Full download ebook at ebookname.

com

Oxford Studies in Metaphysics Volume 3 1st Edition

Dean Zimmerman

https://ebookname.com/product/oxford-studies-in-metaphysics-
volume-3-1st-edition-dean-zimmerman/

OR CLICK BUTTON

DOWLOAD NOW

Download more ebook from https://ebookname.com

More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Oxford Studies in Metaphysics Dean Zimmerman (Editor)

https://ebookname.com/product/oxford-studies-in-metaphysics-dean-
zimmerman-editor/

Oxford Studies in Ancient Philosophy Volume XXX Summer

2006 Oxford Studies in Ancient Philosophy 30 Sedley

https://ebookname.com/product/oxford-studies-in-ancient-
philosophy-volume-xxx-summer-2006-oxford-studies-in-ancient-
philosophy-30-sedley/

Oxford Studies in Ancient Philosophy Volume XLII 1st

Edition Inwood

https://ebookname.com/product/oxford-studies-in-ancient-
philosophy-volume-xlii-1st-edition-inwood/

Shakespeare Studies 1st Edition Susan Zimmerman

https://ebookname.com/product/shakespeare-studies-1st-edition-
susan-zimmerman/
Oxford Studies in Metaethics Volume 6 1st Edition Russ
Shafer-Landau

https://ebookname.com/product/oxford-studies-in-metaethics-
volume-6-1st-edition-russ-shafer-landau/

Oxford Studies in Philosophy of Law Volume 2 Leslie

Green

https://ebookname.com/product/oxford-studies-in-philosophy-of-
law-volume-2-leslie-green/

Metaphysics Q A Volume 3 1st Edition Jun Hong Lu ■■■

https://ebookname.com/product/metaphysics-q-a-volume-3-1st-
edition-jun-hong-lu-%e5%8d%a2%e5%86%9b%e5%ae%8f/

Haydn Studies 1st Edition W. Dean Sutcliffe

https://ebookname.com/product/haydn-studies-1st-edition-w-dean-
sutcliffe/

Reality Fundamental Topics in Metaphysics Toronto

Studies in Philosophy 1st Edition Peter J. Loptson

https://ebookname.com/product/reality-fundamental-topics-in-
metaphysics-toronto-studies-in-philosophy-1st-edition-peter-j-
loptson/
OXFORD STUDIES IN METAPHYSICS
OXFORD STUDIES IN METAPHYSICS

Editorial Advisory Board

David Chalmers (Australasian National University)
Tamar Gendler (Syracuse University)
Sally Haslanger (MIT)
John Hawthorne (Oxford University)
Kathrin Koslicki (Tufts University)
E. J. Lowe (University of Durham)
Brian McLaughlin (Rutgers University)
Trenton Merricks (University of Virginia)
Kevin Mulligan (Université de Genève)
Theodore Sider (Rutgers University)
Timothy Williamson (Oxford University)

Managing Editor
Jason Turner (Rutgers University)
OXFORD STUDIES IN
METAPHYSICS

Volume 3

Edited by
Dean W. Zimmerman

1
 3
Great Clarendon Street, Oxford ox2 6dp
Oxford University Press is a department of the University of Oxford.
It furthers the University’s objective of excellence in research, scholarship,
and education by publishing worldwide in
Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam
Oxford is a registered trade mark of Oxford University Press
in the UK and in certain other countries
Published in the United States
by Oxford University Press Inc., New York
ß The Several Contributors 2007
The moral rights of the authors have been asserted
Database right Oxford University Press (maker)
First published 2007
All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, or under terms agreed with the appropriate
reprographics rights organization. Enquiries concerning reproduction
outside the scope of the above should be sent to the Rights Department,
Oxford University Press, at the address above
You must not circulate this book in any other binding or cover
and you must impose the same condition on any acquirer
British Library Cataloguing in Publication Data
Data available
Library of Congress Cataloging in Publication Data
Data available
Typeset by SPI Publisher Services, Pondicherry, India
Printed in Great Britain
on acid-free paper by
Biddles Ltd., King’s Lynn, Norfolk

ISBN 978–0–19–921839–4 (Hbk)

ISBN 978–0–19–921848–6 (Pbk)

1 3 5 7 9 10 8 6 4 2
PREFACE

Oxford Studies in Metaphysics is dedicated to the timely publication of

new work in metaphysics, broadly construed. The subject is taken to
include not only perennially central topics (e.g. modality, ontology, and
mereology) but also metaphysical questions that emerge within other
subfields (e.g. philosophy of mind, philosophy of science, and philosophy
of religion). Each volume also contains that year’s winner of the Oxford
Studies in Metaphysics Younger Scholar Prize, described within.
D.W.Z
New Brunswick, NJ
This page intentionally left blank
CONTENTS

Oxford Studies in Metaphysics Younger Scholar

Prize Announcement ix

I. SYMPOSIUM: HUMAN BEINGS

1. Human Beings, Human Animals, and Mentalistic Survival 3
Denis Robinson
2. ‘‘Human Beings’’ Revisited: My Body is Not an Animal 33
Mark Johnston

II. MODALITY
3. Worlds, Pluriverses, and Minds 77
Mark Heller
4. Analyzing Modality 99
Michael Jubien

III. COINCIDENT OBJECTS AND TEMPORAL PARTS

5. Four-Dimensionalism and the Puzzles of Coincidence 143
Matthew McGrath
6. Time Travel, Coinciding Objects, and Persistence 177
Cody Gilmore

IV. MEREOLOGY
7. Theories of Location 201
Josh Parsons
8. Brutal Simples 233
Kris McDaniel
viii | Contents

9. Contact and Continuity 267

William Kilborn

V. FREE WILL
10. Metaphysical Compatibilism’s Appropriation of Frankfurt 283
Ted A. Warfield

Index of Names 297

THE OXFORD STUDIES IN METAPHYSICS
YOUNGER SCHOLAR PRIZE

Sponsored by the A. M. Monius Institute* and administered by the

editorial board of Oxford Studies in Metaphysics, the essay competition
is open to scholars who are within ten years of receiving a Ph.D. or
students who are currently enrolled in a graduate program. (Independ-
ent scholars should enquire of the editor to determine eligibility.) The
award is $2,500. Winning essays will appear in Oxford Studies in
Metaphysics, so submissions must not be under review elsewhere.
Essays should generally be no longer than 10,000 words; longer
essays may be considered, but authors must seek prior approval by
providing the editor with an abstract and word count by 1 November
2007. To be eligible for next year’s prize, submissions mailed within the
United States or Canada must be postmarked by 15 January 2008.
Authors mailing submissions from elsewhere should ensure that they
arrive before 20 January. Refereeing will be blind; authors should omit
remarks and references that might disclose their identities, but enclose a
cover letter with contact information. Receipt of submissions will be
acknowledged by e-mail. The winner is determined by a committee of
members of the editorial board of Oxford Studies in Metaphysics, and
will be announced in late February 2008. At the author’s request, the
board will simultaneously consider entries in the prize competition as
submissions for Oxford Studies in Metaphysics, independently of the
prize.
Previous winners of the Younger Scholar Prize are:
Thomas Hofweber, ‘‘Inexpressible Properties and Propositions’’, Vol. 2;
Matthew McGrath, ‘‘Four-Dimensionalism and the Puzzles of
Coincidence’’, this volume;
Cody Gilmore, ‘‘Time Travel, Coinciding Objects, and Persistence’’,
this volume;
Stephan Leuenberger, ‘‘Ceteris Absentibus Physicalism’’, forth-
coming in Vol. 4.

* The A. M. Monius Institute is a non-profit organization dedicated to the revival of

traditional metaphysics. Information about other activities of the A. M. Monius Institute
may be found at <http://www.ammonius.org>.
x | The Oxford Studies in Metaphysics

Enquiries and submissions should be addressed to:

Dean Zimmerman, Editor
Oxford Studies in Metaphysics
Philosophy Department
Rutgers University
Davison Hall, Douglass Campus
New Brunswick, NJ 08903
Part I
SYMPOSIUM: HUMAN BEINGS
This page intentionally left blank
1. Human Beings, Human Animals,
and Mentalistic Survival
Denis Robinson

introductory remarks
The debate over personal identity is a complex one, involving many
contrasting views, and ingenious and controversial arguments. But
standing back from the complexity, we can see a couple of major and
strongly contrasting groupings into which many recently defended
views may be seen as falling. At one extreme, we find diverse forms of
Psychological Reductionism, and at the other, views we may group
under the label ‘‘Animalism’’.1 Though Psychological Reductionism is
(deservedly, I think) more popular,2 Animalism also has its following,
and has fairly recently been given an unusually blunt, succinct, and
passionate defense by Eric Olson in Olson (1997).
The doctrinal opposition between these main groupings tends to line
up rather neatly with a general methodological opposition which has
ramifications in metaphysics generally: one where the key issue is,
roughly, the relative priority in metaphysics of folk intuitions and
a priori judgments, as against the findings of natural sciences.
But the very distance between these two main doctrinal groupings
makes it natural to look for some kind of intermediate position, and
for some corresponding methodological middle way for defending it.
My aim in this paper is to inspect this intermediate terrain, and in so
doing to draw doctrinal and methodological conclusions of my own.
My sympathies are with Psychological Reductionism, rather than with

1
An account of the various positions mentioned in these opening paragraphs will
follow shortly.
2
Johnston calls ‘‘dominant’’ a genre of Psychological Reductionism which he names
‘‘Wide Psychological Reductionism’’, citing Sydney Shoemaker, Anthony Quinton, and
David Lewis as amongst its defenders (Johnston 1987: 61). Olson (1997: 170 n. 13) lists
over thirty philosophers, including nineteen whom he refers to (p. 20) as ‘‘big names’’, as
holding some kind of Psychological view.
4 | Denis Robinson

Animalism, but it is not my aim here to defend the former, nor to attack
the latter. Rather, I aim to bring out the importance in this area of very
general ontological assumptions or presuppositions, and the way in
which the outcome of applying a particular methodology may be radic-
ally affected by them. I shall build my discussion around one well-known
attempt—Johnston (1987)3—to defend a middle way between the ex-
tremes of Animalism and Psychological Reductionism. I shall argue that
the attempt is unsuccessful, that the middle ground it attempts to occupy
is unstable, and that the methodological proposals invoked in its defense
cannot do what is asked of them.
Not only is Johnston’s view in Johnston (1987) (henceforth, ‘‘the Human
Beings view’’4) one which might be thought to combine elements
characteristic of our main opposing camps, it also offers what might
be seen as an attempt at a methodological middle way in support, since
(as we shall see below) it invokes a priori intuitions about personhood
on the one hand, while expressing caution about such intuitions, and
prioritizing naturalistic ontological categories, on the other. I shall
argue, however, that any substantive alternative to Psychological Re-
ductionism which that methodology can be made to support will be
even closer to Animalism than the Human Beings view, which is itself,
despite the appearance of compromise, already quite close to an Animalist
view.
To move on, we must begin by saying more about the kinds of
positions referred to above. I shall say quite a bit about Johnston’s
position, the kind of ontological standpoint which underpins it, and
the errors, as he sees them, which he wishes to avoid, before finally
examining his methodological proposals, and the positive arguments for
his view, to which they lead.
Psychological Reductionists agree that the constitutive criteria for
personal identity over time are predominantly psychological. Various
kinds of psychological connection may hold between a person’s earlier

3
It is important to realize that Johnston (1987) alone does not give an accurate
impression of Johnston’s overall position, as he in effect points out in Johnston (1987)
n. 8, p. 64. Fully to appreciate his subtle and complex view one must read also Johnston
(1989a), Johnston (1989b), and Johnston (1992b). Of these Johnston (1989b) is perhaps
both the most important and the most difficult. A critique of Johnston’s whole view would
be a much larger undertaking than space here permits. My own take on the personal
identity debate may be found in Robinson (2004).
4
I capitalize the phrase ‘‘Human Beings’’ throughout, to signal that I am using it in
Johnston’s precisely articulated sense, not merely in an everyday sense.
 Human Beings and Mentalistic Survival | 5

and later mental states: combinations of such connections constitute

the essential conditions for personal identity over time. Psychological
Reductionists differ over the relevant connections and combinations,
and their required causal underpinnings. (Johnston speaks of ‘‘Wide
Psychological Reductionism’’, where ‘‘Wide’’ connotes a relatively lib-
eral attitude to the permissible kinds of causal underpinnings.)
Animalism is the view that the familiar persons of our acquaintance
are strictly and literally identical with members of a particular animal
species, Homo sapiens. Members of this species are typically persons for
most (but perhaps not all) of their lifespans. There is no special problem
of personal identity: identity-conditions for a person of this familiar
kind are just identity-conditions for the relevant kind of animal, much
like the identity-conditions for other animals. An animal is born, it
lives, it dies. For these events and processes, there are no essential
mental prerequisites, certainly no requirement of psychological con-
nectedness or continuity. Thus, Animalism is less a view about the
nature of persons in general, more a view about the nature of those
with which we are mostly acquainted. If there are angels, thinking
automata, Martians, or dolphins which are people, then the identity-
criteria for those persons will just be those appropriate to such beings:
there are no identity-conditions for persons per se.
As noted, the doctrinal opposition between Animalism and Psycho-
logical Reductionism aligns naturally, though not inevitably, with a
methodological opposition. Psychological Reductionist views are typic-
ally justified largely by appeal to intuitions about the loss or preservation
of personal identity in various kinds of real and imaginary cases: by the
‘‘method of cases’’, as Johnston calls it. But this methodology has been
called into question in a variety of ways.
Quine popularized the view that the analytic/synthetic distinction
is at least hazy, and at best a matter of degree—some truths in which
a term figures merely being more centrally constitutive of the term’s
meaning, or the concept it expresses, some less. This casts a shadow on
any correlatively understood enterprise of conceptual analysis, of which
the ‘‘method of cases’’ is typically a core component. The shadow falls
particularly on our ability reliably to recognize a priori the genuine
(‘‘metaphysical’’) possibility or impossibility of some circumstance—as
opposed, for instance, to its degree of familiarity or bizarreness, or its
relationship to well entrenched empirical assumptions. Complementing
both this aspect of Quine’s work, and his epistemological naturalism, we
6 | Denis Robinson

have Kripke’s and Putnam’s work on rigid designation and natural

kinds, suggesting that the essence of a kind of thing might need to be
discovered by empirical inquiry, with conceptual analysis being, not
merely epistemically suspect, but quite beside the point.5
These trends lead easily to the thought that we should reject the
method of cases in favor of seeking where possible to identify the
category of familiar persons with some prima facie appropriate natural
kind. Biological kinds, and in particular animal species, are standardly
taken as paradigms of such kinds. With a little help from evolutionary
theory, Animalism emerges from these considerations as a natural rival
to Psychological Reductionism.
In Johnston (1987) Johnston mounts a qualified critique of the
method of cases, specifically in its application to the topic of personal
identity, and proposes an alternative to Psychological Reductionism
which, like Animalism, gives crucial precedence to the notion of a
biological kind.6 But he uses the notion of a biological kind in a more
subtle way than does Animalism, in arriving at an account of the
persistence-conditions for persons. Appealing to an alternative method-
ology, he argues that familiar persons are to be identified with ‘‘Human
Beings’’: entities having persistence-conditions similar to, but not quite
the same as, those of human animals. Indeed, though the kinds ‘‘Human
Being’’ and ‘‘human animal’’ differ in their persistence-conditions, the
crucial differences are rarely manifest in ordinary life, concerning as
they do extreme and mostly imaginary circumstances.
Johnston’s argument for this view appeals to the special role of certain
psychological states, namely experiential memories, in our knowledge
of personal identity; in this it is reminiscent of Psychological Reduc-
tionism, particularly in those neo-Lockean versions which make experi-
ential memory crucial. Nevertheless, Johnston (1987) rejects the idea of
straightforwardly mentalistic identity- or persistence-conditions. On
the Human Beings view, psychological states and relations have a

5
Though popular, the idea that Quine, Putnam, and Kripke provide us with reasons,
good or bad, to replace conceptual analysis in the service of metaphysics with appeals to the
findings and categories of the natural sciences is not inevitable. Frank Jackson’s recent
Jackson (1998) is one work which defends the view that these supposedly disparate
methodologies should actually be seen as complementary.
6
I shall not much discuss Johnston’s attack on Psychological Reductionism as based on
the method of cases. It revolves in part around a well-known conundrum for the method of
cases due to Bernard Williams. For some criticisms of it, see Oderberg (1989).
 Human Beings and Mentalistic Survival | 7

merely evidential rather than a constitutive relation to personal iden-

tity. Any possibility that a person’s survival might be sustained by
psychological connections which transcend certain biologically based
limitations, as in the fantasies of teletransportation, or of machine-
assisted transfer of psychological states (or the informational states
which realize them) from one brain to another, is ruled out. In all
this, the view resembles Animalism. The key difference is that a
Human Being, unlike a human animal, could persist through trans-
plantation of its living brain into a different body, or in the form of the
proverbial ‘‘brain in a vat’’.

‘‘human beings’’, ‘‘human animals’’,

and the constitution relation
Johnston’s view is that the (normal) relation between human beings and
human animals is one of constitution, using this word in a technical
sense commonly employed, with minor variations, by contemporary
ontologists. The constitution relation holds between material things
which occupy the same place at the same time and are composed of
the same matter but which are nonetheless not strictly identical.7 An
important special case is the relation between a material substance and
the matter which composes it. Notoriously, the constitution relation is
a time-relative or changeable relation, since the matter composing
material substances may change.
It’s important to notice that the constitution relation may hold
between material substances—bona fide, everyday objects—and not
just between a substance and its composing matter. For instance, a
mountain might be composed of a single gigantic rock; subsequently
thermal stress may cause the rock to fracture into small pieces, yet the
mountain persists. The mountain is at one time, not at others, consti-
tuted by the rock. Having different histories, they are never identical.
Similarly, a cloak composed of a button and a piece of cloth may lose its
button, which may be destroyed. Subsequently the cloak is constituted
(solely) of the piece of cloth. Cloak and cloth are never identical, since at

7
At least not in typical cases. There is a debate about whether identity can ever be a case
of constitution. Cf. n. 9.
8 | Denis Robinson

an earlier time the button is part of the cloak, though it is never part of
the piece of cloth.8
Thus, the constitution relation is importantly different from identity,
despite being readily confused with it.9 An important instance of this
general distinction, is that between cases in which one thing ‘‘is’’
another, in the sense of constituting it—as the boulder, at certain
times, ‘‘is’’ the mountain, or the cloth ‘‘is’’ the cloak—from cases in
which an entity falls, temporarily or contingently, under a ‘‘phase-
sortal’’ predicate. The latter are cases of genuine identity. A person
begins life as an infant, passes out of childhood into adolescence, and,
with luck, grows out of adolescence into adulthood. For a limited time,
the person ‘‘is’’ the adolescent. But this is an ‘‘is’’ of identity, rather
than of constitution. The person ceases to be an adolescent, but is
nonetheless identical with that adolescent. The adolescent, on ceasing
to be an adolescent, does not on that account cease to be.10

8
These examples are from Robinson (1982: 319). Though Olson has elsewhere con-
tributed to the literature on the constitution relation, in Olson (1997: 101–2) he is
remarkably curt and dismissive about views appealing to it. But if it is hard to deny
there may be clear examples of ordinary things standing in this relation, appeal to it ought
to be a prima facie option in the personal identity debate.
9
Another point of which these examples may remind us is that, even after we set aside
instances of the ‘‘is’’ of predication, it is neither necessary nor sufficient for the concept of
identity to be invoked by a given statement that terms appear linked by ‘‘is’’ or other
forms of the verb to be. In fact, I believe that even such everyday locutions as ‘‘is identical
with’’ need not signal identity in the philosophers’ sense: for they, too, may be used by the
folk to express the constitution relation. It will be evident that I do not in general take
speakers’ introspective judgments as to their meanings or concepts as an infallible guide,
since ordinary speakers will not automatically distinguish particular utterances which
attribute the constitution relation from assertions of identity in the strict, philosophers’
sense. Failure to note these points makes it spuriously easy to impute to common sense a
strong commitment to Animalism.
As a four-dimensionalist I believe that the constitution relation can be analyzed in terms
of identity of temporal parts: x constitutes y at t iff the appropriate temporal part of x is
identical with the cotemporaneous temporal part of y. In special cases this means that
identity is a limiting case of the constitution relation. Johnston however rejects such an
approach (see Johnston 1992a). It follows that, on his view, no Human Being is ever
identical with a human animal, nor with any temporal part of a human animal.
10
For present purposes, we may characterize a ‘‘phase-sortal predicate’’ as follows. A ‘‘sub-
stance predicate’’, first of all, is one which applies to all and only members of a kind of entity,
where members of a kind in the relevant sense share, inter alia, identity- and persistence-
conditions. Applying a substance predicate to an entity thus carries implications for how it
and members of its kind are to be individuated and tracked through time, and for which
kinds of events they can and cannot, logically, survive. A phase-sortal predicate (such as
‘‘adolescent’’, ‘‘blonde’’, or ‘‘widow’’), finally, is logically equivalent to the conjunction of a
substance predicate with attribution of some temporary or accidental characteristic, and
thus, by the rules for conjunction, carries similar implications. See Wiggins (1980).
 Human Beings and Mentalistic Survival | 9

The important contrast here is with the case of the boulder which,
I say, does indeed cease to exist when it fragments. Thus we should not
see ourselves as referring to the mountain itself, under the description
‘‘boulder’’, when we refer to the boulder (though in many contexts
it may be a matter of indifference which we refer to, so that we leave
it indeterminate). Even more clearly, when we speak of the piece of
cloth, in the example of the cloak, we are not merely speaking of
the cloak itself under a phase-sortal description comparable with ‘‘ado-
lescent’’.
Here I follow Johnston in adopting the vocabulary of what I call a
substance-ontology. Wiggins (1980) is a well-known exposition and
defense of principles typical of such ontologies (though with, inevitably,
a number of idiosyncrasies and optional extras). According to these
ontologies, the world contains a limited variety of kinds of substances.
Each of these kinds has distinctive criteria associated with it, criteria
which, if articulated, would specify what kinds of events could count as
commencing or terminating the existence of members of the kind, and
how members of the kind are in principle to be identified, distinguished,
counted, and traced over time. Important amongst the substance kinds
are the natural kinds, including the biological kinds. Substance-theorists
who prioritize natural kinds typically view the associated persistence-
conditions as naturally, as opposed to conventionally, determined.
Consistently with these views, substance-ontologists typically deny
the reality, or disparage the status, of arbitrary mereological aggregates
of substances or other entities. They also typically deny the reality of
temporal parts of substances, and insist on sharply distinguishing issues
of change and identity for substances from parallel issues involving
either events, or portions or quantities of matter.
Substance ontologies may differ widely in the range of kinds they
recognize, and how they see those kinds as related. Roughly speaking, we
can distinguish the form of a substance-ontology, from its content—the
particular view it takes as to what the constraints on substance-concepts,
and on substances, are. Natural kinds will typically be included, but
that leaves open what other kinds might also be admitted. Suitably
well-defined or determinate conditions for identity and persistence will
always be required, but there is much room for debate as to what‘‘suitably
well-defined or determinate’’ means here. But however these issues are
resolved, questions about substance-identity will for substance-theorists
always ultimately come down to questions about either nominally or
10 | Denis Robinson

metaphysically essential properties of kind-members: those properties

possessed essentially by things belonging to the kind in question.
Thus we may say that on the substance-ontology approach, persons
will comprise all or some of the members of all or some of the kinds of
substances which typically have what it takes to be a person. For present
purposes, and compatibly with all Johnston says, we can sum up what it
takes to be a person, as the capacity for a sufficiently rich, sufficiently
integrated kind of mental life. Consistently with all this, one might take
it that the kinds of substance there are include none whose persistence-
conditions coincide with anything which could plausibly be considered
as psychological unity-conditions—conditions for the unity and per-
sistence of a mind considered as such. An upshot of such an attitude
would be that ‘‘person’’ might not qualify as a substance-concept, nor
even as a phase-sortal concept, being more akin to a mere general concept
like ‘‘fast mover’’, or ‘‘red thing’’. Given such assumptions, Animalism
is a natural (and ‘‘naturalistic’’) conclusion.11
Johnston adopts, not just the vocabulary, but the stance, of substance-
ontology, as sketched above, including, it seems, the rejection of men-
talistic persistence-conditions for bona fide substances.12 Certainly
his argument at a crucial point seems to rest on some restrictive set of
background ontological assumptions—assumptions which effectively
load the dice against any psychological account of personal identity.
Johnston recommends a standpoint according to which psychological
continuity and the like should be viewed merely as providing under
normal circumstances relatively good evidence of personal identity,

11
In addition to Olson, Animalists include Paul Snowdon, Peter van Inwagen, and
David Wiggins. See Snowdon (1990), van Inwagen (1990), Wiggins (1980), and Wiggins
(2001). Note that Wiggins’s Animalism is different in its details from, for example,
Olson’s. Wiggins views the concept ‘‘person’’ itself as a substance-concept and indeed a
natural kind concept, where the kind in question is precisely ‘‘human animal’’. Neverthe-
less, he has been prone to suggest, obscurely, that the psychological nature of persons
might form part of the essence of this kind, and so have some bearing on the individuation
of persons, thus diminishing slightly the distance between his form of Animalism, and a
Psychological view of personal identity, but creating a degree of distance between identity-
criteria for this kind, and for other kinds, of animal. But as compared to his view in
Wiggins (1980), in Wiggins (2001) this distance from other Animalists seems to have been
lessened.
12
It should be emphasized that, as I understand it, this is no implication of substance-
ontology per se, though it accords well with the pre-eminent status typically accorded by
substance-theorists to natural kinds.
 Human Beings and Mentalistic Survival | 11

much as do fingerprints or physiognomy. It would be as serious an error

to take normally reliable psychological criteria as constituting personal
identity, as it would be to take sameness-of-fingerprints as constituting
personal identity.
Suppose we ask: could a machine, by reading the information from
the brain of a human person and depositing it in another brain (real or
synthetic), thereby bring about a continuation of the person’s mental
life and so provide a means of continuing the original person in exist-
ence in another body? Johnston suggests an attitude to this idea, much
like that which an Animalist might take. From such a standpoint, the
suggestion would be as bizarre and loopy as the idea that a machine
duplicating fingerprints, or a plastic surgeon reproducing physiognomy,
could accomplish the same effect. The most charitable view we could
take of such ideas about fingerprints or physiognomy would be that they
depended on radically mistaken overgeneralizations. Reflection on what
is actually done by a fingerprint-duplicating machine, or a plastic sur-
geon, should be enough to make the mistake plain, since they disturb
the very regularities which make this evidence normally reliable. Simi-
larly, Johnston suggests, with the information-transfer machine.
Johnston’s views thus run along similar lines to those of the Animal-
ist, however they diverge at a crucial point. His views are, (i) that the
relationship between a human animal, and a Human Being (in his
technical sense), is not the relationship of identity, but the relationship
of constitution, and (ii) that a Human Being could in principle outlive a
human animal which formerly constituted it, though this possibility is
one which does not get actualized in real life as we know it.13 Indeed we
may pretty much say that according to the Johnston of Johnston (1987),
every actual individual Human Being is a human animal—provided that
we understand this ‘‘is’’ as the ‘‘is’’ of constitution.

projective errors and the rejection of the

method of cases
Johnston does not think that the error of those who misguidedly under-
stand as constitutive criteria what are merely normally reliable evidential

13
Setting aside, I would presume, various gruesome and (literally) short-lived border-
line cases involving beheadings.
12 | Denis Robinson

criteria for personal persistence arises simply from egregious overgen-

eralization. The relevant overgeneralizations gain spurious plausibility,
Johnston thinks, from our tendency to commit what he calls ‘‘projective
errors’’. We are susceptible to a broadly Cartesian view of the self
as a ‘‘bare locus’’ of consciousness: an enduring entity, distinct from
anything physical, distinct too from any particular mental events, but
underlying, sustaining, and unifying such events, simply by being a
constant that-to-which-psychological-presentations-are-presented. No
amount of physical continuity, nor even of psychological continuity, is
absolutely necessary to the continued existence of such an entity. It is
this view of selves as ‘‘soul pellets’’14 which Johnston describes as a form
of ‘‘projective error’’:15 we are unaware of the physical underpinnings
and complex processes which give rise to episodes of consciousness, and
fall into the trap of projecting our lack of awareness into a lack in reality,
taking it that there are no such underpinnings and complexities.
Johnston suggests that even those philosophers whose views about
personal identity revolve around the explicit rejection of any such
account of the self, nevertheless take seriously intuitions which
amount only to silly overgeneralization, in part because they retain an
unconscious residue of the soul pellets view.16 For these and other
reasons, Johnston proposes to reject the ‘‘method of cases’’, which
endeavors to extract criteria for personal identity from, principally,
our intuitions about real and imaginary examples of personhood and
its vicissitudes.

reidentification, mentalistic survival, and

the brain as the natural organ of mentation
Johnston does not totally renounce all appeal to intuitions about imagin-
ary cases. But he advocates a strict methodological regime according
to which only those intuitions which cannot be traced to misleading

14 15
Cf. Johnston (1989a: 372). Cf. Johnston (1992: 593).
16
It’s worth noting here that Locke’s inauguration of the modern philosophical discus-
sion of personal identity, used, in the specific context of an argument supposed to show the
irrelevance of sameness of soul (or of organism) to personal identity, the example of
the soul of a prince coming to inhabit the body of a cobbler. In other words, Locke might be
accused of appealing to intuitions grounded in precisely the view he was attacking.
 Human Beings and Mentalistic Survival | 13

origins like those he identifies can lend support to an account of personal

identity. Johnston’s alternative proposal instead rests centrally on the
observation that we normally identify and reidentify human persons
relatively unproblematically, and proceeds by asking what kinds of
entities could be unproblematically identified and reidentified in normal
circumstances by the practices we habitually use.
Given this methodology, we might have expected Johnston to opt
for Animalism, but in fact he opts for the Human Beings view. His
argument includes two crucial steps. First, he asserts that ‘‘if anything
deserves the name of a conceptual truth about the relation between
persons and minds, it is the claim that a person cannot be outlived by
(what once was) his own mind.’’17 (I’ll call this claim MS, for ‘‘mental-
istic survival’’: it says survival of the mind suffices for survival of the
person.) Secondly, Johnston urges ‘‘a properly naturalistic view of our
mental functioning’’ as the characteristic functioning of a particular
organ, the brain, claiming in this connection that ‘‘talk of a mind is
overly reified talk of an aspect of some minded thing.’’18
These are the two crucial steps from which Johnston concludes that
we should view ourselves as Human Beings, entities normally consti-
tuted by a human animal, but such that in extraordinary (and so far
imaginary) cases such as brain transplants, or brains-in-vats scenarios,
they might be constituted merely by a living human brain, or by a
creature resulting from transplant of such a brain. On this view, the
death not of a human animal, but of a human brain, constitutes the
death of a human person.19
How, we may wonder, do these two steps mesh with Johnston’s
proposed alternative methodology? Even if MS and the conclusion he
derives from it strike us intuitively as plausible, by Johnston’s lights
that in itself counts for little. Is his appeal to this principle really an
improvement on a direct appeal to intuition? This is probably the point
17
Johnston (1987: 77). Olson cites this remark (Olson 1997: 12) in a context which
suggests he takes it to show that Johnston supports a Psychological account of personal
identity—a debatable reading which seems to me to take the remark very much out of
context. Johnston’s account is ‘‘psychological’’ in what is at best a stretched sense, and he
explicitly opposes typical psychological theories. Cf. n. 29, below.
18
Johnston (1987: 78).
19
In discussion, it is often suggested that Johnston’s view here amounts to the view that
a person is identical with their brain. This claim, which Johnston rebuts, certainly does not
follow from the view that a person is in certain extraordinary circumstances to be traced
by tracing their brain. The fact that my car could survive destruction of its doors does not
mean that they are not really part of it!
14 | Denis Robinson

where Johnston’s argument seems most to resemble traditional appeals

to conceptual analysis. Does he merely capitulate to the method of
cases at this point?

mentalistic survival, experiential memory,

and non-mentalistic individuation
Johnston does attempt to tie his appeal to MS to his proposed alternative
methodology. The link is meant to be the role of experiential memory in
self-identification. To see how, we must examine MS and the use
Johnston makes of it. To this end, let’s first compare MS directly with
comparable claims about body organs.20
The claim ‘‘a person cannot be outlived by (what once was) his own
heart’’ is false: we are familiar with cases in which the heart remains
alive and is transplanted into the body of another person, after the death
of its previous owner. Does the same possibility exist, in principle, for
the brain? Johnston urges, plausibly to my mind, that we could not say
the human animal as such would survive if the brain alone were kept
alive and transplanted.21 So a human animal can be outlived by what
once was its own brain. Now if the person were (identical with) the
human animal, it follows that the person could likewise be outlived by
what once was its own brain. The denial of this possibility, conversely, is
the upshot of Johnston’s argument in Johnston (1987). He does not
deny the surgical possibility, but claims rather that survival of the
transplanted brain would be sufficient for survival of the original person.
There is a peculiar subtlety involved in differentiating Johnston’s view
from Olson’s particular version of Animalism, since Olson argues—
contrary to the claim I made above—that an animal cannot survive
without a brain, nor vice versa, taking the view that where the living
animal brain goes, there the animal goes. So Olson does hold that an
animal could in principle survive being reduced to the condition of a
brain removed from the skull which once housed it, and kept alive

20
A full discussion of whether MS is true or a priori would require a detailed discussion
of what might or might not count as ‘‘reidentifying a mind’’. We may reasonably dispense
here with such discussion because Johnston’s next move substitutes, rightly or (as I think)
wrongly, the issue of reidentifying a brain.
21
But note that Olson disagrees on precisely this point. See Olson (1997: ch. 6).
 Human Beings and Mentalistic Survival | 15

through artificial means.22 It may appear therefore that whereas John-

ston takes persistence for Human Beings, on his account, to differ from
persistence for human animals, according to Olson’s account it precisely
coincides with the latter!
This does not threaten the generic difference between the Human
Beings view and Animalism, since Animalists might have other views
than Olson’s as to necessary and sufficient conditions for animal sur-
vival. Furthermore, though Johnston, as we shall see, associates human
mental life with the human brain, Olson’s view is that it is not the
whole brain, but the cerebrum, which is ‘‘that organ which is most
directly responsible for your higher mental capacities such as reasoning
and memory’’.23 If Olson is right about this, then Johnston should say
‘‘cerebrum’’ wherever he in fact says ‘‘brain’’. In addition, Olson does
not consider it necessary that the human animal, to survive, should
retain a functioning cerebrum—the remainder of the brain, if it is alive
and performing its normal functions, will do. In other words, for Olson,
the cerebrum is an inessential part of the brain qua organ essential to
survival. Thus the appearance of coincidence between Olson’s and
Johnston’s views disappears once we take these subtleties into account:
the part of the human animal essential for survival of Johnston’s
Human Being is in fact disjoint from the part essential, according to
Olson, for survival of the animal itself.
Finally, and most importantly, this coincidence would at best be only
an extensional equivalence. Even were Johnston’s reasoning and
Olson’s to arrive at exactly the same conclusion as to what it minimally
takes for a human person to survive—as it might be, continued survival
and functioning of the brain as a whole—their views would remain
intensionally different, since they differ as to which functions of the
brain would earn it this crucial status.
This is particularly relevant in the light of the fact that Johnston
views the relation of human animal to Human Being as one of consti-
tution, and in addition takes the view that constitution always excludes
identity. It seems then that even if human animal and Human Being
always exactly coincided, due to the brain’s functional versatility, Johnston
would be required by his own lights always to see them as distinct,
whereas Olson would see the allegedly distinct substance-category of
Human Beings as entirely spurious.

22 23
See especially Olson (1997: 131–42). Olson (1997: 9).
16 | Denis Robinson

Returning to Johnston’s main line of argument: he derives the claim

that a person can survive through survival of their brain alone, from
MS: the claim that it is a priori that survival of mind is sufficient for
survival of person. And the second main step of his argument proceeds,
essentially, by claiming that the truth of MS is preserved (though not,
presumably, its a priori status!) if the word ‘‘brain’’ is substituted for
‘‘mind’’. He says in support of this: ‘‘A human mind is just a mode of
functioning of a natural unit (e.g. a human organism or a human brain)
whose conditions of persistence are statable in nonmental terms. This is
the sense in which talk of a mind is overly reified talk of an aspect of
some minded thing.’’24
Note how an implicit commitment to a restrictive ontology, most
plausibly some kind of substance-ontology, emerges in the reference to
natural units whose conditions of persistence are statable in nonmental
terms. This is a pivotal point in Johnston’s argument, and one at which
his general ontological presuppositions come into view. Given the
aim of discovering what kind of thing it is which we are most readily
seen as unproblematically identifying and reidentifying in our everyday
person-identificatory practices, it transpires that the candidate kinds
are to be ‘‘natural units’’ such as organisms or body organs, or composite
entities whose persistence-conditions may be formulated, in principle
non-mentalistically, in terms of such natural units. Mentalistic persist-
ence-conditions as such are ruled out.
Here a major general moral emerges. The results yielded by
Johnston’s methodology will inevitably be relative to the options the
world is taken to serve up to us in the way of candidate kinds for our
identificatory practices to be seen as tracking. Since different ontologies
will take different stands on this, there is a danger that the choice of a
particular ontology already tendentiously limits the outcome of the
inquiry before any appeal to identificatory practices can enter the picture.
I believe this actually happens in the case of the answer Johnston
supplies. Indeed I believe that a central weakness of Johnston’s preferred
methodology is its inevitable need for some pre-established restriction
and ranking of the kinds of entities which are to be seen as candidate
targets of our identificatory practices. The more general moral, however,
extends beyond the context of Johnston’s own methodology.

24
Johnston (1987: 80).
 Human Beings and Mentalistic Survival | 17

Consider for instance the defense of conceptual analysis mounted in

Frank Jackson’s recent Jackson (1998). Johnston’s and Jackson’s meth-
odological proposals share a common structure. Faced with the meta-
physical question ‘‘What are Fs?’’ each in effect proposes a test to be
applied to the members of some favored basic ontological pool. Those
members of the pool which best pass the test, thereby qualify to be
counted as the Fs. Jackson’s test may be thought of as generated by
Ramsifying an implicit term-defining ‘‘theory’’ of Fs, taken as revealed
piecemeal by conceptual analysis; Johnston’s (in the specific case where
the Fs are the persons), as generated by asking what kind of thing it is
which we are most readily seen as unproblematically identifying and
reidentifying in our everyday person-identificatory practices. Thus
Jackson and Johnston fish with different nets. But what entities you
can fish out of an ontological pool depends not merely on what net
you use, but on what the pool contains in the first place. If mentalistic-
ally individuated entities are not to be found in it, no net will fish them
out. In other words, any metaphysical methodology which has this
broad structure, identifying the Fs by applying a net or filter to some
predesignated ontological pool, will only be as good as the reasons
available for preferring that pool to begin with. A central weakness of
Johnston (1987) is the lack of any overt argument in favor of the
preference, albeit a fashionable one, for a pool which contains biological
and other natural kinds, but which excludes mentalistically individuated
substances per se.
Johnston’s choice here (to allow only non-mentalistically individuated
entities as bona fide substances) is not altogether arbitrary or unmoti-
vated. He claims that the method of cases fails in this context, in part
because ‘‘the’’ concept of a person, taken as what is in common to all the
different conceptions of personhood which have historically been held or
which may emerge from reflection on imaginary cases and the like, is
impossibly vacuous. Nothing but a soul pellet could fit it, yet a capacity
unproblematically to reidentify soul pellets would be impossible to
account for; nor would it make sense to attach the kind of importance
to the vicissitudes of soul pellets which we in fact attach to issues of
personal identity. Thus it is natural for him to assume that his alternative
method must to some extent rely on nature, rather than on our concepts,
to demarcate the persons: to assume, in other words, that the kind
‘‘Human Person’’ must be constructible relatively straightforwardly
18 | Denis Robinson

out of natural kinds. And in this context, as noted, Johnston identifies the
relevant natural kinds as the biological kinds.25
Nevertheless, it would be possible to meet the demand that the
persons be in some broad sense ‘‘naturally’’ individuated, and yet
provide a greater range of options in seeking a kind to match with our
ordinary person-identificatory practices. Discounting hopelessly gerry-
mandered or arbitrarily or extrinsically individuated ‘‘kinds’’ need not
oblige us to reject mentalistic or psychological persistence-conditions as
Johnston does.
Johnston refers to ‘‘the naturalistic framework taken for granted
by scientifically validated common sense’’: but I doubt that respect for
scientifically validated common sense demands that we reject the pos-
sibility of mentalistic persistence-conditions. It is admittedly a matter of
considerable philosophical debate how to integrate the mental into a
naturalistic world view: but scientifically validated common sense is
happy to admit all sorts of entities which are not, but which supervene
on, those which figure explicitly in actual science. There is for instance
no science of bricks as such, but given that bricks supervene on mol-
ecules, and are relatively naturally individuated, scientifically validated
common sense has nothing to say against them. Similarly, I can com-
monsensically count the ripples on a pond and see how far they travel,
with a clear scientific conscience, even if I acknowledge that ripples
cannot exist independently of the quantities of water which succes-
sively, by their causally interrelated movements, constitute them. It’s
just not true that every bona fide materially constituted entity recog-
nized by the union of common sense and science, and which persists and
can be reidentified over time, counts as an independent substance which
meets the stringent criteria laid down for substance-concepts by writers
such as Wiggins, or which belongs to the biological natural kinds with
which they are so taken.
In Johnston (1987) Johnston says nothing which specifically addresses
this issue. Since Locke,26 though, we should know better. Locke’s
most profound and original contribution, I believe, in his discussion
of personal identity, is not his proposal of a psychological criterion for

25
Johnston does acknowledge the need for further argument in support of his priori-
tizing of non-mentalistically individuated natural units, as preferred targets for our
reidentificatory practices, in his n. 19 to Johnston (1987).
26
All references are to Locke (1961, Book II, ch. 27).
 Human Beings and Mentalistic Survival | 19

personal persistence, but something which is logically prior to that

proposal—his recognition of the fact that, given an ontology of funda-
mental natural kinds (even including ‘‘supernatural’’ kinds such as
spirits), we may also recognize the existence of additional well-indi-
viduated substance-kinds, distinct from more fundamental and autono-
mous kinds, but at all times supervening on them, because constituted
out of them. This is the lesson he draws from consideration of the case of
living things, and applies, in a brilliant and original analogy, to the case
of persons. It is a lesson of which materialists should stay specially
cognizant, since the alternative historically was to believe in an entire
category of fundamental immaterial substances, ‘‘souls’’ or ‘‘spirits’’,
having their own primitive and unanalyzable persistence-conditions,
and somehow guaranteed by their intrinsic nomic essence to sustain
mental life and to individuate mental lives.
Locke in many ways fits the model of a substance-ontologist,
attempting to argue for the cases of living things, and of persons,
alike, that persistence-conditions for things of the relevant kind should
be deducible from a careful statement of what it takes merely to qualify
as a thing of that kind, and in effect arguing that in each case it is
the persistence of an appropriate kind of process which is essentially
constitutive of persistence for the relevant entities. Once again we see
that the general form of a substance-ontology by no means automatic-
ally dictates the kind of prioritizing of biological kinds at the expense of
mentalistic individuation, which Johnston must assume if his argument
for the Human Beings view is to stand up.
Needless to say, there is no way fully to adjudicate the issues here
touched on without a thorough discussion of the constitution relation,
identity over time, and a host of other controversial topics in ontology.27
Similarly, it would require a lengthy digression properly to defend my
take on Locke. For my purposes it is sufficient to point out that Johnston
is relying on assumptions which are by no means inevitable, which he
does not defend, and which crucially determine the outcome of his
argument.
27
For instance, I am steering clear of the issue of four-dimensionalism which, as it
happens, I endorse. Given unrestricted temporal partition and unrestricted mereological
fusion, there will be no special problem about mentalistically individuated entities. But
Johnston, as noted, in Johnston (1987) employs the language of substance-ontology. It is
therefore dialectically more useful to point out that substance-ontology per se does not
rule out mentalistic individuation, than to invoke the kind of apparatus—temporal parts
and mereological fusions—which substance-theorists characteristically reject.
20 | Denis Robinson

Let’s return to that argument. It is not just outright Psychological

Reductionism which Johnston renounces in renouncing mentalistic cri-
teria for individuation or persistence, and the correspondingly richer set
of kinds which might otherwise be regarded as eligible candidates for the
title ‘‘Human Persons’’. Consider the intermediate or hybrid position
I call the Human Animal Stage view. On this view, even if a person
cannot be outlived by their own mind, their mind can be outlived
by their body and brain. It holds that a person continues to exist so
long as their mental functioning is substantively unimpaired, but that a
person who suffers total permanent and radical amnesia and personality
loss thereby ceases to exist. Subsequent development of new memories
and personality will, on the Human Animal Stage view, constitute
the coming into being of a new person, not the continued or renewed
existence of the previous person.
Animalism differs from the Human Animal Stage view in denying
that a person ceases to exist at such a moment of mental disintegration
and amnesia. A Human Animal Stage theorist will hold that a single
human animal may in such circumstances successively constitute two
different people. But Johnston seems precluded from saying that radical
psychological discontinuity would necessarily count as terminating a
person’s existence, or inaugurating the existence of another person, and
hence (though in fact he does not consider such a view in Johnston
(1987)) seems committed to rejecting the Human Animal Stage view.
A contrary impression may be given by the fact that Johnston tends
to say merely that on his view a person may survive, albeit in a
mutilated form, so long as their brain survives, and at one point in his
introductory remarks he suggests a qualification, saying that ‘‘a human
being could be reduced to the condition of a mere brain so long as
that brain continues the human being’s mental life’’.28 In a similar
vein, he says ‘‘ ‘human being’ names a partly psychological kind’’,
given that ‘‘the tracing of a human being gives primary importance to
mental functioning among the various life functions exhibited by
human beings’’.29
But Johnston repeatedly insists on giving the brain the basic indivi-
duative role when tracing Human Beings, as when he says ‘‘it is crucial

28
Johnston (1987: 64). My italics.
29
Johnston (1987: 79). Johnston’s words at this point are reminiscent of Wiggins’s
version of Animalism.
 Human Beings and Mentalistic Survival | 21

to the tracing of a human being that there be something that is the

continued functioning of that human being’s brain.’’30 In the light of
this remark, it is clear that he cannot accept the Human Animal Stage
view. The case raised above, in which the Human Animal Stage theorist
would see a single human animal as successively constituting distinct
people due to loss of mental continuity and connectedness, is one in
which the brain does continue to function, hence by Johnston’s
lights must be seen as a case in which a single Human Being persists.
Thus, Johnston should be seen as denying the necessity, as well as the
sufficiency, of any sort of trans-temporal mental connections, for the
survival of a Human Being. If mentalistically individuated entities as
such are to be ruled out, we cannot make it any sort of condition for the
survival of a Human Being or of its brain that the brain subserve any
particular kind or pattern of mental functioning. The logical destination
of Johnston’s discussion is the unqualified claim that a person simply
does survive just so long as their brain survives.
If this is the logical destination of Johnston’s discussion there is
something rather odd about his route to it. Let me explain.
Given his declared methodological assumptions, and his evident onto-
logical assumptions, Animalism is and ought to be Johnston’s natural
starting point. The biological kind that at least our third-person, person-
identificatory practices are most readily seen as tracking is the kind:
Human Animals. But Johnston notes that we also habitually
and unproblematically identify ourselves by the use of experiential
memory. The deliverances of experiential memory, he says, are deliver-
ances, true or false, about personal identity as such. As a fact of phe-
nomenology, experiential memory always represents remembered
experiences as experiences had by oneself: experiential memory does
not deliver some ‘‘halfway house’’ judgment short of a belief about
personal identity. Furthermore, its seeming aptness for delivering
knowledge about personal identity would seem to survive even a Carte-
sian suspension of belief in one’s body and in anything external to one’s
own mind.31
But if on the one hand we are already committed to finding a variant
on the Human Animals view, in the sense of taking a non-mentalistically
individuated biological kind of entity, or some composite constructed out
of such entities, as the target of our ordinary modes of reidentification of

30 31
Johnston (1987: 79). See Johnston (1987: 77).
22 | Denis Robinson

persons, whereas on the other hand experiential memory seems suited

by its phenomenology to detecting merely mental connections, we seem
in danger of having to conclude that it is after all merely minds, not
persons, which experiential memory is best seen as unproblematically
identifying and reidentifying. It is specially suited to detecting mental
connections: it is not specially suited to detecting physical or biological
connections.
Hence Johnston’s appeal to MS. He needs this principle to rule out
such an option. Any case of reidentifying a mind must, necessarily and a
fortiori, be a case of reidentifying a person. And finally—to complete
the resolution of this impasse—the putative substance-kind ‘‘mind’’ is
revoked, and the organic kind ‘‘brain’’ is substituted for it.
In short, minds enter the picture because of the relative unsuitability of
experiential memory for the unproblematic identification of an organic
entity such as a human animal. But once talk of a mind has done its job,
the excessive reification is cancelled by reverting again to talk of an
organic entity, but a different one—a human brain. The puzzle is that
experiential memory has already been declared to be relatively unsuited—
phenomenologically, at least—to reidentifying any such kind of thing
unproblematically! I smell a rat.
In fact, it seems to me, Johnston’s appeal to the role and peculiarities
of experiential memory simply has the effect (unintended I am sure) of
an exercise in misdirection. For all Johnston says, I believe, we might
reasonably conclude that the things which can be ordinarily and un-
problematically reidentified over time, in just the way in which we
reliably and unproblematically reidentify ourselves and each other
over time, are human animals. Johnston’s preferred view gives a special
status to the human brain, which provides the indispensable essence of
a Human Being in his sense. I think it is important to see why his
discussion of the role and character of experiential memory does not and
cannot lend support to that conclusion. This is what I next attempt to
show.

What is ‘‘experiential memory’’ best suited, phenomenologically,

to reidentifying unproblematically?
Let’s begin with the ‘‘first-person’’ phenomenology of experiential
memory, and the claim that it represents an experience not simply as
having happened, but as having been experienced by oneself. This
 Human Beings and Mentalistic Survival | 23

appeal to phenomenology seems to me at best a complete red herring.

There simply is no logical room for an alternative phenomenology here:
it follows just from the fact that experience is what experiential memory
is memory of. One cannot recall an experience without recalling it as if
it were one’s own experience, because one cannot have an experience
without having it as if it were one’s own experience. A form of memory
which informed one barely that some experience had occurred, without
recreating that aspect of the phenomenology of experience, and thereby
representing it as an experience had by oneself, would not be what we
call ‘‘experiential memory’’. The phenomenology comes with the job
description. What ought to determine the suitability of experiential
memory as a normally reliable form of reidentification of a mind or of
a self is something like the reliability of the information it delivers,
interpreted as information about a mind or a self: not the form that
information is packaged in. At least, not if the packaging is inevitable.
We can go beyond this criticism: further complexities arise concerning
the phenomenology of experiential memory, obscure in their implica-
tions, relevant to Johnston’s claims, but none of them supportive of his
conclusions. At the very least, they highlight the lack of clarity of the
precise claim which Johnston is making here.
For one thing, experiential memory very frequently provides us with
information about the past states and behaviors of our bodies: the sleazy
catch-phrase ‘‘let’s go build some memories’’ suffices to remind us of
this. Indeed, one might say that experiential memory frequently has a
phenomenology which seems peculiarly well suited to the reidentifica-
tion of a body as such. (Compare Descartes’ Sixth Meditation remarks
on the peculiar suitability of the faculty of imagination for providing
us with information about the interactions of an embodied being with
its environment.) Phenomenologically speaking, experiential memory
often stops no more short of a judgment about the identity of a body
than it does about the identity of a person (in contrast, be it noted, to
the fact that it typically delivers nothing in particular in the way of
judgments about the identity of brains).
For another thing, there is what I shall christen ‘‘the displaced view-
point’’ effect. It may readily be confirmed by cross-examination that for
many people, memory and imagination are alike in that remembering
or imagining themselves performing some action often involves imagery
of the action as seen from a displaced point of view. In remembering
some particular occasion of going swimming, for instance, one may find
24 | Denis Robinson

that what comes into one’s mind is an image of oneself, seen from above,
swimming in the water, rather than an image of the surface of the water
as seen from an inch or two away, that is, from the point of view of the
swimmer.
Now when one’s memory takes such a form, it remains true that
there is a particular first-person standpoint associated with the image:
the remembered scene is viewed from a particular viewpoint. It could
hardly be otherwise. The trouble with the displaced viewpoint effect, for
Johnston’s purposes, is not that in such cases experiential memory does
not have an associated first-person standpoint, but that this standpoint is
not that of the person whose experience is being remembered!
So it all depends on what we mean by ‘‘phenomenology’’. I may on
one occasion remember seeing someone else swimming, actually having
viewed them from above; or I might remember an occasion on which it
was I who was swimming, but experience the displaced viewpoint effect,
forming in my mind a memory image which duplicates that of the
previous case. Typically, I will know perfectly well whether it is my
own swimming, or my viewing of another’s, which I am remembering.
But I will not know this on the basis of the memory images per se. We
have all learnt, after all, the Wittgensteinian lesson that imagistic
accounts of thought are inadequate, since the content of such thoughts
will depend crucially on the interpretation of the images, something
which they themselves, qua images, cannot supply.
Perhaps, then, the proper way to understand ‘‘phenomenology’’ here
is to take it as including, crucially, not just the character of memory
images, but along with that, the interpretation of those memory images.
Even so, given that memory images cannot supply their own interpret-
ation, it’s quite imaginable, whether or not it actually happens, that a
person’s experiential memory should be, in the following way, a little
hazy: on a given occasion one might be confident enough one is remem-
bering an actual event, yet, for all that, be in doubt whether the
swimming figure in the memory image represents oneself, as in the
displaced viewpoint effect, or someone else one actually watched swim-
ming. Viewed in this light it’s hard to see just how the phenomenology
of experiential memory does fit it particularly well for the role of
reidentification of selves or minds.
Perhaps this talk of memory images—or more generally, the idea that
what is crucial here is recall of past experiences—is out of place. Despite
his talk of phenomenology, Johnston himself says that the deliverances
 Human Beings and Mentalistic Survival | 25

of experiential memory are propositions32 involving claims about per-

sonal identity. Perhaps his talk of ‘‘phenomenology’’ here is no more
than dressed up talk of beliefs or judgments, and the claim is simply
that usually when we remember some event, we remember that we
were this participant rather than that? But this interpretation makes
Johnston’s talk of ‘‘experiential memory’’ and ‘‘phenomenology’’ in-
appropriate—I might find myself believing some proposition about my
past actions in the absence of any particular phenomenology. Worse, it’s
hard to see how this claim on its own can carry the significance Johnston
is seeking here.
At any rate, to repeat, it surely is not the phenomenology of experi-
ential memory which makes it specially suitable for the reidentification
of this, that, or the other, thing: it is the reliability with which its
deliverances correlate with the facts concerning this, that, or the other,
thing, which should be relevant here. Furthermore, as noted, Johnston
himself says that what the phenomenology of experiential memory
makes it specially well-suited for establishing is the existence of psy-
chological connections per se—though he promptly transmogrifies this
into the claim that it is only a mind per se which it is phenomenologic-
ally specially well-suited to reidentifying. And even this transmogrified
claim cannot survive his subsequent, naturalistically motivated sub-
stitution of ‘‘brain’’ for ‘‘mind’’. (This is arguable, it seems to me, even
if we set aside any merely generic objection to apparent substitutions
within intentional contexts, which is what claims to know identity,
whether of mind or of brain, surely are.)

The contingency of experiential memory as a means of self or brain

reidentification
In the light of these considerations, and of Johnston’s own insistence
that we should not too hastily promote normally reliable evidential
criteria into metaphysically essential constitutive principles, let’s reflect
further on the contingency of the fact that this form of memory
normally gives us reliable information as to who we are, or indeed of
any of a number of related facts. If someone is having a putative
experiential memory, there are a number of questions which may be
raised about it.

32
Johnston (1987: 77). My italics.
26 | Denis Robinson

a) Did an experience such as that which is apparently being remem-

bered ever actually occur, and cause the apparent memory?
(A putative experiential memory cannot be genuine unless it
sufficiently matches some experience, in the sense of providing
sufficient true information about the character of some real
experience.)
b) Supposing (a) answered in the affirmative, did the experience
which sufficiently matches the memory, and which had a role in
causing it, have the right kind of role in causing it, for the putative
experiential memory to count as genuine?33
c) Did the same person have the original experience, as is now having
the experiential memory?
d) Did the same human animal have the original experience, as is
now having the experiential memory?
e) Was the original experience realized by states of the same brain as
the brain whose states are now realizing the having of the experi-
ential memory?
How these questions are related to one another will be a matter for
dispute between those who hold differing accounts of personal identity.
But I claim that it ought to be agreed by all parties that in any given case
of someone having a putative experiential memory, it is only contingent
on certain in-principle-alterable facts about the world as we currently
know it that any of them may be answered affirmatively. This ought
already to be obvious from the fact that people are capable of psychotic
delusions in which they imagine themselves to be, for instance, Napo-
leon. It follows again from the possibility in principle, exploited in such
works of fiction as the movie Total Recall, starring Arnold Schwarze-
negger, of technology capable of implanting false memories. Experien-
tial memory indeed gives those of us who are not psychotic a highly
reliable source of knowledge of who we are. But this is contingent on
facts which could in principle change.

33
If A has an experience and describes it, and B hears, reads, or encounters that
description and is thereby caused to have an apparent memory matching A’s original
experience, this is not a case of genuine experiential memory, whether or not B is, as in the
case of reading one’s own old diary entries, identical with A. Which other causal connec-
tions would or would not be sufficient for genuine experiential memory must remain to
some extent moot here, but clearly we must count the normal neurophysiological pro-
cesses which underlie the formation of experiential memories in the human brain as
sufficient.
 Human Beings and Mentalistic Survival | 27

Suppose we set aside the issue of psychotic delusions and other appar-
ent memories which are false or fictitious in the sense implied by a
negative answer to question (a). We might imagine, for instance, that
we come to know that the world has been made free, through hand in
hand advances in technology and social norms, from psychotic delusions,
the depradations of stage hypnotists, spurious cases of ‘‘recovered
memory syndrome’’, and the like. Consistently with that assumption,
a variety of further scenarios are possible.
First case. Suppose that it becomes technologically possible to per-
form brain transplants without damage to the normal processes by
which memory-traces are laid down and preserved; and suppose that
in other relevant respects not so far touched on, the world goes on much
as it does at present. Then someone’s having a particular experiential
memory would normally provide reliable evidence that the brain whose
states realized the original experiences was the same brain as the one
whose states were currently realizing those experiential memories. At
the same time, it would (pace Olson’s views about animal persistence)
provide somewhat less reliable evidence for identity of the relevant
human organism if any. This is roughly in accordance with Johnston’s
view, though of course the memories would not have the evidential
status just described, for the particular person having those memories,
unless that person was in possession of the relevant facts—unless
they actually knew about the existence and functioning of brains, the
possibility of transplanting them, and so forth. But we can imagine
other scenarios.
Second case. Imagine instead that the technology of brain transplants
stays in its current undeveloped state, but that the technology of
prosthetic brain implants takes off. These implants, we might imagine,
substitute modules relying on silicon chip technology for groups of
brain cells, relevant information being copied from the original brain
cells, stored temporarily, and then entered into the chip cluster with
which those cells are replaced. Eventually an entire brain can be replaced
in this manner. Thus degenerative brain diseases like Alzheimer’s may
be defeated. Under such circumstances, experiential memory would
provide a more reliable indicator of identity of human organism than
of identity of brain: though once again, only those in the know about the
relevant facts would be able to employ it as such an indicator. (Olson,
once again, would dissent: his view is that under these circumstances the
human organism would no longer exist.)
28 | Denis Robinson

Third case. Imagine, finally, that technology was created capable of

recording information from one person’s nervous system and implant-
ing it in the form of experiential memories in another’s. It would then
be entirely in order to accept the veracity of an experiential memory in
the sense of believing it to be faithful to some actual experience—
always to answer question (a) in the affirmative—whilst at the same
time doubting or wondering whether it was indeed one of one’s own
experiences. If use of such technology was rife, the easy and uncompli-
cated ways in which we use experiential memory in tracing people
would no longer be adapted to tracing things of that kind, and this
would equally be the case whether one held the Human Animals
view, Johnston’s own Human Beings view, or even—it is important to
note—some reasonably sophisticated version of Psychological Reduc-
tionism. An affirmative answer to question (a), even an affirmative
answer based on experiential memory, would provide no reliable basis
for answering questions (c), (d), and (e)—questions about identity,
respectively, of experiencing person, of experiencing animal, and of
experiencing brain.
In fact, of course, none of these possibilities obtains. As things stand,
the actual role of experiential memory in unproblematically reidentify-
ing persons accords well enough with both the Human Animals and
Human Beings views of what a person is, but on each of these views its
reliability in this role must be seen as contingent and a posteriori rather
than as necessary or a priori. There are scenarios whose impossibility is
merely technical—that is to say, scenarios which are not even nomically
impossible, let alone metaphysically impossible—in which the special
suitability of experiential memory for reidentifying either, or both, of
these kinds of entities would be undermined. Furthermore, as things
stand, experiential memory is also equally good for reidentifying
persons conceived as the Psychological Reductionist conceives them!

concluding remarks
The upshot is that, as I see it, there is no way to combine the claims

(i) that experiential memory is specially well-fitted to identifying

minds;
(ii) that as a matter of a priori necessity, minds can’t outlive the
persons whose minds they are; and
 Human Beings and Mentalistic Survival | 29

(iii) that the natural unit whose functioning underlies the phenomena
overly reified as ‘‘minds’’ is the brain, to arrive at the conclusion
(iv) that experiential memory is specially well-fitted to identifying
persons, conceived as entities whose persistence-conditions are
derivative on those of brains, as opposed to persons conceived as
human animals.
Claims (i) and (ii), if true at all,34 are true only of minds qua minds:
substitute ‘‘brains’’ for ‘‘minds’’ and any special suitability of experi-
ential memory for the job of reidentification is undermined, and with it,
any specially compelling reason based on that suitability for abandoning
the Human Animals view in favor of the Human Beings view.
Experiential memory is, as (i) asserts, indeed well-suited as things
actually are for the unproblematic reidentification of minds. Even this,
I believe, is quite contingent. Still, if any of these kinds of entities—
minds, human brains, human animals, Human Beings—may be seen as
distinguished by the special suitability of experiential memory for
unproblematically reidentifying its members, it will be the category of
minds as such. And if as good materialists or naturalists, we really want
to avoid reifying minds, we should here introduce as our surrogate not
the materialistic category of brains (nor cerebra!), but the category of
mentalistically individuated persons, conceived as Psychological Reduc-
tionists conceive them: possibly discontinuous physical entities which
realize various mental states at various times, but such that those
mental states collectively exhibit both synchronic and diachronic unity
of a kind and degree appropriate to personhood.
Now of course such entities cannot be acknowledged if mentalistically
individuated entities are ruled ineligible at the outset. Nor can they be
acknowledged if one assumes a background ontology which is in some
other way too restrictive. In principle such entities could coincide with
mere temporal parts of human animals; numbers of them could be
alternately and discontinuously constituted—or continuously and
multiply constituted—by human animals, as in Jekyll and Hyde and
other multiple personality cases; they could undergo fission and fusion;
they could be discontinuously constituted first by some human animal,
later by some android; and so forth. I’ve never found a convincing

34
In what sense (ii)—which is to say, MS—is true, if any, is an interesting question,
but one which space does not permit me to examine further here.
30 | Denis Robinson

argument purporting to establish that such entities cannot be acknow-

ledged consistently with the main tenets of substance ontology, provided
that mentalistic individuation, which Johnston baulks at, is seen as
acceptable. Nevertheless, substance ontologists (in line with their ten-
dency to extend exclusive special privileges to natural kinds) are typically
reluctant to admit such entities into their ontologies (though four-
dimensionalists would have no problem with them).
It has by no means been my purpose here to defend Animalism, even
though I believe that Johnston’s argument in favor of his Human Beings
modification of it does not succeed. Rather, my aim has been to point
out that Johnston’s sole argument in favor of modifying the Human
Animals view does not stand up, and in particular that Johnston’s pro-
posed methodological alternative to the method of cases provides no good
grounds for preferring his ‘‘Human Beings’’ modification. First, his
methodology does not actually rule out his principal target, Psycho-
logical Reductionism, without having to be bolstered by an unargued
and, I believe, indefensible ban on mentalistic individuation, or some
equivalent form of ontological conservatism. Without good arguments
for such assumptions, which he does not supply, Johnston’s argument
is essentially question-begging. But secondly, even if some suitably
conservative ontological perspective is adopted, Johnston’s methodology
does not actually favor the Human Beings view over the Human Animals
view. The kind ‘‘human animals’’ is certainly a more natural kind than
the kind ‘‘Human Beings’’ which, as he admits elsewhere,35 could include
a creature consisting of the living body of a tiger onto which a living
human brain had been successfully transplanted.
It is noteworthy, finally, that Johnston should purport to derive
from his natural-kinds-and-normally-reliable-evidence methodology a
conclusion which it does not favor, though it is relatively strongly
favored by the alternative methodology of appeal to normal intuitions,
which Johnston rejects. Derek Parfit, for instance, says ‘‘suppose that
my brain is transplanted into someone else’s (brainless) body, and that
the resulting person has my character and apparent memories of my
life. Most of us would agree, after thought, that the resulting person is
me.’’36 I think it is easy to go along with Johnston’s argument because of
these intuitions, which I am perfectly happy to accept. The point is that
Johnston himself, officially, has no right to them.

35 36
Johnston (1989b: 456). Parfit (1971: 4).
 Human Beings and Mentalistic Survival | 31

In a nutshell, Johnston’s alternative methodology asks the following

questions. What sorts of characteristics do we take as providing
good ‘‘fingerprints’’ for persons? And what kind of thing is it which
such characteristics would indeed provide good fingerprints for? Literal
fingerprints work well, as we all know. Experiential memory does too.
But experiential memory provides a reliable fingerprint for a human
brain—hence for any kind of entity of which a human brain is an
essential part—only contingently. And under the contingencies which
in fact obtain, it provides exactly as good a fingerprint for a human
organism. In one respect, indeed, it provides a better one. One needs
much less special knowledge—in particular, one needs much less know-
ledge of human anatomy—to be in a position to know of the general
reliability of experiential memory as a fingerprint for human organ-
isms, or to be in a position to know that one is dealing with an
exceptional or unreliable case—than one needs in order to know of the
general reliability of experiential memory as a fingerprint for human
brains, or for entities whose nominal essence includes them.
The normal intuitions about brain transplants reflect, I believe, not a
view about which kinds of fingerprints provide the most reliable means
of identifying which kinds of entities, nor merely a residue of bygone
philosophies or religions, but our sense that what we really care about in
caring about personal survival is psychological unity, so that we prefer
where possible to identify personal unity with psychological unity.
I believe, but shall not argue here, that this is a preference which neither
the world, nor our conceptual commitments, once shorn of false meta-
physics, prevent us from indulging.

references
Jackson, Frank (1998) From Metaphysics to Ethics: a Defence of Conceptual
Analysis (Oxford: Clarendon Press).
Johnston, Mark (1987) ‘‘Human Beings’’, The Journal of Philosophy 84: 59–83.
——(1989a) ‘‘Fission and the Facts’’, Philosophical Perspectives 3: 369–97.
——(1989b) ‘‘Relativism and the Self’’, in Relativism: Interpretation and
Confrontation, edited with an Introduction by Michael Krausz (Notre
Dame, Indiana: University of Notre Dame Press).
——(1992a) ‘‘Constitution Is Not Identity’’, Mind 101: 89–105.
——(1992b) ‘‘Reasons and Reductionism’’, The Philosophical Review 101:
589–618.
32 | Denis Robinson

Locke, John (1961) An Essay Concerning Human Understanding, ed. J. Yolton

(London: Dent).
Noonan, Harold (1993) ‘‘Constitution Is Identity’’, Mind 102: 133–5.
Oderberg, David (1989) ‘‘Johnston on Human Beings’’, The Journal of Phil-
osophy 86: 137–41.
Olson, Eric T. (1997) The Human Animal: Personal Identity Without Psych-
ology (Oxford: Oxford University Press).
Parfit, Derek (1971) ‘‘Personal Identity’’, The Philosophical Review 80: 3–27.
Robinson, Denis (1982) ‘‘Re-Identifying Matter’’, The Philosophical Review
91: 317–41.
Robinson, Denis (1985) ‘‘Can Amoebae Divide Without Multiplying?’’,
Australasian Journal of Philosophy 63: 299–319.
——(2004) ‘‘Failing to Agree or Failing to Disagree?: Personal Identity Quasi-
Relativism’’, Monist 87: 512–36.
Snowdon, Paul (1990) ‘‘Persons, Animals, and Ourselves’’, in C. Gill (ed.) The
Human Person and the Human Mind (Oxford: Clarendon).
van Inwagen, Peter (1990) Material Beings (Ithaca: Cornell University Press).
Wiggins, David (1980) Sameness and Substance (Oxford: Blackwell).
——(2001) Sameness and Substance Renewed (Cambridge: Cambridge Uni-
versity Press).
2. ‘‘Human Beings’’ Revisited: My Body
is Not an Animal
Mark Johnston

Twenty years ago, in the paper ‘‘Human Beings’’ and elsewhere, I

defended an alternative methodology for arriving at an answer to the
question: What kind captures our essence and so determines our condi-
tions of survival over time?1
Previously, when it came to philosophical theorizing about personal
identity, the popular methodology—‘‘the method of cases’’—had been
to collect ‘‘intuitions’’ about real and imaginary cases of personal sur-
vival and ceasing to be, and then bring those intuitions into some sort of
reflective equilibrium that bore on the question of the necessary and
sufficient conditions for an arbitrary person’s survival. Imagined cases
were treated as more or less on a par with real cases; for the then natural
idea was that we should not restrict our evidence base to the adventi-
tious experiments of step-motherly nature, when we could also avail
ourselves of the ingenious thought experiments in the philosophy
journals.

1
There are a number of reasons why this simple method of cases, with its
implied parity of the real and the imaginary, should be rejected. First,
the specific necessary conditions on our survival, conditions that are the
upshot of our common essence, need not be available to armchair, or
‘‘a priori’’, reflection. Certainly, it would be strained to require that we
must grasp such de facto necessary conditions in order to count as

1
‘‘Human Beings’’, Journal of Philosophy 84 (1987), reprinted in M. Tooley (ed.),
Metaphysics (Garland Press, 1993); J. Kim and E. Sosa (eds.), Metaphysics: An Anthology
(Basil Blackwell, 1999); and Postgraduate Foundation in Philosophy (Open University
Press, 2002).
34 | Mark Johnston

semantically competent with terms or phrases like ‘‘person’’ or ‘‘is the

same person’’. That thought generalizes to put pressure on a once
popular model of philosophical analysis as the teasing out and codifying
our semantic competence. This model, which makes philosophy look
and feel like a sort of advanced lexicography, might be replaced with the
idea of philosophical analysis as aiming for the real definition of the
item or phenomenon in question. This involves using all of the relevant
knowledge and argumentative ingenuity we can muster in order to say
what it is to be the given item or phenomenon. Clearly, much of that
knowledge will be a posteriori, even if not particularly recherché.
A large part of my reason for returning to the topic of personal
identity after so many years is to provide an illustration of the viability
of the method of real definition.
Real definition is not inherently at odds with conceptual analysis.
Concepts are themselves among the legitimate targets of real definition.
For given any complex item, we can enquire after its constituents, and
the relevant manner of combination that yields the complex in question.
So a concept may be a conjunction of other concepts, or a quantification
over other concepts, or built up from other concepts in still more
detailed ways. Articulating any such conceptual structure looks as
though it deserves the name ‘‘conceptual analysis’’, even though we
now know that the results of cognitive science and empirical linguistic
semantics will properly condition most of the interesting claims about
conceptual structure. Even in the case of the analysis of (read: the real
definition of) concepts, a posteriori matters of fact will play a crucial
role.
Still, it will be urged that the application conditions of a concept must
be at least implicitly understood by possessors of that concept, unless
those possessors are disposed to defer either to experts (as with the
concept of a tort in law) or to reality itself (as with the concept of water)
to settle the extension of the concept. Just so; none of that need be
denied in order to put real definition at the center of the philosophical
stage. When the pattern of deference goes all the way to reality itself, as
with the concept of water, then, in determining the extension of the
concept across possible situations, we would do well to note our partial
understanding of the concept, and get on with the inquiry into the real
definition of the phenomenon it picks out. So when investigating the
concept of water and its extension across possible situations, we quickly
find ourselves considering the question of what it is to be the stuff water,
 ‘‘Human Beings’’ Revisited | 35

and whether, for example, the manifest form of water enters into its
nature or essence along with its chemical composition, or is instead no
more than a contingent appearance of water.2 So if, in the same way, we
defer to reality in our use of the concept of personal identity then we
would do best to go straight to the real definition of a (human) person.
Even so, it may plausibly be held that in possessing the concepts of a
person and of personal identity we are not deferring to experts, or even
to a reality that is beyond the ken of ordinary users of these concepts.
We all know well enough what persons are! This may be the best
remaining defense of the old a priori method of aiming to articulate
the application conditions of the concept of personal identity simply by
considering our intuitive judgments in imagined cases.
Against this defense of the method of cases, ‘‘Human Beings’’ argued
that our common concept of a person is ‘‘highly determinable’’, a polite
way of saying that it does not carry very much content. This is shown
first by the enormous range of detailed and conflicting conceptions
(theories, ideologies, theologies) of persons and personal survival that
have been expressed by using the common concept. Secondly, when we
let the concept of a person run free in the full range of imaginary cases,
the resultant picture of what it picks out is that of a ‘‘Bare Locus’’ of
mental life, something that in principle could survive any amount of
psychological and physical discontinuity. It is quite implausible to
suppose that we are tracing any such things when, in the easy and
offhand ways that we do, we trace ourselves and others through time.
So the method of cases, understood as a way of articulating the
application conditions of our concept of personal identity is, at the
very least, in need of reining in by a more robust sense of what is really
possible in the way of human survival. Where could that come from if
not from the real definition of a (human) person, an account of what it is
to be a (human) person? Hence the somewhat non-standard way of
putting the question of personal identity: What kind captures our
essence and so determines our conditions of survival over time?

2
There is a second worry about the appeal to the method of cases in
articulating the necessary and sufficient conditions of the survival of
2
For more on this approach see ‘‘Manifest Kinds’’, Journal of Philosophy 94 (1997).
36 | Mark Johnston

something over time. We do not find much evidence that in tracing

objects and persons (as opposed to things in other categories, such as
performances, for example) through time we are actually deploying
knowledge of sufficient conditions for cross-time identity. Instead, as a
matter of empirical fact, it appears that nature saves us cognitive labor
by having us ‘‘offload’’ the question of sufficiency onto the objects and
people themselves—if I may put it that way. Objects of various kinds
are salient to us, they attract our attention, and we trace them through
space and time. As long as they do not manifest changes in respects we
know to be important for their kind, we are ready to credit them as
having survived, even if we remain properly agnostic about what their
persistence actually consists in. So we should be prepared to discover
that in tracing objects we are deploying knowledge of some necessary
conditions for their survival over time—the thing can’t explode into
smithereens, for example—but not of any non-trivial sufficient condi-
tion. The objects just take care of themselves in this regard, they either
persist or cease to be; to witness such outcomes we need not know any
sufficient condition for their persistence.
If this plausible empirical conjecture as to ‘‘offloading’’ turns out to be
true, it should be very bad news for those philosophers who suppose that
in tracing objects or persons we are in the cognitive situation of having
to stitch together short-lived entities like phases or temporal stages into
significantly persisting wholes by means of ‘‘gen-identity’’ relations.
A gen-identity relation is a necessary and sufficient condition for two
temporal stages (or phases, or temporal parts) to be stages of the same
persisting object. But as to the sufficient conditions, they are precisely
what we do not need to know when we are tracing salient objects which
are naturally individuated anyway.
An example might bring out the point more clearly. Here I confess
immediately to using an imaginary case, but only in the service of
clarification. (I also hope the example shows that rejecting the parity
between real and imagined cases doesn’t have to be a puritanically
joyless exercise; we need not banish the small pleasures of cartoon-
like science fiction from the topic of personal identity. Everything
depends on how the cases are used.)
Suppose we are in a civilization where androids are, on the surface,
and in terms of speech and behavior, indistinguishable from humans.
Unlike ours, however, their heads just contain air-conditioners; that is,
their heads contain ‘‘organs’’ for cooling their bodies. (Aristotle’s view
 ‘‘Human Beings’’ Revisited | 37

of us would actually be true of them.) The androids’ heads, unlike ours,

are not at all where the action is when it comes to their survival over
time. An android can function very well for days without a head.
Everything important is going on in the torso of the android. Now
suppose that you have become familiar with someone, whom you take
to be an android, while he is in fact a human. And suppose you have
been unwise enough not to rely on offloading, that is, on just letting this
salient person capture your attention as he does in fact persist through
time. Instead, you concentrate on short-lived stages of your acquaint-
ance, and knit them together in terms of some criterion of identity;
which, because you take him to be an android, amounts to something
like: one stage of such a being is a stage of the very same being if and
only if there is a pattern of continuity connecting the torsos of the
stages.
Then, the inevitable takes place. Your acquaintance loses his head.
But the torso, or ‘‘head complement’’, remains intact. Even though we
have stipulated that your acquaintance is human and has not survived,
you will count him as surviving, at least until the manifest signs of
death are undeniable. For the gen-identity criterion you were using
knits together ‘‘headless torso’’ stages with the earlier stages of your
acquaintance. For you have been treating continuous preservation of the
torso as sufficient for identity.
In doing this you are led to the wrong judgment; in fact, your
acquaintance does not survive the destruction of his head. He is a
human, not an android.
You would have been much better simply waiting and seeing just
what kind of thing he was. You had no business deploying a suffi-
cient condition for his identity over time. In fact, your using that
sufficient condition did not help you trace him forward in time. All
the previous evidence for his survival, the evidence of gross bodily
integrity and consistency of dispositions to speech and behavior,
would have been just as good evidence that he survived. You did not
need to settle what exactly he was in order to trace him through time.
Your fault was unnecessarily to settle in advance, a priori as it were,
what kind of thing he was. You could have ‘‘offloaded’’ that cognitive
task onto him; his nature, which you do not fully know, settles what
kind of thing he is.
My hypothesis is that our cognitive system does ‘‘offload’’ in precisely
this way. For the very simple purposes of tracing and re-cognition, we
38 | Mark Johnston

need not know, and do not know, sufficient conditions for identity over
time. The developmental facts, concerning among other things the early
age at which we in fact trace objects and persons, seem to support this
hypothesis.
Notice that something about the case of the ‘‘android’’ who was really
a human suggests that we can be aware of persisting items and trace
them through time even if we are ignorant of the sufficient conditions
for their identity. In the case of the ‘‘android’’ who was really a human,
you made a mistake about your acquaintance. That implies that you
still had your acquaintance as a topic of your thought and talk, and
that you were able to trace him through time, even though you were
ignorant of the sufficient conditions of his persistence through time.
This suggests that tracing objects through time does not involve draw-
ing on sufficient conditions for their persistence.
Trying to make our implicit knowledge of sufficient conditions for
identity explicit by examining our reactions to merely possible cases will
be a particularly pointless task if the hypothesis of offloading is correct.

3
Both ‘‘Human Beings’’ and ‘‘Reasons and Reductionism’’ played up a
third worry about the implied parity of real and imaginary cases.3 In the
massive core of real cases, many sources of evidence for personal
survival, such as persistent bodily integrity and mental continuity,
converge and agree; whereas the whole philosophical charm and utility
of the imagined cases in the literature on personal identity lies precisely
in teasing these elements apart. The obvious question arises: Might
we not have thereby undermined our ability to make good judgments
about personal identity when considering these very cases?
That question acquires a sharper bite if we make the distinction not
exactly between real and imaginary cases, but between the massive core
of ordinary cases in which bodily, behavioral, and mental evidence
converges, and the fringe cases, real or imaginary, where those sorts
of evidence come apart. Bear in mind the moral horrors that await
many of us around the end of our lives, thanks to the perverse incentives
of modern medicine; mental death, coma, and absurdly protracted

3
‘‘Reasons and Reductionism’’, The Philosophical Review 101 (1992).
 ‘‘Human Beings’’ Revisited | 39

languescent shade-life. Who is to say that such all-too real conditions

present very much simpler objects of philosophical diagnosis than Derek
Parfit’s most outré imagined cases? The same goes for pre-fetal matters.
Neither the beginning nor the end of life present the best starting points
for a theory of personal identity. It’s in the massive core of ordinary
cases of viable healthy human life where we can best trust our judg-
ments of personal identity.
That the relevant distinction is between ordinary and fringe cases and
not between real and imagined cases should have been obvious upon
reflection. For imaginary cases can sometimes presage real cases; our
imagination is sometimes the source of future reality. But where the
imagined case is also a fringe case, it will remain so even if it becomes a
real case.
In order to press home the worry about the alleged parity between
ordinary and fringe cases, ‘‘Human Beings’’ noted three potentially
distorting influences that could condition our judgments in fringe
cases. The distorting influences were the psychological and social con-
tinuer effects, and the effect of a quasi-Cartesian illusion that we are
only essentially just that of which we can be certain of being.
It is possible to imagine cases in which a person ceases to be, but
nonetheless has an excellent continuer of his psychological life. In my
view, teletransportation, in which a person’s body is destroyed and
immediately copied at a remote location, is just such a case. The person
who appears at the receiving end remembers the acts and the experi-
ences of the previous person, he will carry out that person’s intentions,
and he has beliefs and desires that almost precisely correspond to those
of the person he replaced. Yet, those of us who think of teletransporta-
tion as a method of copying people which has the singular defect of
destroying the original copy, will have to admit that teletransportation
would serve up, at least on the mental side, the best sort of evidence we
usually have for personal identity. There is therefore, in such imagined
cases, a natural tendency simply to accede to such evidence, and trace a
person through the events which preserve such psychological continu-
ity. This I dubbed ‘‘the psychological continuer effect’’. It could very
well just be an understandable overgeneralization from the ordinary
run of cases, where such psychological evidence is highly correlated with
the preservation of the bodily integrity of the person in question.
A further possible distorting factor in the fringe cases—‘‘the social
continuer effect’’—arises from our practical interests in the relation of
Another random document with
no related content on Scribd:
small donations ($1 to $5,000) are particularly important to
maintaining tax exempt status with the IRS.

The Foundation is committed to complying with the laws regulating

charities and charitable donations in all 50 states of the United
States. Compliance requirements are not uniform and it takes a
considerable effort, much paperwork and many fees to meet and
keep up with these requirements. We do not solicit donations in
locations where we have not received written confirmation of
compliance. To SEND DONATIONS or determine the status of
compliance for any particular state visit www.gutenberg.org/donate.

While we cannot and do not solicit contributions from states where

we have not met the solicitation requirements, we know of no
prohibition against accepting unsolicited donations from donors in
such states who approach us with offers to donate.

International donations are gratefully accepted, but we cannot make

any statements concerning tax treatment of donations received from
outside the United States. U.S. laws alone swamp our small staff.

Please check the Project Gutenberg web pages for current donation
methods and addresses. Donations are accepted in a number of
other ways including checks, online payments and credit card
donations. To donate, please visit: www.gutenberg.org/donate.

Section 5. General Information About

Project Gutenberg™ electronic works
Professor Michael S. Hart was the originator of the Project
Gutenberg™ concept of a library of electronic works that could be
freely shared with anyone. For forty years, he produced and
distributed Project Gutenberg™ eBooks with only a loose network of
volunteer support.
Project Gutenberg™ eBooks are often created from several printed
editions, all of which are confirmed as not protected by copyright in
the U.S. unless a copyright notice is included. Thus, we do not
necessarily keep eBooks in compliance with any particular paper
edition.

Most people start at our website which has the main PG search
facility: www.gutenberg.org.

This website includes information about Project Gutenberg™,

including how to make donations to the Project Gutenberg Literary
Archive Foundation, how to help produce our new eBooks, and how
to subscribe to our email newsletter to hear about new eBooks.