in depth, as
far as historical evidence is available. Thus we must not be
satisfied with, say, French or German; we must go back as far
as we can, to Old French and Latin and in many cases back to
Greek and even Hebrew, and certainly to Middle High German
and Middle English and even Old English and Scandinavian;
and we must add this diachronic view of the object, as far back
in time as is practicable, whatever the language under consider-
4 Comparative Semantics

ation; nor must regional varieties of the standard languages be

neglected, for they can all bring valuable material.
Even when there is a total absence of historical records, the
comparative method carefully applied to the modern evidence
is still available to try to reconstruct the way things came to be
and by what process a particular object came to be seen in a
given way. This is the essential nature of comparative semantics,
which places the object or concept as the central point of linguistic
inquiry, a method of inquiry that inevitably turns out to be an
exploration of the probings and progress of human thought down
the ages, the most humanist science that could be conceived and
the most illuminating of the spirit of man reaching out towards
light and understanding. There are untold treasures buried from
time immemorial and waiting to be brought to light, and count-
less breathtaking discoveries to be made in this fascinating field
of the human mind.
1
Semantic Developments
from “At Once”

The semantic field associated with temporal expressions is a

particularly rich one, and in what follows there can, of course,
be no attempt to deal exhaustively with the subject. However,
it is hoped that the material to be presented will give a fair idea
of the many avenues open for exploration within this field, and
also stimulate some further research. In the light of remarks
made in the Introduction, it hardly needs to be pointed out again
here that a combining of European and Oriental evidence offers
many possibilities for interesting semantic studies. Consequently
the present chapter, like those that follow, will alternate between
these two sources of material.

Let us first consider Vietnamese. Here we find that moi, “‘recent-

ly”, is used normally to express the immediate past:
No moi di
He has just left. (lit. he recently leave)

However, from this meaning it develops the sense “‘not until’

(much like the German erst), and this secondary use enables it
to express the future. Thus the semantic range is from an im-
mediate past to an immediate future, and from this to an indefinite
future. Mal. baharu and Id. baru, “new’’, have developed pre-
cisely the same semantic range. Such hinge words are funda-
mental, not so much in their primary as in their derived sense.
We must not be misled by the fact that they have become what
may appear to be mere particles, devoid of substance; actually
they are tremendously important, and this is why I prefer to call
them hinge words, a term that ccrresponds more closely to their
real value.
6 Comparative Semantics

If we now turn to French, we find that certain time expressions

become much clearer in the light of what has just been observed.
Consider, for example, tout a l’heure, originally a l’heure, “at
this moment’’, i.e. ‘‘at once’’, later tout a I’heure, lit. right at
once, with an intensive tout. This expression gradually developed
into a less immediate future, and along parallel lines into the
past. It is now limited in time to the same day, and only the
tense of the verb shows whether reference is to the past or the
future. Hence, je l’ai vu tout a l'heure cannot refer to yesterday,
nor can je vous le dirai tout a l’heure to tomorrow. It will be
appreciated that from “‘at once” to “within a few hours (and not
just a few minutes) before or afterwards” is quite a substantial
semantic change.
Tantét, whose early sense was also “‘at once’, has had exactly
the same semantic history as tout a l’heure, and now refers to
past or future within the same day, though it points mostly
to the afternoon: it is more natural to say je le ferai tant6t in
the morning, and je /’ai fait tant6t in the evening.
It is interesting to note that on the other hand bientét, which
originally also meant “at once”, has developed along more
orthodox lines: from an immediate to an indefinite future may
almost be called a natural semantic change. The use of the word
“natural” in this connection may appear surprising, but it seems
justified in view of the frequency of this particular phenomenon,
which may be summed up in the formula “at once” > “soon’’.
Anyone who has been asked to wait un momentito in a Spanish
shop or un petit moment in France, or has been told “I won’t
be a moment” or “I won’t be a sec(ond)”, and has discovered
how long the wait actually turned out to be, will readily see that
the word “natural” is quite apt in this case. Je suis une seconde
dehors is common in spoken French, and here the use of the
present tense is significant as it helps to stress immediacy.
The semantic history of bientét is exactly parallel to that of
Eng. “soon” (OEng. séna, “‘at once”), and further examples of
the same change are not unusual, thus “presently” and “anon”.
Ger. bald also appears to have followed an identical course, and
there is no reason to doubt that these developments are all
independent, since there is additional evidence in the Romance
languages. Sp. Juego and Pt. logo (lit. on the spot, i.e. “‘at once”)
 At Once 7

are cognate with OFr. ues (Lat. /oco with a probable adverbial -s),
which also meant ‘“‘at once’’, whereas hasta luego and até logo
are equivalent to Fr. a bientét. In Italian, presto generally
corresponds to “‘soon”, while subito is the usual term for “at
once”’.
The frequency cf the change from “‘at once” to “‘soon’’, as
contrasted with the few occurrences of the corresponding change
from an immediate past to a more distant past, may well be
the underlying cause of an interesting semantic phenomenon
that can be found almost everywhere. “In the morning” in
the phrase “I'll see you in the morning” obviously means
“tomorrow morning”,-and by extension it readily acquires the
sense “tomorrow”. Hence the widespread change from
“morning” to “tomorrow”: for example, It. domani and Fr.
demain, ‘tomorrow’, are originally de mane, “in the morning”
(OFr. main is the normal word for “‘morning”); and similarly
Sp. manana, Pt. amanhé and Rum. maine, the usual words for
“tomorrow”, are all from Lat. mane, “in the morning’. In
Dutch and German this relation is even more evident, since
morgen means both “morning” and “tomorrow”; in English,
the fundamental identity of ‘““morning”’ and “‘tomorrow” is seen
in the corresponding Middle English terms morwe (or morwening)
and to morwe, and the Shakespearean “‘good morrow!” shows
precisely the same relation. In Scandinavian languages this
point is even clearer.
On the other hand, if we seek evidence of a corresponding
change from “evening” to “yesterday”, only a few sporadic
examples are found: in French /a veille, “the previous day’,
and in English “eve”, as referring to the day preceding a feast-day.
Why “‘eve” is thus used will be apparent by reference to the
Middle English terms in the last paragraph, which show that
“eve” is to “evening” exactly what “morrow” is to “morning”,
both in form and sense. Only in Slavonic and Balto-Slavic
languages do we find regularly the complete pattern that could
have been expected, as for instance Rus. zavtra, “tomorrow,
for earlier zautra, from utro, “morning”, and similarly vchera,
“yesterday”, from vecher, “evening”, and so elsewhere in
Slavonic.
These observations are at once confirmed in the light of the
8 Comparative Semantics

Far East. From the native Japanese asa, “‘morning’’, are derived
asu, “tomorrow”, and ashita, “morning” or “tomorrow”.
Sino-Jp. mydnichi, 948A “tomorrow”, means literally “the
bright day’, i.e. “the day of the dawn’’, and the Chinese either
use the same characters, pronounced mingri, or mingtian, 11K
lit. the bright sky. Moreover Jp. akuruhi, #4 “‘the next day’,
pronounced yokujitsu in Sino-Japanese, also means literally “‘the
day of brightness or dawn’’. By analogy, 44F Jp. myonen, Ch.
mingnian, “‘next year” and Jp. #4F akurutoshi or yokunen, “the
next (following) year” are further examples of the same semantic
development. Since the year originally began in March, it is
normal to find that the spring is looked upon as the dawn of the
year, eagerly awaited after the long night of winter.! The impor-
tant point is that, just as “morning” turns into ‘tomorrow’,
so does “‘the dawn of the year” turn into ‘“‘next year’’.
However, just as in the West, there are only a few traces here
of the opposite change from “‘evening”’ to “‘yesterday’’, and these
actually refer not to the day but to the year. It is common enough
in Japanese to speak of the end of the year as toshi no kure (cf.
kuroi, “black, dark’), lit. the evening of the year, especially
in the expression kyonen no kure ni, “‘at the end of last year”,
but it goes no further and there is no sign that it is at all likely
to develop into a Sino-Japanese compound—in this case a
hypothetical *bonen, “last year’”—which would make it exactly
parallel to myonen, “next year”. The presence of such phrases
as have just been quoted confirms, however, that myonen and
similar year-terms are formed on the analogy of the corresponding
day-terms, such as my6nichi, ““tomorrow’’.
It will be remembered that, while in Vietnamese and in
Indonesian the starting-point of the semantic change under
study was the immediate past, in French (tout a l’heure, tantot)
it was the immediate future, and it seems that in Europe this is
generally the case.
The temporal use of Du. pas is especially arresting. There is
little doubt that it developed from Fr. de ce pas, “‘at once’,
1. An interesting point to be observed is the common element in the Chinese
characters % Jp. higashi, “the east”, and # Jp. haru, “spring” (the season):
in each case the sun (f) is clearly visible. Ed.: comments made in the Intro-
duction, together with those above, can be better appreciated if the visible
link in Japanese between “the east” and “spring” is borne in mind.
 At Once 9

originally used normally with verbs of movement, as in j’y vais

de ce pas. \t is worth noting that the French negative pas was
also used in the first place with such verbs of movement and
meant “(not) a step’, until it became generalized as a particle
strengthening the negative ne, which it ultimately supplanted
in the colloquial language. From je ne puis to je (ne) peux pas
sums up this development. In Dutch, pas first shifted to an
immediate past “just now’’, as in pasgeboren, ‘‘new-born’’, and
from this has come the sense “not until’:
Daar kan ik morgen pas aan beginnen.
I can’t start on it until the morning [i.e. tomorrow].

When it means “just now”, pas is often strengthened in the

form zopas, in line with zojuist and zoéven, both of which are
also used in this way. This use of zo may seem rather paradoxical
when it is realized that zo by itself frequently means ‘‘at once”’,
as will be seen later. However, it also reflects the close semantic
relation between these two meanings, which are exact opposites.
There is another Dutch word that throws further light on
the early history of the temporal pas. In English, “straight”? was
originally the past participle of the verb that has become the
modern “‘stretch”; hence its early sense was “‘stretched, tense”.
The Dutch cognate strak is widely used as an adjective or adverb:
Hij keek haar strak aan.
He looked at her intently.

With the adverbial -s, i.e. in the form straks, it becomes a temporal
adverb meaning “soon” (< “at once’’), and cognates are found
in Dan. straks, Sw. and Ic. strax, all of which have the same
sense, ‘‘at once’. There is nothing unexpected in this semantic
development, since it is precisely what led to the temporal use
of Eng. “straight” and “straight (a)way’. There are similar
words for “at once” in many languages, and the Dutch terms
y
direkt, direct and regelrecht, used in the same sense, are obviousl
means “at
comparable in form with Eng. “directly”, which also
it is
once”. The curious point about Du. straks, and this is where
, is that,
strikingly parallel to the pas we have just examined
te
while all the other terms quoted point clearly to the immedia
develops into
future “at once”, straks in addition occasionally
an immediate past:
 10 Comparative Semantics

Hij was straks hier.

He was here just now.

There is further confirmation of this trend in German. Let us

consider another series of the type “‘shortly’’, i.e. based on a
word meaning “short”: Pt. im breve, It. in breve, fra breve,
OF r. briefment, Middle French de bref, de brief, regional French
(North) tout court, ‘‘at once”, Ger. in Kiirze, in kurzer Zeit, Lat.
brevi (tempore), or again Rus. korotko, which corresponds
exactly to “shortly” in both form and sense. It is noticeable
that these all refer to the future, immediate or indefinite, unless
some specific precision is added, such as “shortly before” (Ger.
vor kurzem). Similarly, Ger. kiirzlich also referred to the future,
as for example kurzliche in Middle High German, but this usage
appears to be obsolete and has now been largely supplanted by
an immediate past, kiirzlich, “‘recently’’.
In every case the change is from an immediate future to an
immediate past; a further instance of this occurs in Rumanian
where curand, which usually means “at once, soon’, can be
used in the sense of “‘recently’’.

If we now consider ‘“‘at once” in itself, rather than in its relation

to an immediate past “‘just now” or an indefinite future “soon”,
we find everywhere a great variety of words and expressions
with this meaning. Some of these are transparent enough, like
“immediately” (‘without intervening time”), Ger. unmittelbar
or Du. onmiddellijk, and parallel forms occur elsewhere; the
words based on “‘straight” or “short” that have been mentioned
earlier also belong to this category. However, there are others
that are much less obvious, like Fr. de ce pas, Eng. “at once’,
or Lat. statim, and finally there are yet others that are so
insubstantial as to be almost beyond the reach of normal etymo-
logical explanation. Let us examine one of these and attempt
to shed some light on its real nature.
There is in Old French a remarkable construction which may
be illustrated by the following examples:
Et quant il le voient, si li demandent dont il est.2
And when they see him, they ask him where he is from.

2. Anon., La Queste del Saint Graal (Paris, 1949), p.

87.
 At Once 11

Endementres que Perceval menoit son duel en tei maniere, si escoute

Ota
While P. was grieving in this way, he listened and heard ...

When a sentence begins with a temporal clause in Old French,

the main clause is usually introduced by si (Lat. sic, “‘so”’). And
similarly if the sentence starts with an “if” clause:
S’il te voient entr’ex, si defenderont il mix lor cors .. .*
If they see you amongst them, they will defend themselves better...

This si seems to be unimportant, and is almost invariably left

untranslated; but if some rendering must be found, the most
satisfactory one is probably “‘then’’. A particle of this type could
conveniently be called a “then-word”: others are Fr. alors,
dans ce cas, Ger. dann, in diesem Falle etc. If conditional and
temporal clauses have this effect on the following principal
clauses, it is because ‘if’ and “when” are semantically very
close. Since Ger. wenn is cognate with Eng. “when”, it is not
surprising that “‘if’’ and “when” overlap in a given context.
In German, the same constructions are usual under the same
conditions, only here the ‘“‘then-word” is so:
Wenn du Zeit hast, so schreibe mir.
If [when] you have time, write to me.

Similarly, when the initial condition is expressed by inversion:

Kommt aber wieder eine Versuchung ... so wende dein Haupt gen
Sonnenaufgang ...°
But if temptation comes again ... turn your head towards the East

The following temporal example is worthy of close observation:

Kaum war er fort, so kam sie zuriick.
Scarcely had he left when she returned.

Consider in English: “As soon as [when] we appeared, the enemy

fled.” Especially with ‘as soon as”’, but also to some extent with
“when”, “at once” is implicit. In the last German quotation it
is clear that, partly because of kaum, but only partly, the particle

3. Ibid., p. 89.
4. Anon., Aucassin et Nicolette (Paris, 1965), II, ll. 20— Dae
5. J.V. von Scheffel, Ekkehard (Boston, 1893), p. 17.
12 Comparative Semantics

so comes very close to acquiring the meaning ‘‘at once’. This

is possibly how so came to be used in sogleich and sofort, the
most frequent terms for “at once” in Modern German. To
appreciate the precise value of so here, it should be remembered
that gleich preceded sogleich in this sense, and that gleich is
still the normal word for ‘‘at once”’:
MOglich, dass wir gleich wieder umkehren!°
It may be that we’ll go back at once!

Any word or expression meaning “‘at once”’ tends to be stressed:

this is the true role of tout in tout a l’heure, as we have seen, or
“right” in “right at once’’; tout in tout de suite is of the same type,
and thus, despite the Académie, de suite is well able to express
this meaning, like Sp. enseguida and Cat. de seguida. This
tendency is especially striking in the following:
Wollet Ihr mich verlassen, so gehet auch gleich und ganz!’
If you wish to leave me, [then] go at once!

There is nothing in English that corresponds to this alliterative

phrase, but comparison with ganz und gar, “totally”, is in-
structive.
The most illuminating parallel with this use of Ger. so is to
be found in Chinese. Here the range of what is probably the most
widely used particle, #t jiu, corresponds remarkably well to
that of so, and it is normally used after the equivalent of our
conditional and temporal clauses. But there is one important
difference: the Chinese frequently leave out the introductory
conjunctions, and we are left with what appear to be co-ordinate
clauses connected by jiu, which thus becomes essentially a
hinge word of tremendous importance. Chinese grammarians
may call such words empty, but they are in effect very full and
should rather be looked upon as key words.
Ch. jiu belongs fundamentally to the “then” category, and
it occurs in sentences ofue type observed in Old French and
German:
PRBe PS HL ni neng zuo jiu zuo
If [when] you can do it, do it. (lit. you can do then do)

6. Ibid., p. 10.
7. Ibid., p. 120.
 At Once 13

A most striking aspect of this temporal use is that it gradually

acquired the meaning “‘at once’”—a much fuller meaning than
would be expected. One example will suffice to illustrate this:
fiBE 3 ta jiu lai
He will come at once.
It should be noted that the first clause is absent and jiu follows
the subject, its normal position when the subject is present.
Moreover jiu is also used after causal clauses. To express
“As (since) the weather was fine I went to the park” in Chinese,
the causal conjunction is omitted and the second clause will
run: wo jiu wang gong yuan. In German too, da, “‘as’’, is usually
followed by so, which introduces the principal clause. Here
again there is a close semantic relation between causal and
temporal conjunctions: “since” has both senses; Ger. weil,
“because”’, is cognate with Eng. “while”, and in French puisque,
“since” (causal), and depuis, “‘since’’ (temporal), both contain
puis, “after, then’’.
Finally, jiu is used after a great variety of time expressions,
ranging from an immediate to an indefinite future; the most
widely used of these is certainly AA (bu jiu jiu), lit. not long
then, equivalent to “‘it was not long before’, Fr. il ne tarda pas a,
or simply to “before long’. In a Chinese translation of an
English text, this with an occasional variant will be the most
frequent rendering of such words as “presently’’, “soon”,
“before long”, “shortly” etc., but even here jiu will almost
invariably follow the time expression. And once again German
is remarkably close: while other European languages have a
number of syntactical links to introduce the principal clause,
in German it runs: es dauerte nicht lange, so ... This use of so
after a non-subordinate clause should be noted, and moreover
it is fascinating to observe that Ch. jiu, besides its particular
meaning “‘at once’, reaches well into the future, a point that
could have brought it sooner into our discussion.
While it is clear that Ch. jiu and Ger. so are used in similar
contexts, there is a striking difference in those cases where jiu
means “‘at once”. It can of course have this sense after a clause
introduced by the equivalent of our “as soon as .. .””, just as it is
implied in German after kaum, but in addition it has developed
14 Comparative Semantics

in Chinese into a completely independent adverb, without the

expected temporal clause or expression. It is curious to note that
in Dutch too zo, obviously the cognate of Ger. so, has come to
be used independently for “at once’, although in Modern
Dutch dan (‘‘then’’) is the particle corresponding most closely
to Ger. so in the contexts examined above. Zo, ‘“‘at once’’, is
especially frequent in the colloquial: /k ben zo terug (lit. I am
so back), “I'll be right back”, is heard constantly. On leaving
somebody for a very short time, one would say tot zo or tot
straks, which are comparable in form to tot morgen (Fr. a
demain), “‘see you tomorrow’, and there are numerous leave-
taking expressions of this type.
When zo is used in this sense, it usually bears a special stress,
often written z6:
Ik ben zo bij je.
I'll be with you straightaway.

This stress is expressed in a different way in the following example,

where zo and meteen are synonymous:
Ik ga zo meteen terug.
I am going back at once.

Du. dan seems to be used rather more freely than Ger. so, but on
the whole it corresponds closely to it:
Als je even wacht, dan ga ik met je mee.
If you wait a moment, [then] I’ll go with you.

Further, since an imperative can contain a virtual condition,

the type below is a natural extension:
Wacht even, dan zal ik je helpen.
Wait a moment and [then] I'll help you.

The meaning is clearly: “If you wait I’ll help you.” The same
holds in the next example:
Kom maar kijken, dan weet je het meteen.
Just come [and] look and you’ll know at once.

Compare further this fragment of conversation:

 At Once 15

“Hoe laat is het?’ “What time is it?”

“Over twaalven.”’ “Past twelve.”’
““O, dan moet ik zo weg.” “Oh, then I must [go] away at once.”

These last two quotations illustrate how “then” and “‘at once”
can live distinctly side by side, and they show too that zo and
meteen exist independently; yet, as we have seen, they can unite
into zo meteen, which in turn sheds some light on the probable
history of Ger. sogleich. A final example will show how Dutch
compares with Chinese:
Ogenblikje, ik ben zo terug.
Just a moment, I’ll be right back.

It may be felt that ogenblikje (diminutive of ogenblik, “‘moment’’,

lit. glance, Ger. Augenblick) implies “‘wait a moment” and so we
may expect dan. But Dutch allows much freedom in this respect:
a clause beginning with als, “‘if’’, need not be followed by a dan
clause. On the other hand, if we attempt to express this Dutch
sentence in Chinese, we first have a choice of words or phrases
ranging from —[ yi shi (lit. one time) to others where the twin-
kling of the eye is clearly visible, e.g. % shun, whose radical on
the left is the eye 8 .® But next to this the Chinese will almost
certainly say BEEN wo jiu hui lai (lit. 1 then turn come) for
“Tl be right back’’, with the inevitable jiu.
In addition to the above words and expressions used for “‘at
once”’, there is a particular type that could easily be overlooked.
In Caesar’s famous dispatch to the Senate after his swift victory
over Pharnaces, Veni vidi vici, such a term is all the more sig-
nificant for its absence. Here vidi is almost winged beside the
pedestrian primo statim conspectu, where “at once” is explicit
in statim but already implicit in primo.
A curious type occurs in a translation into French by C.-M.
Garnier of:
And thither will I straight to visit him:°
Et je vais de ce pas de suite a sa rencontre:

8. Japanese, too, possesses a number of such picturesque expressions based

on the eye: e.g. isshun —WB¥; hitomatataki ni —\§® \c; matataku ma ni
Be< RIC ; mirumani HARI.
9. W. Shakespeare, Julius Caesar, III, ii, in The Complete Works of William
Shakespeare (London, n.d.)
16 Comparative Semantics

At first sight it seems that “straight” is rendered by the cumber-

some de ce pas de suite, but when the line is spoken with due
respect to the caesura, it becomes quite acceptable.
This last example is related to the well-known type, already
referred to, characterized by the addition of an intensive. Thus
in Middle English, sone (“‘soon’’) is frequently found in such
phrases as ful sone, eftsone, wel sone or right sone, and one also
meets streght anon or right anon. Comparable too are Fr. tout
aussitot, tout a ’heure and tout de suite, but in the latter tout
is still largely independent, while the common use of a [’heure,
“on time”, shows that tout a l’heure is a far more closely knit
phrase. Similarly, while beaucoup and bientét are parallel in
form (in so far as they both contain an integrated intensive),
bient6t is a much freer combination than beaucoup, and on
reflection the reason will turn out to be essentially a semantic
one. For another reason, Eng. “‘alone”’ is felt to be one word,
whereas Du. alleen and Ger. allein are rather two: they are all
formed in the same way, with the intensive ‘all’ + ‘“one’’,
but while both parts are clearly recognizable in the last two, in
“alone” they have deviated phonetically and have long ceased
to be perceptible. This is why “all alone” is not felt to be tau-
tological in English.

In exploring the semantic field of ‘‘at once’, it is essential to

keep in mind a number of related notions that could be re-
presented by “without a break”, “off hand”’, “‘suddenly’’, “‘in a
flash”, “now... now” (i.e. the correlatives, Fr. tantét... tant6t;
Ger. bald... bald), “‘at one stroke’’, ‘“‘at one fell swoop”’ etc.
Observe too the French tendency to use tout d’un coup as an
equivalent of tout a coup. “At the same time’’, in so far as this
is not adversative, and “at the one time” are also relevant. This
is the sense of “‘at once”, or rather its earlier form at ones:

I saugh nat this yeer so myrie a compaignye

Atones in this herberwe as is now.}°
It is immediately clear that “suddenly” is very close to “at
once’’. English gives the best proof of this: the difference between
“at once” and “all at once” is illuminating. In fact, most of
10. G. Chaucer, The Canterbury Tales, General Prologue, \\. 764—5, in The Works
of Geoffrey Chaucer, ed. F.N. Robinson (London, 2nd edn, 1957).
 At Once 17

the terms nearest in form to ‘“‘at once’”’ mean “‘suddenly’’: Ger.

auf einmal, Du. opeens, ineens, eensklaps (where eens corresponds
exactly to ones, later “‘once’’), and Rum. deodata, based on
odata, “‘once”. Rum. data is almost certainly a calque on Rus.
raz, “time” (originally ‘“‘stroke’”’), hence Rus. razom and Rum.
indata, which both mean “‘at once’’. This is the ‘‘stroke’’ that
is found in other “‘suddenly”’ expressions, like Fr. tout a coup,
Sp. de golpe and It. di colpo. It is also arresting to compare It.
subito, “‘at once”, and Fr. subitement, “suddenly”.
There are of course a number of other types meaning “‘sud-
denly”: It. improvvisamente, Sp.. de improviso stress the un-
expected (lit. unforeseen); Eng. ‘“‘suddenly”, Fr. soudain, are
derived from the original Latin word represented by It. subito
(Renaissance French soubdain gives a hint of this relation);
Ger. plotzlich, Du. plotseling (cf. Ger. Blitz) beside Du. eensklaps,
are all based on onomatopoeia, and there are yet others. This is
a field where languages are particularly exuberant, and the
ubiquitous intensive is here very much in evidence: “all of a
sudden” beside “‘suddenly”’ etc.
One final example will suffice. It. (tutto) ad un tratto, “‘sud-
denly”, is identical in form with Fr. (tout) d’un trait, “at one
stretch”, ‘‘at one gulp” (speaking, writing, drinking) and (cout)
d’une traite, “without a break or a pause”’ (especially travelling),
which meaning is in turn occasionally expressed in English by
a corresponding temporal use of “‘straight”’: “I did two hours
[work] straight.” This use of trait, traite is clearly a survival of
the time when OF. traire was a verb of motion, later supplanted
by tirer, and it is remarkably close to some of the fundamental
senses of Ger. Zug and ziehen.
Reference has been made to Du. straks, “‘at once’, and its
English and Scandinavian cognates. In fact “straight” and
“straight-way”, both common in this sense in Shakespeare
(apart from the rarer phrase streght the wey in Chaucer'’),
besides the usual “straightaway”, “straight off”’ and other words
of Romance origin such as “directly”, are constantly used to
express this meaning. In French, (tout) droit and directement
are equally frequent, and there are corresponding terms in other
Romance languages which it is unnecessary to enumerate. In
11. Chaucer, Troilus and Criseyde, Book V, 1. 292, in Works, ed. Robinson.
18 Comparative Semantics

effect, ‘“‘straight” and droit acquired the sense “‘at once” by a

series of semantic steps: whether we say “‘he went straight to the
station” or il est allé tout droit a la gare, no one thinks of a
straight line; the sense is “the shortest way’’, 1.e. “the quickest
way’’, and it is at this point that the expression becomes temporal,
“at once’. This is a further illustration of the widespread passage
from space to time.
There thus appear to be two distinct types: Germanic
“straight” and Romance “direct”. But they are linked in a
remarkable way: just as “straight” was originally a participle,
“stretched”, so Lat. di-rectus was the participle of rego, ‘draw
out”, and Eng. “right”, Ger. recht, are the corresponding
cognates. The earlier physical sense of “right” and recht,
‘straight’, shows this relation even more clearly.
One of the most widespread semantic developments from
this basic “‘straight’”’ is to the figurative “‘right, honest”. Just
as Gk edb¢ and Rus. prjamoj both went from “straight” to
“at once” and “right”, so “right”? and recht acquired a meta-
phorical meaning which finally became dominant, and the
result of this semantic overload was the need for a new word
for the physical “straight”: Eng. “‘straight’’ and Ger. gerade.
“Straight” has itself developed the same figurative sense: it is
common enough to speak of a man as being “‘fair and straight”.
Besides this derived meaning, “right” has survived as an
intensive: “right honourable”, “‘right here”, ‘‘right against”,
“right in the middle” etc., and the alternative “right away” side by
side with “straightaway” is especially relevant here; to this
must also be added “‘he’ll be right back” (or “right along”’). There
are comparable uses of Du. and Ger. recht: recht bose, ‘‘very
angry”’, recht zufrieden, “‘quite satisfied’, which appear to have
prompted a parallel use of OFr. droit:
les flors des margerites . . . estoient droites noires avers ses piés et ses
ganbesieee
The flowers of the daisies ... were really black compared with her
feet and legs ...

The Icelandic cognate rett is also usual in this sense: rett vid,
“very near”, and there is the even more arresting rett strax,
“right at once’’.
12. Anon., Aucassin, XII, Il. 25-8.
 At Once 19

However, the closest parallels to all this are to be found in

the Far East. In Chinese, A zhi, ‘“‘straight’”, developed the
metaphorical senses “honest, right’’; the Japanese use the same
character for both meanings, but in addition, when read sugu
it means “at once”, and when read naosu the sense is “‘straighten’’,
hence “correct, mend”. Sugu, precisely like its European equiva-
lents, attracts intensives: massugu, ima sugu. Moreover just as
we say “‘straight after’, “immediately before’’, “‘right in front”
(OFr. droit avant) etc., sugu is regularly found in expressions of
this type: sugu ushiro ni, “right behind’’, sugu chikaku ni, “right
near”’ (i.e. “close by’’), sugu mae ni, ‘‘just before”’, as is the Sino-
Japanese reading of the same character, CHOKU: chokugo,
“just after”, chokuzen, “‘just before”, chokka, “‘right under’’ etc.
In some of these, sugu (CHOKU) itself becomes an intensive,
just like Eng. “‘right’’.
If we now consider the character #H Ch. qu, Jp. magaru, “‘to
be bent, crooked, dishonest, wrong’, and compare this with 4
(which has just been examined and which shows the change from
“straight” to “‘right’’), we are confronted with an opposition
“straight/bent”’ side by side with “right/wrong’’. In Chinese and
Japanese the physical and figurative meanings still exist together,
whereas in English the original physical sense has long ceased
to be perceptible: several centuries have elapsed since a nose
could be described as right or wrong. However, Eng. “wrong”
is seen to contain a clear reminder of its earlier value if it is
associated with its true family: ‘“‘wring’’. It can then be ap-
preciated that ‘“‘wrong”’ literally means “‘twisted”’.
This in turn helps us to see the corresponding French terms
in their true perspective: the present opposition raison/tort was
in OFr. droit/tort:
Paien ont fort e chrestien ont dreit.'*
Pagans are wrong [in the wrong] and Christians are right.

Like Eng. ‘wrong’, Fr. tort (from Lat. tortus, the passive par-
ticiple of torqueo, “‘twist’’) also means “twisted”. The relation
between English and French is here so close that “wrong” may
well be a calque on tort.

13. Anon,. La Chanson de Roland, ed. T.A. Jenkins (Boston, 1924), 1. 1015.
20 Comparative Semantics

One of the most widespread sources of terms for “‘at once” is

of the type ‘on the spot”, i.e. based on a word meaning “‘place”.
Equivalents are to be found practically everywhere: Fr. sur-le-
champ, Ger. auf der Stelle and Lat. ilico, which has found its
way into familiar Fr. illico, are clear examples. However, Lat.
extemplo, ‘‘at once”, is not nearly so obvious. That it belongs
here will readily be admitted when the ultimate identity of tem-
plum and tempus is recognized, tempus being originally a space
of time. This relation becomes especially clear when it is observed
that in Chinese and Japanese = templum is also the basis of iy
tempus: in the latter character the radical H “sun” gives a
semantic hint of the change from space to time. These two Chinese
characters are still the ones most commonly used to express these
notions in both China and Japan, thus showing that the templum/
tempus relation is still very much alive in the Far East today.
With Lat. ilico(i.e. in loco) may be compared OFtr. lues:
S’il est malades /ués revient en santé.!4
If he is ill he at once returns to health.

Parallel too are Pt. Jogo and Sp. Juego, whose significant change
from “at once” to “soon” was mentioned earlier in this chapter.
As a result, “at once” is usually expressed in Spanish by en-
seguida (cf. Cat. de seguida), comparable in turn with Fr. (tout)
de suite. Lat. locus is also widely represented in Modern Provengal
leu, “‘at once, soon, quick”’.!>
Expressions of this kind are also quite usual in the Far East.
Jp. sonoba (de), ‘‘at once’, is literally ‘‘at (de) that (sono) place
(ba); Ch. %¥% dangchang corresponds exactly to sonoba. Even
Korean kot, ‘‘place’’, is used for “‘at once, soon”’.
Particularly worthy of note is Jp. imakoko de (ima “‘now’’,
koko “‘here’’), ‘‘at once”, formed on precisely the same pattern
as Eng. “here and now’’, “there and then’’, which are, after all,
further equivalents of ‘“‘at once’, closely related to the “‘place”’
type we have been examining.
14. Anon., Huon de Bordeaux (Bruxelles, 1960), 1. 3254.
15. It tends to be repeated, /eu-/eu, when it means “quick” and is thus equivalent
in form to Prov. plan-plan and to It. pian piano which have the opposite
meaning. It is curious to note that in regional Belgian French, piyane-piyane
or pyame-pyame is quite frequent for “taking it easy, going slow’, and it
only occurs in this doubled form. Piyane does not exist as such.
 At Once a

Jp. tachidokoro, lit. on the spot (tokoro), introduces the notion

of standing (tachi). To this corresponds exactly Ch. 17#8 Ji di,
“at once” (lit. stand place), and a number of synonymous terms
like 17 ¥ /i shi (lit. stand time). It is clear that we have here the
true explanation of Lat. statim and Du. op staande voet, “‘at
once” (lit. as you stand). The basic sense “‘at once’’ is equally
well expressed by “‘as you sit” etc., that is, without altering your
stance, whether you happen to stand, sit or ride. Parallel to Fr.
séance tenante, we find Japanese terms based on sitting, all
meaning “at once”: sokuseki ni, sekijo de, both used especially
of things improvised “‘as you sit”, and sokuza ni, which is curious-
ly defined in a purely Japanese dictionary as meaning sugu sonoba
de, lit. straight on the spot, a particularly insistent way of saying
“at once’’, which may well be confined to dictionaries.
The Chinese frequently use & mashang, lit. on horseback,
for “‘at once’’. Needless to say, this horse has long since vanished,
but its visible presence in the character was no curb to semantic
change.'!° There have been comparable developments in the
West, right from the Medieval Roman de Renart where the
various animals, when running, are usually described as riding
a charger and holding the reins. Rabelais too, referring to Gar-
gantua, Says:
Puis dormoit sans desbrider jusques au lendemain huict heures.'”
Then he slept without a break [lit. without unbridling his horse]
till the next day at eight.

The original horses have become conspicuously absent in Mod.

Fr. desargonner, (ne pas) dételer, piquer des deux, a variant of
which occurs in Gide:
Aussit6t nous piquions de l’avant, au galop ...'8
At once we dashed forward ...

16. This also applies to the horse in Jp. eki #2 (now abbreviated to §R ), “rail-
way station”, while in Chinese this same character, pronounced yi, still
retains its original sense, “a post-house”—-where fresh horses could be
found for the next stage of a journey.
17. F. Rabelais, Gargantua, chapter XXII. in Oeuvres completes (Paris, 1962).
18. A. Gide, Si le grain ne meurt (Paris, 1928).
22 Comparative Semantics

There is a well-known expression for improvising in German

that is based on the same image:
So tibersetzte ich ihr aus dem Stegreife solche Homerische Stellen
19

So I improvised for her translations of those passages from Homer...

“As you ride” means “‘without breaking your journey, without

stopping, immediately”. The precise sense of Fr. de ce pas and
its exact equivalent in Japanese, sono ashi de (ashi, “‘foot; foot-
step’’), should now be clear. In fact, de ce pas is the successor of
OFr. eneslepas, “at once” (enes = in ipso, and /e is still strongly
demonstrative). Closely related too is OFr. erranment, lit. as
you go, based on OFr. errer, earlier edrer, “go, travel’’ (quite
different from the modern errer, “‘wander’’). Consider the fol-
lowing:
li rois ... lour commande qu’il viegnent erramment a Camaloth sans
nul autre delay.?°
The king . . . orders them to come at once to C. without further delay.
Et il estaint les cierges erranment .. .?}
And he puts out the tapers at once ...
Dis et uit hanas biut de vin tot erranment.?2
Eighteen tankards of wine he drank at one go [rather than “‘at one
gulp”).
These three examples show the gradual semantic development
of this term, first with verbs of motion, like de ce pas, then in a
wider temporal sense, and finally “‘at a stretch, without a break”,
even though there are inevitably some short breaks. This is
probably the true meaning of boyre a tyre larigot,23 whose tirer
has already been considered in its relation to boire d’un trait.
It seems relevant here to compare Fr. sans cesse, paradoxically
used only when there is “cease”’; i.e. the meaning is “continual”
rather than “continuous”. Similarly, it is likely that a man who
had travelled a great distance in one day would have asserted

19. J.W. von Goethe, Dichtung und Wahrheit (Berlin, n.d.), Part II, p. 97.
20. Anon., Le Roman de Balain (Manchester, 1942), p. 4.
21. Anon., Queste, p. 170.
22. Anon., Le Poéme moral (Bruxelles, 1929), 35, 1. 139.
23. Rabelais, Gargantua, chapter VII.
 At Once 23

that he did it d’une traite, ‘‘at one stretch’’, despite the numerous
halts to change horses.
Fundamentally akin to de ce pas are expressions based on
“away”, which are especially well represented in English:
“straightaway’”’ (for earlier “‘straight’’) and “right away” are the
most usual expressions for ‘“‘at once”. The synonymous “forth”
is present in “forthwith”, far more peremptory and now ex-
clusively literary. On the corresponding German cognate is
formed Ger. sofort, ‘“‘at once’’, with the same so as in sogleich,
but forming a much closer unit since, unlike gleich, fort is not
used by itself in this sense. Expressions with “off” such as
“straight off’’ are also of the same type. So is “off you go”, not
in its literal meaning, but as used in prompting somebody to
speak out when time is short. This is used only in the spoken
language, and within very restricted circumstances, and so may
hardly seem worthy of standing beside the other terms men-
tioned. Yet there is in Japanese a remarkably similar usage: the
expression }###1 CG tochii de, ‘‘on the way” (to a place), is applied,
in addition to its physical sense, to someone engaged in speaking.
Of course, ‘“‘off you go” represents only the inceptive aspect of
the continuous action expressed by tochi de. This Japanese
phrase is outwardly comparable to, and yet very different from,
“in the course of’, au cours de or en train de, for these are far
more abstract.
Fr. au pied levé certainly belongs here, and La Fontaine must
have been clearly conscious of this, judging from the opening
lines of the fable “La Mort et le mourant”’:
La Mort ne surprend point le sage;
Il est toujours prét a partir,

and further on in the same fable:

Un mourant, qui comptait plus de cent ans de vie,

Se plaignait ala Mort que précipitamment
Elle le contraignait de partir tout a l’heure, [“‘at once}
Sans qu’il ett fait son testament,
Sans l’avertir au moins: “‘Est-il juste qu’on meure
Au pied levé? dit-il: attendez quelque peu:

The humour is that this unwise man is more than ready; in fact
he is “off”, but to go elsewhere. Here au pied levé means “at a
24 Comparative Semantics

moment’s notice, without warning’’, and from this it has also

developed the more active sense of improvising. The literal
meaning is probably ‘‘at the moment when the foot is raised to
the stirrup in order to get on horseback’’, and this is reminis-
cent of Ger. aus dem Stegreife, discussed earlier.

In contrast to the English expressions based on “‘away”’, there is

a small group of comparable phrases for “‘at once”’ in Renaissance
and seventeenth century French whose common link is the op-
posite notion of “arrival’’. For example d’abord, which in Modern
French is normally used in correlation with ensuite (or puis) and
enfin, “first .. . next (or then)... . finally’, is found in La Fontaine
with or without the intensive tout:
Si quelqu’une de vous touche 4 la quatriéme,
Je Pétranglerai tout d’abord. [“‘at once”’]?4

The corresponding phrase de prime abordée, “at the first on-

slaught”’, in Rabelais?> is still close to its literal sense. So is de
pleine arrivée, ‘“‘as soon as he arrived’’,2° and further on in the
same chapter we find the following:
De premiere venue accoururent droict 4 moy .. . et me eussent
devoré 4 l’heure si mon bon ange ne m’eust bien inspiré .. .
Here “at once” is implied in de premiere venue and droict, but is
explicit in @ ’heure, the earlier form of tout a ’heure, already
referred to.
On the other hand, in Montaigne, d’arrivee27 and de belle
arrivee*® have acquired the purely temporal meaning ‘‘at once’’.
It will be observed here again that all these expressions tend to
be stressed by the addition of an intensive. The most common
is tout, as we have seen, but with a noun asuitable adjective plays
the same role: prime, premier, plein, beau, as used in the preceding
phrases, are listed here in increasing order of abstraction, and of °
frequency as intensives in Modern French. The precise relation
between aborder and arriver lies in the original sense of “reaching

24. J. de la Fontaine, Fables (Paris, 1962), Book I, vi.

25. Rabelais, Gargantua, chapter XXXIII.
26. F. Rabelais, Pantagruel, chapter XIV, in Oeuvres completes (Paris, 1962).
27. M. de Montaigne, Essais (Paris, 1962), vol. i, Book I, chapter XX.
28. Ibid., chapter XXVI.
 At Once 25

the shore” (bord or rive): only boats could arriver or aborder in

the first place, and all the examples quoted show later extensions
of this meaning.

There are many other types of terms for ‘“‘at once’’, most of them
being reasonably transparent. But consider the less obvious Du.
dadelijk, which has come to mean “‘at once”’. It is based on daad,
‘““deed’’, and a comparison with Fr. actuel, originally “‘reai’’, now
“present” (cf. Eng. “actual”, which has retained the earlier
sense), suggests that dadelijk has followed the semantic change
of Fr. actuel and may well be a calque on it.
Further, the history of Fr. maintenant can be found simply
by setting the famous proverb Un “‘tiens’’ vaut mieux que deux
‘ty l’auras’’ beside the English equivalent “A bird in [the] hand
...” This is very relevant here, as in Old French maintenant
meant “‘at once” and so tended to be stressed as tot maintenant.
Comparable is the use of “hand” in Eng. “at hand” and Middle
High German zehant, “at once”, whose Dutch cognate thans
(with adverbial -s) is still usual for “now”. There are other ex-
pressions of the same type, such as “off-hand”, “‘out of hand”’.
A few equivalents of ‘‘no sooner said than done” with its
latent ‘“‘at once” well illustrate varying degrees of economy of
expression: nearest in form to Fr. aussitét dit, aussitot fait is
Du. zo gezegt, zo gedaan. The most compact are Ger. gesagt, —
getan, It. detto fatto, and Sp. dicho y hecho. All these expressions
seem to be largely fixed and hardly capable of variation. Aussitot,
which normally points to the immediate future, acquires in such
a context the sense ‘cone moment(before) ... and the next mo-
ment...”, and this helps us to understand how ‘“‘at once” can
so easily come to mean “just now”, as we saw previously.
Classical Chinese possesses precisely the same kind of cor-
relative pattern to express “no sooner . nas
wee lidalisi
At once beaten, at once died. (lit. stand beat stand die)

Li, “stand”, is seen here to exist independently and corresponds

exactly to Lat. statim. And similarly :
BESBE4 sui de sui shi
Lost as soon as gained. (lit. follow gain follow lose)
 26 Comparative Semantics

Sui, “follow”, has here the form and sense of Sp. enseguida and
Fr. (tout) de suite.
Correlation is only one of several possibilities to express this
meaning. The following Dutch is strikingly condensed:
Denken en doen was een bij hem ...
Thinking and doing was one [thing] to him ...

The use of the singular ‘“‘was” stresses that both thinking and
doing practically become one single action.
In the last example there appears to be no word for “‘at once”,
but on closer inspection this sense is seen to be contained in een,
“one”. The following, also from Dutch, brings this out still
more clearly:
Eenmaal in haar suite, wierp ze onmiddellijk haar bontmantel op
een stoel.
Once in her suite, she immediately threw her fur coat on a chair.

Here onmiddellijk is added, but even without it eenmaal is suf-

ficient to convey the full meaning, as ‘“‘at once” is implied in it.
The same observation may be made about parallel phrases such
as It. una volta a casa, Sp. una vez en casa, Fr. une fois rentré,
Eng. “once home”, and the comparable It. appena seduto, Sp.
apenas sentado, Fr. a peine assis:
Cela fait deux fois que Mme F.O. d peine partie, le ““Fléau’’ arrive.
Ce soir encore. un peu plus, elles se trouvaient ensemble.2°

There is a similar sense of urgency in the bare use of the past

participle (i.e. without ‘“‘once” or “hardly”) in It. giunto nella
piazza, Sp. llegado a la plaza, Fr. arrivé a la place. The choice
of the last example is no accident, for this construction is especial-
ly frequent with arrivé:
Fr. Arrivé la, on a appris que ...
Ger. In der Stadt angelangt, beschaftigte ich mich in den friihesten
Stunden—?°
No sooner had I arrived in town than I occupied myself in
the early hours—

29. P. Léautaud, Journal littéraire (Paris, LOG) avoir oans

30. Goethe, Dichtung, Part II], p. 13.
 At Once 27

Another construction, typical of Renaissance French, also ap-

pears to imply “‘at once” and it is significant that Gougenheim?!
refers to it as Le tour arrivé que fut, though he gives no example
with arrivé. However, he is quite right in describing it in this
way, for all the examples I have found (descendu, arrivé, entré,
achevé) belong to the semantic field of arriver. This turn of phrase
also exists in Modern Italian: vestita che fui?* (des que je fus
habillée; sitot vétue)—“‘as soon as I was dressed”’.

Inevitably, while considering “‘at once”, we are led to examine

‘“‘as soon as” and related expressions. Fr. aussit6t, “‘at once’,
and aussit6t que, ‘“‘as soon as”, need no comment. In Icelandic,
pegar, ‘“‘when, as soon as”, is also used for ‘“‘at once’, though
in this case it is usually accompanied by some additional rein-
forcing adverb, nu pegar or pegar i stad (nu, “now”; i stad, “in
stead’, ‘‘on the spot’). Eng. ‘directly’, besides its normal
adverbial use, is also found familiarly in the sense of “‘as soon as”’.
Not unlike the use of ‘“‘once’”’ and une fois examined earlier,
there is in Chinese a strikingly simple and direct way to express
“tas soon as’”’: the numeral — yi, ‘‘one’’, before the verb is suffi-
cient, normally followed by jiu in the principal clause. Thus:
{th— [alaKOrg Rh «ota yi hei lai ni jiw gao su ta
As soon as he returns, you tell him [at once]. (lit. he one turn come
you then tell him)
This construction can be compared with the well-known Chinese
colloquialisms 4 —€@ kan yi kan, “have a look”’, [i] fi] wen yi
wen, “have a sniff”, which mostly concern the senses, just as in
English we find a parallel series: “have a look’”’, “‘have a taste’,
‘“thave a listen’ etc., which are all semantically related. A com-
parison with Ger. hor ’mal (where mal is an abbreviation of
einmal, “once”), Du. luister eens and the widespread Brusseleer
(i.e. Brussels French) écoute une fois is tempting, but would
hardly be relevant here.
Another aspect of “as soon as”, especially conspicuous in
Portuguese, can throw further light on the precise nature of “‘at

31. G. Gougenheim, Grammaire de la largue frangaise du seiziéme siécle (Paris,

1951), p. 177.
32. A. Moravia, La Romana (Milan, 1965), p. 247.
28 Comparative Semantics

once”’. In addition to the expected equivalents (Jogo que, apenas,

or uma vez que), there is one that may at first appear surprising:
mal chegou
as soon as he arrived>
mal olhou para ela...
the moment he looked at her...

There is a corresponding use of no bien in Spanish:

No bien llego la hora sefalada, corrio...
L’heure fixée était d peine arrivée qu’il courut...

It may be helpful to compare mal in Fr. on dormait mal (on

dormait a peine or on ne dormait pas bien), where mal means
“hardly” or “badly”, and further:
Nous voyons mai André Gide s’intéressant 4 une Gertrude qui ne
serait pas aveugle. ..
We can hardly imagine...

There is obviously a vital difference between Fr. mal and the

Portuguese, for the French word is purely quantitative while Pt.
mal is temporal. It is easier to understand this Iberic phenomenon
if we think rather of “hardly” or d peine, which are used equally
well in reference to time and quantity. However, a truer under-
standing of Pt. mal can be gained from Turkish:
Ben gelir gelmez
As soon as I came (lit. I coming not coming, i.e. when I was on the
edge between arrival and non-arrival)

There is a remarkably similar construction in Japanese:

SO omou ka omowanai ka ni michi ga kyi ni magatte. . .
Scarcely had the thought crossed his mind when the lane took a
sudden turning. . .*5 (lit. whether thinking so or not thinking so...)

This throws immediate light on mal; the difference between jai

mal dormi, j'ai a peine dormi and je n’ai (presque) pas dormi is
very slight. At this point the distinction between presence or
absence of sleep almost ceases to exist.
Herein lies the value of un peu plus in the Léautaud quotation

33. R.L. Stevenson, New Arabian Nights (London, 1925), p. 274.

 At Once 29

on p. 26: “If Mme F.O. had left just a little later, they would
have met”, and the imperfect elles se trouvaient stresses this
inevitability much more forcibly than if Léautaud had written:
elles se seraient trouvées ensemble. Elsewhere the same writer
refers to a man who was nearly asphyxiated by carbonic acid,
and he adds: Un peu plus il y restait.*+ To put it more obviously,
but far less well: il serait mort.
We are here in a transition zone between event and non-event:
un peu plus corresponds to our “‘close shave’’, “narrow escape’,
“all but”, “hardly”, “scarcely”, “‘barely” etc. and equivalent
French phrases, with all the associated “lack” and “want”, and
it is here that the precise relation between Fr. faillir and falloir
(peu s’en fallut que) becomes clearer. There is only the slenderest
line of demarcation, a mere hair’s breadth, between narrowly
missing and having barely the time, between just making it and
just missing.
A further aspect of the same hesitation between being and
non-being is seen in the presence or absence of the negative in
It. non appena beside the simple appena. Consider also the
following:
Dhésiiai, le temps d’un éclair, si je ne quitterais pas le trottoir, pour
n’avoir pas a passer prés d’elles.*°

Several important types of temporal expression have been dealt

with in this chapter, and their relation to the central notion of
“at once” has been constantly stressed. It has been seen that
certain of these had a spatial origin, and that they have since
come quite naturally to acquire temporal associations. (The ex-
tremely close relationship between many spatial and temporal
terms has been shown elsewhere in this work.) What is especially
significant is the surprising mobility of these expressions; this is
clearly demonstrated by the fact that they have undergone notice-
able shifts of meaning in various directions. Hence, in the light
of all the doubt associated with ‘‘as soon as”, it is not surprising
that the closely related “at once” should either affirm itself by
shifting to a further future, or again give way to the tendency
towards an immediate past.

34, Léautaud, Journal, vol. x, p. 90.

35. Gide, Si le grain ne meurt.
Z
Semantic Groups in
Malay — Indonesian

When I first looked into Malay a good few years ago, my atten-
tion was soon attracted to certain semantic phenomena which
seemed of special interest, and as I have never seen the remotest
reference to these in any book that happened to come into my
hands I have long been wondering whether this matter could
possibly have escaped the notice of scholars, incredible though
this may appear.
It must first be stressed that we are here concerned exclusively
with words of the original native stock. The innumerable borrow-
ings from Sanskrit, Arabic, Portuguese, Dutch or English, how-
ever interesting they may be and well worthy of study on their
own account, are utterly irrelevant in this study, and in any case
they would be recognized at once as intruders.
A few points of more general interest and some theoretical
considerations will soon emerge from the examples to be pre-
sented, but it is best to let the groups first speak for themselves.
The most striking point to be observed from the outset is to
what extent the look (or rather sound) of a word suggests its
meaning.
1. Kawan, “friend”, seems to be just an ordinary word, but it
becomes remarkable as soon as it is placed alongside Jawan,
“enemy”. The two appear to be opposites, but there is a deep
identity underlying this apparent opposition. Lat. hostis,
“enemy”, is cognate with “guest”. Going towards (with friendly
intent) or against are two senses frequently combined in the
same word: Ger. gegen has both meanings, and contre was often
indistinguishable from vers in Old French. Similarly, the line of
demarcation between “fighting with” and “fighting against” is
often blurred. Jp. aite 1+ combines very smoothly the senses
 Malay-Indonesian 31

“companion” and “opponent”, with “partner” as middle term.

Reference may also be made here to the German pair Freund
and Feind (: “friend” and “‘fiend’’), which were originally present
participles of the older Germanic verbs for “love” and “‘hate”’;
hence the parallel endings. “‘Friend”’ is thus closely comparable
to Fr. amant, ami, especially if it is remembered that Lat. amare
turns into OFr. amer. Later, when analogical aimer became usual,
ami was cut adrift, and much the same can be said of “friend”
and Freund as a result of the disappearance of the original verb,
whereas in Latin amare, amicus and inimicus formed a closely
knit family. OFr. ami was the normal term for “the beloved
one’’, while in Modern French it is used more often than not as
a euphemism for amant. Much of course depends on context
and situation to throw light on the precise meaning of such terms.
2. Sempat, “opportunity”, is really “the right time”, whereas
tempat, “‘place”’ is actually “‘the right or proper place’. It is a
sheer coincidence that tempat is outwardly so much like tempus,
though time and place happen to coincide ever so often. The
very notion of time ultimately goes back as a rule to a measure
of space, and there are innumerable survivals of this identity
in most modern languages. It is precisely in the sense of “right
time (and place)” that numerous expressions of the type en temps
et lieu gradually came into being, though a number of these
are probably calques on an earlier phrase. Opportunity occurs
where time and place happen to meet.
There is further evidence of the semantic identity of sempat
and tempat in that both have developed the secondary senses
“convenient, suitable, fitting”. Eng. “tidy” originally meant
“timely” and is thus exactly comparable to Ger. zeitig, at least
in form. For the present sense of “tidy” a fruitful comparison
may be made with Fr. propre, “clean”, whose earlier usage has
survived in Eng. “proper”. There is a comparable topic for
semantic exploration in Ger. schén, “beautiful”, and its relation
to schon, “already”, and further in the Dutch cognate schoon,
“beautiful” and “‘clean”, and it is surely no coincidence that
both these latter meanings are found combined in Jp. kirei. It is
also in Japanese that we find the most arresting parallel with the
derived sempat, tempat sense we have just been considering: Jp.
koroai, 4» “suitable”, means literally “‘meeting or fitting the
 32 Comparative Semantics

time”,.as can be seen from the characters; koro, “‘time’’, is in

fact a shortened form of tokoro, “place”, another example of
time from place.
3. A further pair worth considering at this point is samping and
damping, whose original sense “‘side’’ is still clearly discernible
in their present use, from “‘side”’ to “beside, near, close”, just
as in French from cété to a cété, and Eng. “‘beside’’. However.
we gain a better insight into the semantics of this group if we
take in another pair, remarkably close to the previous one in
both sense and form: banding and tanding. The basic meaning of
these is “equal’’, and from this it will be readily understood how
the compound perbandingan develops into ““comparison”’, while
the corresponding pertandingan, “‘comparison”’, goes further
along the same semantic path to “contest, match’”’. Considering
_ now the sense of both pairs, it will be clear that there is a fun-
damental unity connecting them, particularly in the relation
between “side” and “comparison’’, for if there is an almost
universal method of comparing objects it is by placing them
“side by side’’; hence the numerous expressions of this type in
many languages. One example will do out of many possibilities:
on se sent bien ignorant aupres de lui.
Perhaps the relation between “‘equal’”’ and ‘‘compare” may
not be quite so obvious, but a glance at the usual Russian word
for “comparison”, cpaexenue , based on paenoiit , “equals;
leaves no room for doubt. In contrast to this Russian term and
Ger. vergleichen (gleich: “‘like’”’) or Ic. samlikja (sam-lik-ja),Fr.
comparaison and Eng. “‘comparison”’ are quite abstract and
unmotivated, especially if Lat. comparare is looked upon as a
compound of parare, “prepare” (or similarly if comparer is
associated with parer or préparer), whereas it is actually built on
par, “equal”. Then and only then can it be seen that c-pae-nenue
corresponds so exactly to com-par-atio that it may well be a
calque on it; but whereas the Russian and Germanic words are
magnificently alive, the French and English terms have long
ceased to be inwardly intelligible and must be accepted blindly.
As to the senses “‘contest, match’’, it is clear that the side-by-
side method of comparison is suitable mostly for physical
features, but that other aspects have to be discovered by a test
or contest. Even where all are equal, there is usually a primus
 Malay-Indonesian 33

inter pares. The word “confrontation” further illustrates these

relations and brings us back to the field mentioned in Section 1
of this chapter: facing with friendly or hostile intent. Here it
may be useful to recall that the normally friendly Fr. rencontrer
(especially in aller a la rencontre de ...) is derived from contre,
“against”’.
What is valid in the West is equally so in the East. A common
Japanese verb for “‘compare’’, naraberu, means literally “‘place
side by side’’. Again, Ch. dui ¥{ and the same character as used
in Japan (xt) shows almost exactly the semantic range observed
above. The basic sense seems to be “a pair’’ and “‘side by side’,
hence “‘face and oppose”, with the further development in
Chinese, “‘suit and agree’’. Once more, friends and enemies are
seen to be not just opposites as would commonly be assumed,
but rather to be complementary notions.
4. The student of Indonesian is likely to meet tulis, ‘‘write”’,
very early in his reading. He is unlikely to come across lukis,
“paint”, until a few months later, and by that time t/is will long
have lost its novelty. By the time he meets guris, “‘scratch”,
both other terms will have been so thoroughly assimilated that
it will probably not occur to him to bring together these three
terms which form such a well-defined semantic family. Here
again, the important point to observe is that sense and form are
related. Or to put it another way, the sound of a word may be
said to suggest its meaning. This is almost as if people were
named after their looks, which may have been true in many
cases originally, but now we are left with names utterly unrelated
to the looks of the bearers and which we accept blindly. How
can a man possibly be called Lamartine or Larousse?
The ultimate identity of “scratch” and “write” is clear if it
is recalled that the original sense of “‘write”’, Lat. scribo and Gk
ypadaw was ‘‘scratch”. But there is the additional evidence that
the most obvious cognates of “write” are Ger. reissen and Du.
rijten, ‘‘tear”. There is moreover the fact that beside the usual
Japanese verb for “‘write”’, kaku, #£< there is another kaku,
“scratch”, written with a different character, 4% <. In view of
what has been observed in Indo-European, it is reasonable to
assume that both these Japanese words are one and the same
and that ‘‘scratch”’ is simply the earlier meaning from which
34 Comparative Semantics

‘“‘write’’ has developed. Both senses continue to coexist har-

moniously side by side, and to the Japanese they are merely
different words that happen to be homonymous.!
The relation between “paint” and “write” has similar his-
torical validity, especially in Slavonic languages, and there is
the further corresponding modern survival in Rus. nucamo, the
usual term for both “‘write” and “‘paint”. Moreover the Gothic
melyan, “write”, is clearly cognate with Ger. malen, “‘paint’’.
It is interesting to note that Old Norse skrifa, ‘“‘write”’ (an early
borrowing from Lat. scribo, and thus parallel to Ger. schreiben,
Du. schrijven etc.), has also developed the meaning “paint”’.
There is no need, however, to go so far back to discover the
identity between “‘writing”’ and “‘painting”’. We are all conscious
of the writer using words to paint or depict what he visualizes
in his imagination; besides expressions such as “‘pen portrait”’,
“word picture” etc., there are a host of words that belong equally
well to both painting (or drawing) and writing: ‘“‘sketch” (and
Fr. esquisse), Ger. Grundriss, “‘outline’’, “‘draft’’ etc. Incidentally,
the riss of Grundriss is obviously the noun form of reissen, “‘tear”’,
the cognate of ‘‘write.”’.
However, it is in the Far East that this identification is most
striking. Even the mere physical action of writing (or copying as
distinct from composing) is here on almost the same level as
creative painting. It is well known that both calligraphers and
painters are held in high honour as artists and use the same
brush & (Ch. bi, Jp. fude), and whereas a few Chinese characters
will add beauty and distinction to aChinese painting, our own
artists almost apologetically relegate whatever has to be written
to the most inconspicuous part of their work; our prosaic letters,
they obviously feel, do not enhance their art.
As can be expected, there is no lack of terms in the East that
bridge the gap between painting and writing. Thus Jp. bydsha
45 is used mostly for portrayal with words, but one reading
of the first character #% taken by itself is egaku, ‘“describe” or
“draw”, which is made up of e, “picture”, and kaku, “write”
(or “scratch” if preferred). The second character & is the usual

1. On this point, see my ‘Introduction to Comparative Semantics”. Aumla, 29

(May 1968): 14-15.
 Malay-Indonesian 35

Chinese word for “write”, xie, and the following object will
differentiate between writing as a scribe or a calligrapher BY
xie zi, and creative writing. There is, however, a further compound
%/£ xie sheng, “draw life’, i.e. “‘draw living objects”, and the
same compound in Japanese, read shasei, is used of a sketch,
written or drawn. A curious aspect of this Chinese compound
is that it is so tantalizingly close to Rus. »#cueonuce, “painting”
(lit. life, write and paint), which itself is probably a calque on
Gk Cwypadia, and we have of course our own ‘“‘paint from life”
and peindre sur le vif. The original sense appears to have been
an explicit contrast between painter and writer, a distinction
that so often vanishes in Chinese writing, while in our own the
difference is far more clearly marked.
All this implies a great deal of synesthesia. Just as a composer
will often set himself the task of making audible what is normally
perceptible to the physical eye alone (Clair de lune, Reflets dans
Peau), so the painters, especially since the Impressionists, seek
to go far beyond the plastic limits of their art, and attempt to
make visible various abstract notions that belong to other
senses or to that more general inner perception which is not
localized in any particular sense, i.e. what is sensed to be, as
distinct from what is seen, heard, smelt and so on.
Creative writing occupies an intermediate position between
painting and music; it is akin to the former in so far as it is
written for the eye, but these visual impressions are usually
transformed spontaneously into sound images, and especially
so with poetry. The writer too will attempt to push words far
beyond their normal! limitations, and thus poetry, or even prose,
readily turns into incantation.
5. The Indonesian semantic group we are now about to examine
will show clearly the precise relation between the form and mean-
ing of the terms under consideration.
Id. kaja, ‘‘wealthy”’, is obviously related in sense to daja,
“nower’’. If there is any doubt on this point, it will be sufficient
to cite kekajaan (i.e. ke-kaja-an), a compound of kaja, which
means both “wealth” and “power”. Even more illuminating
is the expression kaja raja, “‘very rich”, where raja, “‘greater’’,
has developed into an intensive for kaja, “wealthy”, solely
36 Comparative Semantics

because of its form. However, the most remarkable member of

this semantic family is djaja, “victory”, clearly a borrowing
from Sanskrit, which should therefore be excluded from these
semantic groups, together with all other borrowings, as was
stressed at the beginning of this chapter. A purely fortuitous
accident has brought it about that djaja has a form identical
with the terms now being examined; the fascinating result is
that its compound kedjajaan (cf. kekajaan for this formation)
means not only “glory” but ‘“‘wealth’’ as well.
It should be clear without much further elaboration that
this semantic family corresponds exactly to one that is well
known to European linguists. OFr. riche? meant not simply
“rich”, but rather “rich and powerful’, and often the best
rendering is “mighty”. This is equally true of OEng. rice and
Middle High German rich. It will also be observed that the
semantic development into the modern sense, with the dominant
“money” notion, corresponds precisely to the history of such
words as “‘purchase’’, acheter, “‘sell’”’ etc., which had nothing
to do with money originally.
A particular extension of this same Indo-European word is
to be found in Lat. rex, regem, “‘king”’, in such Celtic names as
Ambio-rix, Vercingeto-rix etc., and furthermore in the Indian
raja. Here a note of caution must be sounded. It is certainly
tempting to equate this raja with the Id. raja, especially in view
of the Sanskrit example just quoted and the sense relation
underlying all these terms, but we are actually faced here with
a pure coincidence: there is not only a difference of quantity
in the original Sanskrit vowels, but the j of raja is not just a yod
as in the Indonesian words mentioned earlier. The point will
be immediately clear when it is realized that Sk. raja has itself
been borrowed by Indonesian and is normally spelt radja in
conformity with the pronunciation, and not raja. Thus radja
by its form is excluded from this sense group, whereas raja
belongs.
There is another group of terms in Germanic languages
that unites in a specially striking manner the notions of
“wealth” and “might”. Ger. Vermégen (clearly a compound of

2. See, for example, Anon., La Chanson de Roland, ed. T.A. Jenkins (Boston,
1924), and Anon., La Chanson d’Aspremont (Paris, 1923).
 Maiay-Indonesian an

mogen:“‘may”; cf. Macht:‘‘might’’) has both senses, “‘power”

and “‘wealth”, and the usual meaning of vermégend is “rich”
rather than “‘powerful’’. Moreover practically all compounds
built on Vermégen have to do with wealth and finance, despite
the form of the word, which rather prompts the power aspect.
So, too, vermégensios is simply “‘impecunious, poor”’.
The same phenomenon is found in Du. vermogend (adjective)
and vermogen (noun), which differ from the German usage
mostly in the fact that both “wealth” and ‘“‘power” are equally
present. There is also Sw. formogen, “‘wealthy”, which should
be further examined, and the remarkable Dan. Formue (noun),
“wealth”, together with the corresponding adjective formuend.*
It is of course possible that we are here faced with calques or
borrowings from German, but whatever the starting-point it is
beyond all doubt that the relation between “wealth” and “power”
remains particularly clear and striking in all these Germanic
terms.*
A final remark concerning the sense link between king and
wealth: the French say not only heureux comme un roi but also
riche comme un roi to describe superlative luck and wealth.
No doubt the alliteration plays some part in this phrase, as it
so often does, but clearly the essential relation involved is a
semantic one.
6. The semantic relation between baris, “row” and “line” (in
a book) and “line, queue”’ (of people), and garis, “line” (especially
in its geometrical use), is obvious enough and here again we
see how form and sense are associated. To these two may be
added Japis, occasionally “‘row”, but more usually “layer’’ or
“‘lining”’.
A geometrical line has length and ideally no width. A row is
generally thought of crosswise, i.e. it belongs to width rather
3. G. Brandes, Cajus Julius Caesar (Copenhagen, 1921), vol. i, pp. 157, 256.
4. On this point it is hard to understand how C.D. Buck in his Dictionary of
Selected Synonyms in the Principal Indo-European Languages (Chicago, 1949),
a work outstanding in many ways, could possibly have missed completely
the semantic relation which is the fundamental subject of this discussion:
under MIGHTY there is no trace of “rich”, nor is there any hint of the
existence of the vermogen type under RICH. We could very well ask how
Buck went about compiling his remarkable dictionary. He is certainly much
into
more reliable on Greek than elsewhere, despite his interesting probes
the earlier aspects of English and German.
38 Comparative Semantics

than length. If we consider, say, the arrangement of desks in a

classroom or seats in a theatre, it will be seen at once that line
and row are normally related in this way. However, much
depends on the position of the beholder: he need only go side-
ways and the former rows turn into lines and vice versa. All this
is fairly evident, and a further confirmation of the close identity
between the two senses is to be found in the lines of a book, where
the letters stand in a row side by side. So it is no surprise to find
that in Japanese gy6 {7 is used both for “row” and “‘line in a
book”.
As for /apis, “layer’’, it is clearly the next step in the series.
While lines belong to one dimension and the combination of
lines and rows has to do with plane surfaces, layers (or linings)
add an extra dimension and bring us definitely into the realm
of solid geometry, where the notion of volume becomes dominant.
7. The group djahit, ‘‘sew”’ and pahit, ‘“‘bitter’’, is far less obvious
than the previous one, yet the mere fact of setting these two
side by side amounts to an assertion of semantic proximity,
and the reader may have seen the relation at once, or he may
require some explanation. Actually there are clear pointers
in both Western and Eastern languages, as is usually the case
when we are faced with a fundamental unity. Instead of the
procedure adopted previously, looking first at European evidence
and then passing on to the Far East, let us reverse the process.
Jp. nuu, “‘sew’’, is written #, with the radical * which has
to do with vegetal fibres, threads etc., and a so-called phonetic
which itself exists independently in the sense ‘“‘meet’’ where the
radical is i. The remaining element # originally meant “point”
(not a dot, but a sharp point) and still means “butt” (with the
horns) in Chinese. Add the mountain radical and we have Ité
‘‘a mountain peak”’; with the fire radical #& the sense is ‘“‘beacon”’
(i.e. a signal fire at the top of an eminence), and with the metal
radical # it still refers to “the point of a spear”. On the other
hand, # hachi covers both “bees and wasps’’: if we analyse the
character we find an insect th which stings. It is of course possible
to differentiate if necessary: Jp. mitsubachi is literally ‘“‘the
honey bee”, just as Jp. yagi, “goat”, and hitsuji, “sheep”, help
to distinguish between sheep and goats, both comprised in the
character ¥.
 Malay-Indonesian 39

By this time it is probably clear that everywhere “‘the point”

is the original sense of the Chinese word feng 4 and that the
radicals were later added to show the direction of the various
semantic changes. Thus 3 ‘“‘meet’’ comes from the “‘meeting
point” and # “‘sew”, the character that we started with, is
now seen to be admirably descriptive of the process of sewing
with a needle. French can bring this into still clearer perspective:
a stitch made with a needle is un point, the same word as “a dot”
and originally also the same as une pointe, ‘‘a sharp protruding
point’. I need hardly add that these go back to Lat. punctum,
and that the corresponding infinitive pungere, “‘pierce, sting”,
which turned into poindre, has been supplanted in most of its
uses by piquer.*
So much for djahit, ‘sew’. Let us now try to apply the same
method of exploration to pahit, “‘bitter’’, starting from Japanese
and then going on to Europe. The usual term for “bitter” in
Japanese is nigai 7. Here again, as in the case of nuu, “sew”,
which has just been examined, the main interest lies in the
character itself, which is more commonly read kurushii, UW
“painful, distressing”, or kurushimu, tte “to worty 5 Or
again kurushimi, %& L & “affliction, agony”’.°
There are two points worth noting here and they both belong
to semantics. The first can readily be gathered from the above
relation between “bitter” and “affliction”. “Bitter” is the original
sense from which the notion of distress has developed, precisely
like “bitter” in English and amer in French. So too in Indonesian,
pahit has acquired a metaphorical meaning. It is almost inevit-
able: if we contrast “bitter taste” and “bitter experience”’ the
relation is obvious. The physical is the starting-point of the
figurative.
The second semantic aspect confirms what we have just
observed, but takes us much further back into the past, to the
very source of this bitter taste. There are two approaches: the
Chinese language, where the physical “bitter’’ is clearly the
earlier sense, and, what is far more illuminating, the character
itself, whose radical *" seen at the top of 1 refers to herbs.

5. Rabelais uses poincture for pigiire in Le Tiers Livre, chapter XXII.

verb
6. The final syllabic kana sign distinguishes between the adjective (-i), the
(-mu), and the noun (-mi).
40 Comparative Semantics

The bitter taste in question is undoubtedly that of medicinal

plants (or simples as they used to be called), and there are
several Chinese compounds where this early pharmaceutical
sense is still plainly discernible. Perhaps I may also refer to
various childhood experiences when I was made to swallow
certain undrinkable potions. In those days practically all
medicines were liquids that had to be ingurgitated, no matter
what agony the child suffered in the process. To me at least,
and no doubt to many of my generation, bitterness and medicines
are indissolubly associated.
However, there is a more direct and much simpler way of
discovering the essential unity between “‘bitter’’ and ‘“‘sew”’, and
English happens to provide the immediate solution. For “‘bitter”’
is that which bites, and the relation between “bitter” and ‘“‘bite”’
is clearly seen in all Germanic languages. A glance at various
synonyms of “bitter” will confirm this: sharp, acrid, stinging,
pungent etc. Here it will be noted that “pungent” brings us
into direct contact with Fr. point, ‘“‘stitch’’, which we met earlier.
There is the further evidence that Gk aikpoc, “‘bitter’’,
originally meant “pointed, piercing’. When Rabelais refers to
Picrochole’s cholere pungitive’ he is actually transposing Picro-
chole’s name into Renaissance French. But there is no need to
be so learned. The relation between “‘sewing’’ and “bitterness”
or, to put it now in its simpler form, between “‘sting” and “‘bite”’
is so obvious that it requires no further proof. We could look
at it in a very homely way. There is no doubt that bees and wasps
do sting. But when it comes to ants and mosquitoes we are not
so sure: “a mosquito bite” is more likely to be une pigire than
une morsure de moustique to a Frenchman. On snakebite there is
more general agreement, and yet this is rather a stinging or
piercing by isolated fangs than biting by a row of teeth. So it is
not just a simple matter of observing whether the danger comes
from the tail end or from the front end. There is a doubtful
transition zone between the two types, and many an unfortunate
insect has lost its life because its abdomen happened to have a
slight motion downwards, which suggested a sting where there
was none.

7. F. Rabelais, Gargantua, chapter XX XVIII, in Oeuvres completes (Paris, 1962).

 Malay-Indonesian 4]

Language sometimes adds to this confusion. For example,

French uses /e dard in referring not only to the sting of bees
and wasps but also to the forked tongues that can be seen
darting forward from the mouth of many reptiles. Now this
forked tongue, thrust forward as it is so conspicuously, also
gives the impression of a threat, precisely like the insect’s down-
ward abdomen, and many innocuous lizards have suffered
because of this imaginary danger.
French, as we see from /e dard, can be most misleading. But
as if to make up for this unfortunate confusion, it is in French
too that we find the most arresting proof of the essential unity
between “bite” and ‘‘sting”: morsure and pigure plainly bei:ong
to the same group as blessure, écorchure, éraflure, égratignure etc.
In all of these the semantic affinity between mordre, “‘bite’’,
piquer, “sting”, blesser, “wound”, égratigner, “scratch”, is
brought out by the suffix -wre. This serves to illustrate the fact
that suffixes too can be bearers of meaning and be especially
important.
8. (a) The reader may have noticed by now that the most
constant and significant aspect of the semantic groups examined
so far is to be found in the vowels. The vowel i in particular is
well worth special attention. It has so far occurred in this chapter
in Section 3 (samping, “‘side’’), Section 4 (tulis, ‘“write’”’), Section 6
(baris, ‘“‘row”) and Section 7 (pahit, “‘bitter’’), i.e. as a final, while
the first vowel has always been a back one.
We are now going to consider what happens when the first
vowel is also a front vowel, but it may be even more revealing
at the outset to try to isolate the precise suggestive value of this 7.
A glance at the Indonesian pair tjelepik and tjelepuk is at once
illuminating: to an Indonesian, tjelepik suggests the sound of a
light object falling, while tjelepuk refers to the fall of something
neavy, that is, a fall marked by a thud or a plop. So too the
parallel contrast between memantjit, “squirt little by little’, and
memantjut, “gush’’, will be readily appreciated. Here it is clear
that the opposition resides entirely in the contrast between the
vowels—i for a light sound and wu for a heavy one. (Notice
incidentally that this u sound is not that of “thud”, but is more
like the vowel in “foot”.) If a closer examination were made of
the final vowel u, it is likely that quite a number of new points
42 Comparative Semantics

would emerge which would by contrast further confirm the

respective suggestive value of i and u, a value that corresponds
closely to the real nature of the sounds themselves and may
therefore be considered as natural.
In French, just as in Indonesian, a heavy fall is likely to be
suggested by the same close back vowel, the dull, muffled sound
heard in plouc or flouc (cf. flou and un bruit sourd), a sound that
is often vaguely felt to be rather dark in contrast to the clear
ringing 7 heard in tinkle, tingle, clink or jingle, or again in Fr.
déclic or cliquetis.
In English, on the other hand, it is rather the vowel heard
in thud, thump, dump etc. that suggests these heavy objects
(cf. dull, muffled). Many other examples will readily come
to mind.
It is also worth observing that the notion of collapsing
heavily is frequently expressed in various Romance languages
by a word descended from plumbum, “‘lead”; thus It. piombare,
Sp. desplomarse and still more striking Pt. chumbar: passou na
escrita, mas chumbou na oral, “‘he passed in the written (exam)
but flopped in the oral’’. Note too that when a French play turns
out to bea flop it is usually called un four.
The vowels of Eng. “little” and “much” stand in practically
the same relation as those of the Indonesian pair tjelepik and
telepuk: i small and wu big. When I first heard the proverb
“many a mickle makes a muckle” I was very puzzled, since to
me “‘mickle” meant “much”, consistent with earlier English.
Now, this old saw could only make sense if “‘mickle” had the
opposite meaning, and closer investigation soon revealed that
““muckle” was merely a variant of “‘mickle”. Later on I came
across MEng. muchel beside luttel, but the significant point is
that “much” ultimately triumphed, with a vowel that may to
some extent be due to the labial influence of m, whereas Juttel
soon disappeared, as could well be expected from the suggestive
value of the two vowels.
It can now be seen how the proverb came into being. From
an earlier “many alittle (or pickle) makes a mickle’’, which is
consistent with the true sense of “‘mickle”, the existence of
“‘muckle” beside “mickle” suggested a corresponding contrast
 Malay-Indonesian 43

in meaning, muckle “much” and mickle “little”, prompted

solely by the respective vowels.
It will be recalled that Jespersen wrote an absorbing article,
“Symbolic Value of the Vowel i’’,® in which he gives a wealth
of illustrations in various European languages—illustrations
that it would therefore be superfluous to mention. I will limit
myself to a couple of recent examples. Not long ago I happened
to overhear a neighbour say she would give “‘those flowers a
teeny little wee bit more water’’. There is also the great vogue
of the miniskirt, which has been responsible for an extraordinary
number of mini-compounds, and for some time even succeeded
in transforming all minuscules into miniscules. It may be relevant
here to recall that “‘miniature” owes its present sense to the
accidental fact that its initial syllables were reminiscent of
“minimum”, and so suggested something small. Contrast the
opposite “‘maxiskirt” and “maximum”. Obviously the much
lesser popularity of this maxiskirt could do little to induce the
creation of maxi-compounds.
However, we are here concerned in the first place with
Indonesian. What is most arresting is the large number of
examples that could be quoted to show that the vowel i has here
a symbolic value closely comparable to that found in our Western
languages, but with a small difference: the i is final and the
preceding vowel is usually also a front vowel, mostly i or e; this
is in marked contrast to all the groups previously considered,
where the final i was preceded by a back vowel, generally a or u.
Let us first consider the type E-I.
(b) There can be no doubt that the most obvious member
of this family is ketjil, the usual Indonesian adjective for “little”,
which can profitably be compared with Fr. petit, especially
since the form of this French word is also largely prompted by
its suggestion of smallness. The fact that Id. kegjil is usually
pronounced ketjik does not affect this essential relation; on the
contrary, it stresses again that the vowels, rather than the con-
sonants, are semantically significant. What is still more curious
is that the Indonesian vowel written e in the terms now under
consideration represents a short indeterminate neutral sound

8. O. Jespersen, “Symbolic Value of the Vowel i”, Linguistica: Selected Papers

in English, French and German (Copenhagen, 1933): 283-303.
44 Comparative Semantics

that is in fact strikingly similar to the e of petit. The remarkable

resemblance between ketjil and petit.can hardly be said to be a
pure coincidence, since it is to some extent due to the same
natural promptings, though it is debatable whether “natural”
is the right term to use in this case.
Even more arresting is menit, ‘‘a minute” (of time). Whether
this is borrowed from Dutch or English, and should as a borrow-
ing be excluded from our semantic groups, its original form was
sufficiently close to the type now being studied for the word to
be attracted into it, and it thus became motivated by its form
within the Indonesian semantic framework. The corresponding
term for ‘‘a second”’ (of time), detik, is seen to have exactly the
same form. Menit and detik are both being used with increasing
frequency, especially in assessing sporting results, so that the
link between these two words is constantly being strengthened.
It may be of passing interest to recall that our own “‘minute”’
was originally (pars) minuta prima, in contrast with minuta
secunda, minuta tertia etc., each being one-sixtieth part of the
preceding unit. Though we don’t yet hear of a “‘third’’ as the
next subdivision of the second, I distinctly remember that when
a father was asked the name of the next step towards smaliness
in the series heure, minute, seconde, he answered: “‘C’est une
tierce.” With that the child must have been satisfied, for I don’t
recall any further attempt to probe into the next possibilities,
quarte, quinte etc., but this may have been due to language
deficiency.
It would be tedious for those who know neither Malay nor
Indonesian to read a mere enumeration of the many terms that
belong to this E—I family, and those who do know these languages
will readily find quite a number for themselves, now that this
semantic feature has been brought to their notice. So I will
merely mention a few typical examples chosen among many:
desik and lesit both suggest “‘a rustling sound”; renik “fine,
minute”; getis “brittle”; kerik “the sound of ticking or typing”
(hence the derived ketika, “‘a point of time’’, and from this
“when”’); kedip “a wink of the eye, a flicker”; ketip “the sting
or bite of a small insect” (attention has been drawn in Section 7
to the fact that the distinction between bite and sting vanishes
as the insect becomes smaller); ketis “to flick or flip; a speck of
 Malay-Indonesian ~ 4$

dust” etc. It will have been observed incidentally how frequently

the same vowel i occurs in the corresponding English terms.
(c) Transition zone between E-I and I-I: there are four
main groups forming this transition and they are of considerable
interest. ;
Benih means “‘seed, seedling; semen; germ”, hence “‘origin,
cause”, and precisely the same range of meanings is found in
bibit, a very interesting use of which occurs in kuda bibit, ‘“‘a
pedigree horse”. However, the most remarkable member of
this group is undoubtedly bidji, “‘seed”, not only because it
happens to be a borrowing from Sk. bija, “‘seed’’, whose form is
again made to fit the sense, but especially because of its present
widespreau use as a numerical classifier.
The classifiers are probably the most distinctive characteristic
of Far Eastern languages, and all students of Chinese or J apanese
are well acquainted with them. However, it is worth stressing
how closely English and French correspond in this respect. It is
of course no coincidence that “five head of cattle’? and cing
tétes de bétail are so peculiar in precisely the same way, or that
*un bétail is just as un-French as “‘‘*one cattle” is impossible
in English, provided the French wn be understood as the numeral
and not just as the article.
It is curious to observe that Turk. bas, “‘head’’, is used in
exactly the same way in counting livestock; and so is Jp. 10, 5
“head”. While this is possibly a simple coincidence, it may well
reflect the twofold function of cattle as domestic animals and
as capital. “Cattle” is of course the same word as “capital”.
Lat. capitalem, OF r. chattel, which passed into English: ““goods
and chattels’. When it is recalled that Ger. Vieh, “‘cattle”’
(cognate with Eng. ‘‘fee’”’”), shows precisely the same peculiarity,
but with a different classifier, Stick, “piece”, this explanation
becomes quite plausible. Ger. zwei Stiick Vieh, “two head of
cattle”, is closely comparable in form with Fr. deux piéces de
monnaie, where monnaie cannot be accompanied by an exact
number. Similarly, the comparison of “old furniture” with Fr.
de vieux meubles, and “‘this piece of furniture”, “three pieces
of furniture” with ce meuble, trois meubies, shows that Eng.
“furniture” requires a classifier.
On the other hand, Fr. gens differs from Eng. “people” in
46 Comparative Semantics

that it cannot be used of a definite number: while it is usual

enough to find peu de gens, beaucoup de gens, bien des gens etc.,
it is not possible to say *cing gens like “five people’, and there is
no classifier available to solve the problem as in the case of cattle
and bétail or furniture. “People’’, incidentally, seems to be a
plural without the mark ofa plural: ‘“*one people” is not possible
in the sense of “‘one person’’. It is rather strange that many
grammarians who give so much space to gens and its eccentricities
in gender do not so much as mention this remarkable aspect
of the word.
In Indonesian, bidji, “‘seed’’, is the normal classifier used for
all small objects; it is also used frequently in the language of
sport: where we would say “‘he was declared the winner with
161 points’, the usual Indonesian word for “point” is bidji.
For school marks too bidji is quite common, just as the French
use point. Buah, “‘fruit”, is the corresponding classifier for large
objects. Seed and fruit form such an obvious semantic contrast
by their size that no further comment is necessary.
It may be useful to illustrate briefly the use of these grammati-
cal tools in Indonesian:
tiga bidji telur
three eggs (lit. three seed egg)
dua buah rumah
two houses (lit. two fruit house)
se-orang budak
one child (lit. one human-being child)

In this last example, se-orang is exactly parallel to Ch. — A,

where A Ch. ren, Jp. hito, means “human being’. This may
appear rather cumbersome, but the apparent awkwardness lies
solely in the literal translation.
The next transition group includes djepit, “‘to be squeezed”
and djentik, “‘pinch”. Pedih, ‘‘smarting; poignant”, and sedih,
“sad”, also belong here, and if there is any doubt, reference to
Section 7 (djahit, pahit) will throw further light on the semantics
involved. It is to be remembered also that “poignant” was
originally “piercing”, being the present participle of poindre,
Lat. pungere, mentioned in elucidating this same group.
Lindih and lindis, “press, crush’, are also relevant here.
Compare further pidjit, “‘press, squeeze’, and getil, “‘pinch; hold
 Malay-Indonesian 47

with two fingers”. The latter is reminiscent of Jp. hasamu, “hold

between two objects” and especially the corresponding Chinese
character #; whose form is particularly illuminating since 7
means “press, squeeze’, and the radical on the left is the hand.
Thus the original sense of this Chinese character must have been
very close to that of present-day Id. getil.
While we have this character fresh in our minds, let us consider
sempit and pitjik, both meaning “narrow”; the former is more
usual in the physical sense, though both have developed a
number of figurative meanings: “narrow-minded; mean; hard-
pressed; be ina fix” etc., much as in our own Western languages.
Compare “‘dire straits”, whose ‘“‘strait’’ is borrowed from OFr.
estreit (Lat. strictus), Fr. étroit, “‘narrow’’.
The relevant character is #%, whose radical on the left is the
dog. This is still the usual way of writing Jp. semai, “narrow”,
but in Chinese this character has lost its original sense and now
generally signifies ““mean’’. We may well wonder what the dog
has to do with this, but on reflection it seems to be derived from
hunting. In the East as in the West, the dog has long been closely
associated with the hunt. In Chinese, the normal character for
“hunting”, 3%, is also seen to be based on the dog, and this is
equally true of the Japanese equivalent kari 4#. “Cynegetics”
may not be normal English, but the French refer quite commonly
to exploits cynégétiques; this goes back to the Classical Greek
term where the presence of the dogs is unmistakable.
To come back to our present semantic problem, “narrow”,
the actual image seems to be that of the hounds in hot pursuit
closing on their quarry, and the quickly narrowing gap. A careful
examination of the approaching death of Roland in the Chanson
de Roland will show that the underlying imagery is that of death
as a pursuing hound, though this is not specifically mentioned.
Asa curious coincidence it may be worth recalling the Modern
French colloquialism il est trés chien, “he is very mean”, or
again faut pas étre si chien,” (you) mustn’t be so mean’’, whose
chien has exactly the sense of Ch. # xia, “mean”, already
discussed.
A further example belonging to these transition groups is
petik, “pick”, whose semantic association with smallness may
not be at all obvious—that is, until it is realized that the fun-
48 Comparative Semantics

damental sense of ‘“‘pick” is closely related to that of “peck”.

Think of the hen and the basic notion will be immediately clearer.
I place petik in this section because of pilih, “fastidious”, or
better still ‘““choosy”, whose association with petik is to be
found in the common phrase “pick and choose”’. Far from being
haphazard, expressions such as these contain a quintessence of
semantic virtue, as if the wisdom of centuries past had been
slowly distilled into a lasting form, for the illumination of
future generations.
With these too must be linked fitik, “‘a drop; dot, point’,
though here again the reason is not clear at once. However,
there is a definite hint of a fundamental identity in the Chinese
characters jf “‘a drop” and 4 “to pick’’, which have the same
phonetic, but with the water radical for the former—as
can be expected—and the hand radical for the latter. It was
pointed out earlier that the phonetic is often of great semantic
significance.° Jp. tsumu, “‘ta pick; pinch; nip; clip, trim’’, is
normally written #%t, and the Sino-Japanese compound #4#
tekiyO designates “a digest, extract, abstract’’; literally it is
the result of picking and choosing (teki) the essentials (yé).
Coming back briefly to the “pick-peck”’ relation, it is probably
no accident that a woodpecker is normally called in French
un pic, which suggests that a number of these terms have an
onomatopoeic origin; this fits admirably with the symbolic
value of the vowel i and suggests how it may have partly come
about that this phenomenon is so widespread.
Before leaving this section, brief mention should be made of
a final group of words that are all closely related in sense.
Retjik and pertjik both mean “spot, stain”, and with the prefix
me-, “sprinkle, spray” is the usual sense. To these may be added
rentjis, especially with the latter meaning.
With these should be compared bintik, ‘dot, speck”, hence
berbintik, ““be spotted, have spots’’, as can be expected with the
prefix ber-; and also rintik, “‘fine spot’, which, with this same
prefix, develops the senses ‘“‘with fine dots; fall in small drops”
and is more specifically used of “drizzling rain’’.
As to the semantic relation between dot, spot and stain, it is
immediately illuminating to compare spotted, speckled and
9. See Sect. 7, p. 38.
 Malay-Indonesian 49

dotted, where the spots are rather more ornamental than “‘stains”
would suggest. This distinction is strikingly obvious in the
French pair derived from tache, “‘stain, blot’’: taché, “stained”,
as contrasted with tacheté, “‘spotted”. For the sense value of
the suffix -té here, a useful comparison can be made with mou-
cheté, ‘dotted, speckled”’.
(d) Type I-I: a good few examples of words of this pattern
have already been quoted in the preceding section, where both
E-I and I-I terms were found in closely associated semantic
groups.
Bibir not only means “‘lip’” (part of the body) but in addition
‘hem, edge”, and this sense may suggest to some extent how it
fits here. From gigi, “tooth”, is derived gigit, ‘a bite”, which will
be clearly seen to belong here, especially if we recall the words
meaning “crush; squeeze; pinch” already cited. The eligibility
of pipi, “‘cheek”’, to be added to this group may well be questioned,
though lips, teeth and cheeks are naturally close together. At
first sight it seems impossible to discern any relation between
smallness and cheeks, rather the opposite-—witness Fr. joufflu,
““chubby-cheeked”’, whose ending -u is particularly significant.
Actually this case of the cheek is an absorbing semantic
problem, which we could first attempt to solve by considering
the East. As soon as we recall the usual Japanese word for
“cheek”, hd, or better still the Chinese character for it, A,
we catch a fascinating glimpse of the truth. For as we have already
seen, 3€ “press, squeeze” is found as a phonetic in a number of
characters, among which we have had occasion to quote # and
#&. In Chinese, #1 retains much of the original sense “‘jaw”’, and
here we meet Id. gigi, “tooth”, and gigit, “bite”, or at least we
are very close to them.
It may now occur to the reader that this same relation between
“saw” and “cheek” is not limited to the East. Chaucer refers to
Sampson in “The Monk’s Tale” in the following arresting words:
A thousand men he slow eek with his hond,
And hadde no wepen but an asses cheke.*°

This “cheek” is unmistakably a jawbone, and the original Hebrew

of Geoffrey Chaucer,
10. G. Chaucer, The Monk’s Tale, ll. 2037-8, in The Works
ed. F.N. Robinson (London, 2nd edn, 1957).
50 Comparative Semantics

lehi++ was used for both “jaw” and “‘cheek’’. It can readily be
gathered from this Middle English usage that “cheek’’ itself
originally meant “‘jaw” and that the modern sense is a later
development.
Nor is this in any way limited to Chinese and English. It is
interesting to consider incidentally that both “jaw” and Fr. joue,
“cheek”’, though of obscure origin, certainly appear to be de-
scended from the same parent term. Lat. gena, “‘cheek”’, is the
same word as Gk yévuc, ““cheek’’, whose original sense “‘jaw”’ is
still the only one found in Homer. Besides, from Lat. maxilla,
“Yaw’’, are derived a number of learned borrowings which have
to do with the jaws, but what is far more significant is that maxilla
turns into the usual Spanish term for the cheek, mejilla; maxilla
has also survived in OFr. maisele and in various Walloon patois,
and in all these cases it refers to the cheek and not to the jaw.
This particular point clearly illustrates once again how much
light can be thrown on semantic problems ‘‘au niveau populaire”’
and that learned borrowings are of little use in this respect. What
seemed to be an odd and isolated Indonesian problem again
turns out to be of great significance in understanding not only
Chinese and Japanese, but the true history of our European
languages as well.
There is really nothing surprising in the close association
between “‘jaw” and “‘cheek” both in the East and in the West,
and if we look at the matter a little more closely it is found that
the chin can hardly be separated from these. Behind it all is the
physical fact that there are two jaws and that the cheek is to the
upper jaw very much what the chin is to the lower.
Let us now apply to the chin the same procedure, and move
from West to East for a change. It is at once found that Eng.
“chin” and Ger. Kinn turn out to be the same word as Lat. gena,
“cheek”, and Gk yévuc, “jaw” to “cheek”, already examined,
and from this Greek word is derived yévelov, ‘“‘chin”. However,
the most remarkable piece of evidence is undoubtedly the usual
German term for the jaws, Kinnbacken, since this is a compound
of Kinn, “‘chin’”’, and Backen, “‘cheeks’’.
Again, Jp. ago, “chin” and “jaw”, is generally written with
the character 3, which in Chinese designates both ‘“‘the cheek-
11. Judges xv.
 Malay-Indonesian at

bones” and “the jaws”, and any possible ambiguity is removed

by the addition of a suitable indicator, such as “upper” or
“‘lower’’.
These considerations are limited to the relation between the
chin, the cheeks and the jaws, and therefore exclude the numerous
names of the chin based on the beard, some of which will readily
occur to the reader. The most obvious examples of these are to
be found in Spanish, Rumanian, Greek and various Slavonic
languages.

We shall now consider an entirely different type of semantic

problem. At first sight didik, ‘educate’, seems utterly out of
place here, despite its vocalism, since it has no obvious asso-
ciation with smallness. Yet on closer examination it is found to
belong. If we recall titik, ““drop’’, already discussed, we may soon
discern some light, and these placed beside tetek, “‘female breast”’,
and dada, ‘“‘chest’’, should be quite illuminating. Notice inciden-
tally the suggestive symbolism of the various vowels. The im-
mediate link between titik and didik is further evidenced in firis,
“‘leak’’, beside diris, “‘to water (plants)’’.
It will be realized by now that didik, ‘“‘educate’’, ultimately
goes back to the nursling suckled by its mother, precisely like
our own “education’’, whose earlier sense is clear from Varro’s
famous dictum: “‘Educit obstetrix, educat nutrix, instituit paed-
agogus, docet magister...” This is a notion we are very familiar
with: such expressions as “‘mental diet” hardly attract attention,
while “‘intellectual pabulum’” is more learned. More closely
allied to didik are our alumni, originally ‘“‘nurslings” (Lat. alo,
“‘nourish”’), and alma mater, who is literally ‘‘the mother suckling
her child”. Terms belonging to this family are not always as
learned as alumni. In Portuguese, for example, al/uno seems to be
the most usual word for “pupil”, and is actually used at times
to explain the sense of discipulo. In Old French, nodrir (Lat.
nutrire, “‘feed”’) usually means “bring up, educate’’, and Eginhard
in his Vita Karoli describes Charlemagne as his nutritor.
In French, while élevage refers exclusively to the breeding of
animals, é/éve mostly belongs to education, though it retains
some lesser-known associations with élevage. The normal term
for “education” in Dutch is opvoeding, whose literal sense is
transparently a ‘“‘feeding up”.
52 Comparative Semantics

We discover precisely the same notion underlying the usual

terms for education in Chinese and Japanese ## (Ch. jiaoyu,
Jp. kyoiku), where the second character’s fundamental sense 1s
‘nourish’. Much more could be written on this topic, but this
should be sufficient to show that Id. didik, “‘educate’’, belongs
here.
As in the case of E-I words in Section 8(b), a sprinkling of
examples of I-I terms will be sufficient for further illustration:
nipis or tipis, “thin”; bisik, “whisper”; kitik or kilik or gili,
“tickle”; sisik, ‘fish scales’; rintjih, “‘small bits or thin slices”;
sibir, ‘‘small slices”; tjilik, ‘“‘small’’; silir, “gentle breeze”’; tindik,
“pierce” and fisit, “darn, mend”; djingkit, “hop along”, and
djingdjit, “walk on tiptoe’’; pitjink, “wink”; pirik, “pulverize”;
kikih, “giggle”, which may well be onomatopoeic (cf. Ger.
kichern).
However, the most interesting member of this group is un-
doubtedly dikit, “‘little’”, which occurs with great frequency in
the derived Mal. sadikit and Id. sedikit, ‘‘a little, a bit’, an
adverb of quantity and thus essentially different from kezvjil,
“little, small’’, discussed in Section 8(b), which can better be
looked upon as the corresponding adjective. Sedikit has prac-
tically the same sense as Fr. un peu, while ketjil is similar to petit
in form and sense, and it also possesses other features that
make it well worth devoting a special section to.
(e) Indonesian sedikit (Malay sadikit): what must surely be
the strangest aspect of this word is the prefix se-, ‘“‘one’’. It is
clear that its original form is Mal. sa-, since this is an abbreviation
of the numeral satu, “‘one’’. Let us glance at the numerals from
one upwards in both Malay and Indonesian:
Mal. satu dua tiga empat lima enam tujoh delapan
Id. tujuh
sembilan sa-puloh
se-puluh

Both languages have identical sets of numerals, except for the

minor differences shown. The shortened prefix is usual with
numerical classifiers, as was pointed out in Section 8(c), but its
use is not limited to this.
What is most startling is that the first two numerals are so
remarkably Indo-European. Dua is practically identical with Sk.
 Malay-Indonesian 33

dva, ‘two’. It is true that “one” in Sanskrit is eka and moreover

that sa- is short for satu, but the prefix sa-, ‘‘one’’, is also quite
common in Sanskrit. It can be clearly discerned in Gk &-Katov
(i.e. he-katon, for earlier *se-katon), ““one hundred” and in Lat.
semel, “‘once’’. However, such basic numerals are hardly the
kind of words that would be borrowed, and this must be an
extraordinary coincidence.
At this point I must mention that, on first looking into Malay,
I quickly came across kepala, “‘head’”’ and mata, “‘eyes”’,'? un-
mistakably borrowed from Mod. Gk Kedadr and pati respec-
tively. The p of kepala was only to be expected since Malay has
no f: similarly, our “film” turns into pilem when passing into
Indonesian. As I went on and no further Greek borrowings
turned up, I realized that it was a sheer coincidence. However
this, in addition to the first two numerals, must surely be one of
the most incredible series of coincidences to be found.
Incidentally, Id. mata, “‘eye’’, is known to most of us from
mata hari, “sun” (lit. eye day, i.e. eye of day), and it is even
more curious to find that the usual word for the sun in Viet-
namese is also mat gioi (lit. eye of sky). There is no need to point
out the ultimate identity of “day” and “‘sky’’, equally well known
in the West as in the East, but there is the further arresting fact
that the Vietnamese “eye” is mat, so tantalizingly close to Id.
mata'3 and Modern Greek.
Our present concern, however, is the prefix sa-, and it can
best be examined in Id. saudara, the normal form of address
between people of the same age. The original sense “‘brother or
sister” (cf. Ger. Geschwister), which still survives, shows it to be
unmistakably borrowed from Sk. saudara, i.e. sa-udara, “‘of the
one or same womb” (udara: “uterus”’), exactly parallel to Gk
adedpos, “brother”. Chinese and Japanese have a strikingly
similar compound [Al fi, Ch. tongbao, Jp. doh (lit. same womb),
“brother and sister’, but with the further development “‘com-
patriots”, where the native country is clearly regarded as the
motherland.
Let us now consider briefly the semantic range of the prefix

p. 36.
12. See M.B. Lewis, Teach Yourself Malay (London, 1947),
13. Mata, “eye”, is found also in Tagalog, Maori, and Samoan.
 54 Comparative Semantics

se- in Indonesian. There are four main developments to be

considered:
(i) “‘one’’: this is the basic meaning and has already been suffi-
ciently discussed.
(ii) “‘same’’: the phrase “‘one and the same” gives a good hint
(cf. “pick and choose’’, examined earlier); e.g. Id. se-wmur (lit.
one age), as in “‘she is the same age as I am”’. If instead of a noun
like umur we have an adjective, as in se-ketjil, the sense will be
literally ‘‘as small as’’.
At this point it will be useful to return briefly to Sk. saudara,
“of the one (and same) womb”. The Sanskrit prefix sa-, besides
its identity with Indo-European se-, ‘“‘one”, could well be con-
sidered as a shortened form of Sk. sama, ‘“‘same”’, itself derived
from this prefix.'* It is curious to observe that Id. sama, ‘‘same’’,
and nama, “‘name”’, are remarkably similar to the corresponding
English words. These go back of course to Indo-European, and
Indonesian has simply borrowed the Sanskrit terms sama and
nama. The interesting point to notice is the striking survival of
nasals, a phenomenon readily observable in Chinese where the
word nan, FA “‘south”’, remains intact, not only in Sino-Japanese
but in practically all Chinese dialects as well.
(ili) “all” or “whole”: the phrase ‘“‘one and all” is a useful guide
here; this sense is close to that found in “God is one’”’, i.e. indi-
visible. It is easy to think of ‘‘one” as “whole”, contrasted with
the fractions of unity, or again “whole” (healthy) as opposed
to “defective”’.
This particular use is present in Chinese, for example — [>
yi xin (lit. one mind), “whole-heartedly; like-minded”, and is
especially frequent in Japanese: — ichiza (lit. one sitting), “‘the
whole party, all those present”: —% ikka (lit. one house, i.e.
one family), “the whole family”; — 4 ichinichi (lit. one day),
“the whole day”, similar to Id. se-pandjang hari (lit. one long
day), “‘all day long” etc.
From the sense of ‘‘one’’, ‘‘all”’, a concessive value is but a
short step: se-ketjil, ‘“chowever small”. This semantic develop-
ment is well known in European languages. Du. al, “‘all’’, is

14. The relation between “one” and “same” is still more obvious
in Indonesian,
where “‘as small as” can be expressed not only by se-ketjil but also
by sama
ketjil dengan (lit. same small with).
 Malay-Indonesian BD

widely used by itself for “although”. Fr. tout is also quite usual
as a concessive, nor is it an accident that “although” is found
as an alternative to “though’’. Other parallels will readily occur
to the reader.
(iv) “as soon as’, hence ‘“‘when”: this use may be puzzling at
first sight; in fact, most of those asked to explain a semantic
change from ‘“‘one” to ‘‘when” would probably be utterly at a
loss. Yet the clues are all within our grasp in the West. It does
not take long to discover that “‘at once’ is based on ones, the
old adverbial genitive of “‘one’’, slightly disguised in spelling,
just like ‘“‘hence’’, “‘since” etc. Even without this it is obvious
that “‘once”’ is essentially “‘one’’. If further European evidence
is needed, compare Du. meteen (lit. with one) and Cat. tot d’una
(lit. all of one), both of which are usual for ‘‘at once’”’.
The most immediate parallel, however, is to be found in
Chinese:
fth— [8 aKOrg 45a th «ota yi hui lai ni jiu gao su ta -
As soon as he comes you tell him (lit. he one turn come you then
tell him).
But this could also be expressed in the following way: “When
he comes you tell him at once.”
As to the passage from “‘as soon as” to “‘when’’, this is pre-
cisely parallel to the change from ‘‘at once” to ‘“‘soon’”’, which
occurs in ever so many languages and has been explored in depth
in Chapter I. Let it be sufficient here to point out that OEng.
sona, later “‘soon’’, meant “‘at once’’.
(f) Further developments from sedikit: it has just been
shown that sedikit is made up of se-, ‘“‘one’’, and the basic dikit,
which appears to be used otherwise only with the prefix ber- in ber-
dikit?, i.e. berdikitdikit, ‘‘little by little” or “‘be thrifty”. However,
owing to its great frequency, sedikit is felt to be an indivisible
unit and it forms a new pattern on which a number of other
terms have been built. If we remember the link between form
and sense, it will not be surprising to find that these secondary
formations also belong here.
From selidik, “accurate”, is formed with prefix pe- and suffix
-an the word penjelidikan, ‘“‘research”’, while penjelidik is ‘“‘an
56 Comparative Semantics

investigator”.!5 Selisit, with the prefix me-, means “‘to inves-

tigate’’. It is clear that teliti, “accurate, careful”, also belongs to
this semantic group.
That sedikit is the new basic pattern is evident from the fact
that there is no such form as lidik behind selidik or lisit behind
selisit. Nor is there any possibility of further analysing ‘eliti:
it is clearly a spontaneous creation on the analogy of sedikit.
All these terms can be considered as a sub-group based on the
E-I pattern, which is ultimately linked with the I-I type.
The notion underlying this closely knit semantic group is un-
mistakably that of smallness. The relation between “‘accuracy”’,
“‘research”’ and “‘smallness”’ is something we are already familiar
with: expressions such as Fr. un examen minutieux and Eng. a
“‘minute” examination, close and “minute” scrutiny etc. are
common enough, and “‘minuteness” is a near synonym of “‘ac-
curacy’.
Even more revealing is the Chinese /J\.{) xiaoxin (lit. little
mind), “‘careful’’, while the same characters read by the Japanese
shoshin either mean “prudence” or ‘“‘cowardice” (i.e. not much
heart), though the latter sense is less frequent. Similarly, 4.0
Ch. xixin, Jp. saishin, whose literal meaning is practically that
of the preceding compound, is used in both languages for “‘care-
ful, cautious’. Again, in #£#4 Ch. xiangxi, Jp. shdsai, ‘details’,
the second character “‘fine’”’ helps us to see the cutting process
basic to “detail’’, Fr. détail, derived from tailler, ‘‘to cut’’.
The relation we have just observed in both the East and the
West between “‘small’’ and “‘care”’ is strikingly confirmed by the
common Japanese expression ome ni miru, “‘to overlook”’, which
means literally “‘see with big eyes”. This sense may not be ap-
parent because of all that is lost in the transliteration into our
script, but the characters for 6me, AH “big eyes’, leave no
doubt as to the literal sense of this expression.
If we consider the evidence carefully, the semantic develop-
ment from “‘smallness” to “research’’ is admirably clear. A fur-
ther example should be convincing. The normal term for

15. The internal phonetic change from s to nj after certain prefixes is regular
in Indonesian, and it would be beyond our present scope to examine it here,
since we are essentially concerned with meaning.
 Malay-Indonesian oY

“research” in Chinese and Japanese is ff 3% Ch. vanjiu, Sp.

kenkyii, and here the sense of “‘research’’ comes directly from
the basic notion of ‘“‘grinding to powder”, a meaning that the
first character still has in Modern Chinese. Its radical on the
left is the stone 4 used to crush the substance to be investigated.
It is also striking that it is the character most often used for
writing Jp. togu, “to sharpen, whet”. At this point it should
be recalled that ‘“‘whet” and “grind”’ are practically identical,
and the important word here is “‘practically”’, i.e. “‘in practice’:
Eng. “grind” means both “grind (the grain) to meal” and
“sharpen”, and any lingering doubt should vanish when it is
realized that Fr. meule is equally frequent for both “millstone”
and “grindstone”. The same sense development can be observed
in # Ch. yan just mentioned, since it means “grind; rub fine;
powder” and it is from these meanings that the notion of “‘re-
search’”’ develops quite naturally.
The reader could now profitably refer to my article “An In-
troduction to Comparative Semantics’’,'® where I drew atten-
tion to the fundamental relation between ‘‘flour” and “‘intelli-
gence’, a relation that has been denied by a very learned
Classical scholar. Yet this same scholar was within immediate
reach of the truth when he used the expression “choicest flour”,
whose suggestive force is as striking as “pick and choose”’ etc.
It is important to remember that French gives the key to the
sense development in English from “flower” to ‘flour’, ob-
viously the same word originally, as can readily be observed in
the Chaucerian use of flour, “flower”, a borrowing from OFr.
flour, later fleur (Lat. florem). In Dutch too bloem, cognate with
Eng. “bloom” and Ger. Blume, “‘flower’”’, is commonly used for
‘Hout’?
The other examples I quoted in my article, especially Jp.
konasu (lit. make flour or powder), “reduce to powder’’, hence
“digest”, hence “master (a skill)” can now be more fully appre-
ciated.17 And it is worth while pondering again the other par-

16. Aumla, 29: 12-13.

compound
17. Much more revealing than the simple verb konasu are the various
ly
verbs based upon it; the most commonly used of these is undoubted
yomikonasu, “to master/digest a text”, i.e. read with understanding. (Ed.)
 58 Comparative Semantics

allels I quoted on that occasion, particularly the relation between

Lat. scio and scindo, and the semantic legitimacy of tracing Fr.
trier, “pick and choose’’, and Eng. “‘try”’ back to tritare, “‘grind”’.
The further evidence of Ic. skilja, “part, break up; separate”,
hence “‘discern”’ and ‘“‘understand”’, should now be seen in clearer
perspective.
The exploration of the semantic field of ‘‘flour’’ is one of the
most absorbing and rewarding that could be undertaken, and it
would be hard to find a more suitable and promising starting-
point than the phrase “choicest flour”, which is so rich with
suggestion. We shall have occasion to return to this point later
on.
Our present concern is with the change from ‘‘smallness” to
“accuracy” and “‘research’’, and there is striking evidence of
precisely the same development in Arabic. From a basic daqga,
“be fine, thin’, is derived the sense “make fine’, i.e. “erind”’,
and there is moreover a second derived form dagqaqa, “grind
fine’, hence “‘be exact, accurate”. There are numerous deriva-
tives from the original dagqa, whose three consonants d q q con-
stitute its fundamental semantic basis, while the vowels play
merely a secondary role, as is normal in Semitic languages.
Clearly relevant here is daqiqa, ‘“‘a small particle’, hence “‘a
minute of time’, comparable with Id. menit and Lat. (pars)
minuta, on which the Arabic term is possibly a calque, especially
since Ar. thaniya, ‘“‘a second of time’’, is built on the three con-
sonants thn y, whose primary sense is “two”. However, in view
of our great debt to Arabic civilization in medieval times, the
truth may well be the other way round, and (pars) minuta could
turn out to be a calque on the terms used by Arab mathematicians.
This is a point that could easily be solved by reference to the
original Arabic material.
The most arresting member of this d q q family is certainly
daqiq, “fine, thin”, hence two developments: “accurate” and
“flour”. This is the usual Arabic word for “flour” and it is
illuminating to find it here again, developed spontaneously and
independently from the notion of smallness. There is grist for
our mill in the most unexpected places. Nor is this surprising.
“The choicest flour” or /a fleur (de la farine), as can now be seen,
 Malay-Indonesian 59

is not just English or French, but is what may almost be called

a “natural” semantic change. After all, flour cannot be anything
but choice or fine: this is precisely why it is called “‘flour”’’, i.e.
fleur, “‘flower”. There is no such thing as coarse flour, unless
this be applied to a mixture of flour, meal and bran. A little
reflection will show that Eng. “meal” has acquired a coarser
sense precisely because of the presence of flour, whereas the
German cognate of “‘meal’’, i.e. Mehl, is still ‘flour’? because
there was nothing corresponding to flour for the more refined
product. This shows that it is obviously futile to explore meaning
except within its semantic environment. In my assessment of the
sense of the term “‘flour” as equivalent to ‘“‘choicest flour” I am
of course referring to present-day usage, for it was not so long
ago that flour used to be sifted on the eve of festivals for making
specially fine bread or cakes.
Another member of this fascinating group of d q q terms in
Arabic, the word tadgig, “‘accuracy”, should be carefully ob-
served for it is well worth special attention, particularly for its
form. Hebrew also possesses the same basic dagagq, “‘to beat” or
“srind small’’; i.e. the substance can be pulverized by beating
or crushing, and it is from this notion of “beating” that we find
further Arabic developments from the d q q group, such as daqqa,
“type” (on a typewriter), and dagga, “‘stroke; tick” (of a clock),
which are reminiscent of some of the Indonesian terms we have
examined.
Before we leave Arabic, let us consider ‘‘research’’. One of
the most usual words for this is tangib, whose pattern is identical
with that of tadgig, ‘“‘accuracy’’, just mentioned: there is the
initial ta, a sort of prefix, then the three basic consonants without
vowel between the first and the second, and a long 7 between
the second and third consonants. Arabic is largely built on
patterns that can be analysed almost like mathematical formulae.
Actually an Arabic term may be looked upon as a meeting-point
of two co-ordinates, one corresponding to the three basic
consonants and the other determining the patterns. But while
this description may appear rather abstract and theoretical, the
words themselves are very much alive.'®
18. To illustrate in a practical way the principle underlying those patterns,
consider for example najjar, “carpenter”’, and jazzar, “butcher”. The doubling
60 Comparative Semantics

This preamble is intended to make it clear that tadgqiq, “‘ac-

curacy”, is to the basic dq q exactly what tanqib is to the new
basic n qg b. The simplest word of this family is nagaba, “to bore
through”, hence “‘seek, investigate thoroughly” and “do research
work”’. It is evident that drilling a hole is often the only practic-
able method of exploration, for example in probing deep-lying
geological strata, but it could also simply refer to the initial
approach in attempting to discover the secrets of a certain mass
of matter by perforating it.
What is especially significant here is that Ar. tangib, “‘re-
search’’, throws light on the semantic development of the notion
of research in the Far East. It will be recalled that in discussing
the usual word for “research” in Chinese and Japanese, ff2%,
the first character was carefully examined while the second, 7,
was rather taken for granted.
Now, the reason was that %% is simply found to mean “ex-
amine” in both China and Japan. However, in the light of
Ar. tangib it is possible to discover a clue as to the earlier sense
of this Chinese character. Its radical is 7% which, when used
independently, designates ‘‘a hole’”’ and in Chinese it also means
“make a hole’. It is therefore reasonable to assume that “‘ex-
amine” is a secondary sense which has developed from an
earlier one of “boring a hole’. If we cast a rapid glance over
the most current characters based on this radical, always found
on top and rather squashed like so,“ , there is no doubt that

of the second consonant and the two vowels, short and long a in that
position, is the primary pattern used to designate a person engaged in a
certain activity. Thus, given hamala, “to carry”, “a porter” will be hammai,
and similarly on samak, “‘fish”’, is formed sammdk, “fisherman”; in precisely
the same way, on daqga, “grind”, is built dagqdg (lit. one who grinds), ‘‘a
dealer in flour” or possibly ‘‘a miller’.
Another clearly recognizable Arabic pattern with initial ma-, used to
designate a place, can be readily discerned in the following Indonesian bor-
rowings: madjelis, “‘council” (based on dj | s “‘sit”); makam (q w m “‘stand”,
cf. Ger. Stelle beside stehen and Indo-European -stan as in Pakistan etc.);
mahkamah, “court of justice” (cf. hakim, “judge’”’); madarsah “institution
of learning” (drs “‘teach”); masdjid, “‘mosque” (s dj d), whence our ‘“‘mos-
que”, from Egyptian Arabic, through Spanish, Italian and French. The
Maghreb or Moghreb (gh r b) is literally “the place where the sun SEtS;—1.e.
the Occident, as seen by the Arabs, and so refers to Algeria or Morocco.
 Malay-Indonesian 61

the fundamental sense “‘hole”’ is still clearly discernible in most

of them, even though some secondary sense developments have
gone quite a long way from this basic notion.
It could finally be observed that this Far Eastern word, the
normal one for “‘research’’, is made up of two characters, more
or less synonymous, which represent two aspects of the physical
exploration of matter. This is the most usuai type of Chinese
compound and they are innumerable. The passage from the
physical “grind” and “bore” to the metaphorical ‘“‘search,
investigate” is so “natural’’ and obvious that it needs no further
elaboration. The point worth recalling is that, in this particular
case, it was through Arabic that we were led to a fuller under-
standing of Chinese and Japanese ff 3%
Another obvious method of exploring matter is simply to
cut it up. Our ‘‘anatomy”’ (ana-tom-y) was originally a “‘cutting
up’’, as must be clear to anyone with even a slight knowledge of
Greek. Hence it came to mean “dissection” and it finally de-
veloped its present sense. In a famous letter from his father
Gargantua, Pantagruel is thus urged to discover the truth
about man:
par frequentes anatomies, acquiers toy parfaicte congnoissance de
laultre monde, qui est ’homme.!®

The sense here appears to be literally ‘“‘dissection’’, but it prob-

ably goes further, and Rabelais may well have meant that this
process of first discovering the secrets of the human body could
possibly be the way to probe the mysteries of man’s behaviour.

19. F. Rabelais, Pantagruel, chapter VI{I, in Oeuvres completes (Paris, 1962).

In the same letter Rabelais had just been urging a close study of Greek,
Latin, and Arab physicians: “‘Puis songneusement revisite les livres des
medicins Grecs, Arabes, et Latins. ..”. It is odd to reflect that even a man
as enlightened as Rabelais based his teaching of medicine on a study of
texts written in antiquity. He at least did know Greek and could explain
and discuss the Greek originals, which was more than most of the other
professors of medicine could do. Whether Rabelais practised dissection
himself is not certain, but he deserves infinite credit for recommending it
at a time when it was still strictly forbidden by the Church, a prohibition
that had utterly blocked all progress in medicine right through the Middle
Ages. What little progress was made came mostly from what happened to
dribble through from the Arabs.
62 Comparative Semantics

At any rate, there seems to be little doubt that Shakespeare

went far beyond mere dissection when he wrote:
Lear. Then let them anatomize Regan; see what breeds about her
heart. It there any cause in nature that makes these hard
hearts ??°

At that time the sense of “anatomy” was still very close to that
of ‘‘analysis’’ (also from the Greek), which is literally a resolution
of matter into its original components.
It is of interest to record that the Chinese and Japanese for
“dissection”, #3] Ch. jiepou, Jp. kaibd, the first character
close in sense to ‘‘analysis”’ and the second based on “‘cutting’’,
has also acquired the meaning “‘anatomy’’. This certainly shows
to what extent anatomy depended in the first place on dissection,
unless it happens to be a calque.
In the light of what we found in the Arabic dq q concerning
the relation between “grind” and “‘accuracy”, or if preferred
between “‘smallness” and “‘research’’, or again between “flour”
and “intelligence”, I can hardly do better than to quote a point
I made in the article previously referred to:
Actually the clearest evidence is in Japanese. ‘“‘I understand”’ is here
usually expressed by wakarimashita, lit. “it has broken up”. The
person vanishes as being obvious in the given situation and the tense
is past. Present understanding is the result of past ‘‘analysis”.?!

Let us examine this example more closely. The Chinese character

is 4 fen, “divide; share’’, hence “‘distinguish’’, and as a noun it
designates “‘a share” or “‘a part”, which may be quite a small
one, such as ‘“‘a minute of time’. With the rice radical #} it
acquires the meaning ‘‘flour”’. The character 4 is itself built on
the knife radical /J, so the original sense was clearly “to cut”’. It
may be useful to recall at this point that Eng. “share” belongs
with “‘shear’’, “‘cut’’ and “‘shore’’, and as an incidental but very
relevant exercise the reader could work out the essential semantic
relation between ‘‘Algiers’’, ‘‘a butcher’’, “‘an island” and “a
decree” (all from the same basic Semitic consonants g z r).

20. W. Shakespeare, King Lear, III, vi, in The Complete Works of William
Shakespeare (London, n.d.)
21. Aumla 29: 13.
 Malay-Indonesian 63

All the Chinese meanings we have noted are present in

Japanese, but we find here a further development towards
“‘understanding’’. Special note should be taken of both the past
tense wakarimashita and the construction hinted at in the transla-
tion, since the matter understood is the subject of the Japanese
verb, while in English it would be the object: contrast ‘‘(I)
understand this” with Jp. “‘this has (been) divided, broken up”’.
Perhaps we could best compare with this the French colloquia-
lism compris! which can stand either for j’ai compris or c’est
compris; only the latter corresponds to the Japanese, though
instead of ‘“‘a grasp” as in French, we find here a subdivision
of the substance into its components.
For the past tense there are obvious parallels in the West,
such as Lat. cognovi, memini etc. and especially Sk. veda, “I
know” (lit. I have perceived), Gk oida, “I know” (lit. I have
seen), and the Germanic cognates Eng. “wit”, Ger. wissen,
Gothic wait, all of which mean “know’’, except that Icelandic
has retained in addition some of the original sense, “‘see”’.
These Germanic verbs show the same peculiarity that we have
observed in Sanskrit and Greek: they are past in form but present
in meaning. This explains Eng. “God wot” (and not “wots’),
Ger. er weiss (and not weisst). “Wot” corresponds in form to
the past ‘“‘sang”’, and thus “he wots” would be just as barbarous
as “‘he sangs’’.
The one vital point of difference is that while all these “know”
verbs go back to an original verb of perception, “see, perceive”,
which is the direct ancestor of Lat. video, the Japanese 4 *4
wakaru, “understand”, goes back to an original cutting or
subdividing of the substance, which is still visibly present in
the written character.
Before we finally leave this fascinating aspect of smallness,
we should observe in passing Id. risik, ‘“confidential inquiry”,
and sidik, which usually occurs as menjidik, “investigate”, and
in the phrase menjidik midik, “‘minute scrutiny”. Midik seems
to be found only in this expression, a typical bound form. These
unmistakably belong here, both because of their form and because
of their sense.
Moreover Id. sipi, ‘“be just off the mark’’, can now be seen
to be relevant to our present investigation, and more specifically
 64 Comparative Semantics

to this section. For surely the difference between being just on

or just off the mark is minimal. It is true that Fr. il est a c6té de
la question means “‘he misses the point”, and there is an ex-
pression “‘wide of the mark’’, but in practice this differs only
slightly from “‘just off the mark’’. It is mostly a matter of personal
appreciation: a paper that is a bad failure to one examiner
may have only just failed in the eyes of another and even had a
bare pass in the opinion of a third. Nevertheless there is a vital
difference between pass and failure, as there is between just
missing and just catching a train, and this may well be why
we are sometimes told that ‘‘a miss is as good as a mile’’, an
expression that possibly owes its birth to the alliteration.
(g) Partly aberrant cases, A—-I and U-I: it will be seen
from this heading that the terms to be discussed now are not
fully aberrant but only partly so, since the final vowel is still i
and the final vowel is always the more significant. Let us first
consider a few examples of the type A-I.
Id. sakit is the usual term for “ill” and it is found in a large
number of combinations for different kinds of ailments. What
is curious is that the derived bersakit?, i.e. bersakitsakit, means
not only “be sickly” but also “‘bend every effort”, and on the
other hand pesakitan refers not merely to one who is “sickly”
but to “a prisoner” as well. At first sight there does not seem
to be any clear link between these various meanings; but Eng.
“disease” gives a clue as to the likely developments, for its
original sense “dis-ease” clearly suggests being ‘“‘in a tight spot”,
a point that may be better seen through Fr. étre a I’aise with its
association of ample space, freedom and relaxation. On the other
hand “bend every effort”, like Fr. bander un arc or simply
bander, implies the opposite, ‘“‘tenseness”, and here we are
within close range of “prisoner” and “‘illness’’.
Id. parit, “ditch, trench”, also means “slit, groove’’, and this
could well be the original sense. It is instructive to compare
Fr. fossé, ‘ditch’, fosse, “pit, grave’, and fossette, “dimple”,
but there are even more revealing parallels that could be adduced,
as will be seen a little further on.
Id. pasir, “sand”, also has the sense ‘“‘beach’’, which is likely
to be a secondary development from “sand”, just as Fr. gréve
must be taken together with gravier, a useful point since it
 Malay-Indonesian 65

strikingly illustrates the sharp difference between the attractive

sandy plage and the inhospitable shingly gréve, which is closely
associated with shipwrecks in the popular mind.?” That pasir,
“sand”, is related to smailness is obvious enough, especially if
it is recalled that the corresponding Chinese character # is
literally made up of small fragments of stone® . Id. pesisir,
“‘beach’’, is also based on pasir, but on the sedikit pattern.
The problems encountered so far do not present any difficulty
comparable to Eng. “big” and “small”, which can so easily
be quoted against any attempt to discover whether i is a symbol
of smallness. Yet it is undeniable that this vowel has the same
symbolic value in English as elsewhere. On this point, Jespersen
has marshalied impressive evidence in the article already men-
tioned.
Some of the topics that have been dealt with separately may
be brought into line with those now under discussion. Thus
Id. tepi, “edge”, and bibir, “lip” and “edge”, strongly suggest
that the terms examined in Section 3 (samping, ‘“‘side’’) and in
Section 6 (baris, “‘line”) are secondary semantic developments
and originally belong here.
Similarly, apit, “pinched”, is to be taken with other words
meaning ‘“‘pinch’”’, “squeeze”, discussed in the Section 8 (c)
transition group, and the relation of these with Section 7 (pahit,
“bitter”) has already been pointed out. To these must now be
added tjubit, “pinch’’, and this brings us to the type U-I.

There can be little doubt that Id. butir, “grain’”’, must be placed
beside bidji, “seed”, and ali the more so since, like the latter,
it is used as a numerical classifier for small round objects, though
not nearly as widely: dua butir telur, “two eggs’. Id. bulir,
“ear of corn”, obviously finds its place here as well. Perhaps the
’’,
most interesting member of this group is nuklir, “‘nuclear
group
clearly a borrowing, but a borrowing shaped to fit into the
more
concerned. How exactly it belongs here will be appreciated
fully when it is recalled that Lat. nucleus is literally “the kernel
“nut”),
of a nut” (cf. Fr. noyau, “kernel”, and its relation to noix,
which
and that Id. bidji is not only “‘a seed” but also “a kernel”,
is only to be expected.
22. See, for example, Victor Hugo, Oceano Nox.
66 Comparative Semantics

Id. kulit, ‘‘skin’’, may not at first be seen to be relevant here,

though “‘skin deep” gives a good hint of its real sense. It is true
that kulit also means “‘hide” and that we often speak of ‘“‘a thick
hide’, but this is surely a secondary development. In fact,
kulit ranges all the way from “skin” to “‘hide’’, “leather”, and
on to “the bark of a tree’. However, there is no doubt that
“skin” is the starting-point. I admit that in referring to a man
“under whose rough exterior there is a heart of gold”, a French
writer would quite naturally think of sous cette rugueuse (rude)
écorce, but this is a picturesque expression that no one would
take literally.
Any hesitation on this point will be at once dispelled if the
reader will take the trouble to glance at the sense of the Chinese
characters based on the skin radical % : quite a number of
these have acquired the meaning “chapped, cracked”, with
particular reference to the skin. It may seem odd to interpret
French through Chinese and Chinese through French, but it
is all the more valid as neither the French nor the Chinese know
anything about it. There is a great deal in every language that
the native speaker cannot know, and only a real poet catches
an occasional glimpse of a great truth, as when Shakespeare
saw the mountains as warts upon the face of the earth.?5
As to the fact that “‘skin”’ is the starting-point of the various
senses mentioned, it will be sufficient to observe that K Ch.
pi, Jp. kawa, also ranges from “skin” to ‘“‘bark’ and even to
“fur” in both China and Japan. A little reflection will show that
fur/feathers/scales is actually an extension of the skin/hide,
itself an extension of the flesh, all of which make upaseries of
wrappers to cushion the essential creature concentrated in the
vital organs. Furthermore vegetation is to the earth’s surface
very much what hair is to the skin. What is perhaps curious is
that ordinary people all over the world discovered a number of
these fundamental truths quite independently and in the most
unrelated languages and cultures.
If reference is now made to Section 4 (tulis, “‘write”, lukis,
“paint” and guris, ‘“‘scratch’’), it can be appreciated that these
words are likely to be secondary developments from ““scratch”’,
which was then shown to be the most probable original sense
23. Shakespeare, Hamlet, V, ii.
 Malay-Indonesian 67

both in the East and in the West. Further evidence towards

this interpretation can be found in Id. wkir, ‘“‘carve, engrave”,
and the even more illuminating tutik, “‘a nick, a notch’, and
as an added point of interest tutik also refers to “the woodpecker’’.
It will be recalled that the French woodpecker, /e pic, was men-
tioned in connection with Eng. “‘pick-peck”’.
Similarly, if we keep in mind what has already been discussed
on the fundamental relation between research, investigation
and smallness, and if we also take into account the point made
in the last paragraph on “write” and “‘scratch’’, it will come as
no surprise that here again we find U-I terms that unmistakably
belong to this semantic family. Thus Id. ulik, “investigate”,
which needs no further comment, and another term that is
used with great frequency, udji, “test, experiment’’, from which
is derived udjidan, ‘‘an examination” (at school).
The relevance of Id. bukit, “hill”, might well be questioned,
as it seems to bear little or no relation to smallness, unless it be
contrasted with a mountain. Yet if we recall that Id. parit ranges
from a mere “‘slit” to a “ditch”, and that the latter meaning
has now become by far the more usual one, we find in this a
hint of what may have happened to bukit. It is noteworthy that
there is a similar distance between the little “‘molehill” and the
normal “hill” in English. A close parallel can also be found in
It. poggio, OFr. pui (later puy, surviving in Puy-de-Déme), and
Cat. puig, pronounced [put], a term especially common in the
Eastern Pyrenees, as can be gathered from a map; all of these
are from Lat. podium, whose sense is to a real hill very much what
a molehill actually is. An additional example that can throw
more light on this phenomenon is Fr. fertre, which designates
a small mound such as marks the presence of a grave ina cemetery.
Cognate with it are a number of regional Belgian terms like
tiene at Nivelles or tiér at Liége, which refer to a steep rise. The
equivalent tertre, “mound”’, also suggests that Eng. “‘mound”’,
of obscure origin, could well be a variant of “mount”, despite
the semantic distance between them. After all, the Old French
tertre was considerably larger than the present-day tertre, as is
clearly shown in the Chanson de Roland, where it designates a
rise comparable to our present “‘hill’’ or colline.
9. Let us now consider an entirely different type of vowel
68 Comparative Semantics

system, E-A in native Malay-Indonesian words, to try to

discover whether the remarkable similarities in thought and
culture that have been observed so far between East and West
will obtain. A useful starting-point is the word most commonly
used for “near, close”, Id. dekat, a term to which /Jekat is obviously
closely related, especially in its me- derivative, “attached, in-
timate, devoted”’.
From “near” to “intimate”? needs no elaboration, but it is
instructive to compare Ar. garuba, ‘“‘be near’, for on these three
consonants grb are based among others garib, “near” and
“relative”, tagriban, “about”, and dn garibin, “shortly”, and
exact parallels occur in Hebrew. Ar. garib is found in the In-
donesian borrowing karib, “relative, close, intimate’, and from
this to “‘close friendship” is but a short step. Friends, as was
pointed out in Section 1, are naturally close. There are “distant
relatives” but not ‘“‘distant friends’’, for the obvious reason that
distant relatives are still relatives, while distant friends, unless
their distance were merely geographical, would cease to be
friends altogether. A further striking example of the intimate
relation between closeness and friendship is to be found in the
characteristic Arabic comparative of garib, “‘near’’, i.e. agrab,
“nearer”, which was itself borrowed in the form of Id. akrab,
“close friend”. Of course, Indonesians do not feel this to be a
comparative at all, any more than akbar, “‘great” beside kabir,
“mighty”, both also borrowed from Arabic k br, where the
term akbar means “‘greater’’ or “‘greatest’’.
Another semantic development based on the same vowels
is found in Id. tepat, “‘exact’’, and selat, ‘“‘narrows’’, i.e. the
geographical strait(s). This sense relation is so clear that it
can almost be taken for granted. However, it may be worth
pointing out that the Romance names for the strait all go back
to Lat. strictus, “tight, close”: It. stretto, Sp. estrecho, Pt.
estreito, Ofr. estreit (whence is borrowed Eng. “‘strait’’), later
étroit, then détroit. It is of some interest to observe that the
present expression “straight and narrow” was originally “‘strait
and narrow’’, i.e. “‘strait” in the sense of “narrow”, a synonymic
couplet formed to prevent confusion with the other “straight”,
which had become homonymous with it. From Lat. angustus,
“narrow”, is derived angustiae, “‘a narrow passage’’, hence the
 Malay-Indonesian 69

notion of being “‘in a tight spot’’, with all the resulting “‘per-
plexity, poverty” etc. Lat. angustia turns into Fr. angoisse,
which in Old French had remarkably close associations with
destreit (Lat. districtus), “mountain pass’’, the later Fr. détroit,
which in the sixteenth century supplanted the earlier estreit,
i.e. étroit as the name of the geographical strait.
In Germanic languages there is an even more illuminating
example of the relation between “exact”? and “narrow” (.e.
Id. tepat and selat, just quoted). Du. nauw, the normal term for
‘narrow’, is the same word as Ger. ge-nau, “‘accurate”’, and is
used for the Straits of Dover. It probably represents a calque
on Fr. Pas de Calais, where pas was the mountain pass still
found in disguised form in il est dans un mauvais pas, “he is in a
tight spot’; the Dutch also say quite naturally hij is in ’t nauw,
and it is thus possible that all these phrases contain a lurking
memory of Roncevaux and the Thermopyles. Haven’t we all
had that feeling when passing through a narrow street, and
didn’t we generally hurry along, whereas we would take it much
more easily in a wider street? This explains why the Dutch and
the Germans use the cognate of “street” for “the strait”: Ger.
die Strasse von Calais, Du. de straat van Gibraltar, and it also
strengthens the probability of a French calque in the Dutch name
for the Straits of Dover.
It is of more than incidental interest to note that from Id.
selat, ‘“‘strait’’, is derived Id. selatan, “‘the south’. Originally this
may well have been the Straits of Malacca seen from further
north. But whatever the starting-point, “‘the strait’” was in the
south. For a similar phenomenon we could compare the Chinese
character for the east #{ , the sun H seen through the trees X.
It is idle to speculate as to the precise location on which this is
based, but fundamentally this development is not unlike that
of Id. selatan, i.e. it is true of a particular geographical locality
and was later extended, whether geographically valid or not.
It has thus become an abstraction, completely severed from
its original sense, and this brings up another fascinating topic:
to what extent should words be motivated?
Some linguists, like Dauzat, consider French to be superior
to German because it is “‘une langue abstraite”, whereas German
is a strongly motivated language, with words closely linked to
70 Comparative Semantics

the original “‘roots’’. The fact is that the study of Germanic

languages other than English can be an exhilarating experience
precisely because of this quality, whereas French and even
Latin were already largely abstract even at the earliest times.
Against Dauzat’s view it could well be pointed out that phi-
losophy flourished in German and Greek, two languages that
happen to be far more motivated than French or Latin. Indeed
it might well be argued that philosophy is not likely to be viable
unless it be built on a physical basis, i.e. that the abstract is
sound only in so far as it is based on the non-abstract. I have
a feeling that the philosophers thought first in terms of images
and that the abstractions came later, just as in general facts
are observed first and rules elicited later. However, I leave this
to philosophers to debate among themselves.
The notion of pressure is dominant in Id. tekan, from which
are derived terms meaning “press; oppress; stress”’, all of which
can readily be seen to have a common semantic core. To these
may be added peras, whose me- derivative has the sense of “‘press,
squeeze’”’. How closely this links up with what has just been
mentioned will be easily appreciated if we recall the Chinese 3%
“press, squeeze” and %& Jp. semai, ‘“‘narrow’’, with its dog radical
on the left. If instead of the dog we have the mountain radical,
this gives !& xia, ‘‘a mountain pass’’, which, together with ¥
hai, “sea”, i.e. Ch. haixia, Jp. kaiky6, is the Far Eastern geogra-
phical strait, whereas with the earth sb di, i.e. Ch. dixia, Jp.
chikyo, it refers to an isthmus. After all, a strait is squeezed
between two land masses or often two capes, and is to water
precisely what an isthmus is to land. What is especially worthy
of note is that the Chinese strait has had the same semantic
history as Fr. détroit, from ‘“‘a mountain pass” (still surviving
in the Pas de Calais) to “‘a strait’’.
Immediately comparable to the previous notion is Id. desat,
whose ber- derivative has the sense of “crowding”, and with
me- the meaning extends from “push; urge, press” to “urgent”.
These developments are well worthy of a separate chapter in
their own right. European and Eastern equivalents are extra-
ordinarily numerous, and it is difficult to pick and choose among
so many claimants, since they nearly all seem to have special
relevance to the topic under discussion.
 Malay-Indonesian nt

Thus if we take “‘crowd” and Fr. foule as a basis, they both go

back to the notion of pressure, crushing: in English, “press”
was usual enough in this sense, but now the common term is
“crowd’’, together with ‘“‘crush” as a more forcible alternative;
similarly in French, Ja presse is now rather old-fashioned and
has been largely supplanted by /a foule, whose original crushing
sense is still present in fouler aux pieds, “trample under foot’’.
Eng. ‘“‘throng” appears to be isolated, but it also exists in
Ic. préng, “crowd”, a shortened form of the usual Icelandic
term for ‘‘narrow’’, Dréngur, and it can now readily be seen that
German possesses a host of their cognates based on dringen,
“press, urge’’, especially the participial dringend, “‘urgent’, itself
so remarkably similar to “urgent” in both form and sense that
it could easily be a calque, and which is most noteworthy in
such expressions as dringende Geschdfte, ‘urgent business”, or
dringend notig, ‘urgently needed’’; there is further the corre-
sponding transitive drdngen, whose neutral use is particularly
relevant here: es drdngt die Zeit with its strong sense of impelling
urgency, and finally there is the derived Gedrdnge, next to Menge,
the most widely used German term for “crowd’’, whose prefix
ge- makes it especially striking and apt, and even more suggestive
of pressure than Eng. “crush”’.
At this point special mention must be made of a Japanese
derivative from semai, “‘narrow’’, the verb semaru, written, not
as might be expected, like semai 3 '1, but with a totally different
character, 34 4 . The difference in characters is due to the peculiar
history of Japanese writing, borrowed wholly as it was from
the Chinese and with little regard for the true nature of the
Japanese language, the strange result being that the Chinese
characters often obscure the precise relations between the native
terms, though they do so in an extraordinarily beautiful way.
One quaint and rather unexpected result is that there is as yet
no etymological dictionary of Japanese; this, however, is by the
way, for our main concern is with meaning.
What makes semaru especially worth noticing is that from
the literal sense, “be near, close”, it extends to ““be imminent”
(i.e. near in time) and further to “‘be distressed”, and this par-
ticular distress is frequently due to poverty. There are many ways
of being in a tight spot, but whether you are in the West or in
 72 Comparative Semantics

the East makes little difference fundamentally, as will now be

evident. This may be the right time and place to remind the
reader that Eng. “‘anxious, anxiety” are either borrowed from
French or directly from Lat. anxius, itself from the same base
as angustus, ‘narrow’. A further illuminating comparison can
be made with Ger. eng, “narrow”: the corresponding noun
die Enge, besides its literal physical sense “narrowness”, frequent-
ly comes to mean “difficulty” or “dilemma”. Similarly, Ger.
bang, ‘“‘afraid”’, will now be easily seen as a compound of eng
with the prefix be-, and its relation to Angst should be obvious.
These are just a few examples picked almost at random from
ever so many, but at least they should make it clear that there is
little essential difference between the Eastern and the Western
modes of thought. There is very little room left for the famous
inscrutability of the East. Whether you are in a tight spot or
just in a bit of a spot, you are in a spot anyway and it makes
no difference whether this spot be in the East or in the West.
In fact, perhaps the best way to become more deeply aware of
the significance of Roncevaux and particularly that it is funda-
mentally Roncesvals, i.e. Ronces-vals, namely a valley, is
to discover that the usual Chinese character for a valley # has
developed in both China and Japan the sense of being hemmed
in: it is precisely the kind of place where either advance or
retreat often becomes impossible, and one is stuck in it.
We could acquire a clearer conception of this notion of being
in a spot and related semantic phenomena by casting a brief
glance at Japanese, and here I don’t mean Japanese influenced
by Chinese, i.e. by a learned borrowing, but purely Japanese
terms. There are a number of compound verbs such as norikomu,
He OIA, “get in” (a train, car etc.), or nomikomu, FAAD
“take in, swallow” (food) etc., all of which end in -komu, whose
general sense is motion into a place; -komu also exists as an
independent verb, komu, iAts “‘be crowded, jammed”. From
it are derived the transitive komeru, 3A 4 “put in’, and the
intransitives komoru, #&% 4 ‘“‘be confined to a place’’, and komaru
Al 4 , this latter being the most usual term to express the notion
of “being distressed; in straitened circumstances; in a fix or
a dilemma’, i.e. some of the most characteristic results we
have already observed that come from being in a spot; this is a
 Malay-Indonesian 73

linguistic or if you prefer a semantic proof that it is no more

comfcrtable to be in a tight spot in the East than in the West—
bustle as much as you like, you are stuck fast anyway.
It is curious to observe that even members of such a clearly
recognizable semantic family as komu, komeru, komoru and
komaru are all scattered, to the extent that they hardly know
one another, by the use of utterly unrelated Chinese characters
which obscure the essential association of these native Japanese
terms, with the result that we have already commented on. There
is no doubt that the Japanese would understand their own lan-
guage far better if the accidents of history had first brought
them into contact with a people possessing an alphabetical
system of writing, instead of the Chinese who have brought
them beauty at the expense of clarity. Chinese characters are
extremely valuable to the linguist, and in many cases very illu-
minating, but the point made here is that they were about as
unsuitable for writing Japanese as anything that could possibly
be found, considering how utterly different from Chinese Japan-
ese is in every conceivable way. In fact, Chinese is infinitely
closer to English than to Japanese, and it would be hard to
imagine a language more demonstrably unlike Japanese than
Chinese. One of the most remarkable aspects of the whole
history of writing is that the Japanese could have succeeded in
adapting a system of writing so totally unsuited to their needs
as that of the Chinese. However, the gain is undoubtedly ours
as linguists, and this is why we can learn so much from a study
of Chinese and Japanese if we would discover the fundamental
truths deeply imbedded in many of our Western semantic prob-
lems.
Perhaps we could pause a moment to consider the suitability
of Eng. “jammed” in this semantic context. Although it may
appear incredible at first sight, a traffic jam is very much like
the jam we put on our bread. In both cases there is the funda-
mental sense of particles coming closer together until they be-
come stuck. It is not so long ago?‘ that a friend and colleague
of mine, who possesses an exceptionally keen sense of language,
described a traffic jam as a “coagulation”, and this immediately
‘“‘a
reminded us both of Ger. gerinnen, “clot”, which is literally
24. This was written in May 1972. (Ed.)
 74 Comparative Semantics

running together’. This opens up another fascinating semantic

field, but we must resist the temptation, however great, to make
a further exploration in this direction. At least this brief apparent
digression is undoubtedly relevant, since it shows a new aspect
of the notion of being stuck fast in a (tight) spot.
It is useful to reflect that “‘distress’’, or rather its immediate
source Fr. détresse, referred originally to ‘‘a narrow place”, and
that this “‘distress”’ is fundamentally the same as “stresses and
strains”, and our old friend “strait”. A comparison with Fr.
étreindre, étreinte, and also with étroit, and their ancestors Lat.
stringere, strictus throws immediate light on those relations.
A point that deserves special attention is the notion of
urgency and imminence, so noticeable in Jp. semaru, derived as
we have seen from semai, “narrow”. A further point of striking
relevance here is that in a tight spot there is little space, which
means little time, hence the development towards desperate haste
and simply “haste”. This is how Fr. pressé, “in a hurry’’, is derived
from presser, “press”, and prés, ‘‘near’’. These relations are
obviously neither accidental nor coincidental; in fact, innumer-
able parallels will be found to confirm them in many languages.
Even an examination of such common English expressions as
“pressing on” or “closing in on” will reveal much of their
original contents, which are clearly related to our present topic.
If any doubt remains as to these fundamental truths, it will
be sufficient to glance briefly at the opposite concept of ample
space, ease, and the resulting slowness: il travaille toujours tout
da son aise means that he takes it very easily in his work, that
he takes his tirne, that he is slow. Especially significant is the
usual Spanish term for “slow”, despacio, with its basic sense of
space, and the underlying notion that where there is space there
is time, and where there is time there is no hurry; all this neatly
packed up in one word. A fruitful comparison can be made here
with Lat. spatior, “walk about’, It. spaziare, “‘wander about’’,
and in particular the various combinations found with Ger.
spazieren, such as spazieren-gehen, -fahren, -reiten etc., in all of
which the visible space stands for leisure; these words are clearly
motivated, whereas the corresponding Fr. se promener is utterly
abstract. Similarly, in both Chinese and Japanese, “going for a
stroll” is #4, where the second character has to do with
 Malay-Indonesian vs)

walking while the first means “scatter, disperse’, a notion that

again suggests that time and space are ample. These are but a
few examples chosen among many.
It is important to observe that there is only an apparent con-
tradiction between hurrying and being stuck fast, and this is
how Eng. “fast”? has come to develop both senses—“‘stuck”’,
hence “‘firm’’, and on the other hand “‘quick”’.
In view of all these considerations, it should come as no
surprise to find in Malay-Indonesian a series of E-A terms
related to “‘fast” (“‘stuck”’, hence “‘firm’’), such as erat, “tight;
firm’”’; tegat, “‘stiff; stubborn’’; tegap, “firm; well-built”; tetap,
“fixed, permanent”; kepal, “clench tight; fist; lump” etc., and
a further series comparable to “fast’’ (“quick’’), for example
pesat, ‘“‘quick’’, and deras, ‘“‘speedy”, as well as a number of
others worth examining more closely because of interesting
developments.
Among these tjepat, ‘fast, quick’, has come to mean “early”,
and similarly /ekas, from a basic “‘quick’’, has developed the
senses ‘‘soon” and “early”. Here again we are very close to the
semantic field of “‘at once’, since it is well known that “(you)
go quick” is often used as the equivalent of “you go at once,
straightaway”. Actually if we look at the words used for “early”
we find that many of them go back to “quick”. The very first
word we met in a semantic study of this concept?> was Pt. cedo,
which is obviously Lat. citus, “quick”. Other examples occur
everywhere: Fr. 16, ‘“early’’, goes back to an earlier tost, “quick”.
A corfirmation of the same relation is found in the opposite
notion from “slow” to “late”. It will be sufficient to remind
readers that Lat. tardus, “‘slow’’, turned into Fr. tard, “late”.
Similarly, Id. Jambat, “‘slow’’, also means “‘late’”’, and beside Jp.
hayai, “quick” and “‘early’’, can be quoted Jp. osoi, “slow” and
“ate”; this is equally true of Chinese. You just can’t escape
from the evidence, however painful to your pride, that Orientals
think in precisely the same way we do.
Further absorbing semantic extensions are to be observed in
Id. hemat, “thrifty”, which can be compared with our “being

25. This is a reference to the first topic dealt with in the 1967 series of lectures
on Comparative Semantics. (Ed.)
76 Comparative Semantics

tight’, in its literal as well as its figurative senses. Again, Id.

lebat, ‘‘packed, thick, dense”’, is often used of “heavy rain”,
while tebal, “thick, dense”’, is more readily applied to fog. More-
over berat, “heavy; grievous; serious; dull”, and keras, “hard;
taut, tight; strong; stiff, firm’—especially noteworthy in the
combination sakit keras, ‘“‘gravely ill’’—are reminiscent of
numerous Western semantic developments which the reader
will enjoy discovering for himself.
Of particular interest is Id. telah, “already”, whose sense is
very relevant here, especially if we recall the concessive use of
Ger. schon beside its temporal meaning “already”. It is also
helpful to bear in mind that schon is the adverbial form of
schon, “beautiful”, a fact that throws much light on a number
of French semantic enigmas. Jp. mohaya, “‘already’’, is derived
from hayai, ‘“‘quick”’ and “‘early’’, just mentioned, a further point
which helps to see that it does belong here, not merely because
of its form but by reason of its sense. Moreover setelah (i.e. se-
telah), ‘‘after”’, should enlighten us as to the essential relation
between Fr. prés and apres, and the use of Fr. prochain and Eng.
“next”, “nearest”, in referring to an immediate future. From
“near” to “after” is certainly worth pondering, particularly
by those French scholars who are so ready to accept these words
at their face value.
Id. kerap, which from ‘‘close” has developed the sense
“often”, is also of considerable interest. In fact, most of our
terms for “‘often” go back to an original ‘“‘thick”. Even Lat.
saepe is from an earlier “‘packed together’’. It. spesso is from Lat.
spissus, “thick, dense’’, and Rum. adesea (i.e. a-des-ea) is derived
from des, Lat. densus. Du. dikwijls is literally “‘thick-while-s”’,
with the adverbial -s found in such words as “‘once”’, which was
an earlier ones; “‘hence’’, “‘since” etc., where the final -s is
similarly disguised. Even the Slavonic and Balto-Slavic terms for
“‘often”’ have practically all come down from an original “thick”
in the sense of “dense”’.
Id. tjegah, whose me- derived form means “‘restrain, prevent”’,
is clearly very close in sense to “tight”, a meaning central to
our inquiry. Id. semat, “pin, peg, hook’’, is also used like Ger.
stecken of something inserted or enclosed between two objects.
Compare such expressions as “‘he can’t get off the hook” and
 Malay-Indonesian FH

our “stuck” from “‘stick’’, the cognate of stecken. That Id. sepah,
whose ber- derivative means “‘cluttered up, littered about’’,
belongs here too will be evident from a comparison with Fr.
encombrer, “‘clutter up’, and encombre, a bound form occurring
in sans encombre, the traditional phrase used to announce arrival
at one’s destination safe and sound, an echo of the time when
travelling was still very dangerous and one could be stuck in a
number of ways—by bad roads, being waylaid etc. It is note-
worthy that Fr. combre passes into Ger. Kummer, “‘sorrow’’,
and that another close parallel is OFr. pesance, “grief” (lit.
heaviness, as Chaucer would often put it). Eng. “‘grief”’ is itself
from OFr. grief, “heavy”, by a similar sense development. We
may also compare the many expressions of the type “‘it weighed
heavily on his mind”’ or again “‘it is a great weight off my mind”,
whose equivalents are readily to be found in most languages.
This clearly comes from vertical pressure, while lateral pres-
sure leads to distress, as we have observed repeatedly in the
course of this discussion. Hence the close relation between
“grief”? and “distress”, both of which result originally from
pressures, in the first case “crushing from above’’ and in the
second “‘squeezing from the sides’.
There is another extension of meaning from the basic “‘near”’
and “‘press”’ in Id. tjetak, whose me- derived form is the normal
term for “to print” and “to make coins’, two processes that
on further reflection will be discerned to be essentially identical.
Pressure, if it is to be very powerful as in the case of producing
coins, is often the result of a blow. Here lies the key to the
ultimate identity of Eng. ‘‘coin” and Fr. coin, “corner; wedge”’,
and cogner, “strike violently’’. The French phrase battre monnaie,
which used to be the normal way of expressing the sense of
“coining money”’, contains a significant battre. Parallels in other
languages are not hard to find. To some extent a phrase such as
“cuneiform inscription” provides a link between “‘making coins”
and “printing”, and the Modern Greek word for “print”,
tunmva, is derived from témocg (hence our “type’’), which
ultimately goes back to tata, “strike, beat’”’.
Usually, however, the notion of printing is designated by a
simple variant of ‘“‘pressing”’, a much gentler process than that
required for stamping an effigy on metal in order to produce
78 Comparative Semantics

coins. Thus Fr. imprimer and other Romance terms are from
Lat. imprimo (i.e. im-premo), “‘press on’’, and it is the corre-
sponding noun empreinte that passes into English as “imprint”
and the shortened ‘“‘print”. It is worth mentioning that printing
was called in Renaissance French Jes impressions, which was
later supplanted by l’imprimerie, but the original sense is still
clearly discernible. Incidentally, the Classical Greek meaning of
tTbTOG Was “impression’”’.
A close parallel is found in Germanic languages other than
English; for example in German, drucken, “print”, derived from
Druck, “pressure”, stands side by side with dricken, “‘press”’.
The Russian term for “printing” is derived from neyamo, “a
seal’, a word that has passed into German as das Petschaft,
thereby attracting a great deal of grammatical attention because
of the apparent conflict between the gender and the feminine
ending. That nevamp originally had to do with pressure is further
strongly suggested by the related neyaio, “sorrow”, which can
now be compared with Ger. Kummer just mentioned, and other
members of the same family.
When reference was made earlier to stamping as the process
of making coins, it may also have occurred to the reader that
we have here another term used equally well of printing, as in
It. stampa, and it is obviously the same word that is found in
Fr. estampe, “‘a print, an engraving”, and in Eng. “stamp”,
whether it be rubber or postage. This group of words is unmis-
takably of Germanic origin and identical with Ger. stampfen
and Eng. “stamp” (one’s feet); this is closely comparable to
Fr. fouler aux pieds, ‘trample under foot”, which has already
been mentioned in this E—A section in its basic relation to foule,
‘“‘crowd’’. The seal, which may be considered to some extent as
the forerunner if not the immediate prompter of the newly
discovered process of printing, is not only the basis for the
Russian terms for “printing” but also appears in Sp. sello,
whose meanings “‘seal; rubber stamp; postage stamp”’ are very
significant in view of present considerations.
Let us finally consider another member of this Indonesian
semantic family, gelap, ““dark”, whose presence here may at
first appear rather puzzling. There does not seem to be any
discernible link between “near” and “‘dark”’. However, we can
 Malay-Indonesian 79

obtain a hint of their relationship from Id. tjelah, “‘gap”, and

especially a particular use of this word in the sense of ‘“‘threading
one’s way” (through a crowd). If we further recall what has
already been said about traffic jams and dense fogs, two notions
that fundamentally belong to this family, the solution cannot
be far away. If a large number of units move closer together
there takes place a gradual thickening, a growing denser and
denser, and ‘“‘dark”* or “black” will be the resulting impression
if it be expressed in colour terms. At the other end of the scale,
where units are scattered far and wide, there is “‘light’’, and
Fr. clairsemer is a particularly significant example of this phe-
nomenon.
A novel aspect of this image, with a striking figurative exten-
sion, occurs in the following example:
Women workers are thick on the ground and thin at the top...*°

Further on in the same article, those rare women who manage

to reach the top are said to be ‘“‘few and far between’.
Perhaps we may see more easily the relation between “‘near”
and ‘‘dark” if we consider Ger. dicht, the cognate of “tight’’: it
is quite natural to refer to ein dichter Wald, “‘a dense wood”’, just
as dicht is used of foliage or fog. But there is another sense of
dicht that should be considered beside this: dicht am Rande,
“close to the edge”, or again dicht bei ihm, “near him’’. The fact
that Eng. “disclose” will often correspond fairly well to Ger.
erklaren, Fr. éclaircir, or to “elucidate” throws additional light
on this. Again, since a wood is normally thought of as “dark”,
it becomes quite understandable that in Aucassin et Nicolette
the fairy is so beautiful that the whole wood is brightened by
her presence: tos cis bos en esclarci.?”
Although I cannot lay my hands on the relevant passages, I
have a distinct recollection of coming across a number of in-
stances where the sky was darkened by an extraordinarily large
quantity of arrows. However, as if to make up for this, I happened
recently?® to light upon the following passage:

26. Australian, 30 January 1970.

27. Anon., Aucassin et Nicolette (Paris, 1965), XXH, 1. Sill:
28. This was written in May 1972. (Ed.)
80 Comparative Semantics

Tanti di qua, tanti di la fur mossi

E sassi e dardi, ch’oscuronne il cielo.*?

The picture here is unmistakable: rocks and spears were hurled

in such profusion from both sides, by Christians and Infidels
alike, that the sky was darkened as a result. Observe too the -ne
in oscuronne, which has exactly the same value as en in the
French quotation just given: ‘“‘as a result’ or “thereby”’.
It is actually in Portuguese that we find the semantic relation
under study in its purest form: perto, ““near’’, is beyond doubt
the same word as preto, “black”, and there is further confirma-
tion in the Spanish cognate prieto, “tight” and “blackish”,
though the latter sense is obsolescent or dialectal. Moreover
the corresponding Spanish terms apretar, ‘“‘to press’, apreton,
“anxiety”, apretura, ‘a narrow passage’, and aprieto, “‘em-
barrassment; oppression’’, are seen to be very relevant.
It is quite natural for the French to say Ja place était noire de
monde, and comparable expressions can also be found in German
and English. In Dutch too there are remarkably close parallels,
as for example de markt zag zwart van de menigte, lit. the market
looked black with the crowd.
Exactly the same relation can be observed in Rus. moma, a
member of a widespread Indo-European family, whose sense
ranges from its basic “darkness” to ‘“‘an immense number’’, and
moreover in Japanese there is the compound kuroyama, lit.
black mountain, where yama, |l| ““mountain’’, refers to a large
number of people jammed tight to the point of blackness (Kuro
2). This is a very common word, and there is a quaint naivety
in the Japanese “‘scholar’’ who tried to explain away this little
problem by asserting that the blackness was due to the colour
of Japanese people’s hair. One point common to all the examples
I have met is that these expressions always refer to people,
except in Portuguese where the meaning has become far more
abstract, since it is “black” in a general sense.
It is possible to discover the same essential relation between
“near” and “black” in another way: if we set side by side “‘in
broad daylight” and its French opposite d /a nuit close, “‘after
dark”’, it is evident that ‘“‘close’”’ is associated with ‘‘darkness”’

29. T. Tasso, Gerusalemme liberata (Verona, 1956), Canto XI, sect. 48, Il. 1-2.
 Malay-Indonesian 8]

precisely as “broad” is with “light”. And this in turn leads to

an insight into the relation between “open” and “light” as
opposed to “shut” and “‘dark”’. To some extent, no doubt, we
feel that there exists such a relation, and it even seems almost
“natural”. A forest is looked upon as closed in contrast to the
open fields, and we have already noted that the forest is seen as
being dark. There are a number of other points that will readily
occur to the reader.
The most interesting of these are undoubtedly the more
specific linguistic ones, for they are also more convincing. There
is the fact that in Rumanian, to distinguish between light and
dark colours (as in light blue and dark blue), the corresponding
terms are deschis, lit. open, for “‘light’’, and inchis, lit. shut, for
“dark”. There is moreover the obvious fact that “wide open”
forms the usual contrast with “tightly shut”, and we have
observed that “tight is very close to “dense” and “narrow”
and to “close” as well; a further point is that “‘air-tight”” means
“hermetically sealed”’.
To the Dutch, the identity between ‘“‘shut” and ‘“‘near” must
be evident, since the Dutch cognate of “‘tight’’, i.e. dicht, is
constantly used in both senses: dicht is the most widely used
intensive to accompany bij, “near”, in such combinations as
dichtbij, dichterbij, “nearer” etc., which correspond exactly to
“close by’’; furthermore, dicht is very frequently used for “shut”,
in almost precisely the same way as the Germans use the particle
zu-: Du. dichtmaken and Ger. zumachen, “‘shut’”’, are identical
in parts and as a whole. Compare also Du. hij trapte de deur
dicht, “she kicked the door shut”, or de gordijnen waren dicht-
geschoven, “the curtains were drawn (closed)”, which corre-
sponds to Fr. Jes rideaux étaient tirés. It will be noted that in the
last example, while Dutch specifies dicht, ““closed’’, this is 1m-
plicit in Fr. tiré and Eng. “drawn”.
However, there was no need to go so far as Dutch, for the
best solution to the relation between ‘“‘near” and “shut” is to
be found in English, since ‘“‘close” (‘‘near’’) and “close” (“‘shut’’)
are one and the same word. Eng. “‘close” is a borrowing from
Fr. clos, Lat. clausus, the participle of claudere, “‘shut’’, and the
sense “near” has therefore developed out of “shut”.
i0. As we proceed with oui exploration of vowel symbolism in
82 Comparative Semantics

Malay-Indonesian, its most characteristic features are gradually

emerging more clearly, and the semantic value of the group we
are about to consider could almost be anticipated from the
nature of the vowels involved, A—A.
The vowel a is the most open of all vowels and it could thus
be assumed a priori that the sense of this group will have to do
with openness: the remarkable point is that this is the very sense
that strikes us from the outset. It will be remembered that the
final vowel is especially significant, and here again as elsewhere
this principle holds. The most frequent ending is found to be
-ang.
It will be observed that we are here in a field which is prac-
tically the opposite of that dealt with in the last section. Whereas
in the E~A group we examined concepts such as being stuck in
a tight spot, urgency, anxiety etc., we are here concerned essen-
tially with the notions of space and leisure. Notice too that the
word “‘relief’’, which also clearly belongs here, ultimately goes
back through French to Lat. /evis, “‘light’’, as opposed to gravis,
“heavy”, and implies an absence of vertical pressure.
Id. /apang, “open, wide, spacious’’, is typical of this new
group, and from these basic meanings it acquires the further
figurative senses “‘free; unoccupied”; the derived kelapangan
(i.e. ke-lapang-an) means “space; leisure”, and moreover “‘re-
lieved”. There is also the derivative /apangan, ‘‘a field’, and
lapangan rumpat (lit. field of grass) is “‘a lawn’’.
Id. padang, ‘‘a plain’, obviously belongs here too. It is com-
monly used in referring to “‘a playing field”, and padang rumpat,
“a meadow”, should be compared with lapangan rumpat, “a
lawn’”’, just mentioned. At this point it may be worth pausing to
reflect that our “plain” is from Fr. une plaine, itself from Lat.
plana (terra?), “‘an even, level, flat (piece of land)”. The Ger-
man plain, die Ebene, is formed from eben, cognate with Eng.
““even’’, in precisely the same way, and so is the Russian plain,
paenuna, built on paexeiii, whose sense is exactly that of Lat.
planus and Ger. eben. When we take into account the identity of
formation, and further that all these nouns are feminines, it
becomes apparent that they could well be calques; but of course
this does not exclude the possibility of independent creations,
since the passage from “level” to “a level space”’ is so natural,
and the gender could be purely coincidental.
 Malay-Indonesian 83

There is a derivative from padang that attracts attention:

pepadang, “‘information’’. If we look carefully at the relation in
form and sense between pepadang and padang, it must be evident
that pe-padang is built precisely like our ‘“‘ex-plan-ation”, which
is to planus, “level”, exactly what pepadang is to padang. Informa-
tion, like explanation, is basically ‘“‘a smoothing out, a levelling,
an ironing out’ (of problems), and this will be all the more
convincing to those whose Greek enables them to see “‘problems”’
as obstacles in the way which must be removed, an image that
is made more immediately intelligible in expressions of the type
“bristling with difficulties’, or Fr. hérissé de difficultés. Another
illuminating aspect of this phenomenon is that Gk mpobdAnua,
from a “jutting out’, developed the sense “‘cape’”’; capes have
long been notorious as presenting grave dangers to navigators,
and it is no accident that the French commonly use expressions
like franchir un cap, where the ancient obstacle is still plainly
discernible. Of course, defeatism is not a good way to overcome
obstacles, so the former Cape of Storms was renamed Cape of
Good Hope, as if to soothe the lurking, threatening powers by a
flattering euphemism.
Id. djarang, “‘wide apart; rare; sparse, thin”’, clearly finds its
proper place beside the previous terms. With the prefix memper-
it is used of “thinning plants (in a field)’’. Here again there is a
marked contrast with the terms examined in the last section,
and in particular “thick” as opposed to “thin” and “‘few and far
between’’. The cognate of Eng. “wide” is Ger. weit, “far’’, and this
strongly suggests that the true etymology of “narrow” is rather
to be sought in its relation to “‘near’’; there is moreover the
common phrase “far and wide’’, now seen to be closely compar-
able with “‘straight and narrow”, and “time and tide’, besides
countless others which were originally synonymic couplets and
which later development separated, as so often happens.
Id. ladang, “‘a dry field” (as opposed to an irrigated field) is a
further relevant term. And with the occasional first vowel
variation that has previously been observed, so is bidang, “a
plane, level surface; spacious, broad”, used specifically of a
piece of land or field, for example in sebidang kebun, ‘‘a garden”’,
where bidang is clearly a numerical coefficient. Another type of
first vowel variant can be seen in /uang, “empty, vacant; free”
84 Comparative Semantics

and the derived peluang, “opportunity”, whose sense relation

to freedom and leisure requires no elaboration. Note too ruang,
“large room; space; leisure’’.
There are numerous Indonesian terms worthy of inclusion,
but I shall confine myself to citing just one more: kadang,
“sometimes, once in a while”. I think its relevance will be
appreciated when it is considered that the fundamental sense is
‘cat distant intervals”, i.e. Fr. de loin en loin, similar to “few and
far between’, another instance of the transfer from space to
time. Many additional examples will readily occur to the reader.
The passage from space to time is ever-present in most
languages, and it is no exaggeration to say that the notion of
time is ultimately derived from that of space, a fact for which
linguistic evidence produces abundant proof. It will be sufficient
here to recall what has been said about Ger. spazieren and Sp.
despacio, “‘slowly’’, discussed in the previous section, and here
could be added Pt. devagar, ‘‘slowly”, whose basic relation to
time is at once observable if we set it beside divagation (in English
or in French), either in its literal physical sense of “wandering
about” or in the figurative “‘divagation of the mind”’. It should
also be helpful to compare “‘his mind kept wandering from the
subject” and Ger. Zerstreuung, zerstreut, “‘absent-minded”’,
beside Fr. distraction, distrait, and there are innumerable
parallels observable in Western languages. It would bea fascinat-
ing exercise in practical comparative semantics to try to discover
the fundamental space relation underlying Ger. zerstreut and
Jp. sampo, #4 “‘stroll’’, which was quoted previously. This is
immediately relevant to our present inquiry, and we shall have
occasion to return to this point further on. In the meantime, the
reader could most profitably ponder the matter.
Before we pass on to another aspect of this semantic field, it
may be useful to reflect that in the East too space and time stand
in precisely the same relationship as in the West. When discussing
Id. sempat and tempat, it was pointed out that Jp. koro, “‘time’’,
is a shortened form of tokoro, “‘place’’. (Tokoro itself is frequently
used with a temporal sense.) Similarly, Jp. ma, “‘space, room;
leisure”, written [Hj], is also read aida, “interval; between;
during; since’, i.e. it designates an interval in space or time,
and in Chinese the transfer from space to time is so usual that it is
simply taken for granted.
 Malay-Indonesian 85

Let us now examine Id. kandang, ‘‘stable; corral”, and its

me- derivative “‘fence in’. Beside this can be set Id. halang,
“block, prevent’’, and /arang, whose me- form means “forbid,
prohibit”. The relation between this group and that previously
studied may not be clear at first sight, but if it is remembered
that Eng. “fence” is a shortened form of “‘defence”’ and that Fr.
défendre not only means “‘defend” but also ‘“‘forbid’’, it will
probably be realized that a space may be and remain free only in
so far as it is enclosed by a fence or a wall for defence or security.
The most typical example is the English “‘town’’, which was
originally surrounded by a wall, and thus a town is a fortified
city. The Dutch cognate tuin, “garden”, and the German cognate
Zaun, ‘‘fence’, are especially convincing. Moreover OEng.
tun, the earlier “‘town’’, designated “‘an enclosed piece of land;
a field or farm’’, and even “a village’’.
Precisely the same relation can be observed if we start with
the Russian town 2opod or its Slavonic equivalent grad. Byelgorod
and Belgrade are both literally “White City”. The Slavonic
names of the “garden” are mostly based on grad, for instance
Rus. ozopod and Czech. zahrad, cognate with Rus. 3azopoodxa,
‘‘a fence”, and zaeopodump, “enclose, fence in”. The Rumanian
garden, gradina, is unmistakably a Slavonic borrowing, though
strictly speaking all these Slavonic terms are ultimately of
Germanic origin and it is from this Germanic ancestor that Fr.
jardin, Eng. “garden” and It. giardino are in turn ‘borrowed.
If we now take the French “hedge”, /a haie, another Germanic
borrowing, it will be clear that The Hague (La Haye, ’sGraven-
hage) is also named after its enclosing walls. A curious point is
that this word even wormed its way into “Copenhagen” and Fr.
Copenhague, whose Danish name is K¢ébenhavn, lit. Merchants’
Harbour. Havn is obviously the same word as “haven”: compare
Newhaven and Le Havre.
Again, from the ancestor of Ger. Burg, “castle” (i.e. a fortified
place), comes Fr. bourg and Eng. “borough’’, Edinburgh etc.,
and similarly Lat. castrum, “a fortified place”, is present in
OEng. ceastra, “city”, now ‘‘-chester” as in Manchester. Fr.
chateau, found in numerous names of towns, has come down
from Lat. castellum, derived from castrum, a diminutive in form
86 Comparative Semantics

but not in sense. Le Cateau in Northern France is the local

Picard form of chateau.
However, here again, despite the wealth of European words
confirming this semantic development, it is in the Far East that
we find the most arresting evidence. The Chinese character i
ch’éng, while still referring to the walls of a city, is more commonly
used of the city itself. On the other hand, in Japanese this character
is read shiro, “‘castle’’, or in Sino-Japanese jd, in which form it is
used for the names of a number of cities, for example Keij6, 5 jk
‘Seoul’. What is especially worthy of note is that 4 is based on
the earth radical + , which shows that those walls, in China as
elsewhere, were originally built of earth or mud, and we retain
a memory of this in Eng. “earthwork” as used of fortifications.
The Great Wall of China is still normally called 44% ch’ang
ch’éng, and thus at a glance we can see all the way from the
earliest mud works to the city walls and the modern city.
The terms of the A—A type so far considered have ended in
-ang, but there are also some members of the same semantic
family ending in -an.
First a brief transition between the two groups: while Id.
parang designates ‘‘a chopping knife’, Id. halaman refers to
‘the yard” (of a house) and moreover to “‘the page” (of a book).
These two may at first appear incongruous, but a courtyard is a
flat enclosed space near a house, and is closely akin to the
notions we have been exploring in connection with the names of
towns and gardens descended from original enclosures. In fact
“yard”, earlier geard, is a cognate of the Germanic terms that
have given “garden” etc. and the Slavonic towns grad etc.
If we now set “the page” and “the knife” side by side, and we
realize that the essential part of a knife is the blade, the link
between them is clear: they are both flat surfaces. That the blade
is characteristic of the knife hardly needs elaboration, but there
is the further linguistic fact that a sword is frequently referred to
by its blade (cf. “‘a good blade” for ‘“‘a good swordsman’’); this
goes back to medieval times when the Germanic brand, ““blade’’,
even became quite usual for the sword, as in OFr. brant and
earlier It. brando. However, the fundamental identity between
“blade” and “‘page” can best be observed in German, where
Blatt, “leaf”’, is the cognate of “blade’’, and there is the further
 Malay-Indonesian 87

vital fact that Blatt also extends its meaning to cover ‘‘a sheet of
paper” and of course “‘a newspaper’’, all senses in everyday use.
A close parallel is Fr. feuille, “leaf; sheet of paper’, and the
newspaper sense is also found here occasionally, as in feuille
quotidienne, “‘news-sheet”’, and the familiar feuille de chou, which
refers to a third-rate newspaper or local rag.
Id. Jahan frequently occurs with the prefix per- in the form
perlahan, “‘slowly’’. It will be recalled that among the unpleasant
accompaniments of being in a tight spot there is “urgency’’, and
that from this comes “‘haste”’. If we now take the opposite notion
of space, especially the level and flat space that is our present
field, there is “‘peace”’, “‘ease”’, “leisure” and “‘slowness’’. Since
space is plentiful, time too is plentiful, and it is thus possible to
do things ina leisurely way. The French expressions étre a l’aise,
se mettre a l’aise, and the corresponding English phrases with
“ease”, will also be seen to be the opposite of the cramped
conditions and narrowness of tight spots. And from the notion
of ease we pass on smoothly to easy circumstances, peace and
comfort, abundance and wealth, whereas in a tight spot there
was turmoil, discomfort, dearth and want. “Dearth” should be
noted, for it is the noun corresponding to “‘dear’’, from which it
has completely cut its moorings in English, but a moment’s
reflection will.reveal that high cost is inevitably associated with
scarcity, and that where things are plentiful they are cheap.
Thus, from these a priori considerations we can expect some
relation between all these notions that form a complete system
based on the amount (plentiful or lacking or short) of the space
and thus of the time available.
“Rasy” and “difficult” as a rule go back to words meaning
“light” and “heavy”. It is so in Old French and in several
Germanic and Slavonic languages. From a basic sense “heavy”’,
OFr. grief (hence Eng. “grief”’?) acquired the meaning “difficult”.
So does Ger. schwer combine both the senses “heavy” and
“difficult”, and Schwermut is what Germans feel when they are
depressed. Both “difficult” and “heavy” are clearly seen to be
particular cases of vertical pressure. In a remarkably compact
phrase, Hast und Last, Thomas Mann?° combines admirably
the figurative results of both lateral and vertical pressures.
41.
30. T. Mann, Der Tod in Venedig (Hamburg, 1954), p.
88 Comparative Semantics

It is but a short step from “‘easy” to “slow”. When somebody

works very hard he may be told not to flog his inside out, to take
it easy, i.e. to take his time over it, not to hurry so much. The
same semantic developments can be noted in Fr. a son aise,
“in a leisurely way”, and in It. adagio, where agio is obviously
the same word as aise. Again, as an alternative to ilfait tout cela
a son aise, the French say quite naturally sans se laisser bousculer,
where bousculer, ‘jostle, hustle’, shows unmistakably that
“hurry” belongs to the bustle, the crush and pressure of the
crowd, which we have also found to be associated with tight
spots. Similarly, a person who is pushed around is also in an
uncomfortable spot, and he does what he can to escape.
It is clear that there are innumerable results to be observed as
soon as we take this basic view of things and related linguistic
phenomena. It should come as no surprise to find that it is
language which brings the most immediate confirmation of the
points that have been briefly sketched. An expression like avoir
danger in the sense of avoir besoin, ‘‘need”’, is utterly unthinkable
in French, but it is normal in my own Nivelles Walloon, while
further to the east in the Namur and Liége regions the significant
word for “‘need” is not the local equivalent of danger but of
malaise and mesaise,*! both of which clearly belong to the same
notion of being in a tight spot. It is of undoubted interest to note,
however, that Chaucer uses daunger>? for “‘scarcity” and that
there are parallel examples of dangier in Old French.33 On this
point it may be helpful to bear in mind the semantic range of Du.
nood and Ger. Not, beside their English cognate “‘need’’, in order
to see the intimate relation between urgency, emergency, danger
and want. Parallel phenomena can be found everywhere in
European languages. Thus it is no accident that from the cognate
of It. fretta, ““haste’’, is derived Cat. fretura, “‘need’’. And there
are innumerable expressions of the type ‘“‘pressed for time” or
the regional “‘crowded for time”’, which become more meaningful
as soon as they are seen against the factual background sketched
here, which is after all merely an extension in the opposite
direction of what we observed in the tight and narrow E-A type.

31. In Namur the relevant term is [malo3i], and in Liége [mozoh].

32. Chaucer, The Romaunt of the Rose, \. 1147, in Works, ed. Robinson.
33. See, for example, Chrestien de Troyes, Yvain (Manchester, 1961), 1. 5304.
 Malay-Indonesian 89

If we now concentrate on “‘slowly’’, its relation to level, open

space becomes clear when it is realized that It. piano piano, “tout
doucement, trés lentement”’, goes back to Lat. planus, “‘level’’.
and that the simple piano still retains this sense beside the derived
one just mentioned. Similarly in Provengal, plan-plan is the most
common term for “slowly” and the familiar piane-piane has
penetrated French and is also used in Belgium. Note that every-
where the reduplication is clearly significant. As to the musical
piano as opposed to forte, the key to this has already been sug-
gested in the French rendering tout doucement, where slowness is
inseparable from lightness. While noise and turmoil are natural
concomitants of a tight spot, things are far slower and quieter
in the plain: think for instance of a river rushing noisily through
a narrow gorge, and observe how slow and peaceful! it becomes
when it can spread out alittle on reaching the plain.
Very comparable sense relations exist in Japanese. If we start
from + taira, “level, even”, we find the derived tairageru, “‘be
quiet”, and when read hei (Sino-Japanese), the same character
gives a series of compounds such as heiwa, *¥-All ““peace”’; heigen,
EJB “a plain”; heiki, 2% “calmness”; heii, 4B “easy”’ etc.
Yasui %#%€\. combines the meanings “cheap; quiet; easy”, the
derived yasumu means “‘to rest”, and when read an (Sino-
Japanese) there are further compounds: anshin, & <> “peace of
mind”; anzen, #4 “safety”; anraku, KAR “comfort” etc.
Again, with yurui, #\. “slow; loose, lax; easy’, it is instructive
to compare Sp. Jejos, ‘far’, from Lat. /axus. In each of these
there is found to be the fundamental notion of ample space,
hence ease and all the other pleasant results, in contrast to all
the unpleasantness we discovered in tight spots.
There is good reason to assume that Id. djalan, the usual word
for “road” or “‘street”, is so named from its association with
level space. Let us first recall that Fr. route is originally Lat.
(via) rupta, “a road broken through and opened up”; the German
pioneer, Bahnbrecher, makes this image all the more con-
a
vincing, since the pioneer is essentially the one who opens up
a compara ble
road through a territory as yet untrodden. There is
,
notion underlying (via) strata, “a road levelled and covered”’
‘*street”’ etc.
from which come It. strada, Ger. Strasse, Eng.
It is worthy of note that Ic. braut, “road’’, has had exactly the
90 Comparative Semantics

same history as route, since it is built from the participle of Ic.

brjota, “‘break’’. This verb was present in Old English as breotan
and its only living remnant seems to be Eng. “brittle”, while
Dutch and German have no trace of it. However, it is found in
all Scandinavian languages, where it is the most usual term for
“to break’’. Another interesting aspect of Ic. braut is that it is
quite isolated as a word for “‘road’’, since elsewhere in Scandi-
navia a cognate of Eng. “‘way’’, Ger. Weg, has been preferred.
Nevertheless, braut is still very much alive and in constant use in
all Scandinavian languages in an unexpected form: Ic. burt,
Sw. and Dan. bort are widely used as temporal adverbs cor-
responding to “he went away’’, Ger. er ging Weg etc., but with
the additional characteristic feature that there is a distinction
between motion and rest, as in Sw. bort (‘“‘motion’’) and borta
(“rest”). There is a similar term in Walloon, évoye, due no doubt
to Dutch or German influence. However, évoye is really closer
in form to Eng. “‘away’’, since it is actually é-voye, from Lat.
in via, with an attenuated preposition just as in English. For
example, taper évoye, “‘throw away’’, corresponds closely to It.
buttar via, both in form and sense.
The spatial aspect of Id. djalan is particularly noticeable in
its frequent use with the prefix ber- in the sense of “going for a
stroll”, a notion that we have already examined and found to
belong to space, since it is characterized by leisure and no
definite destination. A further use of this term, closely akin to
the previous one and actually an extension of it, is to designate
‘a trip made for pleasure or relaxation”, and it could be a very
long trip too, such as right across the ocean, provided that it be
done ina leisurely way. This is strikingly like certain uses of Jp.
asobu, it# 3 “play; spend time with somebody”, or just “drop in”
on a casual visit etc., which reflect the Chinese sense (“‘roam’’)
of the character #%. It is noteworthy that Id. mai, “play”, in its
reduplicated form maimai, is used in the same manner for a
casual visit or a pleasure trip. Now it is clear that “playing”
belongs to leisure, whereas ‘‘work”’ belongs rather with difficulty
and “‘want” to the tight spot. It may be useful to consider at this
point how frequently a word basically meaning ‘“‘work”’ is found
to be used for “need’’: Lat. opus, OFr. mestier, Sp. menester
(los menesterosos are “‘the poor’) etc., and to reflect that Fr.
 Malay-Indonesian 91

besoin and besogne are essentially variants of the same basic

form (les besogneux are “the poor’’).
There is a further similarity observable between Indonesian
and Japanese if we take the Sino-Japanese reading yu of the
character for asobu, “‘play’’, in the compounds yireki, 32/E ‘“‘a
pleasure trip’, and yuho, “‘a stroll”; the latter, written #46 , can
profitably be compared with the other Japanese term for ‘‘a
stroll’, sampo #4 , which was examined in Section 9. Since the
second character is the same in both and means ‘“‘walk’’, it can
be assumed that there is a close identity of meaning linking the
first characters: in yuho it is clear that yu is asobu, “‘play’’, while
in sampo it has been shown that it designates “‘a scattering’’, two
notions that unmistakably belong to space and leisure.
Another compound based on 3% is ¥% yusei, “un astre
errant” (i.e. a heavenly .body wandering through space), namely
‘“‘a planet’. The second character is usually taken to mean
“star’’, but in effect it includes both stars and planets. Yusei looks
very much like a calque, since a “planet” is literally (un astre)
errant, a name that has survived from the time when planets
other than the sun and moon were so called in contrast to the
fixed stars. To the Greeks they certainly did appear to be wander-
ing through space, since the Earth was at the centre and the stars
were fixed.
There is another fascinating aspect of the word “planet”,
which links it intimately with our present discussion. Gk mAavyty¢
borrowed as “‘planet’’, is derived from a Greek word meaning
“wander”, cognate with Lat. planus, ‘‘level’’, and thus here again
is found the basic notion of ample space that we have seen to
be inherent in planus and exteriorized in later developments,
since wandering and strolling both imply essential leisure.
A final note: in the light of what has just been discussed and
especially in relation to Jp. taira 7, yasui #\>, and yurui #3,
the following “spacious” character should be illuminating: 5 *P
*® Jp. yuruyaka, “broad, spacious, wide”, hence “not strict;
easy; gentle; forgiving’. When read as the Japanese verb
kutsurogu, the sense is “‘se mettre a aise”. Exactly the same
semantic range is found in Chinese.
Mention should also be made of the Japanese compound
 92 Comparative Semantics

kandai(na), #(%)‘‘generous; lenient”, to which may be

compared Fr. Jargesses.

We have now completed the major part of our study of vowel

symbolism in Malay-Indonesian,** and it is of particular interest
to observe how utterly different this is from Semitic languages,
where the fundamental sense rests essentially on the three basic
consonants and the vowels play only a secondary role. It will be
noted too that Semitic writing, and in particular the Arabic
script which gives the consonanta! outline of words while the
vowels are purely external and easily dispensed with, is admirably
suited to this type of language. However, languages where
vowels play a vital rele—English, French, and Indo-European
languages in general—would hardly be intelligible in Arabic
script. So it is no surprise that Turkish, which is non-Semitic and
~whose vowels are far more important than in Arabic, should
have switched from the Arabic to the Latin script more than
forty years ago. (The Turks had borrowed the Arabic alphabet
together with numerous Arabic words.) Similar situations have
arisen elsewhere, and it is hardly an exaggeration to say that
whatever system of writing is used in a given language is largely
a historical accident. What is true of Turkish is equally true of
Malay, which used to be written with Arabic letters, no doubt as
a result of the large inflow of borrowings from that language
which are still clearly recognizable, but it must be stressed that,
no matter how numerous the borrowings from various languages
and especially from Semitic, Malay-Indonesian is anything but
Semitic, just as English remains essentially a Germanic language
despite the overwhelming number of Latin and Romance
borrowings.

34. As was pointed out in the Preface, the sections on the vowel u (together with
the endings -ut and -ung) are dealt with in Chapter III, since the material
contained in them is semantically relevant to the discussion of ‘‘the bow”.
3)

The Bow

The question of motivation or non-motivation in language

formed the starting-point for the exploration of the semantic
field surrounding ‘“‘the bow’. We looked briefly in class at
the German family Zug, ziehen etc., and traced parallels,
mainly in French (train, trainer, tirer, traire), English (“draw”,
“draught” etc.), and Dutch (tocht). Rather surprisingly, we found
that English and Dutch, despite their common Germanic
heritage, preserve only a few isolated remnants of what in German
is a very important and viable family. The conclusion reached in
the light of this cursory examination was that German is a highly
motivated language, whereas English and French tend to be
somewhat abstract.
Like German, Japanese is often highly motivated, since here
the script used makes meaning visible in many instances. But it
was only as the result of a chance find that the Japanese verb
hiku came into the picture. What is especially interesting is that
the semantic range of hiku corresponds remarkably well to Ger.
ziehen. Thus, from an apparently insignificant starting-point we
were led on to the exploration of a semantic field whose possible
developments are very far-reaching—as will be seen from the
following treatment of this topic.
Underlying the earlier part of the discussion, and constituting
its central thread, will be the two principal and characteristic
notions associated with the bow, namely tenseness and relaxation.
The later developments, however, do not lend themselves to such
a clear-cut delineation, and it has therefore been thought best to
let the material speak for itself as it unfolds, without attempting
to provide an introductory survey of what is contained in the
remainder of the chapter.
94 Comparative Semantics

Let us begin with Jp. hiku 4| <, which has as its radical the bow. 5.
Since Oriental civilization is very ancient, and since the bow has
been known for thousands of years, it is hardly surprising that in
a system of writing based on ideographs a number of characters
should include this element. Hiku 5] < contains within itself
the opposite meanings “draw” and “withdraw”, like MEng.
draw and to a lesser extent Fr. tirer, thus showing that two
contraries can turn out to be very close.
Jp. yumi wo hiku 3#45|< is equivalent to Eng. “bend
(draw) a bow”; Fr. “‘tendre (bander) un arc”. In each case
tension is present. The image behind the Far Eastern bow is
that of a weapon similar to a medieval cross-bow which, despite
its cumbersome nature, was nevertheless very powerful—and
this impression of power is important; the degree of tension
built up in the process of firing this weapon may well go part of
the way towards explaining why, in Chinese and Japanese, the
word for “strong” is depicted by a character based on the bow:
a (Ch. giang; Jp. tsuyoi).
If we consider for a moment the English verb “‘bend”’, it
becomes apparent that its uses also have to do with force, or with
what might be called a concentration of energy. From the original
physical phenomenon of the bow, “‘bend” acquired figurative
meanings, and its ultimate relation with the bow has long
ceased to be recognized. However, we need only bring to mind
the image of a bow, arched or bent to its utmost immediately
prior to being fired, and “bend” takes on quite a new significance:
it once again becomes motivated.
Just as “‘strong”’ 5# is based on the bow, so too is its opposite,
“weak” 9% (Ch. jo; Jp. yowai); the compound #43 (Sino-Jp.
kydjaku) is used in Japanese to form the abstract noun “‘strength”’.
In this connection it must be pointed out that the bow is not
always bent (or stretched): it may be quite limp. This aspect,
which is not so obvious, could easily be overlooked; however,
it will be examined further on, as it adds a new dimension to the
field under discussion.
Two other characters will serve to show how closely related
are felt to be the notions of tenseness (cf. strength) and relaxation
(cf. weakness): Jp. haru, 54 “to stretch, spread’, and Jp.
yurumu, 4ts “to loosen, relax’’. The radical in each case is again
 The Bow 95

the bow; hence it is once more demonstrated that the bow

incorporates to a remarkable degree two seemingly contradictory
elements, which in reality are very close.
Jp. tento wo haru 7 Y + % 5&4 corresponds to Fr. dresser une
tente, and to Eng. “‘pitch a tent’. Note that a tent is literally
that which is “stretched” (Lat. tenta, “‘stretched material’’).
Jp. §&& 4, also pronounced haru, is used to express the idea of
sticking or pasting (Fr. coller) something such as a poster on a
wall. We must not be misled by the fact that it in no way visually
resembles haru 5&4 just referred to: the two words are obviously
related.
In Chinese, zhang #z is equivalent to Fr. tendre (un arc; un
piége), and to Eng. “‘set”’ (a trap). It may be of interest to mention
that “‘set” contains the idea of tension, although this is not
readily apparent. If we think of the phrase “ready set, go”’, it
becomes clear; there is tension when, for example, a runner is
poised and ready to begin a race.
Setting (a trap) is a natural extension of the idea of stretching.
This latter, like “‘bending a bow’’, has also developed a figurative
meaning, being used in the sense of “boasting”. The phrase
“you're stretching things a bit” is well known, as is the expres-
sion “stretching the truth”. We can also say “draw the long-
bow”, i.e. exaggerate, though this is seldom heard nowadays.*
Let us return to Jp. hiku, this time written ## <. Its semantic
range is as follows: “play a stringed instrument” (koto, samisen,
piano etc.); “repel” (Fr. rebondir); and as a noun, RH (Ch. tan;
Sino-Jp. dan) designates ‘‘a bullet’. As a tentative explanation
we might say that in each case the underlying notion is that of a
spring—which is true as far as it goes. By taking in turn each of
the meanings given, we shall see that there is in fact a close
relationship between them, which explains how it is that the
same character can be used for all three.
In the introduction to this section, it was stated that the bow
has been known for thousands of years. Just as this explains the
existence of Chinese characters based on the radical 5, it also
explains much in the history of music. It is well known that the
most ancient and primitive form of stringed instrument was the
(London,
1. An example occurs in H.H. Richardson, The Getting of Wisdom
1960), p. 196.
96 Comparative Semantics

bow string. From the discovery that stretching a chord and

releasing it abruptly (as happens in the case of plucked strings)
produces a slight vibrating sound, stringed instruments were
gradually evolved. Hence it should come as no surprise to learn
that hiku #%< is specifically used of the playing of a stringed
instrument.
In the case of the second meaning “repel” (Fr. rebondir), the
same character is read hajiku (a verb). Anyone who has observed
a cat preparing to spring could hardly fail to have been struck
by its resemblance to a bow being bent: it crouches, draws itself
in, and then makes its move. The accumulated energy, which was
previously concentrated in a confined space, is released—in
precisely the same way as with a spring.
Now consider the following Japanese example, where the
image of a release of tension is strikingly conveyed: not only
does it contain the verb hajiku, but also the bow itself:
Roba wa, genin wo miru to, marude ishiyumi (#) ni demo hajikareta
(fk nz) yo ni, tobiagatta.?
The old woman, with a glance at the retainer, sprang up, just as if
she had been shot from a cross-bow.

Admittedly this story is set in the twelfth century, but never-

theless it is clear that for a modern Japanese writer this is a
powerful image, and one that is particularly appropriate in
the context.
An English equivalent of “repel” is “‘fillip’, which is probably
a simple variant of the verb “‘flip’”. When one flicks a marble
etc., the fingers form what is sometimes called a “fillip” (Fr.
un chiquenaude). It is not difficult to look upon the finger as
forming a bow, since it is joined to the hand and to the thumb;
in this position it is ready to shoot forth the marble. As with
the strings of an instrument, the important point to notice is
that there is a concentration of energy. At the moment when the
projectile is launched, the tension is released.
‘“Fillip” is also used in the figurative sense of “‘a stimulus”.
It is a relatively simple matter to think of instances where the
two opposite aspects of the bow complement each other: for
example, ““The economy is sagging; this will give a fillip to it”.

2. R. Akutagawa, Rashomon (Tokyo, 1968), p. 13.

 The Bow 97

In view of what precedes, it should now be clear that the

compound 544. (Jp. tama; Sino-Jp. dangan) illustrates the same
principle as the string of a musical instrument and the fillip—
namely, a release of tension: when a bullet is fired from a gun,
this is precisely what takes place.
Note incidentally the compound 9%) (Sino-Jp. danryoku),
“elasticity” (lit. spring force). It is seen to be built up in exactly
the same way as Ger. Springkraft, and is in fact almost certainly
a calque on the German term.
We have now examined the character ## in some of its most
frequent uses. As well as possessing the readings hiku and hajiku,
it is also pronounced hazumu. Since the most common meanings
of hazumu are “‘to spring, bound, rebound, bounce’’, it is evident
that the notion of buoyancy—what in French is termed entrain
or vigueur—is very much to the fore, and that hazumu expresses
the opposite of “sag; limp” (Fr. fléchissant). So it represents the
bow when it is tense, i.e. strung, and this is only to be expected
in view of the other uses of the character 5#.
Hazumu is frequently encountered in descriptions of the voice,
when a writer wishes to convey what can best be called a sharp
variation in tone. When someone is happy, or excited, or worked-
up, such emotions are reflected in the voice, which takes on a
somewhat “‘springy” quality:
Takeda Yayoi wa, kyt ni hazunda koe wo ageta.*
Takeda Yayoi suddenly raised an agitated voice.

That hazumu, “bounce, rebound”, is the usual term for a con-

versation going on without flagging might almost be considered
as an extension of the above usage. In Mark Twain’s Tom
Sawyer, where the original runs:
So the talk ran on. But presently it began to flaga trifle . . .*

the Japanese rendering is as follows:

Ko iu fu ni hanashi wa hazunde itta ga, mamonaku sukoshi dare-

hajimete...

3. H. Sukeno, Himitsu Partei (Tokyo, 1961), p. Ws

Boken
4. M. Twain, Tom Sawyer (New York, 1959), p. 110. Tom Sawyer no
(Tokyo, 1943), p. 199.
98 Comparative Semantics

With dareru, “flag”, compare Fr. la conversation languissait,

which suggests a similar image, in this case, of course, the
opposite of “bouncing”’.
Notice finally that OFr. resortide and Mod. Fr. rebondir
correspond very well to hazumu 4; in the Chanson de Roland,°*
the former term is used of a sword which, having been struck
violently against a hard block of stone, sprang back vigorously
owing to the force of the impact. We have seen something of
the English verb “‘spring” (Ger. springen), and its Japanese
equivalent. The noun will be taken up later in connection with
certain other Japanese phenomena.
Leaving the Far East for the present, having by no means
exhausted it, but having seen that it offers many fascinating
things that could be explored in greater depth, let us. now turn
to German, where the bow family is still very well represented.
Ger. Bogen possesses the following senses: “the bow of the
archer” (Fr. arc); ‘‘a saddle bow” (Fr. argon); “‘a bow for a
stringed instrument” (Fr. archet); ‘‘an arch or vault”’ in archi-
tecture (Fr. arc; arche [d’un pont)); “the arc of a circle” (Fr. arc).
Bogen is thus extremely compact and self-sufficient, whereas
English, as has just been shown, normally requires the addition
of an “indicator” in order to render a particular meaning of
this German noun. Notice that in French too a variety of words
is used, but despite the fact that they are based on the arc, it
is probably true to say that there is not felt to be any fundamental
relationship between them, and the same can be said of the
corresponding English terms based on “‘the bow”’: they are simply
taken for granted, and each one is treated in isolation.
There is another noun, Bug, which is used for “bend; bow;
bow (of a ship)”. Two verbal forms, beugen and biegen (whose
past participle is gebogen), both mean “‘bend, curve; bow”.
From biegen is formed the noun Buckel, “‘a hump on the back”.
Clearly there is no shortage of terms in German relating to the
bow, and the foregoing examples do not by any means exhaust
this particular family; we shall have occasion to refer to other
members of it a little further on.
English is not nearly so fortunate: a comparison with German
shows that it has lost many of the features which continue to

5. Anon., La Chanson de Roland, ed. T.A. Jenkins (Boston, 1924), 1. 2341.

 The Bow 99

enjoy a very active existence in that language. It would not be

inappropriate, therefore, to refer to the English survivals as
“orphans” since, for the most part, their origin is no longer
recognized and as a result they appear to be quite isolated. A
glance at several relevant words will demonstrate this point.
In “elbow” the “bow”? is literally “ca bend”’, since cloth was
originally measured with an ell. Precisely because it is such a
common word, no one sees the “‘bow”’ in it: blind acceptance
over the centuries has made it utterly opaque.
“Bout” (formerly written ““bought’’) means “a turn, a bend”’.
It is even more isolated than “elbow’’, as its origin is disguised
by the spelling. The expression ‘“‘a bout of flu” is well known.°®
Several years ago we had occasion to become acutely aware of
this word, since it was frequently used in speaking of the im-
provement in relations between China and the outside world:
journalists and commentators referred many times to “a bout
of ping-pong diplomacy”.
“Bight”, the best-known example of which is undoubtedly
the Great Australian Bight, is perhaps the least transparent of
the three English words that have so far been considered.
However, when one looks at a good-size map of Australia it is
at once apparent that the name is most apt, for the bight’s shape
does indeed resemble that of a large bow. Ger. Bucht, ‘‘an inlet,
bay, creek”, and Du. bocht belong, of course, to the same family.
Mention must also be made here of Sino-Jp. wan #%, “‘to draw,
bend’’. (Notice the bow radical underneath.) When the water
radical 7 is added, the meaning changes to “a gulf, bay” %,
but the pronunciation remains wan, as in Taiwan iA (lit. Table
Bay). Thus it is clear that both in Germanic languages and in
Japanese, the identity between the “bow” and the “bay” is a
fundamental one.

We come now to an examination of the English adjective

“buxom”, perhaps the most interesting survivor of the Bug,
biegen family; it is cognate with Ger. biegsam/beugsam, “flexible,
yielding, supple, capable of being bent”.
Up to the seventeenth century, “buxom” was used to indicate
an attitude of mind, as can clearly be seen in the work of Chaucer,

6. See also W. Shakespeare, Hamlet, V, ii: “Ill play this bout first”.
100 Comparative Semantics

where it means “submissive, obedient”.7 Moreover it seems to

have been applied exclusively to men during this earlier period.
Then in the eighteenth century came the transference of usage
to women (a development that is not at all easy to account
for). The result was that the former “‘figurative”’ use, i.e. referring
to an attitude of mind, was supplanted by a purely physical
one: that of an abundance of flesh—particularly a well-developed
bosom. Hence ‘“‘buxom”’ can now be equally well applied to a
shapely young woman and to a middle-aged matron. This usage
has become so current that in present-day speech reference to
a man as “buxom” would sound decidedly odd. (Yet I can
attest to having heard just such a use in a conversation—by
someone who would certainly never have looked into the history
of the word.)
It is necessary to draw attention to the fact that the difference
between the uses of Chaucer’s time and those of the present day
may be summed up as a change from the relaxed or sagging
bow, i.e. the idea of mental submission, to what might be called
its tense aspect, i.e. the springiness and bulging quality associated
with the flesh. (A comparison with Jp. mukkuri further on will
help to elucidate this point.)
It was noted at the beginning of this chapter that both relaxa-
tion and tenseness are inherent properties of the bow; neverthe-
less it is somewhat puzzling to find that “buxom” has undergone
such a radical change of meaning in the course of its history.
It would seem that this situation can be at least partly accounted
for if we bear in mind that “buxom” (like the other English
terms that were referred to as “‘orphans’’) has long existed in
isolation, having been cut off from its roots in the bow family.
But now that “buxom” has been clearly shown to belong to
this family, there is no doubt that it takes on a new significance,
and once more becomes motivated.

The curious semantic history of “buxom’’, whose fundamental

dualism can only be understood in the light of its hidden bow,
with its potential tenseness and laxity, may become clearer if

7. G. Chaucer, A Complaint to his Lady, 1. 119; The Shipman’s Tale, 1. 177;

The Clerk’s Tale, |. 186 (“buxomly’’), in The Works of Geoffrey Chaucer,
ed. F.N. Robinson (London, 2nd edn, 1957).
 The Bow 101

we consider the following description of a woman putting on

her stockings:
Puis elle s’assit. Il lui fallait maintenant enfiler ses bas. Son front
se plissa. Elle devint plus attentive, plus précautionneuse, sa main
se fit plus flexible, plus enveloppante. Une maille était si vite‘partie!
C’était un luxe auquel elle tenait. Avoir toujours des bas trés fins et
trés bien tirés. On n’avait pas tort de dire qu’elle avait des jambes
parfaites. Elle banda son mollet, \e fit valoir dans sa cambrure la plus
favorable. Oui, sans conteste, c’était ce qu’elle avait de mieux. Non
pas des jambes honteuses d’elles-mémes, non pas les jambes un peu
cagneuses de Caroline, ni fléchissantes de Clarisse, mais des jambes
superbes, piaffantes et spirituelles, longues et déli¢es, douces au
toucher comme 4a la yue et dont elle jouait comme d’un instrument
de séduction. Elle tendit sur sa cuisse bombée la soie fragile.®

Besides the words italicized, there are a number of points worth

noting, and especially from the outset the fact that it is a very
serious operation which requires the utmost care and attention.
This is made at once visible: son jront se plissa, and it is no
accident that attentive and flexible contain the two comple-
mentary notions that characterize the bow.
Observe too the dual sense of firer: it usually means “‘take
off” when referring to a garment, but it is clearly used here in
the opposite sense and is synonymous with tendre, which occurs
further on.
Again, cambrure and bombé are contrasted with fiéchissant,
precisely as “tense” and “bulging” are opposed to “Jax” and
“sasging’’. It is interesting to note that cambrure has preserved
its ancient sense ‘“‘eurved, vaulted”, and that its present Picard
form is remarkably close to OFr. canbre vautie.° The original
chambre, i.e. Gk kapapa, was essentially “vaulted”. The analogy
becomes even more striking when the perfectly stockinged leg
becomes a musical instrument dont elle jouait comme d’un instru-
ment de séduction. Here we are irresistibly reminded of Jp. hiku
m4 < in its secondary sense, “play a stringed instrument”’.
Perhaps the most remarkable expression of this passage is
elle banda son mollet. If we remember that mollet is derived
from mou and if we also bear in mind the expression bander un

8. R. Guérin, Parmi tant d’autres feux (Paris, 1950), p. 69.

9. Anon., Aucassin et Nicolette (Paris, 1965), V, 1. 2.
102 Comparative Semantics

arc, it is but a short step to reach the erotic sense of bander used
absolutely.
The mollet is of course the soft part of the leg, by contrast
with the hard shin. The shin or shinbone is so much in evidence
that it readily becomes the normal name of the leg, as in Ger.
Bein, Du. been, Dan. and Sw. ben, all of which are cognates of
“bone’’. The Italian calf, polpaccio, is literally “‘the fleshy one’’,
again in implied contrast with the bony front. The Spanish leg
is particularly fascinating, as the calf, pantorilla, ‘‘the little
paunch’’, bulges away from the shinbone, espinilla, “‘the little
spine’. It may be relevant here to record that the Germanic
ancestor of the shin passes into Fr. échine, “‘spine”’ and “‘back’’,
and that there are corresponding Italian and Spanish terms.
The calf seen as “‘a little paunch”’ makes it easier to under-
stand why in English the calf of the cow and the calf of the leg
are originally one and the same word. Further confirmation of
this identity is to be found in Du. kuit and Rus. ukpa, both of
which combine the senses “calf of the leg” and ‘“‘fish spawn”’.
Anyone who has seen and felt what fish are like when filled with
innumerable eggs will easily appreciate the accuracy of this
semantic development. “a
Similarly, the Japanese calf, fukurahagi, “‘the bulging part of
the leg’’, is so called in contrast with sune, “shin”, hence, “‘leg’’,
a development comparable with Ger. Bein. The sense of “‘bulge”’
is contained in fukura, which belongs to a large family of words
such as fukurami, the most usual term for the bulges of the
breasts or buttocks, and fukuro, “bag”; fukurettsura, ‘“‘sullen
face” (lit. pout-face), illustrates a further aspect of this field.1°
Mention must also be made of fukufuku, “‘swelling”’, beside
which there exists a parallel form, bukubuku, having precisely
the same meaning: bukubuku fukureta, “bulging, baggy’’.
A fascinating aspect of Japanese expressive words is to be
found in their phonetic variants. Apart from the characteristic
reduplication already noted in fukufuku and bukubuku, the group
under discussion includes the following: fukkuri, “‘plump, well-
rounded, puffy’, with a further variant seen in fukkurato marui
10. The Chinese characters used do not bring out the obvious relationship
between all these terms, and it is for this reason that they have been omitted.
It must also be pointed out that the expressive words subsequently discussed
are all written in kana. (Ed.)
 The Bow 103

masshirona hd, “chubby, rounded, perfectly white cheeks’’, and

again in fukkurashita chibusa, “bulging breasts”. There is an
even more arresting member of this family, with initial p, to be
found in ohara wa pukkurato fukureagari, “the belly bulges and
swells up”’.1! Moreover in addition to the various forms with
initial f, b and p, there is another very common one, with the
corresponding nasal m: mukkuri okiageru, ‘“‘spring up” (sud-
denly); mukkuri koeta, “‘plump”; mukkuri hareru, “‘to bulge,
swell up’.
There are comparable series of expressive words in French:
bouder, “‘pout”, and bedaine, “‘paunch’’, are clearly suggestive
of bulges; bouffi, ‘‘swollen’’, and bouffée, “‘puff”’, can be placed
beside un pouf and pouffer de rire, ‘bubble over with laughter’,
and Rabelais’ moufles,12 ‘empty nonsense”’ (lit. swelling). It
may be worth pausing here a moment to observe about “bubbling
with laughter” that Jp. bukubuku is itself used of “froth” or
“bubble’’, in addition to the senses already mentioned.
This French series inevitably leads to a further parallel one
in English: puff, huff, muff (cf. Mod. Fr. moufle, “‘mitten’’), and
quite possibly guff and chuff. The probability of the relevance
of American “guff” is increased if it is considered that the
derived ‘“guffaw’”’ is very close in sense to Fr. pouffer de rire.
“Guff” is of course “puff, empty talk, hot air”, what the French
would call des blagues. Blague is a Germanic borrowing, ulti-
mately identical with Ger. Balg, ‘‘bag; bellows” and Eng. “belly,
bellows’’; its earlier sense is still found in blague a tabac, “tobacco
pouch”, which recalls the Japanese terms just discussed, especially
fukuro, “bag”, and yet another member of this teeming family,
fukuroguma, “koala” (lit. pouch bear). Lat. follis, “bellows”
(cognate with Ger. Balg, Eng. “‘belly”’), turned into Fr. fou by
a semantic process comparable to that which led to blague,
“nonsense’’, and the synonymous Fr. billevesée has had a similar
semantic history. Eng. “windbag” is a further example of the
same phenomenon.

1. These examples are taken from Japanese magazines dating from around
1960, but the precise references are unknown to me. (Ed.)
2. F. Rabelais, Gargantua, chapter XV, in Oeuvres completes (Paris, 1962).
104 Comparative Semantics

Jp. hara, “belly”!3 (well known to Westerners from harakiri,

“‘belly- cutting”), is the key-word of yet another series of “bulge”
terms, such as haramu, ‘“‘be pregnant’; haremono, “‘swelling”’;
hareru, “be distended’’; and our old friend haru, “‘stretch”’, of
which hareru is now seen to be the intransitive form.
Let us now cast our thoughts back to Guérin’s Marie-Amélie
as she is tensely engaged in the important task of putting on her
stockings, stretching them carefully so that there is no crease,
no faux pli. In this operation, tire and tendre were key words.
If we could now imagine a Japanese woman making the same
kinds of gestures, the corresponding term used is very likely to
be a compound verb made up of hiku, 5| < “pull”, and haru,
i=4 “stretch”, and while this should regularly give hikiharu,
the form in actual use is hipparu 5| D5 4.14

A bow exists only in so far as it is strung, and the bowstring well

deserves our attention. The usual Japanese term for the bowstring
is tsuru, written 5%, whose radical 5 is clearly the bow. However,
there is another Japanese word tsuru, written with a totally
different character, & , whose meanings are “‘tendril” and “‘vine”’.
Both senses coexist in the modern language, but the Chinese

13. It is of course a sheer coincidence that hara is so remarkably close in both

sense and form to Lat. haru-spex, ‘“‘entrail-gazer’’, and what makes this still
more curious is that the usual term for the leech is Jp. hiru and Lat. hirudo.
However, there are other coincidences elsewhere that are quite as amazing.
14. Instead of setting down hiku haru as the starting-point, it would be more
accurate to postulate hik haru or rather an earlier hik paru, where a kind of
assimilation of k to p will give hipparu, just as Lat. noctem turns into It.
notte. The last consonant of a syllable is usually in a weak position when
followed by another consonant, and so is readily assimilated to the latter.
As to the comparison just made with It. notte from Lat. noctem (or It.
atto from Lat. actum or aptum), it may be worth pointing out that other
Japanese sound changes are remarkably like our own. From yomu, “‘read”’,
for example, the past tense should regularly be yomita, but after loss of
internal i and assimilation of consonants, the change to yonda is to be ex-
pected, where ¢ is voiced through the final nasal, itself turned into a dental
by the following ¢. So also from kagu, “‘smell’’, the regular past kagita under
the same conditions turns into kaida, whereas with the corresponding voice-
less stop in kaku, “‘write”, for example, the past tense is found to be kaita.
It is very tempting to postulate palatalization of k after loss of internal i,
i.e. from kakita to kakta and finally kaita, and in the case of kagu, a parallel
change of g with the voice passing on to the final consonant, just as Lat.
iud(i)care gives in French not *jucher but juger.
 The Bow 105

evidence (from “‘tendril” to ‘“‘creeper’’) shows conclusively that

“vine” is a later development from “‘tendril’”’.
Lat. vitis had precisely the same semantic history, from “‘ten-
dril” to “‘vine’’, and both senses were common in the classicai
language. It should also be noted that the present Danish and
Swedish names of the vine, vinranke and vinranka respectively,
literally mean ‘‘vine tendril’’. There is further evidence of the
clinging sense of Lat. vitis in the fact that it turned into Fr. vis,
““screw’’, hence Eng. “‘vice’’, whose essential part is obviously a
screw. Similarly, the Latin diminutive viticula, which should
regularly develop into Fr. *ville (pronounced [vi:j] and so in
no way homonymous with ville, “town’’), is present in Mod.
Fr. vrille, “‘tendril’ and “‘gimlet”, whose r may be due to virer.
Compare It. i viticci della vite, “the tendrils of the vine”.
All this suggests very strongly that Jp. tsuru, “bowstring”’,
and tsuru, “tendril” and “‘vine’’, are originally one and the same
word, and that the earlier sense had to do with tying or binding.
The fundamental identity of Eng. “bind” and its Ablaut-variant
“bend” now becomes clear, as the bow can be bent only if it is
strung or bound. Exactly the same relation is found in Fr.
bander: bander unt plaie, “bind”, but bander un arc, “bend”’.
If we next compare Lat. nervus and Gk veipov, “sinew;
strength; bowstring; string of a musical instrument’, and the
corresponding adjective nervosus with Fr. nerveux and Eng.
“nervous”, their sense “strong” seems to be largely due to
Classical reminiscences. A highly strung person is anything but
strong. Actually we have here again the twin complementary
notions of tenseness and laxity that essentially belong to the
bow.
Tsuru, “bowstring’”, is the native Japanese reading of the
Chinese character 4% mentioned earlier; there is also a Sino-
Japanese reading, GEN, which stands in phonetic correspondence
that GEN shares with Ch. xian. GEN denotes not only “the
as learned, and to some extent this also applies to the senses
which GEN shares with Ch. xian. GEN denotes not only “the
string of a musical instrument”, but in addition “the chord of
an arc” and “the crescent”, This last-mentioned is a further
example of the dual meanings that we have been observing
everywhere in the bow family. The “crescent” is literally “the
106 Comparative Semantics

increasing moon’’, but it is used indiscriminately of the moon

waxing or waning, so that strictly it could be a symbol of strength
as well as weakness, of greatness or of decadence. The native
Japanese yumiharizuki 5 4 5 , lit. bow-stretched moon, similarly
corresponds to the tense aspect of the moon and ignores the lax.
It is, however, as ‘“‘the chord of an arc”’ that GEN is especially
interesting. There is no need to point out how easily the bow
and bowstring could be looked upon as an arc of a circle and
its chord. This identity is even more striking in other languages,
for example in French, where /a corde d’un arc either means
“the string of a bow” or “the chord of an arc’. What must be
stressed is that while the bow forms a curve, the bowstring is a
straight line (and here it should be recalled that straight literally
means “stretched”’); as an arrow is drawn back, it forms two
straight lines enclosing an angle gradually narrowing from 180°
towards 90°. The chord subtends the arc, just as the hypotenuse
subtends the right angle, and it is no surprise to find that Sino-Jp.
GEN is also used in the sense of “hypotenuse’’. This relation
becomes clearer still when it is observed that “hypotenuse” is a
Greek participle, izoteivovow, literally meaning “‘stretching
under’, from the verb 6dzo-teivm, ‘“‘stretch under’’. It will now
be realized that ‘“‘subtend” is in fact a calque based on this
Greek word.
It is interesting to note that in Chinese and Japanese the sinus
of an angle is iE3% (Ch. zheng xian; Jp. seigen), lit. the true
chord, while the cosinus is very properly 44% (Ch. yu xian; Jp.
yogen), lit. the surplus or complementary chord. While our ter-
minology does possess a co- prefix for the three 4: functions
(cos, cosec, cot) which decrease as the angle increases, we have
no corresponding prefix to characterize the three IF functions
(sin, sec, tan), which increase and decrease precisely as the angle
does and can therefore be considered worthier of being described
as “true” functions.
Thus it can be appreciated that in contrast to the very precise
Far Eastern terminology just referred to, our own leaves a
certain amount to be desired. This is particularly evident if we
consider our mysterious “‘sin” (Lat. sinus, “bend, fold, curve’);
it is all the more mysterious as the evidence clearly points to its
ancestor as the bowstring or the chord of an arc, i.e. a straight
 The Bow 107

line and not a curve. It does look as if some early mathematician

confused the bow and its string.
Some etymologists have explained sinus as a mistranslation
of Ar. djayb, but actually sinus is an excellent choice to render
this Arabic word, whose chief meanings are “breast; cavity;
pocket’’, besides its mathematical sense. That the latter is due
to a borrowing from Sk. djiva, “‘string’, must be doubtful,
especially as the Sanskrit word seems to be rather djya, ““bow-
string” (cognate with Gk fios, ““bow’’), which is far less likely
to be identified with Ar. djayb. At least the mistake was made
earlier by Arab mathematicians, who assigned to djayb a meaning
it could hardly bear, and not in the later choice of Lat. sinus as
an equivalent of the Arabic term.
It should be observed that all the senses of djayb are bulges,
positive (“‘breast; pocket’’) or negative (“‘cavity’’), and that the
mathematical use is quite isolated. Notice, by the way, that djayb
is no stranger to Europe: in the form cep it has become the usual
Turkish word for “‘pocket’’, and this in turn has passed into
Modern Greek, Serbo-Croatian and Hungarian. The accuracy
of sinus as the rendering of djayb can best be appreciated from
its later development into Fr. sein, ““breast”’, and It. seno, “breast”
and “‘bay”’, which latter sense is also found in It. insenatura and
Ger. Meerbusen (lit. sea bosom); this in turn reminds us of the
“bight” group and the wan of Taiwan already mentioned.
What makes the “sin” problem even more intriguing is that
there is a Germanic family of this form: Old Norse sin, OEng.
sinu, Eng. ‘“‘sinew’’, Du. zenuw. Ger. Sehne is of course the same
word and means not only “‘sinew”’ but is also the standard term
for “the string of a bow” and “the chord of an arc”. Another
fascinating coincidence! It should be remembered too that
“sinew” is the basic meaning of the Latin name of the bowstring,
nervus. Yet despite all this, the mathematical “sin” is undoubtedly
an abbreviation of Lat. sinus, though it is far less suitable to
express the sense required than the Germanic terms just men-
tioned.
As this is essentially a linguistic, or more precisely a semantic
exploration, only passing reference need be made to the bow in
architecture, i.e. to the arch, though its significance through the
ages is striking enough. The curves of the arch or the vault had
108 Comparative Semantics

obvious virtues in the eyes of the early architects, and the funda-
mental role of the well-named keystone (Fr. clef de voiite) as it
finally jocks or binds the whole structure into perfect cohesion
and stability, where otherwise there would be total collapse,
need not be stressed here, any more than the gradual develop-
ment of the various types of arches, from the round Romanesque
to the pointed arch of the so-called Gothic stv!e.
However, it is worth recalling that Fr. arc and Ger. Bogen
are used equally well of “‘the bow of the archer’’, “‘the architec-
tural arch” and “the arc of a circle’’, and that this identity is to
be found in a number of languages, as can well be expected from
senses that all clearly descend from the same original bow. The
differentiation of terms in English rather obscures this historical
development, and so does the use of “‘chord” as distinct from
“string”. Eng. “arch” is witness of the period (thirteenth to
sixteenth centuries) when French hesitated between arc, the
expected form from Lat. arcus, which finally triumphed, and
arche in this architectural sense, but there is a survival of the
latter in Fr. arche d’un pont, “‘arch of a bridge”. As already
mentioned, there is also behind Fr. chambre an early vault which
has survived in Fr. cambrer, of Picard origin.

We are now going to examine an entirely new aspect of the

semantic field related to the bow—that of hair and grasses.
Lat. penna means both ‘“‘wing” and ‘“‘feather’’, the latter sense
probably being a secondary development from “‘wing’’. Rus.
nepo, “feather” and “pen” (to write with). appears to have had
a similar history, since it also means “‘fin’”’ (ofa fish) and “blade”
(of an oar), both of which are more closely related to the wing
than to the feather.
Turning now to the Far East, 44 Jp. ha or hane, Ch. Vu,
“wing” and “feather’’, shows precisely the same range of mean-
ing. On the other hand, & jp. ke, Ch. mado is not only the usual
term for “hair” and “fur” but includes “feather” as well. How-
ever, in both Chinese and Japanese it extends further yet. as is
clearly seen from the compound 4% Jp. fwmd, Ch. bumao,
lit. no-hair, i.e. ‘hairless’, commonly used to describe a sterile
or barren land, without any growth.
Nor is this surprising. It will probably have occurred to the
 The Bow 109

reader that we have in the West too a long tradition of considering

vegetation on the land as something comparable to hair on the
skin. The Greeks looked upon foliage as a kind of hairy growth:
Koun, “hair”, is used for “‘foliage’”’.1° The corresponding adjec-
tive Koyuntys, “long-haired”, also means “leafy, grassy”. This
extension of meaning was possibly helped by the fact that the
Greeks originally seem to have worn their hair long, this being
locked upon as a mark of birth: the verb kouéqw does not just
mean “be covered with hair’ but rather “wear long hair’, and
it is similarly used of the leaves of trees. Hence the parallel
twofold sense of Lat. coma, which is probably best known in
comata silva, “‘a forest in full leaf’, directly imitated from the
Greek.
So it was quite natural for Ronsard, steeped as he was in
Greek learning, to write of the verte criniere (lit. green mane)
of the forest in his famous Elégie contre les bucherons de la forét
de Gastine.1© This Classical metaphor is still current, and I have
met a reference to les frondaisons chevelues des marronniers in
the work of a contemporary French writer.!7
As is to be expected, such images are also found in Shake-
speare:
her hedges even-pleach’d,
Like prisoners wildly overgrown with hair
Put forth disorder’d twigs. ..18

A parallel relation exists in Swedish: skog, “‘forest’’, is ultimately

the same word as skdg, “‘beard’’, cognate with Eng. “‘shaggy”’.
Likewise, Fr. pelouse, “‘lawn’’, is originally Lat. (terra) pilosa,
“hairy (earth)’’. On the other hand, a field or hill totally devoid
of vegetation will naturatly be called champ pelé or mont pelé in
French, where pe/é oes not have the sense “‘peeled”’ (Lat. pellis),

15. A rather amusing parallel in reverse to this Greek usage occurred several
years ago in a newspaper article dealing with the ban on long hair imposed
by the authorities in Singapore. The writer commented: “‘It is possible to
prune the offending foliage” (Age, 13 January 1972). This clearly shows that
there is a continuing awareness of the very close relation between hair and
vegetation. (Ed.)
16. P. de Ronsard, Elegie XXIV, in Oeuvres completes (Paris, 1950).
17. I have been unable to trace this reference. (Ed.)
18. W. Shakespeare, Henry V, V, ii, in The Complete Works of William Shake-
speare (London, n.d.)
110 Comparative Semantics

but “hairless” (Lat. pilum), and there is a comparable image

underlying Eng. ‘“‘bare hill” and Fr. colline chauve. So too, Lat.
in nuda valle occurs in a Spanish translation as en un valle pelado.*°
While It. and Sp. pelo, like Fr. poil, are normally limited to
“hair” or rather poil (as distinct from Fr. cheveux, It. capelli
and less strictly Sp. cabellos), the corresponding verbs Fr. peler
and Sp. pelar, at least in their participial form, have been shown
to extend to vegetation as well. It. pelare is worthy of special
notice, since it is used not only for “‘peel’”’ and “remove hair”
(Fr. peler) but in the further sense of “remove the feathers”
(Fr. plumer), a semantic range remarkably similar to that of
Chinese and Japanese & discussed earlier. But it should be
noted that whereas in the East this wide extension of meaning
is found in the noun, in the West it is in the verb.
There is nothing surprising in the identification of hair and
feathers. Strictly, Fr. duvet, ““down’’, belongs to birds, but it is
very often used of fine human hair, as are the adjectives duveteux
or duveté and “downy’’, and even plumeux, “feathery”, can
occasionally be used in this way.?°
There is also abundant evidence of the use of plant terms for
hair. Some of these, for example ‘‘a thick growth of hair” or
“bushy eyebrows’’, are so common that they are hardly noticed,
and similarly a thick tuft of hair will quite naturally be called
in French un buisson de poils (cf. sourcils; cheveux en broussaille).
Comparable too is the very frequent use of such French adjec-
tives as dru and touffu, which belong equally well to vegetal and
hairy growth and refer particularly to thick hair, grass or under-
growth. There are, however, more specific terms, such as Eng.
“stubble” for ‘“‘whiskers’”’, Ger. Stoppeln, ‘“‘stubble”’ and “bristly
whiskers’, and furthermore “‘pin-feathers”, all of which make
up a very coherent semantic group.
It is of interest to observe that in Pidgin English, the beard
is frequently called grass belong face, and in French I have come
across les parties désherbées des jambes in reference to those
parts that had been shaved.?} More remarkable examples are

19. T. Livy, 46 Urbe Condita (Madrid, n.d.), Book XXII, sect. XXVIII, p. 105.
20. See, for example, Ph. Hériat, Famille Boussardel (Paris, 1957), p. 292.
21. The author and the work are unknown to me. (Ed.)
 The Bow 111

sometimes found, thus in Mauriac: /a poitrine feuillue de Pan,?”

and in Hériat: Victorin voyait briller le poil mousseux, fourni.
doré de laisselle,*? where the use of the singular seems to be
something of a coquetterie littéraire. Compare the occasional il
a le cheveu sale for the far more usual les cheveux.
There is the further point that Fr. poil, being so sharply
contrasted with cheveux, often possesses a strong erotic sense
which makes it strikingly different from Eng. “‘hair’ or Ger.
Haar. The result is a tendency to use a variety of euphemisms:
for instance, frondaison, mousse etc. are drawn from the plant
kingdom, while toison belongs to the animal world.
The fundamental relation between hair and vegetation can
be observed not only in its negative form, colline chauve, “bare
hill’, and Jp. fumd 7% but in a further negative aspect, which
‘may be looked upon as an intermediate state between presence
and absence of hair/vegetation. Jp. koke ## is the normal term
for ‘“‘the scales” (of a fish), but there is another koke, #, the _
everyday word for “‘moss”’, which is also found to mean “‘fur”’.
On closer investigation it turns out that the ‘fur’ in question
is that on the tongue, which is about as far from a real fur (Fr.
fourrure) made of hair as can possibly be imagined. The French
don’t seem to have a name for it; they express the same notion
indirectly: il a la langue chargée, which is certainly closer to the
true meaning. The odd thing is that it should be called “fur” at
all in English, for it will be noted that it is undoubtedly more
akin to moss than to hair or grass. Actually the Japanese usually
feel the need to specify this less frequent sense by means of what
might be called an indicator: shitagoke & & (lit. tongue moss);
the Chinese write it the same way, but pronounce it shétai.
The reader will probably have realized that the two Jp.
koke are originally one and the same, and that the differentiation
between “scales” and “moss” is a later development which
corresponds exactly to the semantic relation between hair and
grass: moss, contrasted with grass, is as flat on the earth as the
fur on the tongue and as the scales are on the body of the fish,
whereas hair and grass or vegetation stand out boldly.

22. F. Mauriac, Le fleuve de feu (Paris, 1923).

23. Hériat, Boussardel, p. 423.
112 Comparative Semantics

There is yet a further negative way of looking at all these

phenomena. Both Eng. “‘moult’’ and Fr. muer are used equally
well of the fall of hair and feathers, and even of skin: when
referring to reptiles. They thus correspond to Fr. pe/é in the
senses “‘peeled and skinned” and “‘hairless’’, and to It. pelare
in the further sense of “removing the feathers” (Fr. .plumer).
However, we have to turn to the East, and more specifically
to Mal.-Id. /uruh, for the full semantic range from the fall of
hair and feathers to the fall of leaves and flowers, though here
another term is used in the case of reptiles.
This is reminiscent of comata silva (from “‘hair’’ to “‘leaves’’),
which has already been considered, but what seems to be more
characteristic of the East is the corresponding identification of
feathers and leaves. Sk. parna, cognate with Rus. nepo mentioned
earlier, means ‘“‘wing, feather’? and also “‘leaf’’, and Sk. pattra
shows precisely the same semantic development from “wing”
to “feather” and on to “leaf”. Pattra is actually the same word
as Lat. penna (pet-na), Gk nétopuai, “fly”, better known in the
noun form ztepov, “wing”, from which have been created the
names of various classes of insects, among others lepidoptera
and orthoptera etc. These sound learned in English, but Fr.
coléoptére is nearly as common as “‘beetle”’, while /épidoptere is
obviously learned beside papillon and “‘butterfly’’, just as Jp.
ché is the homely name of the butterfly in contrast to the scientific
term rinshirui (lit. scale-wing species), clearly a calque on /epidop-
tera. RIN is the Sino-Japanese pronunciation of the character
koke, “‘scale’’, discussed previously. What appears to be fine
dust on a butterfly’s wings is in fact made up of neat rows of
scales, and these (like moss on the earth etc.) are flat on the wings.
We have already met the usual term for “wing” or “‘feather”’
in Japanese, ha or hane 4%. In the light of what has just been
observed, it is of interest to note that a “leaf” in Japanese is
also called ha, written of course with an entirely different charac-
ter #. If we now compare the Sanskrit words quoted earlier and
these two native Japanese ha, remembering at the same time what
has been elucidated about the ultimate relation between the
bowstring and the vine by way of the tendril (Jp. tsuru), and the
two Jp. koke that have just been examined, it is impossible
not to arrive at the conclusion that both ha are originally the
 The Bow 113

one word and that we find here again a further example of later
differentiation.

Let us now return to Indonesian, this time to an entirely different

semantic group which is clearly relevant to some of the phenom-
ena just discussed in connection with the bow.
Id. rambut is the usual term for “hair”, and Id. rumput the
most commonly used word for “‘grass, weed, straw’’. Id. djanggut,
“‘beard’’, is also in its place here. All these terms share the final
-ut, and they all belong to a closely related semantic group that
includes both hair and grass, two notions whose sense relation
has been shown to be very widespread. To quote an example
from the Russian poet Voznessensky, interviewed on his arrival
in Adelaide: ‘‘One of the reasons I wanted to make this trip to
Adelaide was to see Allen [Ginsberg]’s new beard. When I met
him last in New York, he’d just shaved his beard off in India,
and on his chin was just small elegant grass.”’?*
As to Id. rambut, “‘hair’’, it is true that there is also Id. bulu,
which covers “‘hair”’ as well as ‘‘feathers” and “‘wool’’, a semantic
range reminiscent of Jp. ke , but this is not the normal word
for the human hair, which is rather rambut. Of course, bulu may
well be on its way towards *bulut, but I have not come across.
this form. However, we noticed in the earlier groups quite a
number of parallel adaptations of sense to form.
There are a number of interesting terms connected with
“hair”, and first must be mentioned Id. kusut, ‘“‘tousled’’, which
meaning extends to “rumpled” and even to “complicated”.
This figurative development is especially arresting in phrases
like kusut musut and kusut murut, ‘“‘very confused’’, where musut
and murut are bound forms, since they never occur except with
kusut. Closely comparable with these is kelut melut, ‘confusion;
complicated”, where melut is indissolubly bound to kelut. There
is also karut, “confused, muddled”’, with the further combination
karut marut, “utterly confused’; marut can occur only in this
combination and does not exist independently.
Thus we have a series of bound forms, most of which differ
from the basic term only in respect of the initial m-, which replaces

24. Australian, 9 March 1972.

114 Comparative Semantics

the first consonant and whose role seems to be purely intensive;

to these can now be added midik, mentioned earlier. Incidentally,
there is a somewhat similar use of initial m- in Turkish which,
transposed into English, would give for example police molice,
“‘nolice, spies, secret agents, private eyes etc.” 75 However, when
a Turkish word begins with m- this mode of expression cannot
be used. j
It is noteworthy that in Indonesian, a corresponding limitation
applies when a word begins with a vowel: this can be illustrated
by another relevant term, intjang-intjut, “zigzag; higgledy-
piggledy”, a sense closely related to “confusion” and again
clearly connected with the significant ending -ut. To these may
also be added such odd terms as sarut and sengkarut, which
indirectly acquired the meaning “confusion”, undoubtedly
because their ending suggested just this sense.
Another type of metaphorical extension is seen in Id. sungut,
“feelers; moustache; whiskers’, which, combined with the
prefix ber- and reduplicated bersungut?, i.e. bersungutsungut,
means “grumble, complain’’, and with this can be taken rengut,
‘be morose, sullen, grumbling’. The sense development en-
countered here may well have been along the lines of our figura-
tive use of “‘bristle up” or something like “that’ll make him
bristle”. Relevant here too is Id. temut?, i.e. temuttemut, “to
twitch”, which develops the further sense “to mumble’. In
-all of these the suggestive value of the ending -ut is once again
unmistakable. A comparison with Eng. “‘ruffled feelings” or
“ruffled tempers” and with “unruffled” is also illuminating,
especially in view of the primary sense of “‘ruffle”’.
There are two verbs unggut and rengut which, combined with
the prefix me-, are used for “‘pull, tug”. The latter seems to be
most frequently used of “tugging at one’s beard’’; nor is this
surprising in view of the ending -ut. What is curious, however,
is that every one of the earliest uses of Fr. firer, i.e. in the Chanson
de Roland, is found in reference to the beard. It is well known
that the normal verb for “draw, pull’? was then traire, which
was later supplanted by the more regular and thus more manage-

25. It is an interesting curiosity that Id. mata, “eye”, when reduplicated mata-
mata, has the sense “‘spy, detective’.
 The Bow is

able tirer, just as atraire was by attirer etc., while traire has
survived solely in the limited sense of “milking” (a cow).
Id. perut, “belly, stomach” (hence “‘pregnant”’), appears to
have been so called originally because of its hairy covering.
The combinations perut betis and perut kaki for “‘the calf of the
leg” throw further light on the true history of this particular
calf and its relation to the calf of the cow. The sense “‘pregnant”’
should be carefully noted; it will be recalled that there is remark-
able evidence on this point in Dutch, Russian and Japanese.
If we now turn to rumput, “grass”, there are several terms
that immediately demand attention, for example Id. kasut,
often “‘shoes”’ but actually “‘straw sandals”; with this may be
compared Jp. zori #}&, where the first character means “grass,
straw”, and thus corresponds to the ending -ut of kasut. To
distinguish between leather shoes and straw sandals, whenever
this is called for, it is quite common to refer specifically to the
kasut rumput, which leaves no doubt that the latter are meant.
There is also bubut, used originally of “‘pulling out grass or
weeds”. However, the most striking term is undoubtedly /umut,
which covers both ‘‘moss” and ‘mildew’. We have already
examined Jp. koke, whose meaning extends to both “moss” and
“the scales of a fish”, though with a different character for each.
At this point the difference between hair and vegetation practical-
ly vanishes, and it is therefore not surprising to find that to the
Chinese ‘‘mildew” is 8 = bdimdo, lit. white hair.
The reader would now do well to revise what was said earlier
about kulit, “‘skin’’, its semantic range, and other comments
on the relation between the skin, the hair and the flesh. Just
what hair is to skin will be more clearly understood from further
Indonesian terms that bridge the gap between the two notions:
purut, “rough-skinned”’, and kerut, “furrow; wrinkle’, together
with kerut merut, where merut is another bound form existing
only in conjunction with kerut, “be wrinkled’’, and also “dis-
hevelled”, which is very reminiscent of kusut, ‘‘tousled”’, already
discussed. That the two senses ‘“‘wrinkled’’ and “‘dishevelled”’
can be found in the same word is illuminating and should help
us to understand the fundamentai relation between skin and hair.
Id. selaput, ‘““membrane’”’, obviously belongs here, and it
may be noted in passing that selaput perut (lit. membrane of
116 Comparative Semantics

belly) designates “the peritonium”. Id. ketjut is even more

arresting: from its primary sense “wrinkled, shrunken” it
acquires the meaning “‘sour’’ (said first of the face, then later
of taste by a natural extension), and from the notion of “‘shrink-
ing” there is a striking development towards that of “fear, scare,
cowardice’. It should be helpful to compare our expression
“it gives me the creeps’.
There is another aspect of the skin that will not be difficult
to grasp, since it has been shown that the belly is essentially a
bag. In fact, nothing could be more natural than to look upon
the skin as a bag, a wrapper which covers the body. This view
of the skin is particularly clear if we consider certain membranes
such as the pericardium, in which the heart is wrapped up, and
the pleura, which plays the same role for the lungs; and there
is of course the midriff or diaphragm, and further the peritonium,
which has just been mentioned in its Indonesian form selaput
perut, doubly relevant here.
This should throw some light on the sense and related form
of the following Indonesian terms in -ut: liput, in the derived
form meliputi, “‘cover, wrap’, and salut, “envelope, wrapping”
whose me- derivative menjalut is practically synonymous with
meliputi. Saput, ‘“‘a cover, shroud, veil’’, clearly fits here, as
does the corresponding me- form menjaput(i), ‘““cover, muffle,
drape’.
However, the most unexpected term that must be cited here
is probably semut, “‘ant’’. This seems to be an incredibly long
way from hair or skin, semantically speaking, and yet it is in
its place here. This could well be a fascinating semantic problem,
but the mere fact of putting the ant in this context is a good hint:
any housewife who has seen her kitchen invaded by ants, whose
serried hordes literally cover everything they crawl over, will
have no hesitation about the relation between “‘cover’’ and
‘ant’. It is only a further step along the line of semantic develop-
ment we have been following. Another Indonesian pointer to
this is in the verb menjemut, “swarm, teem’’, closely comparable
in form and sense to meliputi, menjaput(i) and menjalut, just
quoted.
Moreover there are the well-known Romance equivalents
derived from the descendants of Lat. formica, “ant”: Fr.
 The Bow ae

fourmiller; \t. formicolare; Sp. hormiguear etc., all of which

are synonymous with Eng. “‘swarm’’. Another fascinating aspect
of the same phenomenon may be observed in the fact that Id.
orang menjemut ditanah lapang itu, “the square was swarming
with people’’, may be rendered in French as /a place était noire de
monde, a use of noir which in this case is literally and metaphori-
cally true and whose exploration in Chapter II took us from
Japan to Portugal.
Before we finally take leave of the ant, it hardly needs to be
pointed out that the partial phonetic similarity of semut and
emmet, the earlier form of “ant” (cf. Ger. Ameise), must be
regarded as purely fortuitous, though here again there are
absorbing semantic phenomena that could be explored.
Id. sabut, “husk’’, requires no comment, since its relevance
to the “wrapping” terms just examined is clear. The same may
be said of selimut, the normal word for “‘a blanket”, from which
is derived menjelimut, ‘‘cover, hide”. Id. tjantjut, “loincloth’’,
finds its proper place here quite naturally, and so does balut
(or barut), which basically means “a bandage, wrapping”, and
which also acquires the sense “‘red and swollen” in referring to
the eyes. This latter extension probably requires further elucida-
tion, but the reader may profitably speculate on this point until
we come to it.
Just as skin can be smooth or rough, chapped or wrinkled,
so can hair range from the softest fur to the thick hair of the
horse’s tail and mane, or even to the spines and prickles of the
hedgehog and porcupine. So it is not unexpected to find that
beside Id. Jembut, the usual word for “soft, smooth’, there are
a number of other terms that must be interpreted in the opposite
sense of a projecting or protruding object. Thus Id. mulut,
“mouth”, does not seem to be so called from the hair that may
accompany it, but rather from the sense of “pouting”, an
interpretation reinforced by the presence of a derivative of
kerutjut used of “pouting lips”, and the fact that most Romance
words for “mouth” (Fr. bouche etc.) are from Lat. bucca, “puffed-
out cheek”. There is the additional evidence that from Id.
kenjut and sedut are formed me- derivatives meaning “‘to suck”’.
In sucking, as in pouting, the mouth is thrust forward in a
characteristic manner.
118 Comparative Semantics

There is also Id. kedjut, whose usual sense “frightened” can

be traced back to a dog’s ears “cocked up” in alarm. These
various observations suggest that Id. perut, ‘“‘belly’, may not
have been originally so named because of its hairy covering, but
that the sense “‘pregnant’? may well have been the primary
meaning.
In favour of this point of view could also be quoted burut,
“hernia, rupture’, caused by the bulging of an organ through its
wrapping membrane selaput, beside untut, that enormous
thickening of the skin known by the name of “elephantiasis’’,
and furthermore the combination perut gendut, ‘‘big-bellied”’,
whose gendut has itself the sense “‘corpulent, pot-bellied’’, as
well as “pregnant”. It will now be appreciated how balut,
“bandage, wrapping’, comes to mean “‘red and swollen” in
referring to the eyes.
However, the most arresting examples of this type of sense
development are to be found externally. Thus Id. sudut, “angle;
corner”, and kerutjut, ‘a conical object”. From sangkut, a
derived sangkutan refers to ‘‘a hook” or “‘peg’’ or again to ‘“‘an
obstacle”. Actually this particular development extends to
busut, “‘an ant hill”, hence “‘hillock’’, and even to djungut,
“a mountain spur’. The passage from ‘“‘ant hill’ to “hillock”
is reminiscent of the molehill and its relation to the hill, which
was discussed in reference to bukit, ‘“‘hill’’.
There are even more striking parallels with these features if
we consider “‘water’’. The sense of Id. kedut extends from “fold,
crease’, to “ripple”. It is but a short step from “wrinkle” to
“ripple”, and a good example of the transition occurs in La
Fontaine’s well-known fable “‘Le Chéne et le Roseau’”’:
Le moindre vent qui d’aventure
Fait rider la face de l’eau,
This is an image that has become extremely common in Modern
French.
It is, however, in Id. /aut, the usual term for ‘“‘the sea”, that
the most remarkable example of this phenomenon is to be found.
The Chinese character for “the sea”, @£, is made up of the water
radical 7 together with ¥ ‘‘sheep’’. In French the white-crested
waves are also normally called moutons, “sheep”, while in
English clouds are often said to be “‘fleecy”. In a magnificent
 The Bow 119

sonnet, Floridum Mare, by Heredia, the sea is strikingly described

as being blanche de moutons (lit. white with sheep). This recalls
our “‘white horses” for the white-capped waves. The Aegean
Sea is undoubtedly so named because its appearance suggested
herds of goats, Gk at'yec. Rus. eoana, though with a difference
in the position of the stress, refers to both “wool’’ and “‘waves”’.
Still more arresting is Rus. 6apaiyxu , whose basic sense “young
wethers”’ extends to “‘fleecy clouds” as well as to “‘white-crested
waves’.
We catch a tantalizing glimpse of a further aspect of Id.
laut in another Chinese character for “‘the sea’’, #, and its
apparent relation to # “‘every” and & “‘mother’’. Let us think
of the teeming seas and recall the obvious association between
water and abundance, which is borne out by innumerable terms
and expressions in most languages. The reader will find quite
a number of these almost at once: in fact, the word ‘“‘abundance”’
itself is a clear example, and this will suffice among so many.
It is more immediately relevant to note that the names which
designate “the waves” are often identical with “pregnant’’,
especially in the West, and more specifically in Greek. A close
parallel is found in Eng. “‘billow’’, fundamentally the same word
as “bulge” and “belly”. John Orr made the interesting observa-
tion that the “billowing sails’ is a popular or landlubber’s
etymology for the sailor’s ‘“‘bellying sails’’,*° and it is still more
curious to discover that in Japanese, sails swollen with wind are
literally said to be ‘‘pregnant’’. It was mentioned earlier that Jp.
haramu, ‘“‘be pregnant”, is derived from hara, “‘belly’’, and another
interesting point is that it is used not only of sails but of clouds
“big with rain” as well.
In view of all this, it is hard to resist the temptation to speculate
about the possibility of an influence of /a mére on la mer, where
we would normally have expected */e mer, from Lat. mare.
This relation between the belly and the sea may recall a powerful
image in Tristan Corbiére’s famous poem La Fin, which is
unmistakably a retort to Hugo’s even more famous Oceano
Nox:

1953), p. 4.
26. J. Orr, Words and Sounds in English and French (Oxford,
120 Comparative Semantics

Voyez a l’horizon se soulever la houle;

On dirait le ventre amoureux
D’une fille de joie en rut, a moitié sodle...
Ils sont la!—La houle a du creux.

Here it should be observed that /a houle is ‘‘the swell’’, which to

some extent again suggests the “‘pregnant”’ billows.
It has probably occurred to the reader that, by a further
extraordinary coincidence, the Indonesian suffix -wt is remarkably
similar in sense and form to a widespread Romance suffix that
can be traced back to such Latin words as nasutus, “having a
large nose’’, and hirsutus or Low Lat. canutus, which refer to hair.
Hence, on the one hand, Fr. ventru, “‘pot-bellied’’; pansu,
“big-paunched”’; joufflu, “‘chubby-cheeked’’; Jlippu, “‘thick-
lipped”; charnu, “fleshy”, or better “plump” (which can be
contrasted with charnel, and décharné, “‘fleshless, scraggy’’),
as well as the less common fessu, mamelu and boulu,?7 all of
which are semantically to ventre, panse, joue, lippe etc. what
nasutus is to nasus; i.e. there is abundance of flesh.
On the other hand, in Fr. chevelu, poilu, velu, barbu, moustachu
beside herbu, feuillu,. moussu, dru, touffu etc. there is abundance
of hair or grass. It is certainly no accident that here again we
find the same fundamental relation between flesh and hair/grass
that we observed in Indonesian, present in the same suffix. This
is clearly due to an unconscious awareness of the essential unity
underlying those ideas, a point that was mentioned incidentally
in our discussion of Id. kulit, “‘skin’’.
Mlle Durand, in her remarkable study on Le Genre gram-
matical en francais parlé, quotes an oral example from ‘‘une
femme agée de petite bourgeoisie”’ : il sera mieux comme ¢a qu’avec
des cheveux tout hirsuts, and she comments that hirsuts is a
“masculin insolite formé daprés le méme modéle que apostat,
violet, idiot, par fausse interprétation de la forme unique com-
portant une terminaison occlusive’.*® Mlle Durand is essentially
a phonetician, and she interprets this aberrant word in the light
of her phonetic experience. Hirsut is clearly a borrowing, actually

27. This last is found in A. Rimbaud, Les Assis. It is also worth noting a further
possibility, /evres peaussues, which occurs in Villon’s Testament, |. 516. (Ed.)
28. M. Durand, Le Genre grammatical en frangais parlé, a Paris et dans la région
parisienne (Paris, 1936), p. 119.
 The Bow 121

a fairly recent one, and it stands alone among all the other
adjectives relating to hair. Hirsut (which should rather have been
spelt hirsu) is undoubtedly formed on the analogy of all other
-u words associated with hair. It is not a phonetic, but a semantic
phenomenon: the deviant form is brought back to the norm.
I have very good reason for this interpretation, as some time
ago a student of mine did in fact create the form hirsu quite
spontaneously and I was able to verify how it had come about.
Further light is thrown on the same point by the expression des
cheveux bien brossus (brossus for the regular brossés), which I
also owe to another of my students—which demonstrates how
instructive and illuminating certain mistakes can prove to be.
Returning briefly to the French ventru series where the notion
of bigness is contained in the suffix -u, it should be observed that
ventru could not possibly mean “‘bellied’”’, a sense as utterly
unthinkable as “legged” or “eyed”, though it is quite normal
to find such formations as two-legged, four-, six-, eight-legged
etc., while three- or five-legged would be decidedly odd; one-
legged, however, is not so odd, nor is one-eyed. So little odd is it
in fact that the French even have a special word for it: borgne,
while the Latin expression ocu/o captus gives at least a hint of
the sense. All this opens up new semantic prospects and throws
light on the presence or absence of certain words. The underly-
ing notion is of course that it is taken for granted that everyone
must have a belly, eyes, legs etc. Similarly, one can be long-
eared or lop-eared, but not just “eared”, and further, chubby-
cheeked or rosy-cheeked, but not just “cheeked” etc. And it is
another aspect of the same phenomenon that beside white-
skinned, red-skinned etc. there is such a word as skinned, but
with a very different meaning. Here again we have the fact of
presence or absence from a new angle. However, all this is so
obvious that it needs no further comment.
There is no doubt that Rabelais was fully aware of the bigness-
content of the final -u in Pantagruel: le ventre leur devenoit bossu
comme une grosse tonne...?° Next he only just misses membru
and couillu, but goes on to a pun based on jambus and Iambus.
Couillu, which is common enough in regional French (at least

29. F. Rabelais, Puntagrue/, chapter I, in Oeuvres completes (Paris, 1962).

122 Comparative Semantics

in Belgium), may not be found in all French dictionaries, as

most lexicographers tend to feel somewhat squeamish about
such terms; but that is no excuse for leaving them out.
It will be appreciated that dodu, “‘plump”’, and trapu, ‘‘thick-
set’, also belong to this semantic family, though these terms
are not derived from names of parts of the body and their
meaning seems to be essentially determined by their ending. The
form of Fr. bossu, ““hunch-backed”’, will be seen to be especially
apt, though here again a bosse is not normally a part of the
body. Fr. repu, too, originally the participle of repaitre, has been
attracted to this group largely because of its ending and its
meaning ‘‘well-filled with food, sated’’.
There are a number of interesting secondary sense develop-
ments: thus Fr. tétu, which has come to mean “‘stubborn”’ (cf.
qui en fait a sa téte, qui a la téte dure, s’entéter, It. incaponirsi);
goulu, originally ‘““big-mouthed”’ (gueule), hence “‘greedy”’; cossu,
originally ‘“‘whose pods are well filled”, hence “‘very well-off’’;
and bourru, “cross-grained”, whose primitive sense was some-
thing very much like “rough as bags’’.
The semantic history of tétu is exactly parallel to that of It.
cocciuto, “‘stubborn”’, based on coccio, ‘“‘a pot-sherd’’, which is
also the original sense of Lat. testa, the ancestor of téte. The
Italian ending -uto is the same as Fr. -u (Lat. -utus)3° and in
Spanish it has turned into -udo, as can be expected. Thus to Fr.
pansu correspond It. panciuto and Sp. panzudo, and they all
mean “‘pot-bellied”. Similarly, a person who has plump leg-
calves is said in Italian to be polpacciuto and in Spanish pantor-
illudo, where it will be observed that to a Spaniard those calves
are like “little paunches”’, and from this to the pregnant cow

30. The corresponding Catalan suffix ends in -ut (masc.) and -uda (fem.). The
four examples I have met are equally distributed between hair (barbut,
“bearded”; pelut, “hairy”) and flesh (galtut, ‘“chubby-cheeked”; geperut,
“hunch-backed”). With geperut compare Sp. giboso and It. gobbo, whose
first o appears to be due to the following labial (Lat. gibbus).
Latin had two suffixes, -utus and -osus, referring both to flesh and hair/
grass. Thus nasutus beside hirsutus and Low Lat. canutus, villutus etc., and
on the other hand ventruosus, gibberosus beside the more common pilosus,
villosus, herbosus, frondosus etc. Cat. geperut is clearly from a gibberutus as
alternative to gibberosus, just as villutus was created on the model of villosus.
The suffix -osus was more extensively used than -utus: umbrosus, ventrosus,
petrosus, arenosus etc. Herbosus regularly becomes Fr. herbeux.
 The Bow 123

in calf is but a short step. As far as I know Fr. *molletu is still

in the process of gestation, but its birth is almost overdue and
cannot be long delayed.
Fr. velu, “hairy”, goes back to Low Lat. villutus, derived
from the Classical villus, “‘hair”’, which also turned into It.
velluto, ‘“‘velvet”’, and Sp. velludo, “‘hairy’’, whence is formed a
diminutive, velludillo, which refers to a kind of velvet. Pt. veludo,
“velvet”, is found as a borrowing in Jp. birddo, which is the
usual term for “‘velvet’’. Eng. “‘velvet” is itself originally a di-
minutive from Fr. velu, i.e. velu-et.>! It is also relevant to note
that Fr. velours, “‘velvet’’, goes back to an earlier velous (Lat.
villosus) and OFr. velos, whose first appearance is in the dimin-
utive veloset, the earliest name for “‘velvet’’.
If we now return to Id. perut, “‘belly”’, it will be appreciated
that my first interpretation, “hairy”, is not contradicted by my
later one, “‘bulging’’, since they essentially correspond to the
twofold aspect hair/fiesh of the same fundamental unity that
I originally elicited from the European evidence. Flesh, skin, and
hair/feathers represent the successive layers cushioning the vital
organs. An exact parallel can be discovered in Id. Jaut, ‘“‘sea’’,
and in the corresponding Chinese characters 7 (based on the
sheep) and ## (based on the mother). Fr. moutons, ‘“‘white horses’’,
and other examples of the same type already quoted represent
the “‘hair’’ aspect of the sea, whereas /a houle, ‘“‘the swell’, and Ja
mer est grosse are closer to the “mother”’ view.
The notion of successive layers that closely fit the organs
to cover and protect them is further confirmed by a large number
of Indonesian -ut terms whose general sense is “closely joined,
fitting’, and which have in turn developed secondary meanings;
however, an examination of these would take us too far outside
our main topic, though they also are of great semantic interest.
As to the coincidence of Lat. -utus and Id. -ut, it seems clear
that this Indonesian suffix is made up of two distinct parts: the
vowel u and the final ¢. It should be noted that a number of
terms closely associated with hair do not end in -ut, but that

31. It is quite common to find that various kinds of material have names that
are diminutives in form: our own “‘cotonnette’’, “flannelette”’, and “drugget”’,
beside the earlier Fr. blanchet and the original bureau, i.e. burel, from bure,
are examples, in addition to Sp. velludillo quoted previously.
124 Comparative Semantics

they do have this vowel, bare or with another final consonant.

Bulu, ‘“‘hair, feathers, wool”, has been mentioned before, but
here it should be more definitely specified that while Id. rambut
refers to the hair on the human head (Fr. cheveux) bulu, apart
from its extension to wool and feathers, corresponds rather to
Fr. poils. There is also Id. daun, ‘‘tree leaves”, and the relation
between leaves and hair has been shown to be widespread, both
in the West and in the East. A further point, especially relevant
here, is found in the form and sense of Id. Juruh, whose me-
derivative is used of “shedding” (hair, fur), ““moulting”’ (fowls),
and the “‘shedding”’ of leaves and flowers, a fine example of the
link between hair, feathers and leaves. Id. rimbun is equally
frequent in the senses “leafy” and “thick hair’. In each case
observe the ending, with its significant final vowel u common to
all the terms of this semantic group, of which -ut is now seen to
be only a particular variety, though a very important one.
I have long held the view, and practical experience in teaching
and exploring the subject with students has confirmed it, that
semantics, and more specifically this comparative semantics,
while dealing first and foremost with meaning, is inevitably con-
cerned as well with what would normally be regarded as the
preserve of grammarians and phoneticians. And as I began to
think and to write about the Indonesian ending -ut, the same
point was again brought home to me.
It has been mentioned that this suffix is made up essentially
of the vowel u plus a final -t. Now, as I pondered the various
-ut terms I had come across, the significance of Id. Jaut, ‘“‘sea’’,
gradually became more and more striking, and when the Chinese
character came to my mind, with the sheep so obvious beside
the water radical %, I had little doubt that /aut was relevant
and that it clearly belonged with rambut, “‘hair’’.
However, as I wrote, an interesting problem of pronunciation
began to dawn on me. I should mention here that my knowledge
of many of the languages I have explored has been acquired
from books and that I have had little opportunity to hear them
spoken. Naturally I have some notion of the pronunciation as
far as I can gather it from books, and as a practical linguist I
keep this living aspect of language always in mind. But since
 The Bow 125

semantics is my chief interest, the approximate pronunciation I

have thus acquired for myself is satisfactory enough.
This preamble concerns especially Id. /aut, “‘sea”’, and daun,
“leaf”. I had discovered that au is officially pronounced as in
Ger. Haus, which is practically the same as Eng. “‘house’’, but that
in colloquial Indonesian this diphthong tends to become a
simple o. The passage from aw to 9, like the parallel aj to e, is
the most natural imaginable and examples are to be found all
the way from Latin and French to Sanskrit and further.
As I went on writing this I gradually realized that daun, which
I pronounced like Eng. ““down’’, and in particular /aut, which
for me rhymed with “‘shout’’, could only belong here if those
words were pronounced as disyllables and the vowel wu given
its full independent value, like the vowel in “foot”. My first
Malay book, the excellent Teach Yourself Malay by M.B. Lewis,
was rather misleading on this point as it listed au among the
diphthongs, even in /aut, whose pronunciation the author marks
thus with an acute accent. It is true that she mentions “a slight
hiatus’, but the accent on a again suggests a pronunciation
similar to the vowel in “*house”’.
It was not until I happened to glance at Miroslav Oplt’s
Bahasa Indonésia (Prague, 1960)—an excellent book too, and one
that possesses the exceptional merit of being written in both
Czech. and English—that I discovered that in words like /aut
and daun, where au is in a closed syllable, i.e. where it is followed
by a final consonant, both vowels keep their own separate value,
a and u, which admirably fits the semantic requirements that
have been elicited in this piece of research. In fact, this proves
the symbolic value of the final vowel: in this case it remained
intact, despite the natural tendency of the combination au to
be reduced to aw or 0.
I suspect that the stress shown on /dut, as given by M.B.
Lewis, is rather the result of the contrast between the open a
and the darker, muffled u sound, which by its very nature would
appear unstressed, since it would be much less audible than the
a. Further light on this point can be obtained by considering
the French pronunciation of Raoul, mahout, or raout (which is
Eng. “rout” borrowed), and on the other hand saoul or aout,
where the “‘natural’’ reduction has taken place.
126 Comparative Semantics

It is curious to reflect that my pronunciation of daun, “‘leaves”’,

exactly like Eng. “down”, was partly motivated by the feeling
that there is an essential unity between leaves and hair, which
has been shown to be present in the most diverse languages and
cultures; though this feeling was right, it is nevertheless a pure
coincidence that Eng. ““down” has to do with hair and Id. daun
with leaves.
Since the rambut, “‘hair’’/rumput, “grass” couplet offers fur-
ther evidence of the relation between hair and grasses that
was previously examined in connection with “the bow”, it im-
mediately attracted my attention, and it thus became the starting-
point for this last section. Laut, ‘‘sea”, and daun, “‘leaves”’,
only came in gradually as I continued to explore the field, and
the final points concerning the pronunciation of these terms also
proved to be relevant to the central topic, since they demon-
strated again that vowels have a remarkable symbolic value in
Malay and Indonesian. The light that was thrown incidentally
on the true nature of the Chinese character for “the sea’, #,
could not come until Id. /aut was clearly recognized to belong
to the group under examination, and this in turn led to a con-
sideration of the other Chinese character for “‘the sea”, ##, and
related phenomena.
My purpose in including these explanatory remarks is to
point out that while I can see beforehand the general direction
of my research and IJ anticipate a great deal, there is still a large
amount of entirely new material that comes to light as I examine
a field more closely and I become gradually more aware of the
nature of the subject I am exploring.

We are now coming upon an entirely new world, that of bulges

and protrusions. It will be recalled that we have already met
bulges in examining hair and grasses, since we were led on that
occasion to an awareness of the identity underlying the apparent-
ly twofold nature of the Latin suffix -utus and of the correspond-
ing French suffix -u, twofold in the sense of abundance of hair
as in velu, side by side with abundance of flesh as in ventru,
joufflu etc., and we observed similar phenomena in other
Romance languages. We also drew attention to the interesting
fact that hair, grasses and feathers etc. have at various times and
 The Bow 127

in quite independent cultures been looked upon as being largely

the same type of object.
Our present discussion leads us to a vowel system that could
to some extent be discovered from a priori considerations. Thus
it might be said that bulges have depth, and since the vowel u
is the deepest it is no accident that we first came upon bulges in
examining the final -ut of hair and herbs. And it so happens
that our present vowel is u, or rather u together with a back
nasal consonant which seems to give it greater depth, i.e. the
final -ung. Malay, on the other hand, tends to favour the equi-
valent ending -ong, and since this too is fairly current in Indo-
nesian, we shall therefore take -ong as a possible alternative to
-ung.
As our first example of Indonesian bulges, consider Id. pun-
tung, “blunt; butt, stub, stump’’, hence ‘“‘maimed, crippled”, an
interesting semantic extension, and beside this term let us set
Id. guntung, “blunt” and “short”. Though “short” here seems
to be purely physical and is used, for example, of short-sleeved
(shirts), or shorts, meanings that are in turn comparable with
“‘maimed”’, just mentioned (cf. Fr. manchot “one-armed” and its
relation to manche, “‘sleeve’’), it is certainly relevant to reflect
on our own figurative ‘‘blunt” and its near synonym “curt”,
whose particular sense has grown out of the earlier physical
“short”, still seen in Fr. court. To these can be added the
phonetically expressive Id. njonjong, whose me- derivative means
“‘protrude”’.
That which protrudes need not be rounded, but may well
taper to a point. Here belong Id. rantjung, “sharp, pointed”’ ;
luntjung/londjong, “tapering”; tuntung, “point” (of a needle);
gondjong, “tapering”; and the widely used udjung, “end, tip,
point, top’, from which are derived a number of words used to
express such notions as “tip” (of nose, finger, nipple, tongue),
“end” (of road, month), and “cape”, whose special relevance
will be appreciated more fully later. It should be noted inciden-
tally that Eng. “point” or ‘“‘end” and Fr. pointe are quite common
names for this geographical feature.
There are many ways of carrying an object, especially in
the East, but it is highly significant that when the object is
carried in such a manner as to protrude from the person, the
128 Comparative Semantics

word ends in -ung. Thus Id. usung, “carry on the shoulders”,

and dukung or gendung, “carry on the back’’, whence dukung
develops the further sense of “supporting” (one’s parents),
reminiscent of Jp. ou, & 9 “carry on the back” and ‘“‘owe’,
and the relation between “exonerate” and ‘“‘onus”. There is
moreover Id. gotong, “‘carry a heavy burden together’, and
djundjung, “‘carry on the head’’, and from this “hold in high
esteem’’, which is comparable with a number of Chinese and
Japanese terms for “‘high esteem” based on “‘heavy’’, and “‘low
esteem’’ based on “‘light”’: for example, Jp. omonjiru, HA U4
“respect”, from omoi, HV. “heavy”, and karonzuru, #3 4
“look down on”, from karui, #£\. “light”. Anyone who has
been in Japan and observed the strange humpbacked form of
Japanese women due to the bulky obi on the back will readily
appreciate some of the protrusions described here.
Even a lateral protrusion seen as a “leaning” object calls for
a word of this type, dojong or tjondong. The latter is particularly
interesting in its metaphorical extensions towards “leaning”,
i.e. “sympathy for’’, and this probably explains the form of such
terms as gandrung, “‘be in love’’, beside gadung, “‘be madly in
love”. We have obvious Western parallels in the figurative use
of Eng. “leaning”, Fr. penchant and Ger. Neigung, whose ending
-ung bears only a coincidental resemblance to the Indonesian
ending we are now examining.
A more forceful type of protrusion is found in Id. atjung,
“hold up one’s hand; point to”, and ambung, “toss, throw up”,
a derived form of which is used of a boat being “‘tossed on the
waves’, a meaning it also shares with apung and lampung, “‘float”’,
with the implied sense of “being tossed” or “bobbing up and
down”.
Another kind of bulge can be noted inId. rebung, “bamboo
shoot”, and djagung, “‘corn”, both with me- forms meaning
“to sprout”, from which djagung develops the further extension
“appear’’, i.e. “come into view’’. Instances of bulges that are
not so easily recognizable are to be found in Id. sambung, ‘‘joint,
extension, connection”, and hubung, “connection, relation”:
these senses at first seem to have nothing to do with our present
concern, but if we take into account the other terms just men-
tioned (rebung etc.), together with Jp. fushi, Si “joint”, and
 The Bow 129

Sino-Jp. kansetsu, B48 ‘joint, articulation” that have developed

out of the original swollen bamboo joints from which the new
shoots emerge, such a relation becomes very probable, especially
in view of the fact that # is based on the bamboo radical ‘1.
There is quite a different type of protrusion discernible in
Id. agung and ulung, both meaning “excellent, exalted, out-
standing”. On reflection it will be seen that “outstanding” gives
the key to both the form and the sense of these terms, since that
is “outstanding” which literally “stands out” among the many.
The same observation could be made if we took either “excellent”
or “exalted” as a starting-point, for this would bring out the
Latin words from which both are ultimately derived, with the
common notion “height’’.
Comparable with agung and ulung is Id. bung, whose semantic
range from the most respectful to a familiar form of address
seems extraordinary until we remember that we have something
almost identical in our prefix “arch”, which has similarly
preserved terms belonging to both extremes: at the one end we
find the archbishops and archdukes, and at the other the arch
villain and arch fiend.
The honorific use of bung inevitably brings to mind Bung
Karno, a particularly respectful way of addressing the late
President Sukarno, and this in turn throws light on the true
nature of many Indonesian surnames. It cannot have escaped
notice how many names—other than those of Arabic origin,
which are clearly identifiable by their triliteral characteristic
or certain unmistakable Semitic prefixes—begin with su: now
we have Suharto after Sukarno, almost as if the form of his
name had marked him out to become Sukarno’s natural suc-
cessor, a point which, curiously enough, has never been com-
mented on, at least to my knowledge, though I have long been
conscious of this quaint item of apparent determinism. Then
we come across Sunarto and Sukarma, beside Sunarti and
Sumantri etc. These names are obviously of Sanskrit origin and
the initial suis a “‘lucky”’ prefix, since it helps to form innumerable
Sanskrit compounds which all have to do with goodness. The
opposite prefix dus- is also well known in Greek, and to us
through borrowings such as “eupeptic”’ contrasted with “‘dyspep-
tic’. Both particles are found side by side in the Indonesian
130 Comparative Semantics

phrase suka dan duka, “happiness and sorrow’’. That the bulk
of Indonesian names are either from Sanskrit or Arabic is not
surprising in view of the successive layers of borrowings that
have gone into the make-up of Malay and therefore Indonesian.
There is an unmistakable bulge, in fact the most conspicuous
of them all, to be found in the usual Indonesian name for “the
mountain’’, gunung, and beside this must be placed Id. Jombong,
‘“‘mine’”’. The close relation between a mountain and a mine is
well known to the many Europeans who have observed that
the presence of a mine is generally advertised by a hill seen from
afar. This is especially striking in a rather flat country, and I
remember vividly my earliest impression of what I was told
was un charbonnage: no mine was visible, nothing in fact but
those huge hills. Once again, linguistic evidence must be decisive,
and it must be a matter of common knowledge that Ger. Berg,
“mountain”, also means “mine” and that Bergleute covers
both “‘mountaineers”’ and “‘miners”. There is a further parallel
in Slavonic languages, in Polish (where it may be due to the
German usage), and also in Russian where eopuoiii, the adjective
derived from zopa, “mountain”, is applied to mines as well as
mountains, though here too the influence of German cannot
be wholly ruled out, since mining techniques (or know-how
as they say nowadays) frequently come from another country,
in this case Germany, and the relevant words usually follow
or are imitated in the form of calques.
Let us now consider bulges connected with houses, for example
Id. gedung, ‘“‘a big building’’, beside barung, which is a mere
“hut” or “a temporary dwelling”, and the conspicuous tjerobong,
‘‘a chimney” or “‘smokestack” or again “funnel” (on a ship).
Houses close together form the Malay kampong, ‘“‘a cluster of
houses”, making up a “hamlet” or “‘village’’.
If we look at a Malay house more closely, we may first discover
the andjung, “an extension or annex’’, or more precisely a room
that “‘juts out” forming a covered verandah, but our attention
is more likely to be drawn to the bumbong, “roof”, and the
derived Id. bumbungan refers to “the ridge of the roof’’. There
are Other similar names for this protruding part of a house, such
as Id. rabung, “the ridge of a house’, bubung, “ridge, top”, and
more specifically “‘ridgepole”, and the derived bubungan, “‘the
 The Bow 13]

rooftop of a house”’; from bubung is also formed a me- derivative

with the more general sense ‘“‘to rise, go upward”’.
While considering house bulges it may be appropriate to
cite Id. panggung, ‘a platform, stall, grandstand”’ or ‘“‘a scaffold-
ing” or again ‘“‘a balcony” (in a theatre), a range of meanings
to some extent reminiscent of Fr. perron, originally ‘“‘a big stone”
(Lat. petra), whose architectural development in Modern
French, beside its use in Dutch and German as “a railway
platform”, has some semantic features comparable with Id.
panggung.
A kampong is a “cluster” of houses. Similarly, Id. gabung
refers equally well to “‘a bunch” (of flowers), “‘a gathering”
(of people), “‘an association” (of students) and to “a coalition”
or “federation” in a political sense. Again, while Id. bondong
designates ‘‘a crowd of people’, Id. gebung is used of “‘a bundle”
(of notes), gelung is ‘“‘a hairknot” or “‘bun’’, and the derived
gelungan refers in addition to “‘a coil” (of rope). Clearly these
are all various types of clusters, linked together by the common
ending -ung.
We now come to the universally known sarong, that garment
of Malayan origin which has acquired such an immense vogue.
The interesting point about the sarong is that it is not a garment
originally but a sheath, and with a suitable indicator it refers
just as well to ‘“‘a scabbard, a glove, a pillow case”’, and even “a
thimble’’. It is instructive to recall that Fr. fourreau, ‘‘sheath’’,
also designates a garment, and so does Eng. “‘sheath’’, probably
as a result of this particular French usage. Thus a sarong is
found to be both a cover and a container, two notions that
are to some extent complementary.
As a container, sarong may be placed beside quite a number
of Indonesian names of containers ending in -ung, such as
tabung, ‘“‘a cylindrical box of bamboo” or “a money box’’;
tjambung, “a large basin’; rombong, “‘a big rice basket with
cover”; kurung, “a cage; a prison; a cabin (on a ship)”’; /umbung,
“a rice granary”; bujung, “‘a narrow-necked, big-bellied earthen-
ware jug”; gajung, ‘‘a dipper”; and tjongkong, ‘“‘a sentry box’’.
This last-mentioned must be regarded as another kind of box
or container, beside andong, which is a type of four-wheeled
carriage, and gerbung, “a train coach”, all of which, along with
 132 Comparative Semantics

kurung, “prison” or “cabin”, are boxes suitable for human

beings.
For the sarong seen as a wrapper there is another series of
comparable terms in -ung: kelubung, ‘‘a cover”; kerudung, “a
veil”; selubung, “‘a cover, wrapper; veil; envelope”; keredung,
whose ber- derivative means “wrapped” (in a blanket or sarong)
etc. Beside these must also be set kepung, whose derived forms
have to do with “surrounding” (a leader) or ‘“‘encirclement”
or again ‘“‘a siege”. There are further extensions of this sense of
“covering” observable, a physical one in mendung, said of a sky
that is ‘‘cloudy, overcast”, and a figurative one in murung,
“gloomy, depressed’’, which is to the mind what clouds are to
the sky. “Gloomy” is obviously equally applicable to both, as
it is used just as readily of the weather as of the spirit, in contrast
to bright weather or a bright disposition. Human beings are
very apt to see nature in sympathy with their own feelings.
Id. pentung, “‘club’’, and tembung, “‘cudgel’’, represent yet
another kind of bulge. There must be a relation between these
and Id. gajung, ‘‘a dipper”, which has the further sense of “a
blow with the fist or a weapon’’, and also with berondong, “‘fire
a volley of gunfire”, and the metaphorical berongsong, ‘‘flare
up, explode in anger”’.
Comparable with the roof of a house is the umbrella, pajung,
whose resemblance to a hat is especially noticeable in the Japanese
minoboshi, “‘straw rainhat’’, which is practically umbrella, hat
and hood combined, at least the way it is worn in Hokkaido.
The Portuguese, incidentally, call their umbrella chapeu de chuva,
lit. chapeau de pluie, i.e. ‘rain hat’. The Malay kotjong, “‘hood”’,
~ is also relevant here, and so are similar appendages belonging
to birds, such as Id. kuntjung, “‘tuft’’ or “‘crest”, and balung,
“cock’s comb or crest’. However, this last word has another
sense of absorbing interest that will be mentioned later; the
cock’s comb, which is really secondary to it, is a good hint and
well worth reflecting upon.
Much less familiar to women than the sarong, but better
known to zoologists all over the world, is the dugong, that
quaintly bloated sea creature which can be seen to be admirably
and tellingly named after its looks. And beside the huge dugong
let us place the tjebong, equally well named, for its appearance
makes it very much like a dugong on a smaller scale: this is the
 The Bow 133

humble and unobtrusive tadpole, that small swollen lump all

head and tail, whose French name fétard originally meant
“big-headed”’, just as “‘tadpole”’ is literally ‘“‘a toad (tad) that is
all head (poll)””. Now that they are side by side, it can be seen that
the tadpole does look like a little dugong and the dugong like
a huge tadpole.
However, this is rather incidental to our main concern, which
is fundamentally human, though it could rightly be pointed out
that all animals and. objects are relevant inasmuch as they are
seen and interpreted by human beings, wherever and whenever
they lived, and this is precisely what comparative semantics
is about. Of course, the object viewed by different eyes in various
cultures includes first and foremost man himself, and it is to
the human aspect of “‘bloated”’ objects that we are now going
to address ourselves.
t will be appreciated that Id. kembung, “‘puffed up”; lembung,
“inflated, swollen’; busung, “bloated, distended; bulging,
packed”; gembung, “‘swollen, puffed, stuffed”, and gelembung,
“bubble; swollen’’, all form a clearly defined semantic group.
Together with the various terms meaning “swollen” that
have just been cited, Id. sombong, “‘arrogant, conceited”, must
occupy a special place, since unlike kembung etc. it is nct merely
physical: “arrogance” is essentially a human feeling and a
human failing at once. If certain animals are looked upon as
sharing similar feelings, as shown by terms such as Fr. se pavaner,
It. pavoneggiarsi, Sp. pavonearse, “strut”, all derived from the
name of the peacock, Lat. pavo (which turns up in Portuguese
as pavo, “turkey”, another bird fond of strutting with its feathers
puffed out and its tail spread out), this is because man himseif
has attributed such attitudes and feelings to those animals and
therefore a study of those terms is part of the study of man and
very relevant here.
To appreciate Id. sombong more fully, we should think of
such phrases as Eng. “‘be puffed up with pride”, or It. rimpettirsi
and Fr. se rengorger. This again illustrates admirably the methods
of comparative semantics, for if you want to see through every
detail of se-r-en-gorg-er there is no better, more direct or simpler
way than to put it beside r-im-pett-ir-si. Here we see the well-
known relation between “‘chest” and “throat” in Fr. gorge,
found in a number of expressions and notably in soutien-gorge,
134 Comparative Semantics

a euphemism for what should really be called soutien-sein.

What is more to the point is that the image immediately under-
lying all these terms for “arrogant” is that of a bird thrusting
its inflated crop forward while throwing back its head, giving
it what appears to us to be a peculiarly arrogant attitude—to
some extent natural to human beings when they strut and swagger.
In fact, we can hardly imagine Lady Bracknell, Oscar Wilde’s
immortal character in The Importance of Being Ernest, other
than in precisely this attitude.
If proof were needed of the close relation in Indonesian
thought of sombong, “‘arrogant’’, to the “‘bloated” terms com-
mented on, it would be sufficient to quote Id. dia melembungkan
dada (lit. he inflates chest), i.e. “he is bragging’’, where me-
lembugg-kan, ‘“‘inflate’’, is derived from lembung, “inflated,
swollen’’, cited earlier (cf. me-lambung-kan dada, “‘boast’’), or
again Id. dadanja busung (lit. his -nja chest dada is inflated
busung), i.e. “he feels proud’. As we once more come across
busung, ““bulging’’, it is interesting to observe another physical
use of this term in the phrase tanah busung (lit. land bulging),
applied to “hilly ground’’, for which a close parallel can be
found in Fr. mamelon, ‘“‘a rounded hillock’, derived from
mamelle, ‘“‘breast”, where the ending -on has an augmentative
value similar to that in Fr. perron.3?
32. Augmentatives are a conspicuous feature of Italian, where the ending is -one,
and some of these Italian words have passed into French, as for instance Fr.
ballon (borrowed from It. pallone adapted to Fr. balle), hence Eng. “‘balloon’”’,
lit. a big ball, and canon, hence too the English term, whose primitive sense
was “‘a big tube”’.
Canon belongs to an extensive family, worthy in itself of a special study,
since it goes back through Latin to.a Greek original which can hardly be
separated from Gk kavav, “a straight rod’’, hence ‘“‘a rule”, a sense that
is clearly discernible in “canon law’’, where the word has become associated
with the Church. In the Late Middle Ages, “canon” was appropriated by
Medicine, as can be verified in Chaucer’s Pardoner’s Tale (1. 890) and in
Rabelais’ Gargantua (chapter XXIII); this was due to Avicen’s celebrated
Kitab-al-Qanun-fil-Tibb, i.e. Book of the Canon of Medicine, undoubtedly the
most authoritative work on medicine since Hippocrates and Galen. Ar.
ganun was of course borrowed from Greek, and it is interesting to discover
that from Arabic it spread to Turkish kanun and Hindustani gdniin, where
it stands for “law”, whereas Indonesian keeps much closer to the original
Arabic terms for “law” based on the three consonants h k m: Id. hukum,
“law”; hakim, “judge”; mahkamah, “court of justice”. This simple Greek
word for “rod” is thus intimately associated with the history of the Church,
Medicine and Law, and the foregoing is only a small glimpse of what is an
immense family of words.
 The Bow 135

It has been mentioned that Id. udjung, “‘point’’, etc. is used

for a cape inter alia, and now that we have met tanah, “land”,
it can be specified that the exact phrase is udjung tanah, lit. point
of land. However, the usual word for this geographical feature
is rather tandjung, a cape in its own right, and here again the
“Sutting”’ value of the ending -wng can be appreciated. In fact
tandjung is to “‘cape”’ very much what udjung is to “point’’, 1.e.
the latter is one of a number of terms that can be used in referring
to a cape, but not the one that immediately comes to mind.
A cape is really a headland. Fr. cap is borrowed from Pro-
vencal and this accounts for the initial c remaining intact, as
also in It. capo and Sp..-and Pt. cabo, Lat. caput, which turns into
OFr. chief, later chef. In Renaissance French chef occurs in
the sense of ‘‘cape’’. Fr. cap passes not only into English but into
Dutch and German as well. The use of Ar. ras, “‘head’’, for
“cape” may be a calque, but it could well be the other way round.
More relevant to Id. tandjung is Id. hidung, “nose”, and it
hardly needs to be pointed out that the nose “juts out” from
the face. So it is no surprise to discover that a cape often bears
the same name as the nose. Thus on the English side of the
Straits of Dover, Dungeness juts out into the strait facing Griz
Nez on the French side. Fr. nez, ‘‘nose’’, was in Old French
nes (Lat. nasus), indistinguishable from Eng. ness, which is also
a nose originally. As to the addition of Cap in Cap Gris Nez,
this is exactly like Ja pointe de Penmarche, whose Penmarche
is literally the horse’s head (pen, “‘head”’) in Breton. In Russian,
Hoc, “nose’’, is also the normal word for “‘cape’’, and in Turkish
too burun is usual in both senses. How close the two notions
are can further be observed in Finnish, where ‘‘the tip of a cape”’
is called “the nose of a cape” (Finn. nend, “‘nose; tip”); it is
but a short step from this to the idea of the nose as a “cape”.
In all these cases it is essential to note that the sense “‘cape”’ is
secondary to “‘head” or “nose”’.
This is just the right time and place to mention Id. timbrung,
whose me- form means “interfere in somebody’s affairs”, a
rather dull and abstract way of expressing a notion that
timbrung brings out far more vividly, for the actual image
unmistakably present in the Indonesian term ending in -ung
is that of “poking one’s nose into somebody’s business” (cf. nosy
136 Comparative Semantics

and Nosy Parker) or as the French would put it, fourrer son nez
dans les affaires de quelqu’un. Thus this stressed and relevant
proximity to hidung, “nose”, was putting the reader on the way
towards the solution of this little semantic problem.
Another and more obvious aspect of the nose is to be found
in Id. montjong, “snout; elephant’s trunk”, whose me- derivative
is used for “jut out”. A significant example will suffice: mulutnja
memontjong kedepan, “his mouth jutted forward”. This is most
illuminating in view of what we said earlier on the problem of
interpreting mulut, “mouth”. Similarly, Id. muntjung, “snout,
muzzle; beak”, appears to be a simple variant of montjong.
Id. dajung, ‘“‘fin of a fish’’, sticks out from the body of the
fish much as the kuntjung, ‘“‘tuft, crest’, stands out on a bird’s
head; from the sense “‘fin’’ dajung develops the secondary
meaning “oar”, while with the prefix ber- is formed the cor-
responding verb “‘to row’. Comparable with these too is the
comb on the cock’s head, balung.
The most arresting point about Id. balung is that it is the usual
word for “‘bone’’, and the cock’s comb is seen to be a derived
sense. Ger. Bein, ‘bone’, acquired its present meaning “leg”
precisely because the shinbone stands out so much. While Du.
been, like the cognate Danish and Swedish words, has retained
both senses, Ger. Bein has long been used of the leg alone, though
the old bone is still visible in das Gebein, ‘“‘skeleton’”’, Elfenbein,
“ivory”, and fixed phrases of the type durch Mark und Bein
etc. Eng. “‘bone’’, it need hardly be said, has kept intact its
original meaning.
There is another bone in the body that is conspicuously
jutting, and it can easily be discovered through the shin just
mentioned. For this is the same word as It. schiena, “‘back’’,
and Fr. échine, “backbone” and often “back’’. Its protruding
nature is vividly illustrated in Spanish and Portuguese, where
as esquina it designates an external jutting corner, in contrast
to the inner corner, rincon. There is similar information to be
gathered from the German cognate Schiene, “‘a rail’, which
stands out clearly above the ground.
Moreover the relation between “back” and “bone” can
readily be discerned within German itself, since Riicken, ‘‘back’’,
is frequently used of mountains, especially in compounds like
 The Bow 137

Bergriicken ot Gebirgsriicken, and what is still more cogent,

Ricken is actually the same word as Eng. “‘ridge’’. Just as the
German leg got its name from the protruding shin, so did the
German back receive its name from the jutting backbone. The
relation between the spine and the shinbone is, however, brought
out most picturesquely in Spanish, where the normal name of
the tibia is espinilla, lit. little spine, since Sp. espina is “‘the spine’.
At this point it should be mentioned that Eng. “‘spine”’ is
from Lat. spina, “‘thorn’’, hence also Fr. épine, ‘‘thorn’’. It is
interesting to note that Rus. cnuna is the usual name of “the
back”, unlike Eng. “spine” and Fr. épine (dorsale), which
generally refer to ‘‘the backbone’. The “thorn” is also present
in a most striking way in the Indonesian phrase punggung kuda,
lit. back horse, i.e. ‘‘horseback’’, which must be interpreted as
“riding bareback’’; here again, the symbolic value of the suffix
-ung is unmistakable, since one could hardly imagine anything
more painfully thrusting than a horse’s back ridden without a
saddle.
It is true that the Indonesian leg, kaki, which is also “the
foot’, is not named after the shinbone, but here we must keep
in mind what has been said about the manifold ways of looking
at a given object. It is no argument against the etymology of
Bein, which cannot be anything but the cognate of “bone”,
to highlight the apparent discrepancy of the senses and to point
out that the English bone has not turned into a leg, though the
English shinbone protrudes as much as the German. The
semantic change from “shin” to “leg” is always possible, just
as that from “spine” to “back’’, but it is not fully actualized
except under certain conditions which would be worth examining
more closely.
The development of “‘bone” to “‘shinbone” and then to
“leg’’ may well have been encouraged by the fact that the shin
was particularly exposed and needed protection. Let us remember
the Roman ocreae, the Greek Kvnpidec, which occasionally
turn up in Latin as weil, the German Beinschiene and the English
‘“sreaves”’,?3
? which are various types of leg armour, or jambiéres
as the French call them, and which are still in existence in the

33. Rabelais uses greves for the front of the leg in Gargantua, chapter XXVII.
138 Comparative Semantics

modified form of the motorcyclist’s leg shields or the cricketer’s

thickly padded leg-guards (pads).
On reflection it is difficult to imagine a bone more in evidence
than the shin or the spine. Of course the jaws and the chin can
be thrust forward, but only on occasion, whereas the shin and
the spine are permanently and characteristically protruding,
and this explains to some extent the semantic changes that have
been noted.
It was observed earlier that Id. kaki includes both “‘leg” and
“foot”, and similarly Id. tangan covers both “arm” and “‘hand”’.
This particular lack of differentiation is normal in a number of
languages, and especially so in Japanese and Slavonic. Actually
as one moves in Germany towards the east, there comes a point
where Fiiss is used for “leg” as well as “hand” in the regional
speech.
In Indonesian as in English, the shinbone has failed to turn
into the leg, but it only just missed, as can be ascertained in
another way. We have seen that perut kaki (or perut betis) is
one of the names used in referring to “‘the calf of the leg’, and
this means literally “belly of the leg’’. If the calf is looked upon as
the belly, it is but a short step to consider the shinbone as the
backbone of the leg, all the more so as it is conspicuously
jutting. It is certainly no accident that the Indonesian bone bears
the suggestive name of balung, which ends so unmistakably
in a protrusion.
However, there is yet another name for the calf of the leg,
djantung betis, and this one is particularly deserving of attention.
At first sight there seems to be a contradiction, for the ending
-ung, which so admirably expresses the notions of “jutting” and
“protruding”, appears to be used here with quite a different
value. At this point it would be worth revising what was said
before on the subject of perut, “‘belly’’, and mulut, ‘“mouth’’,
and how I was gradually led to the conclusion that they were
not so named because they are “hairy” as I had first assumed,
but rather because of their bulging or pouting nature, an inter-
pretation that was confirmed as further evidence became avail-
able. Here again we find the correspondence with the twofold
nature of the French suffix -u, which has been shown to be
equally well applied to bulges and to hair, both of which represent
 The Bow 139

successive layers of protective covering and are, therefore,

fundamentally one. There is a deep unity behind this apparent
diversity, just as there is between protruding bones or jutting
parts of the body such as snouts, trunks, fins, combs etc. and
the bulging flesh.
Evidence is readily available for this interesting “bulging”
value of the suffix -ung. Thus we have Id. kandung, “uterus,
womb’’, whose me- form means “‘pregnant’’ as well as ‘‘contain”’
in a more general sense; this can be seen in kandung kentjin,
“bladder”, which must rather be interpreted literally as “a
pocket or bag of urine”, and so we must adopt the working
hypothesis that kandung was originally “‘a bag’’. Similarly, from
kempung is derived kempungan, ‘“‘abdomen; bladder’’.
Beside these can be placed Id. indung, “‘mother’’, and especially
the combination indung telur, “‘ovary”, which must be interpreted
as “a pocket of eggs’’, and this again leads us to postulate an
earlier sense of “bag”. Two other terms, not in -ung but with
the same deep vowel ending, need to be mentioned here: Id.
induk, ‘““mother”’ (of animals), and waduk, ‘‘stomach, paunch”
(of animals?), used also in the wider sense of “large container;
reservoir’.
The human stomach is usually Jambung, significantly another
-ung word, whose meaning further extends to “the side” (of
a ship or person) and whose me- derivative is current in both
Malay and Indonesian to express “‘bounce, leap’’. It is relevant
to compare perutnja busung, “‘his belly is distended”, and in
particular Japung, whose reduplicated ber- form means “gobble
up”.
In support of our interpretation of these -ung terms for “belly,
stomach, womb” etc. as originally “bags’’, we must first cite
Id. karung, “‘bag, sack”’, and kantung, the usual native Indonesian
word for ‘‘a pocket’’, whose sense and ending are now seen to
be very significant, together with saku (borrowed from Dutch
and clearly recognizable by its odd ending), “bag; pocket’,
as in sakunja busung, “‘his pockets are bulging”’.
It is worth recalling here that Eng. “belly” is cognate with
Ger. Balg, “‘bag’’, and there must have been a clear awareness
of this sense relation in Middle English, judging by the following
remarkably apt line from Piers Plowman:
140 Comparative Semantics

and pus-gate ich begge,

Withoute bagge oper botel, bote my wombe one.**

Du. zak, ‘pocket’, found in Id. saku just mentioned, is itself

clearly the same word as ‘‘sack’’, which is ultimately of Semitic
origin. Our “pocket” is literallyalittle “poke” or “bag”’, and from
its Frankish cognate comes Fr. poche, which returns to English
as “pouch”, whose senses are equally related to “purse” and
“‘belly’’. There is also the old proverb ‘‘to buy a pig in a poke”
and Fr. acheter chat en poche, whose basic notion “swollen”
(Lat. bucca, “‘puffed-out cheek”, is a cognate of “‘poke’’) is
reflected in Marot’s remark after he had received money from
the king: ma bourse avoit grosse apostume, “‘my purse had a big
swelling in it’’, i.e. it was well filled.*° Similarly, it must be of
some interest in the land of marsupials to record that Rabelais’
Escolier Limousin refers to his purse as his marsupie,*° a direct
borrowing from Latin.
Other parallels, at least as convincing, can be found in the
Far East. A Japanese member of this semantic family is futokoro,
originally “the bulging place’, currently used for ““breast; pocket;
purse’’, to which must also be added “belly”, as is clear from
several of the uses of the corresponding character in both
Chinese and Japanese. The Japanese stomach, i §, is usually
enlarged to ibukuro '§ 48 (lit. stomach-bag) because “‘it looks
like a bag” (so the children are taught, I understand). Then there
is the remarkable ofukuro % § (lit. honourable bag), used mostly
by Japanese men in referring to their mother, a term whose
literal translation cannot possibly do it justice and which can
only be appreciated in its social context. Moreover there are
the “bellying sails”, i.e. when they are blown out by the wind,
for which the Japanese use the verb haramu, “‘be pregnant”.
It can thus be seen that all these bulges constitute a well-defined
semantic group.
There is a further series of terms in -ung designating what
may be looked upon as a “regular’’ type of bulge, more akin to
mathematical forms. For instance Id. lengkung, “arc, curve’’,

34. W. Langland, Piers the Plowman, C Text, Passus VI, ll. 51-2. Quoted in
F. Mossé, Manuel de l’anglais du Moyen Age, (Paris, 1949), vol. ii, p. 297.
35. C. Marot, Epistre XXVU, in Oeuvres completes (Paris, 1919).
36. Rabelais, Pantagruel, chapter VI.
 TheBow . 14]

and lingkung, “circle; circumference”; the derived lingkungan

is quite common for “environment”. Another word of a similar
nature is Id. tjembung, used of “chubby” (cheeks) and which
also has the more “regular’’ sense “convex”, while the opposite
Id. tjekung, “‘sunken’’ (cheeks) or ‘“‘hollow’’ (eyes), has the fur-
ther meaning “‘concave”’.
That both convex and concave and similarly chubby and
sunken (cheeks) should belong among the -ung terms, which
have been found to be practically all “jutting” words, suggests
that “sunken” or “concave” is likely to be a secondary sense
development that arose as a result of the kind of semantic
polarity that is so frequent in language. A curve, like a corner,
can be seen outwardly or inwardly, yet the first impression is
usually that the curve is a variety of bulge. There is a similar
phenomenon in Id. relung, “‘a recess”, whose me- form combines
“dome-shaped” and “hollow”. Kelung, “bent; hollow’ and
lekung, “hollow; curved’, must also belong to this category.
However, beside these is found a small group of -ung words
used solely in the sense of “‘hollow’’, suchas Id. tjenkung, “‘hollow,
sunken” (cheeks) and ‘“‘dimple”, and kempung, which also refers
to cheeks as “‘sunken’’.
Even though the great majority of -ung terms are connected
with bulges, we must not be surprised to find that a certain
number refer to hollows. After all, we have phrases of the type
“over hills and dales”’, Fr. par monts et par vaux. And in Japanese,
there is a most interesting and expressive term, dekoboko fh],
whose visible form both in the characters and in transliteration
is strikingly apt, since deko-boko includes “‘the places (ko for
tokoro) that come out (de-ru) and the places that are hollow
(bo for kubomi)’’, and thus refers to iand whose surface is uneven,
with bulges and hollows.”
It may now be of interest to return to the body and to consider
what other parts come into this picture of bulging objects. We
have already met Fr. pansu and ventru, as well as It. panciuto
and Sp. panzudo, ail of which mean “pot-bellied” or “big-
bellied’’. The Italian augmentative pancione and the correspond-

37. I remember an elderly Japanese lady in Paris describing the appearance of

the buildings of Tokyo as dekoboko; anyone who has been there will readily
appreciate the aptness of this description. (Ed.)
142 Comparative Semantics

ing Sp. panzon combine both the senses “big paunch” and “one
who has a big paunch’”’, and there is also the Spanish term
panzada, which designates the “‘full paunch”.*° As an interesting
contrast, it is worth setting beside these “‘healthy” terms a quo-
tation from Marot, which shows very graphically what can
happen to a belly through illness (or some other cause); in his
case, une lourde et longue maladie resulted in:
la cuysse heronniere,
L’estomac sec, /e ventre plat et vague :3°

Let us now see to what extent Indonesian corresponds to the

West. One remarkable aspect of the body is the nose. There is
no doubt that this is the best example of a regular protrusion
found in all human beings, and it is a bulge that does not depend
on pregnancy or accidental presence or absence of food. Its
characteristic shape is made immediately visible to the eye and
to the mind in Id. hidung, just as in muntjung, “snout of an
animal” or “elephant’s trunk”, for the trunk is really an elongated
nose.
What is curious about the big nose is that despite all its
qualifications for recognition, and the further fact that Lat.
nasutus, “‘big-nosed”’, is the best-known example of -utus for
bulk, there are practically no traces of this formation in Romance
in reference to the nose. There is of course It. nasone, “big nose’’,
but no *nasuto for “big-nosed”, and it may be remembered that
Rabelais searched for just such a term in Pantagruel Chapter 1,
and all he found was the joke Nason et Ovide, which splits Ovid
into two different persons. Relevant too may be the fact that
already in Latin nasutus had developed the secondary senses

38. The ending -ada is well known in borrowings like cannonade, fusillade etc.,
and further in those of the type cavalcade, hence motorcade etc., all of
which are current in English and French; in Spanish and Italian it has
retained its original sense of “‘a blow”: Sp. guantada and It. guanciata, “a
box on the ear”’, une gifle, lit. a blow on the cheek. Thus panzada, “un coup
de panse’’ and from this it develops the meaning beaucoup de panse, which
incidentally affords a fascinating glimpse into the true history of Fr. beaucoup,
well worth a special semantic study in its own right and whose original sense
can still be seen through such phrases as Eng. “a fair whack”’.
39. Marot, Epistre XXVII.
 The Bow 143

“acute, sagacious”’ and even “satirical’’, which is a revealing

hint of the close relation between the nose and the mind and is
also reminiscent of Sophocles’ Ajax, where Ulysses is described
as being ebdpic, “with a good nose”, i.e. ““keen-scented”’, i.e. “keen
at tracking out’’.*°
It should be noted that in French, forms like ventru, pansu
etc. are normally built up from nouns in mute e preceded by a
consonant; from nez this would not be practicable, and so a
word *nasu would not be convincing, but there is no such
phonetic reason for the absence of an Italian *nasuto or a Spanish
*nasudo. Here it will be appreciated that Id. hidung occupies its
proper place in the language in a form that corresponds per-
fectly to its shape.
The phonetic problem that faces us in the case of Fr. nez
does not arise with lippe, which regularly gives lippu (and not
*Jevru from /évre as might be expected, because thick lips are
looked upon as a defect, and so the word is formed from a foreign
borrowing which tends to have disparaging associations), just
as readily as patte gives pattu and membre membru which, though
rare, is nevertheless found sporadically (cf. Sp. membrudo).
As to jouffiu, ““chubby-cheeked”’, there is a slight irregularity
of form, due partly to the fact that joue lacks the consonant
before the final e; *jouu is not only phonetically unsatisfactory
but the two successive u are offensive to a French eye. And so
gifle, formerly ‘‘cheek”’,+! and joue combined into a Mischform
or blend *jouffie, as a convenient base form on which to build
joufflu.
There is another remarkable gap observable if we now con-
sider the ears, and here again Rabelais visibly tries to find a
word like *oreillu which would perfectly suit the huge ears he is
describing in Pantagruel Chapter i. However, though big noses
are common enough, and the lack of a special term for them
all the more curious, big ears are much rarer; in this case we

40. Sophocles, The Ajax, 1. 8, in The Plays and Fragments, Part VIL, ed. R.C.
Jebb (Cambridge, 1896).
41. Note, by the way, that giffe is a dialectal borrowing of Germanic origin and
cognate with Ger. Kiefer, “jaw”’; this is a further instance of the semantic
passage of “jaw” to “cheek”, as in Eng. “cheek” and other examples cited
on pp. 49-50 in connection with Id. pipi, “cheek” whose form also strongly
suggests a parallel sense history.
144 Comparative Semantics

must beware of confusion between big ears as such and big ears
in the sense of long ears, for these are quite normal ina figurative
sense and in fact popular prejudice runs in favour of long ears
for dolts, no doubt because of the long ears of asses. But these
are long ears and not thick ears, and there lies the vital difference.
Thus the need for an *oreillu, which would be an easy formation
on oreille, is not nearly as great as for a *nasu, far less acceptable.
The “breast” is a further noticeable absence. Nothing of
course can be expected from sein, which is even more intractable
than nez, since for the latter a basic nas- is already present in
“nasal” etc., whereas with sein we are helpless. True, there is a
French word mamelu, but it is rare and most French people
would assert with some vehemence that it does not exist, any
more than fessu or membru. And this puts us on the track. For
however protruding and bulging the female breasts may be, there
is a sort of social convention that they are not noticed—at least
in good society.
The absence of an -ung term for “‘the breast” in Indonesian
probably reflects the same uneasiness before those embarrassingly
conspicuous bulges, and so they are treated as if they didn’t
exist. Similarly, implicit social taboos have resulted in French
words like mamelu, fessu and membru on the one hand, and
poilu on the other, becoming far more sexually loaded than
might otherwise have been expected; hence the refusal, or at
least the reluctance, to admit their existence.
While on this aspect of our inquiry, mention should be made
of Id. burung, “‘a bird’’, which has come to be used as a euphe-
mism for “the male organ”. As a rule, euphemisms conceal
rather than reveal, but in this case the opposite is true, as can
be appreciated from the uses of -uwng terms that have been
examined. This would tend to suggest that Indonesians are
largely unaware of what is really a most remarkable feature of
their language.
Just as easy as the passage from Fr. ventre to ventru etc. was
that of téte to tétu. However, there is a difference: the head does
not bulge in the same way as the belly or even the nose. And
this may well be why tétu developed its secondary meaning
“stubborn”, which has long been the only one left. We have
seen the earlier word for “‘a big head”, tétard, also move away
 The Bow 145

from its primary sense and become the normal name of the
tadpole.
These considerations lead us to have another look at “the
cape”. All the names of the cape based on “head” (most Euro-
pean terms, ultimately from Prov. cap, Lat. caput and Ar. ras)
are found in a single geographical area and may all come from
the same source, Romance or Semitic, hence a series of borrow-
ings or calques. The cape names based on the nose, on the other
hand, are scattered throughout many regions—England, France
(at least in that part nearest England), Russia, Turkey, Indonesia
etc. This suggests quite a number of independent formations and
seems to indicate that the nose is seen as a protrusion from the
face more naturally than the head is seen as a protrusion from
the trunk. Note further that the Indonesian head, kepala, is not
an -ung term, i.e. it is not regarded as a bulge. Thus the whole
terminology of the cape is worth reviewing in this new light.
Let us now turn to Id. bokong, “‘buttocks’’; it extends to “the
back” on which a load is carried, and is therefore comparable
to dukung and gendung and better still to Fr. rein, ‘‘kidney’’, in
so far as it belongs to this field, as in il a les reins solides, ‘‘he
has a strong back’’. It is undeniable that buttocks do bulge in
a most characteristic way, and we find a further proof of this in
the frequent use of “‘cheeks”’ in referring to this part of the body;
in French too le joufflu is a very common euphemism for /es
fesses. It should be observed that bokong is a Malay word which
was retained in its Malay form in -ong, possibly because it was
considered unworthy of full Indonesian citizenship.
Id. bagong seems to belong here, not only because of its
ending, but also because its senses “‘big and heavy” or ‘‘cumber-
some”’ are clearly unfavourable. Note that it is used as well of
“a wild boar”’.
As in the case of bokong and bagong, an arresting aspect of
Id. sombong, “arrogant, conceited”’, is its Malay form. Dis-
paraging notions are often expressed by foreign borrowings. This
is especially striking in Fr. hablerie, “boasting”, based on Sp.
hablar, “‘speak” (Lat. fabulare), and the Spaniards return the
compliment when they use parlar, borrowed from Fr. parler,
““speak’’, in the sense of “idle chatter”. Synonymous with hablerie
146 Comparative Semantics

are Fr. gasconnade and fanfaronnade, whose form at once betrays

their Southern origin.
These points should help us to interpret our regional Belgian
French imberné. Surely my father never suspected that such a
humble and insignificant word of his would ever attract attention,
least of all in Australia, which is so far removed from Belgium.
I should stress that my father knew French admirably and would
never allow himself to use a single Walloon word with us children,
lest our French be contaminated. Unfortunately I was never able
to verify whether he had any inkling that this word might be
Walloon, and if so how he could bring himself to use it so fre-
quently with us, at least with me who must often have been dirty
when I came home from school or playing out in the streets.
Imberné unmistakably meant “‘dirty”, and the solution to the
mystery must lie in this meaning: French was fit for the best,
and by implication Walloon was fit for dirt.
There is a special use of Id. Jambung, ‘“‘stomach’’, hence “‘side”’
(of a ship or person) which is worth closer investigation. For the
stomach, like the belly (cf. “tummy” from “‘stomach’’), is a
bulge, and the sides of a ship are also bulging, in contrast to the
keel, which is a ridge closely comparable with the backbone,
and which can in fact be looked upon as the spine of the ship;
hence in a number of languages the keel becomes the normal
word for the ship.
rom the curved, convex, bulging side of the ship, so named
from the bulge of the stomach, comes a semantic development
that is widely used in both Malay and Indonesian for “‘bounce”’
(for example, in referring to a ball) or “jump up” (with joy) or
again “‘praise”. Id. lambung or rather melambung and Mal.
lambong with me- or ter- prefixes are in fact the usual words to
describe “‘the bouncing of a ball’, and one Malay sentence will
suffice to illustrate this:
Bola itu terlambong ka-tengah-tengah orang ramai.*?
The ball bounced right into the middle of the crowd.

Here too there is a relation with -ut words. Just as perut, “‘belly’’,
exists beside Jambung, “‘stomach’’, so there is another term, baut,
used to refer to “a strong-arm man, a bully employed as a

42. M.B. Lewis, Learn to Talk Malay (London, 1954), p. 93.

 The Bow 147

bouncer”. This again confirms that -wt as well as -ung terms can
be used for bulges, and recalls the point made previously con-
cerning the fundamental unity between flesh/skin and hair/feathers
as successive layers protecting the vital parts. We shall have
occasion to return to this point a little further on, but first there
is an interesting expleration to be made in Japanese in relation
to the Indonesian -ung bulge terms.
We have seen that in Indonesian, both “jutting” and “bulging”
notions are present in the same ending (most frequently -ung,
but also -ut);Japanese, on the other hand, would appear to make
a semantic distinction between them. For we have already noticed
that the words normally associated with ‘“‘bulges” belong to the
fukkura and hara families, whereas it will be apparent that the
-ne terms with which we are here concerned, with very few ex-
ceptions, clearly refer to the idea of “‘jutting”’.
Let us begin with Jp. yane, Bt “‘roof’’, whose first character
means “house; shop’’, while #& designates “the root’. Here too
may be placed Jp. kaki(ne), ta (#&) “fence”, and Jp. o(ne),
(#8 ) “‘tail’’. In each of these it can be seen that #& acts as a kind
of intensive, making more visible the essential notion of “jutting”
referred to before. It must be stressed, however, that whereas
with o(ne) and kaki(ne) there is a choice as to whether or not
the significant ending is added, with yane there is no choice,
since #2 alone distinguishes between the roof and the building
itself; hence its role here is really that of an indicator.
Turning now to the other terms belonging to this family, it
is worth pointing out that the Chinese characters that are used
disguise their fundamental relationship. Yet in the spoken lan-
guage the link is unmistakable. An excellent illustration is pro-
vided by ff “‘chest, breast”; 7) #7 “‘the back of a sword or knife”;
and #& ‘“‘roof ridge’. The reading in each case is Jp. mune. This
example should make it possible to appreciate the extent to
which Chinese characters can veil the semantic affinities between
obviously related words.
Closely linked to mune, t% “roof ridge”, is Jp. une, HE “an
earth ridge between fields”, and an even more striking aspect of
this same “jutting” notion may be observed in Jp. mine, W& “a
mountain peak’’; it will perhaps be recalled that the phonetic #
is found in a number of characters having to do with “points”.
148 Comparative Semantics

It is of considerable interest to note that none of the other -ne

terms contain this phonetic, and yet it is undoubtedly more
suggestive of the idea of “jutting” than #&. This may well indicate
that the semantic relation between the Japanese -ne words is
sensed intuitively, rather than consciously perceived.
The right of Jp. fune, fit “boat”, to a place in the present
group might at first be questioned; but there is no doubt that
it is one of the most interesting members of the -ne family. We
saw on p. 146 that Id. Jambung is used for the side of a ship,
owing to its resemblance to the bulge of the stomach. On the
other hand, Jp. fune is the normal word for “‘ship”, and since
its ending rather suggests the notion of “jutting” (in line with
the other terms so far mentioned), the explanation for this
particular meaning must be sought elsewhere. We can obtain a
good indication of the possible semantic development involved
if we consider that a boat may be compared to a house upside
down, surprising as this may at first appear; viewed in this way,
the keel can then be seen to resemble the protruding roof ridge.
However, there is an extension of this comparison, for which
Japanese itself happens to provide the best evidence. Remember-
ing that the keel is a ridge closely comparable with the backbone,
it is most interesting to discover that in Japanese “‘the keel”’ is
ryukotsu, %‘ lit. dragon bone. There can be no question that
the bone referred to is the backbone, a point that clearly shows
that the keel of a boat is seen to be especially protruding.
What makes this even more convincing is that the native
Japanese reading of # is hone, “bone”; and there is also the
striking fact that the Japanese “‘backbone”’ is called sebone 7§'%.
(The visual resemblance with Eng. “bone” is of course purely
coincidental, the b being the result of phonetic assimilation.)
These considerations give some plausibility to the suggestion
that Jp. fune fi was originally the name of “‘the keel’’, and that
it was later supplanted in this particular sense by Sino-Jp.
ryukotsu 7, thus freeing it for use in the more general sense
of “boat, ship’’.
While on the subject of bones, mention must be made of Jp.
sune, I& “shin”, which can also be used for “leg”. This duality
of meaning was examined in some detail in connection with the
relation (particularly in German) between the bone and the leg.
 The Bow 149

It should be noted that the observation made at the beginning

of this section concerning the essential separation in Japanese
between “‘jutting” (-ne) and “bulge” (fukkura, hara) terms is
very clearly illustrated by reference to sune, “‘shin”, and its
counterpart fukurahagi, ‘‘the calf of the leg’’.
In contrast to the “jutting” words that have so far been
examined, let us now turn our attention to one that can be
said to suggest a bulge: Jp. tane, f& ‘“‘seed’”’. The composition of
this character is very revealing, for beside the grain radical #
we find #, which in native Japanese is read omoi, “‘heavy”’. In
connection with the seed it is quite natural to think of pregnancy,
and this brings up some very interesting material which helps
to explain why the seed deserves to be considered essentially as
“bulging” rather than “‘jutting”’.
It is most instructive to discover that the usual Japanese verb
for ‘“‘swell’’, hareru, is written fit 4 (i.e. the flesh radical A [A]
is combined with @ just referred to; cf. also haremono, [£%
“tumour’’). And what is still more striking is that this verb
belongs, as we have seen, to the same semantic family as haramu,
“be pregnant”’.
The English phrase “heavy with young” can be better ap-
preciated in the light of what precedes; it should be pointed
out that this expression tends to be confined to animals, whereas
the equivalent French term Jourd(e) can occasionally be used of
a “pregnant’’ woman. The Indonesian phrase perut gendut is
especially relevant here, since both words possess the sense
“pregnant”, and we have already noted that the ending -ut is
connected with bulges (as well as with jutting).
What makes Jp. tane f& still more fascinating is that, apart
from its distinct association with “‘bulging’’, it also contains the
potential idea of “jutting”. This becomes clear if we consider
the seed of a plant (as opposed to the notion of pregnancy),
for at a given moment the seed will sprout or shoot. A com-
parison with Id. djagung, “‘corn, rice”, and its men- derivative,
which means “‘to sprout”, should be convincing.
Here too may be mentioned Jp. ine, fif “rice plant”. It was
Id. djagung which, by its ending, suggested the relevance of ine,
for otherwise it would almost certainly have gone unnoticed. At
first sight it would appear not to fit into the semantic group
150 Comparative Semantics

under discussion, but if we bear in mind that ine is “‘the rice

plant” (i.e. growing in the paddy field, i.e. sprouting), its in-
clusion here becomes quite plausible.
Jp. sane, #% “‘stone, kernel’, is reminiscent of tane, “seed”’,
examined earlier. However, sane also possesses an erotic sense
which can easily be appreciated if one takes into account the
male counterpart dankon, 4% lit. male root. This latter, it will
be realized, differs from the other -ne terms written with the
character # by virtue of its Sino-Japanese pronunciation. It was
pointed out on p. 147 that, except in a few cases, the Chinese
characters obscure the semantic relation between the -ne words,
whereas the spoken language makes this clear; yet with dankon
Hii the opposite is true. Hence we have the rather unusual
situation that for the ear this term represents a euphemism, since
the pronunciation conceals, whereas for the eye it is anything
but a euphemism. With this must be compared Id. burung which,
in view of its revealing ending, cannot be classified as a euphe-
mism, despite what one finds in dictionaries.
When placed beside the comparable feature in Indonesian,
the Japanese words ending in -ne form a most interesting semantic
group. However, it must not be forgotten that the foregoing
exploration belongs to the wider semantic field surrounding “the
bow’’. This will be immediately appreciated when a term that
has not yet been mentioned in this part of our discussion is
recalled to mind: Jp. ha(ne), “‘wing; feather’. We saw earlier
that ha, “wing”, has an alternative form hane (cf. ha, “‘leaf”’,
which is never written with the ending -ne), and yet in this case
the character 74 alone is used; it should be observed, however,
that it is occasionally written 7448.
From Jp. hane, “‘wing” and ‘‘feather’’, is formed the very
common verb haneru, “‘jump, spring”. The sense development
from “feather” to “jump” may not at once be clear, and in
order to work out precisely how it took place, it is first necessary
to look closely at the phonetic nature of Jp. ha, and more specifi-
cally at what is usually transliterated as an h.
If we consider the h row of the (kata) kana syllabary,
JV E Yi “ 7”
ha hi fu he ho
 The Bow 151

there is an obvious deviation from the normal pattern, fu instead

of the expected hu.*? This f is somewhat misleading, since it
is not a labio-dental sound, but this is none the less as near as
it can be written in our alphabet. It is in fact a voiceless bilabial
fricative, i.e. it is the voiceless counterpart of the Spanish b/v,
and as such it is a clear pointer to the true ancestor of the aspirate.
For there is abundant internal and external evidence that
originally the initial sound was not h but p, so that the earlier
form of the word for the wing was pa rather than ha.
This can be demonstrated most clearly in the use of the so-
called ‘numerical classifiers”. The precise nature of these has
already been illustrated, and it has been seen that classifiers
do occur sporadically in the West; but in Chinese and Japanese
they are used with all “countable” nouns when accompanied
by a number. These classifiers are not haphazard and usually
reflect some characteristic of the object. Sometimes they also
make patterns among themselves, as in Malay-Indonesian,
where bidji, “‘seed’’, is the classifier for small objects and buah,
“fruit”, is used in referring to large ones. What is especially
relevant to our inquiry is the fact that the classifier for birds
in Japanese is ha, ‘‘wing’’.
With hato ni wa (lit. pigeon two-wing), “two pigeons’, it can
be seen that there is no phonetic change; but the result is striking-
ly different with san, ‘three’, for here phonetic assimilation
occurs between san and ha and this gives hato samba, where the
original p survives in its voiced form 5, since nasals are voiced.
On the other hand, with roku, ‘‘six’’, the combination of roku
and ha produces hato roppa. Such phonetic changes also take
place with the numerals ichi (“one”), hachi (“eight”), and
ju (“‘ten’’), when the classifier begins with h (for example, hiki—
used for counting animals); i.e. the original p survives as p or
b. Actually h and p, as well as the voiced form b, and m, the
corresponding nasal, must be considered as belonging together.
All this is revealing enough about the original nature of the
Japanese h. There is moreover other evidence that points in
43. There is also another departure from the norm in hi, though this spelling
does not reveal it: the vowel i disappears after transforming the preceding
consonant into something like the German ich-sound, so that the pronun-
ciation of the word transliterated hito, “human being’’, can be represented
in phonetic script as [Jto].
152 Comparative Semantics

the same direction. If, for example, we consider the corre-

spondence between the Chinese and the Sino-Japanese pro-
nunciation of various Chinese characters, it is found that where
Sino-Japanese has h, Chinese usually has p or b. Among the
innumerable examples that could be selected, a few of the simplest
will suffice: A “white”, Sino-Jp. HAKU, Ch. bai; 4t “north”,
Sino-Jp. HOKU, Ch. béi; f& “protect”, Sino-Jp. HO, Ch.
bao; 7% “not”, Sino-Jp. FU, Ch. bu; %& “break”, Sino-Jp. HA,
Ch. po; & “book’’, Sino-Jp. HON, Ch. ben.
There is further evidence of the earlier nature of the Japanese
h, and it is external too, in so far as all writing can be considered
as external to language. In writing the kana, the same sign is
used for kaand ga, kiand gietc., the voiced sound being identified
by the addition of a small diacritic sign * : 7 ka, #7ga. However,
when ‘the same procedure is applied to the h row, the corre-
sponding result is a b, i.e. 7) ha, 7‘ ba, and the further device of
adding alittle circle ° is used to write the sound »* pa. The
circle is used exclusively for this purpose: of all the voiceless
sounds, p alone has a diacritic sign, an unmistakable indication
that p must be a secondary development.
However, the most fascinating evidence is to be found again
in the kana, but in a totally different aspect of it, and to discover
it we must turn to Sanskrit, and more especially to the devanagari
alphabet.
Anyone who has undertaken the study of Sanskrit must
have been amazed to find that the Indian grammarians, even
before the beginning of the Christian era, had acquired such a
thorough understanding of practical phonetics that they were
able to devise an arrangement of the sounds of their language
and of the corresponding letters that cannot but arouse the
admiration of a modern linguist. Briefly, the order is: first
the vowels, then the stops, semivowels, sibilants and finally the
aspirate. There is extraordinary wealth in this alphabet; in
addition to vocalic r and |, there are five rows of stops: velar,
palatal, lingual, dental and labial, i.e. from back to front, and
each row includes five sounds and so five letters: voiceless,
voiceless aspirate, voiced, voiced aspirate and nasal, which
means there is already an array of twenty-five sounds and letters
in the stops (occlusives) alone. As an illustration, the labial row
 The Bow 153

can be conveniently transliterated as follows: pa pha ba bha ma,

but in Sanskrit there is no common element such as the p in
pa and pha. It is clear from Indo-European historical considera-
tions that both palatal and lingual stops are secondary develop-
ments, and that the basic series are undoubtedly those based
on ka, ta and pa.**
If we discard the letters of the devanagari alphabet which
represented scunds unknown to Japanese in the late ninth
century, we are left with the following:
a iu eo ka nga ta na pa ma ya ra wa sa ha

These are here set out in the Sanskrit (devanagari) alphabetical

order. The Japanese Buddhist scholars who had studied Sanskrit
simply borrowed this arrangement,*> but set it out according
to their own syllabic requirements, which were naturally pre-
determined for them by the demands of Chinese, as follows:
ka” nea ta fa “pa ma ‘ya ta “wa sa “id

=a
ood

44. It is obvious that, beside the admirable order and clarity of the Sanskrit
alphabet, our own is a sad mess, and European linguists, including those
who have little or no knowledge of Sanskrit, have long been aware of this
fact. Of course, we got it from the Romans who got it from the Greeks who
got it from the Semites who got it heaven knows where, and there is an in-
creasingly overwhelming weight of tradition behind it, so it looks as if we
are stuck with it for all time.
45. The introduction of Buddhism into China must have led to extraordinary
linguistic problems, especially in the rendering of Indian proper names.
Often the only possibility was to use Chinese characters purely as phonetic
symbols, entirely devoid of meaning. However, this method was appallingly
cumbersome, and there was little agreement as to which particular character
should represent which sound. Moreover it was frequently difficult to discern
whether a given character was used for its sense or for its sound. Nor did
the Chinese have anything corresponding to our capital letters, so precious
in identifying proper names in contrast to ordinary words.
All these problems were present in an acute form for the Japanese, whose
agglutinative particles were especially intractable. So it is no wonder that
a purely phonetic system, the kana, was gradually evolved, using a number
of Chinese characters in an abbreviated form. The kana was indispensable
for writing the particles, endings etc., i.e. the grammatical features as distinct
from the fuller words which were represented by the usual Chinese characters,
and it was also invaluable in transcribing proper names and foreign borrow-
ings.
154 Comparative Semantics

and filled in ki, ku etc. Clearly this is no longer a blend of alphabe-

tic and syllabic systems as in the devanagari, but exclusively
syllabic, as Chinese written practice demanded.
Several points should now be carefully observed:
1. There was no aspirate in Japanese at the time (i.e. the late
ninth century), so ha was cut out as just shown and discarded
with all the other sounds that were not relevant to Japanese.
A further proof of the absence of the aspirate is that when Sk.
maha, “great”, was borrowed, the Japanese transcribed it as
maka, quite the usual procedure for rendering a foreign aspirate
in a language that does not possess such a sound. (Compare
Rus. gavanna, ““Havanna cigar”; garem, “‘harem’’ etc., and
Medieval Latin michi for mihi). 2. Every stop is followed by the
corresponding nasal. However, doubt was naturally felt about
the right of the velar nasal, shown as nga, to be present at all
in this syllabary, for unlike all others this sound does not occur
initially, any more than in Greek where it is purely internal, as
in «yyedoc, or in English where it can be only internal or final.
3. The final result was the elimination of nga, because of its
very dubious claim, and the blank space was filled by shifting
sa into this position:
a ka sa ta na pa ma ya ra wa

4. All that is now needed is to fill in the other four vowels, and
we obtain the kana system as it was originally. Not until much
later did the pa series turn into the aspirates,*® and this is our
chief concern here. It is significant that the present aspirates
are in exactly the position where pa originally stood in the Sanskrit
alphabet, whereas the Sanskrit aspirate was at the very end,
as we have seen. There has been no shift from ha to pa, but on
the contrary an internal sound change: all p turned into h
(except fu) in the course of time, and this alone is fully consistent
with everything that has been observed elsewhere, and already
discussed.
The presence of the p noted in the case of the numerical

46. This change seems to have begun towards the end of the sixteenth century
or early in the seventeenth century. See ‘Historical Changes in Japanese
Phonology”, in Readings in Japanese Language and Linguistics, ed. J. Yama-
giwa (Michigan, 1965), pp. 23-77, esp. pp. 42, 68-9. (Ed.)
 The Bow 155

classifiers, though it appears to be a change from h to p, 1s

rather to be regarded as an internal survival of the earlier sound.
This is comparable to the French liaison, which goes back to
the time when final consonants were still pronounced; liaison,
it will be noted, is essentially internal, although the way words
are separated in writing obscures this. The point is that liaison
takes place especially between words clearly linked in one sense
group, hence between article and noun etc., and in innumerable
expressions like tout d coup, tout a fait etc. which are really one
unit.
Similarly in Japanese, the same p for h will occur internally
in a large number of ‘cases: empitsu, “‘pencil’’, from en, hitsu,
and happyo, “‘announcement”, from hatsu, hyd (or rather hat
py6), can be quoted among many. Sometimes there is a choice,
as between tsukihanasu and tsuppanasu, ‘‘thrust off”; katahd
and katappo, ‘“‘one of a pair’’, exist side by side, as do also
katahashi and katappashi, “‘one end”’.*7
5. A final proof of the change of pa, pi etc. to ha, hi etc. is that,
whereas there are innumerable native Japanese words beginning
with h or fu in the modern language, the only words with initial
p are either foreign borrowings or onomatopoeic creations.
No doubt the direct influence of the Sanskrit devanagari script
on the kana was greatly helped by its syllabism, forming as it
were a bridge between East and West. Sanskrit is of course an
Indo-European language, and as such it is an essential part of
our European background. But it is at the same time the most
easterly language of that vast family of languages which ranges
all the way'from India to Ireland. When it is recalled that lan-
guages further east are fundamentally syllabic, we may well
wonder whether the syllabism observed in the devanagari script
may not to some extent be due to the same geographical or
regional cause. There is certainly a case for a new kind of linguistic
geography which would examine features common to related

47. It is worth mentioning in this connection that the native term harakiri )¥)
(lit. belly cutting) is not used in Modern Japanese, where it has been sup-
planted by the learned Sino-Japanese equivalent seppuku ji (from setsu
fuku or more accurately set puku), possibly as a euphemism. It should be
observed that both characters are the same but in reverse order, and this
clearly illustrates one of the major differences between Chinese and Japanese:
in Chinese the verb precedes its object, while in Japanese it follows.
156 Comparative Semantics

and unrelated languages in a given area. There is no need to

stress how interesting such a study would be in the Balkans and
elsewhere.
The foregoing sketch of the development of the Japanese
aspirate is immediately relevant to the problems we are examin-
ing. Remembering that from hane, “wing” and “feather”, is
formed the verb haneru, “jump, spring”, it is worth drawing
attention to several compound verbs based on hane which show
very clearly the fundamental relation that is felt to exist between
feather, wing and spring: hanekaeru, “spring back” (Fr. rebondir) ;
haneageru, “‘spring up”; haneokiru, “‘jump (spring) up”; and
tobihaneru, “‘spring, bounce’, do not by any means exhaust the
possibilities.
Two other verbs, not based on hane but semantically insepar-
able from those just mentioned, are hajiku #4< and hazumu
##¢, which were discussed earlier on. The following example
links up “the spring’ (bane) and hazumu in an interesting
comparison:
[kanojo wa] karada wo bane no yo ni hazumase[ta}*®
[She] made [lit. caused] her body [to} rebound/bounce like a spring.
(i.e. doing gymnastics)

This quotation highlights the vital fact that the term used to
designate “the spring”, bane, differs from hane, “wing; feather’,
only in respect of its voiced initial consonant, and here it is
essential to bear in mind the point made earlier concerning the
fundamental relationship between the consonants h, f, p, b, and
m. It can now be seen that the sound relations just examined
occur not only internally but also in the initial position: thus
fukufuku, “swelling”, and the parallel bukubuku, and yet another
form, mukumuku, which is applied both to hair (shaggy/bushy)
and to flesh (plump).
Here we could pause to recall that the French ending -u
has a dual application to flesh/skin and hair/feathers and that
the Indonesian endings -ut and -ung between them cover an
identical range of meaning.
In connection with the flesh, mention must be made of Jp.
bukubuku futotta and mukumuku futotta, “plump, fleshy”

48. S. Hoshi, Bokkochan (Tokyo, 1971), p. 43.

 The Bow By.

(futoru, “be fat”); in this case, the expressive terms bukubuku

and mukumuku may be said to act as intensives.
Closely linked with these semantically are Fr. rebondi,
“plump” and Eng. “bouncing” (used very commonly of a chubby
baby). Eng. “buxom” is also clearly relevant here, for we have
seen that in its modern use it is inseparable from the idea of
abundant flesh. We also saw that “buxom” is in fact a member
of the “bow” family, and that the two complementary notions
of relaxation and tenseness are essential characteristics of the
bow. This helps to explain the “fundamental dualism” referred
to earlier in connection with this term, for it can hardly be
denied that abundant flesh has a “springy” quality, by virtue
of its “bulging” nature. A further Japanese expressive word that
must be cited here is mukkuri (undoubtedly a variant of muku-
muku mentioned earlier), since it is applied equally well to
springing and to bulging: compare mukkuri okiru, “spring up
suddenly”, and mukkuri koeta, “plump”’.
As regards the hair/feathers aspect of this same fundamental
unity, a good illustration is provided by the cushion. What is
significant is that we can still speak of “‘plumping the cushions”;
this ritual is explicable only in terms of the fact that a cushion
yields when sat on, and can only be appreciated if we think
of the various materials used for stuffing that are connected
with the domains of hair and feathers, notably leaves, kapok
and straw. Even though these materials have largely been
replaced, at least in modern Western society, the English usage
just referred to shows that this has not prevented the old ex-
pression from retaining its place in the language, regardless of
whether or not its origins continue to be recognized.
Allusions to “stuffing” are found in Rabelais, who speaks of
la plume (i.e. of a pillow, Renaissance French coussin),*° and
Rimbaud, who in Les Assis says of a chair, la paille cede... No
doubt one could easily multiply the examples. (Note too Ger.
t—
Futter, “fodder”, which is also used of the lining of a garmen
ef. Fr. fourrure—and Futteral, “a sheath’’.)
ce for
A particularly relevant and interesting piece of eviden

Oeuvres completes (Paris,

49. F. Rabelais, Le Cinquiéme Livre, chapter XV, in
1962).
158 Comparative Semantics

the springy quality of hair and feathers occurs in the following

passage:
Je cueille des brassées d’herbe. En les arrachant, je sens la meilleure
idée de la journée sortir, exploser: un lit d’herbe...J’essaie Je plu-
mard, je m’allonge, je fais sauter les ressorts ... Fameux, fameux, me
dis-je en m’allongeant sur le lit a ressorts d’herbe.5°
This quotation helps us to appreciate the very significant point
in Japanese concerning the relation between hane, “‘wing;
feather” and the verb haneru, ‘“‘jump, spring” which, as we have
seen, is derived from it. Only from an awareness (which may
well no longer consciously exist) of the springy quality of hair
and feathers could such a development have resulted. And in
fact it must be pointed out here that it is this particular associa-
tion of ideas, so clearly visible in Japanese, which provides the
answer as to why the field of “hair” and “grasses” was felt to
be of special relevance for the study of “‘the bow’’.
Here it is also essential to take into account Jp. bane, “‘the
spring”, for it could first of all be observed that springs bulge
in a most characteristic way. We have already seen that bane
differs from hane only in respect of its voiced initial consonant,
and perhaps it would not be too fanciful to suggest that it may
have developed out of hane, by virtue of the fact that feathers,
as was mentioned previously, have a springy quality. Not only
do we find that hane and bane are linked in these ways, but
perhaps even more significant is their common ending -ne,
which has been shown to be of such absorbing semantic interest.
(Note incidentally that bane is invariably written in kana: »s %
or |£42.) The relation between these two terms is thus seen to
be both a phonetic one (i.e. the h-b correspondence) and a
semantic one (i.e. the ending -ne).
A consideration of Jp. hane, haneru, and bane enables us
to approach related European terms in a new way, and with a
clearer insight into their semantic significance. So, turning now
to Europe, we find that Du. ve(d)er, ‘‘feather’, and veer,
““spring”’,°? correspond precisely to Ger. Feder, which has both

50. J. Douassot, La Perruque (Paris, 1969), pp. 154—5.

51. In the colloquial language, intervocalic d tends to drop out: cf. broer for
broeder etc.
 The Bow 159

senses. (It should be noted, however, that in order to distinguish

between them, an indicator is used: Ger. Springfeder; Du.
springveer.) The same relationship is found in all Scandinavian
languages: Dan. fjeder; Sw. fjdder; and Ic. fjodur, this last
undoubtedly being a calque on German. English, on the other
hand, has no trace of this usage.
It is interesting to observe that there is a colloquial French
expression, “vous aimez vos plumes’, just as la paillasse is used
(at least in the regional speech of Belgium) in the same general
sense of “‘bed’’. It needs to be stressed that these uses are some-
what ionic; nevertheless, this again proves the persistence of
earlier terms, despite the fact that they refer to a different object,
in this case an inner-spring mattress.
In Italian and Spanish, molla and muelle respectively mean
not only ‘‘soft” (Lat. mollis, “‘soft’’), but refer equally well to
a “‘steel spring’.
Yet despite this evidence, it would seem that there is no longer
any real awareness in European languages of the fundamental
relation between “feather” and “spring”, a fact that is probably
attributable to the very nature of the. modern spring. The gap
between the idea of something soft (contained in “‘feather’’)
and the hardness associated with a steel spring is quite a sizeable
one, and at first sight it is difficult to reconcile the two. But it is
precisely in the idea of resilience, or to put it another way, of
elasticity, which these two things have in common, that the key
to their relationship is to be found. Thus, if the notion of elasticity
(which implies both tenseness and relaxation) is borne in mind,
the duality of meaning observed in the various European terms
quoted ceases to be puzzling. Here Japanese has played an
important role in helping to shed light on Europe.

This completes our exploration of the semantic field connected

with “the bow”, a study that has taken us far beyond the question
of motivation or non-motivation.in language which, it will be
recalled, formed the starting-point of this discussion. By com-
paring the East and the West it has been possible to bring some
light to bear on hitherto unexplained aspects of both groups of
languages.
It will be realized that material related to “the bow” is ex-
160 Comparative Semantics

tremely varied, and it could of course be further explored in

other languages. Paul Canart summed up the open-ended
nature of semantic research, together with his belief in the
importance of comparative semantics in the following words,
with which it is fitting that this book should end:
Past experience of field research in comparative semantics has con-
vinced me that any topic selected will turn out to be practically
inexhaustible in its multiple aspects and that it will remain a per-
manent possession, a constant source of human and spiritual enrich-
ment, a fount of endless joys for ever new and fresh. Postera crescam
laude recens could well be the motto of this infinitely rich and in-
spiring humanist science which has seen the light at a time when
mankind so desperately needs it. Cedat machina homini!
Index

absences in language, 15, 142—44

abstraction in language, 16, 19, 23, 32, 33, 69—70, 74, 80, 93, 99, 135
adverbial -s, 7,9, 17, 25,55, 76
alliteration, 12, 37, 64
“already” and “‘beautiful’’, 31, 76
analogy, 8, 31, 56, 105, 121
“at once’, 6—28 passim
—, future (immediate) -»future (indefinite), 6-7, 9, 13, 20, 55
—, future (immediate) -»past (immediate), 9-10, 25
—, past (immediate) —»future (indefinite), 5, 9
—, >“‘within a few hours” (past and future), 6
=, “arrival, 24—25, “away*, 23; off’, 23
—, “as soon as”, 11, 27—28, 55; “when”, 11, 27, 55
—, correlatives, 16, 25—26
= “eye id; mand:?, 25: “deed, 25
—, “hardly”, 26, 28—29
—, intensive, 6, 12, 16-19, 24, 25
=‘one”,.20=27,,55
—, onomatopoeia, 17
—, past participle, 26—27
—, “place”, 6—7, 20-21
= ‘quick, 755 /shorty,.10
—, “riding”, 21—22, 24; “sitting”, 21; “standing”, 21, 25
_, “right”, 9, 17-18; “straight”, 9, 17-18
—,“‘stroke”, 16—17; “suddenly”, 16—17
—, “then-words”, 9, 11—15
augmentatives, 120—22, 126, 131, 134, 141-42

“bag’’. See “bulges”

“bite” and “sting”, 40—41, 46
“bitter” and ‘‘distress”, 39; “bitter” and “sew”, 38—40
“black”. See ‘‘darkness”’
borrowings, 34, 50, 53, 54, 68, 78, 85, 92, 134n, 140, 145
bound forms, 63, 77, 113, 115
“bow’’, 93-101, 104-8, 156-59
in architecture, 101, 107-8
“bow” and “bay”, 99, 107
“bow” and “‘bowstring”, in mathematics, 106—7
162 Index

“calf” and “shin”, 102, 138, 149

“spring”, 95, 96-98, 156
“tendril” and “vine”, 104-5, 112 ts
and “relaxation”, 64, 93-94, 96 297 1011 LOSS Sia
“tenseness”
“bulges”, 83, 102—4, 107, 115, 118, 127—50, 156—S7
and “pocket”, 139-40;
“bag” and “belly”, 103, 116, 139-40; “bag”
“bag” and “pouch”, 103, 140; “bag” and “skin”, 116
“belly” (“paunch”) and “‘calf” 102, 115, 122—235038
“beily” and “pregnant”, 104, 115, 118-19, 140
“belly” and “sea”, 119—20, 123
“bone” and “back”, 136-37; “bone” and “leg”, 102, 136—37, 148
“bulge” (“bloated”) and “arrogant”, 133-34
“bulge” and “belly”, 118, 123, 138, 146; “bulge” and “mouth”, 117, 136,
138; “bulge” ard “spring”, 157--S8
“cape” and “head’’, 135, 145; “cape” and “nose”, 135, 145; “cape” and “point”,
83, 127, 135; “cape” and “problem”, 83
expressive terms, in Japanese, 102—3; in French, 103; in English, 103
“thollows’’, 141
“mother” and “sea”, 119—20, 123
““mountain” and ‘‘mine”’, 130
“spine” and “keel”, 146, 148; “spine” and “shinbone”, 102, 137, 138; “‘spine”
and “thorn”, 137
“water” and “abundance”, 119

calques, 32, 58, 106, 112, 135

“cape”’. See “bulges”
“cheap” and “plentiful”, 87
“cheek”, “‘jaw” and “chin”, 49-51, 143
“chest” and “throat”, 133-34
Chinese characters, beauty of, 34, 71, 73
—, effect on Japanese, 33, 71, 72—73, 91, 94—95, 102n, 104, 112, 115, 147, 150
—, limitations of, 153n.45
—, meaning made visible, 8n, 47, 56, 62, 63, 147
—, phonetic, 38, 47, 48, 49, 65, 70, 118, 119, 123, 124, 147-48, 149 ;
—, radical, 15, 20, 21n.16, 38, 39, 47, 48, 57, 60, 62, 65, 66, 69, 70, 86, 94,
95,99, 104, 118, 124, 129, 149
“closeness’’ and “friendship”, 68
cognates, 9, 18, 31, 33, 57, 63, 69, 71, 83, 85, 88, 136, 140 :
coincidences in language, 31, 36, 45, 47, 53, 82, 104n.13, 107, 117, 120, 123,
126, 128, 148
comparative semantics, essential nature of, 3—4, 133
—, extension to phonetics and grammar, 124
—, importance for author, 160
—, objects, ways of seeing, 2—4, 66, 88, 120, 126-27, 133, 137, 138
“compare” and “equal”, 32
“compatison”’, “contest” and “side”, 32—33
concessives, 54—55, 76
consonants, importance in Arabic, 58, 59, 60, 62, 68, 92, 129, 134
“cover” and “‘container”, 131—32
“crooked” and “‘wrong, dishonest”, 19

“darkness’’, “cover” and “‘ant”, 116—17

 Index 163

“dark’’/“‘black” and ‘‘a large number”, 73-74, 79-80, 116-17

‘‘dark”’/“‘black” and “near”, 78—81
“dark, “close” and ‘‘shut’’, 80—81
“‘dense”’/“thick” and “‘often”’, 76
diminutives, 15, 85, 105, 123
diphthongs, reduction of, 125
“dot”, “spot” and “‘stain”, 48—49

East and West, fundamental identity, 1, 72, 75

“education” and “feeding”, 51—52
etymologies, 58, 83, 107, 119, 137
euphemisms, 31, 83, 111, 133-34, 144, 145, 150, 155n
“evening” and “yesterday”, 7—8

Far East. See Oriental languages

faux amis, semantic identity, 2—3
“feather(s)”, See “hair”
“flour”. See “smallness” and “‘research”

“grass”. See “hair”

“hair”, 108—26 passim, 150, 156—59

“feather” and “‘leaf’, 112, 124; “feather” and ‘wing’, 108, 150;
‘feather’, “wing” and “spring”, 150, 156, 158—59
“hair” and “belly”, 115, 123, 138
“hair” and “feathers”, 108, 110, 112, 113, 124, 157-58
“hair” and “grass”, 3, 108-11, 113, 120, 126
“*hair’’/“‘grass” and “flesh”, abundance of, 120, 122n, 126, 156
‘hair’ and “leaves”, 109, 112, 124, 126; “hair” and “mouth”, 117, 138;
“thair” and ‘‘sea”, 118—19;
“hair” and “vegetation”, 66, 108—12, 113-15, 124, 126
“‘scales’”’ and “moss”, 111—12, 115
“skin” and ‘‘cover”’/“‘wrapper”, 116, 118
“skin”’/‘‘flesh” and “hair”, 66, 115, 117, 123, 147, 156
“skin” and “‘leather, bark, fur’’, 66
“wool” (“‘sheep’’) and “water”, 118-19
“heavy” and “high esteem”, 128
“height” and “‘excellence’’, 129
“hinge” words, 5, 12
homonymous words, semantic affinity, 33-34, 95, 102, 105, 111, 112—13, 147

indicators, 51, 98, 111, 147, 158

Indonesian, symbolic value of final vowel, 125, 126, 137
Indonesian vowels, A—A, 82—90 passim
—, A-I, 32, 37—38, 64-65
—, E—A, 68—70, 75—79, 82, 88
=, E-I, 43-45
—, E—I and I—I, transition zone, 45—48
—, I-1, 49—56 passim, 63
—, U-I, 33, 65—67
164 Index

Indonesian vowel uw as a final, 113, 124

—, vowel u plus finaln or h, 124, 126; plusfinal k, 139
—, vowel u plus final t, 124, 149, 156
—, ending -ut, 113—20, 123, 126, 127, 147
—, ending -ung (-ong), 127—50, 156
intensives, in Dutch, 9, 18, 81; in English, 18—19; in French, 9, 18, 24; in German,
18; in Icelandic, 18, 27; in Indonesian, 35; in Japanese, 19, 147, 157
isolated words, 31, 87, 90
—, English words associated with “the bow’’, 93, 98—100

Japanese, aspirate, development of, 150—56 passim

—, kana, Sanskrit influence on, 152—55 passim
—, language, Chinese influence on, 33, 71, 72, 73, 91, 94-95, 104, 112, 115,
147, 150
—, sound changes, comparable European developments, 104
—, writing system, development of, 153n.45

languages, range desirable, 3—4

“Jeaf’’. See “hair”
“leisure”. See ‘“‘spaciousness”
“ight” and “‘low esteem”, 128
“light”, “open” and “broad”, 81; “light” and “a small number”, 79
“linguistics without language’’, 2

Malay-Indonesian. See Indonesian

meaning, duality within one word, 9, 30—31, 33, 81, 100, 101, 105—6, 109,
136, 148 ”
—, physical and figurative, 18-19, 39, 47, 57-63, 68-69, 94, 95,95, 113, 114,
116, 117, 122, 127, 128, 132, 133-34, 142-43
—, wide range within one word, 45, 58, 62, 64, 66, 67, 95, 98, 105-6, 108,
110, 112, 118, 129, 131, 147
“morning” and “tomorrow”, 7—8
motivation in language, 32, 44, 69--70, 74, 93-94, 100, 135—36

“narrow’’. See “‘tight spots”

nasals, survival of, 54
“near” and “‘after’’, 76
“need”’. See “‘tight spots”
numerical classifiers, 45—46, 65, 83, 151

objects, ways of seeing. See comparative semantics

“once’’. See ‘‘at once”
“one” and “‘same”’, 53—54
opposites, semantic identity of, 9, 30—31, 33, 94-95
Oriental “‘inscrutability”, a myth, 1, 72
Oriental languages, importance for Western linguistics, 1—2, 73

“page” and “‘blade’’, 86—87

phonetic. See Chinese characters
“pick” and “‘peck’’, 48, 67
“place” and “‘time’’, 31—32, 84
prefixes, Malay sa-, Indonesian se-, 52—55
 Index 165

—, in other languages, 71, 129

presence and absence in language, 28-29, 111, 1210143
“pressure”. See “tight spots”
“printing” and “beat”/“strike”, 77; “printing” and “pressing”, 77—78;
“printing” and “seal”, 78; “printing” and “stamping”, 77—78

“quick” and “‘early’’, 75

tadical. See Chinese characters

reduplication, in European languages, 20n.15, 89
—, in Indonesian, 55, 64, 90, 114, 139; in Japanese, 102-3, 156-57
“relaxation”. See “bow”
“research”. See ‘‘smallness”
“row”, “line” and “layer”, 37—38

Sanskrit. See Japanese

“scarcity” and “‘high cost”, 87
“scratch”, “write” and “paint”, 33-35, 66—67
“sea”. See “bulges” and “‘hair”
semantics. See comparative semantics
“skin”. See ‘“bulges” and “‘hair”
“slow” and “‘late’’, 75
“slow”. See “‘spaciousness”’
“smallness” and “research”, 56-57, 58-61, 62, 67
“accuracy’’, “‘smallness” and “research”, 55 —63 passim
“accurate” and “narrow’’, 69
“bore a hole” and “research”, 60—61
“cutting” and “research”, 61-62; “cutting” and “understanding”, 62—63
“flour” and “‘intelligence”’, 57—58, 62
“grind” and “research”, 57—61 passim
“small” and “‘care’’, 56
“space” —>“‘time”, 8—9, 17-18, 20, 22-23, 31, 75, 84
“space” and “‘time’’, 74—75, 84, 87
“spaciousness”, contrast with “‘tight spots”, 64, 69, 74, 82, 87, 88, 89
“ease” and “‘wealth’’, 87
“easy” and “‘light”, 87; “easy” and “‘slow’’, 88
“Jeisure” and “playing”, 90—91
“leisure” and “‘space”, 74—75, 82, 84, 87, 90-91
“leisure” and “‘strolling”’/‘‘wandering”, 74—75, 90-91
“level” and “explanation”, 82—83
“level” and ‘‘a level space’’, 82; “level space” and “‘road’’, 89;
“level” and “‘slow, quiet, peaceful’’, 20n.15, 87, 89
“slowness” and “‘lightness”’, 89
“spring”. See “‘bow’’, “‘bulges” and “hair”
“straight” and “right, honest”, 18—19
“strait”. See “tight spots”
suffixes, bearers of meaning, 41, 49, 110, 120—22, 126, 138, 156
synonymic couplets, 20, 57, 68, 74, 83

“tenseness’’. See “bow”

“thick” and “thin”, 79, 83
166 Index

“tight spots”, 64, 69—75 passim, 87-88, 89, 90

“dis-ease” and “‘poverty”’, 69, 71, 87
“heavy” and “difficult”, 87
“Jack” and “‘want”(‘‘need’’), 29, 63—64
“mountain pass” and “‘strait”’, 69—70
“narrow” and “distress”, 72, 74, 77, 80; “narrow” and “exact”, 68-69;
“narrow” and “mean”, 47; “narrow” and “strait”, 47;
“narrow” and “‘tight spot’, 47, 68—69; ‘‘narrow” and “urgency’’, 74
“near” and ‘“‘dark”/“‘black”, 73—74, 76, 78-81;
“near” and “distress”, 71; “near’’ and “haste”, 74;
“near” and “‘narrow”, 83; “near” and “‘shut’’, 81
“need” and “‘danger’’, 88; “need” and “haste’’, 88; “need” and ‘“‘work’’, 90—91
“pressure” and “‘crowd”, 70—71, 88;
“pressure”’and “‘grief’’/“‘distress”, 77, 78; ‘“‘pressure’’and “urgency”, 70—71
“time”. See “place” and “‘space”’
“towns” and “‘cities’~— “fortifications”, 85 —86
“towns” and “gardens”
~<«- “enclosures”, 85 —86

“vegetation”. See “hair”

verbs past in form, present in meaning, 62—63
past analysis » present understanding, 62—63
“seen” » “know”, 63
vowel i, “‘lightness’’, 41—42
—, “smallness”, 41, 42, 43-56 passim, 63, 65, 67
—, widespread symbolic value, 43, 48, 65
vowels i and u, semantic contrast, 41—43
vowel u, “bigness”, 41, 42, 49, 120-22, 143-44
—, “heaviness”, 41—42

“wealth” and “‘power’’, 35—37

West and East, fundamental identity, 1, 72, 75
words of obscure origin, 50, 67