2Q0HPRU\$ULVWRWOH

V&RUUHFWLRQVRI3ODWR
+HOHQ6/DQJ

Journal of the History of Philosophy, Volume 18, Number 4, October

1980, pp. 379-393 (Article)

3XEOLVKHGE\-RKQV+RSNLQV8QLYHUVLW\3UHVV
DOI: 10.1353/hph.2008.0252

For additional information about this article

http://muse.jhu.edu/journals/hph/summary/v018/18.4lang.html

Access provided by University of California , Santa Barbara (30 Jul 2015 19:46 GMT)
 On Memory: Aristotle's
Corrections of Plato
H E L E N S. L A N G

THE De Memoria et Reminiscentia is remarkable for a number of reasons, including its

strikingly Platonic language. On the basis of this language, the arguments of the De
Memoria are sometimes judged "Platonic" and accordingly interpreted as a striking
moment of"Aristotle's Platonizing. ''~ Against such heavy reliance on the presence of Plato
in this treatise, one commentator dismisses the issue of Platonism in Aristotle as "a dull
answer" to the philosophic problems in this treatise, and he treats these problems as if
Aristotle's Platonic language were wholly irrelevant. 2
In the De Memoria, Aristotle first distinguishes people possessing a good memory from
those excelling at recollection. The former are "slow ones"; the latter are "quick and good
learners. ''3 Finally, memory is tied to sense-perception and the essential use of images while
recollection rests with intelligence and requires images only accidentally.
Plato, however, closely associates memory, recollection, and intelligence. 4 A good
memory is possessed by all who are clever and quick to learn: Plato castigates the slow
learner as forgetful and possessing a poor memory. 6
Prima facie, then, Aristotle's separation of memory (and being "slow") from recollection
(and being "quick") seems anti-Platonic. But in his constructive account of memory Aris-
totle quotes Plato, apparently utilizing a Platonic model of the relations between copies and
originals. 7 The mysteries of this text how it appears to reach an anti-Platonic conclusion
through the use of a Platonic m o d e l - - require an account both of its content and of its
Platonic terminology. I shall argue that Aristotle first corrects Plato and only then quotes
him. On this account, Plato's idiom is reinterpreted according to Aristotle's views, and so
we find Platonic language impressed into the service of an anti-Platonic conclusion.

Parva Naturaha: Revised Text with Introduction, andNotes, ed. W. D. Ross (Oxford: Clarendon Press, 1955),
p. 16. These arguments are addressed by I Block,,"The Order of Aristotle's Psychological Writings," American
Journal of Philology 82 (1961):50ff. Ross furthe~ argues his point in his edition of the De Anima (Oxford"
Clarendon Press, 196,1)., pp. 7-12. This issue has provoked a continuous, and progressively more unwieldly,
literature. For a most l~lpFal survey and classification of more recent studies, see W. F. R. Hardie, "Concepts of
Consciousness in Aristotle," Mind 85 (1976): 388-411.
2 On Memory'. De Memoria et Reminescentia, trans., with interpretive summaries, by R. Sorabjl (London:
Duckworth, 1972), p. 5. SorabJi himself considers Aristotle almost exclusively m terms of Humean and post-
Humean distinctions.
3 De Memoria 1, 449b6-9. All translations of the De Memorza are those of J. I. Beare from the Oxford
translation reprinted in The Basic Works of Aristotle, ed. R. McKeon (New York: Random House, I941), pp.
607-17.
4 Rep. 6, 486c ft.; Phil. 34 ft.; Phaedo 73b--c, e; Hipp Min. 368-369b.
Phil. 60d, 64a5; Cratylus 437b3; Rep. 6 , 4 9 0 c l 1,494b2; Tim, 74e8-75.
6 Rep. 6, 486c~:1, Theaet. 194d.
7 Here we find a quote from Theaet 191d as well as references to hkenesses, images, and contemplation.

[3791
380 HISTORY OF PHILOSOPHY

This account requires several steps. (l) We must first establish Plato's view of memory
as constitutive of sensation and tied to intelligence. Following Aristotle, I will consider
primarily the Theatetus and Philebus. (2) I will consider Aristotle's arguments--and ex-
plicit criticisms of Plato--that memory relies on sensation and relates to intelligence only
accidentally. Plato, in the Theaetetus, first suggests and then rejects an image of memory as
like a wax tablet; this rejected image appears as the model par excellence of memory in the
De Memoria. I shall argue that Plato's rejected model can be asserted by Aristotle only
because Aristotle first radically reinterprets Plato's doctrine of sensation and memory. (3) I
shall conclude that the Platonic language of the De Memoria is indeed important and cannot
be ignored: it shows the availability of Platonic metaphysics for a thoroughly anti-Platonic
reinterpretation.

1. Let us turn first to Plato's analysis of memory. In the Theaetetus the question, What
is knowledge? turns into its opposite--What are mistakes? The account of mistakes here
rests upon an image of memory as like a wax tablet; the image is rejected, and consequently
the account concerning memory and mistakes is largely negative. 8
In the Philebus, memory is explained as the "preservation of sensation. ''9 A positive
account, not only of memory, but also of sensation and mental images results. Memory
constitutes sensation and so provides the raw matter of mental images. In the De Memoria,
Aristotle first criticizes the conclusions concerning memory reached in the Philebus and
then quotes the Theaetetus. Thus these arguments constitute rich sources for Plato's view of
memory, while at the same time serving as an origin for Aristotle's position on the same
problem.
Ostensibly, the topic of the Theaetetus is knowledge. Socrates asks Theaetetus to define
knowledge. ~0Theaetetus first defines it as "nothing but perception." ~j Socrates refutes this
Protagorean definition of knowledge by shifting the argument from knowledge to its oppo-
site, perception. All perceptible things always change, and hence perception of them
always changes. ~2Because of this flux, there can be no difference between perception and
nonperception; if we identify knowledge with perception, then there can be no difference
between knowledge and nonknowledge. ~3Hence knowledge cannot be perception, and we
must conclude that "mind contemplates some things throught its own instrumentality,
others through bodily faculties. ''~4 Knowledge resides in the former, while perception
through the bodily faculties "has no part in apprehending t r u t h . . , or existence. "15
The conclusion of this argument looks negative: it unreservedly rejects the view that
knowledge is perception. But this negative conclusion involes a positive, constructive
conclusion. It establishes a dichotomy between body and soul and asserts not only the
negative role of body in regard to knowledge, but also the positive role of soul. Knowledge,
whatever it be, cannot be perception or have any association with body: knowledge belongs
wholly to soul insofar as it operates independently of body.

8 Theaet. 191d-197a.
9 Phil 34al0.
io Theaet. 146c, 187b2-3.
u Theaet. 151e2-3.
12 Theaet. 160a-d5, 181c ff.
13 Theaet. 182e-183.
~4 Theaet. 185e6. All translations of the Theaetetus are those of F. M Cornford, Plato's Theory o f K n o w l e d g e
(New York: Harcourt, Brace, and C o , 1935).
15 Theaet. 186e4-5
 ARISTOTLE ON MEMORY 381

Theaetetus redefines knowledge as "true judgment. ''~6 Insofar as judgment belongs to

soul, this new definition responds to the positive conclusion just reached. But the discussion
again turns into its opposite: the question of "true judgment" becomes a question of false
judgment or mistakes. In order to explain mistakes, Socrates likens memory to a wax tablet.
He then criticizes this account of mistakes, the wax tablet image of memory, and the
definition of knowledge as true judgment.
If knowledge is true judgment, then we must ask what judgment is. Judgment is a special
case of thinking; thinking is mind "simply talking to itself. "~7 Thinking is a process taking
place in the mind independently of body. Judgment occurs when "two voices" talking inside
the mind agree. But, on this view, all judgments will be true, for every man, whether
intelligent or. stupid, sane or mad, experiences agreement and affirmation within his own
mind. How, then, can false judgments occur?

Imagine, then for the sake of argument, that our minds contain a block of wax . . . . Let us call it the
gift of the Muses' mother, Memory, and say that whenever we wish to remember something we see or
conceive in our own minds, we hold this wax under the perceptions or ideas and imprint them on it as
we might stamp the impression of a seal ring. Whatever is rubbed out or has not succeeded in leaving
an impression we have forgotten and so do not know.~8

When a man possesses an impression--what we commonly call a memory--either of a

perception or of an idea, and he perceives something, he may make a "false judgment. ''19
False judgments thus occur whenever a perception is attributed to the wrong mental impres-
sion. "It is precisely in the field of objects both known and perceived that judgment turns
and twists about and proves false or true--true when it brings impressions straight to their
proper imprints, false when it misdirects them crosswise to the wrong imprint. "2~ False
judgments arise, not strictly within the mind, but in a comparison conducted by mind
between its contents and perceptions: mistakes occur in the "fitting together of perception
and thought. ''2~
The origin of such "misfits" lies in the genesis of mental impressions. The "wax" may be
too hard or too soft. People with soft wax learn quickly, but their impressions are indistinct
and soon blur together; people with hard wax also possess indistinct impressions because
the impressions cannot obtain proper depth. Because these impressions are faulty, a fortiori
mistakes occur in fitting perceptions together with them. Minds with wax of the proper
consistency "take" deep lasting impressions and so learn quickly, possess good memories,
and think truly. 22
Socrates now criticizes and rejects this model of memory: it fails to explain mistakes
which do not involve perception. 23This failure originates in trying to define mistakes before
establishing the nature of knowledge. A serious "misfit" results: mind and body, or knowl-
edge and perception, are not properly distinguished and afortiori cannot be properly

~6 T h e a e t 187b5.
17 T h e a e t . 190a3
15 T h e a e t 191c8-e.
19 T h e a e t . 191e3~.
20T h e a e t . 194a8-b6.
21 T h e a e t . 191c5-195b5. 195d-2.
2~See R e p . 6, 486d~87a5.
2~ T h e a e t . 196a~1.
382 HISTORY OF PHILOSOPHY

related. The "'wax tablet" takes impressions indiscriminately from both perception and
ideas. Hence, the account fails to distinguish between them, even though the previous
anti-Protagorean argument showed that perception could make no contribution to knowl-
edge.
The image of the wax is too "physical" and hence imperfectly emancipates soul from
body. Because this second account attributes knowledge to soul, it is superior to the first
account of knowledge as sensation. But even though superior, this second account fails to
explain the nature, origins, or certainty of the contents of mind as independent of body.
Therefore, it fails to yield an intelligible account either or knowledge or of mistakes. The
original positive conclusion of the preceding argument knowledge belongs to soul in-
dependently of body--remains and indicates that we need an account of an entirely different
order. Such an account becomes available not in another examination of knowledge, but in
an examination of pleasure.
Ostensibly, the Philebus explores the nature of pleasure. 24 The problem is whether
pleasure properly belongs to body or to soul. The purpose of the dialogue is to show that
pleasure is not a physical but a spiritual activity akin to knowledge, conducted by soul, and
properly understood as "true" or "false. ''25 But pleasure cannot be understood without
grasping the nature of sensation, memory, and desire.
Socrates distinguishes (1) experiences which affect only the body from (2) experiences
which penetrate both body and soul. The first involves no activity of soul and cannot even
be called sensation: it is "non-sensation. ,,26 The second involves both body and soul and is
properly called "sensation. ''2v Sensation "occurs [only] when soul and body come together
in a single affection. ''2s Although body must be present for sensation, only an activity of
soul yields sensation. The activity of soul requisite for sensation is memory, and Plato calls
memory "the preservation of sensation.'29
Before explaining memory as "the preservation of sensation," Plato distinguishes be-
tween memory and recollection. Memory and recollection differ in that recollection is soul
in and by itself, apart from body, recapturing what soul has experienced in common with
body. 30Recollection and memory are thus essentially the same activity of soul; but recollec-
tion is this activity conducted wholly independently of body, while memory is this activity
conducted when soul comes together with body. When soul comes together with body, this
activity is called "sensation," and sensation requires memory as its preserver.
Desire serves as a prime example of sensation and illustrates the role of memory as the
preservation of sensation. When a man becomes thirsty, he desires drink. 3t The body
experiences only "emptying," that is, becoming thirsty; thirst is the opposite of what the
man desires, that is, "'replenishment" or drink. Since the body "experiences" only thrist,
and thirst is the opposite of drink, the body can neither desire drink nor yield any informa-
tion about the replenishment by drink. 32 "Hence, the only alternative is that the soul
apprehends the replenishment, and does so obviously through memory. ''33 Thus neither

24 R. Hackforth, Plato's Exammation o f Pleasure (Cambridge: Cambridge Umversity Press, 1945), p. 5. All
translations of the Phtlebus are those of Hackforth
25 See esp Phil. 38a~40; Hackforth, pp. 72-73
26 Phzl. 34a.
27 Phil 34a5; cf. Phaedo 65a5; Theaet. 160b-c.
28 Phil. 34a2-3.
29 Phil. 34a10
30 Phil. 34b6-8; cf. Meno 98a4; Phaedo 7365ff
31 Phtl. 34e9-35b4.
32 Phil. 34b6-9.
33 Phil. 35b4-c.
 ARISTOTLE ON MEMORY 383

desire nor memory belongs to body; indeed, in order for a man to desire drink, there must be
a memory "of something opposite" to physical occasions such as thirst? 4 "Our discussion
then, inasmuch as it has proved that memory is what leads us on to the objects of our desire,
has made it plain that it is to the soul that all impulse and desire, and indeed the determining
principle of the whole nature, belong. "'35
Memory, independently of body, apprehends what is desirable and so fulfills desire by
leading body to its proper object. 36 The intelligibility of bodily occasions--becoming
thirsty means needing a drink--is provided solely by memory. But "phisical experience"
cannot even be experience except as rendered intelligible by soul through memory. One
cannot experience thirst at all except as the need for drink. Thus, by rendering sensation
intelligible, memory "preserves" it: memory makes sensation possible by knowing some-
thing opposite to what body suffers.
Here we possess a constructive account of sensation and memory. In accordance with the
positive conclusion of the Theaetetus, this account separates body and soul. The account of
the Philebus is successful, where that of the Theaetetus fails, because the emancipation of
soul from body is complete.
According to the Phitebus, sensation is nothing other than soul, through memory, im-
mediately interpreting the affections of body and so rendering them partially intelligible. 37
Although the body may provide an occasion for sensation, it makes absolutely no mean-
ingful contribution to sensation.38 Hence, sensation is a continuous constructive remember-
ing, preservation, of physical occurrences by memory. Because memory interprets and
thereby makes possible all bodily experience, memory is prior to body, and so memory
preserves past, present, and future sensation? 9
Hence, a good memory belongs solely to intelligence and in no sense belongs to body.
According to Plato, soul may be connected to body, but soul is not dependent upon body.
Soul and body, in this sense, are essentially disrelated. Because of this disrelation, Plato
explains sensation as a spiritual activity which may take place in a body but which is not
dependent upon this location. Consequently, his task becomes first to explain sensation and
mental images as operations of soul conducted independently of body and finally to free
soul from all dependence upon body. 4~The examination of pleasure here in the Philebus
takes up the task of freeing soul from body by showing that pleasure is a spiritual, rather
than a physical, activity. 4~

34 Phil. 35c9-10; cf. Theaet. 163b ft.

35 Phil. 35d-3.
36 See Pol. 267d8; Rep. 1,345d-5; Laws 10, 892-b3
37 See K. Lycos, "Aristotle and Plato on Appearing," Mmd n s 73 (1964) :498-99
38 See Phaedo 96b ft,
3,~ Plato mststs on the priority of soul to body m a number of contexts. See Phtl 39d5, Phaedrus 245c6-246e6~
Laws 10, 892a-b.
4o Here. although Plato does not exphcttly make the point m the Phtlebua, we can see the ~mportance (and the
lack of it) in the &stinctlon between memory and recollection Memory is the actv:lty of soul whtch yields
experience when body provides an occasLon; recollection is the same spiritual actLvLtyconducted wholly indepen-
dently of body. See Meno 81 d5, e4, 98a4, Phaedo 72e5. Since body makes no epistemlc contribution to sensation,
there ~s httle difference between memory and recollection as spiritual actlv~ty. For example, one can actually see a
dog, or one can recall the shape of a dog. Soul is fully responsible for both experiences. The major difference
between the experiences is that memory is less free from body than ~s recollection Because body provides the
occasion on which memory y~elds sensation, memory is "enslaved" to preserve th~s or that, as body occasions.
Recollection operates independently from body (from what soul experiences when together with body), and so
recollection can recall either this occasion or that occasion and ~s thus freer than memory Because it ~s freer from
bondage to body, recollection is more desirable than memory and of a higher epistem~c order.
4~ Hackforth, p, 72
384 HISTORY OF PHILOSOPHY

Later in the P h i l e b u s , Plato returns to m e m o r y - - t h i s time in relation to mental images.

M e m o r y is like an internal scribe writing words in our soul as it preserves bodily
affections. 42 M e m o r y preserves sensation by producing a succession of related presenta-
tions. These successive presentations cohere only because m e m o r y , like writing, preserves
them one after another. Such coherence obtains only insofar as successive parts or presenta-
tions fit together into a unity, a3
Socrates n o w asserts that in the soul there must also be a second artist, who paints true or
false pictures from the presentations written by memory. .4 These "paintings" are prior to
bodily occasions and so apply to all experience, past, present, and future. 45
Unified mental images, then, are painted, that is, actively composed, by the second artist
in the soul. Images compare to writing as a whole without parts compares to a whole with
parts. A picture constitutes a single vision which can be grasped as a whole, while writing
must be read through one word after another. Thus, according to Plato, mental images are
not immediately located in sensation (itself a spiritual activity); mental images constitute a
unified intuition of the presentation of m e m o r y during sensation. Thus, sensation is the
spiritual activity of m e m o r y , conducted at one remove from body; mental images arise from
spiritual activity unifying the presentations of m e m o r y at two removes from body. '.6
Thus, in the P h i l e b u s we possess a constructive account of sensation, m e m o r y , and
mental images. The account fully emancipates soul from body and so completes the positive
conclusion of the T h e a e t e t u s . Sensation is nothing other than the activity of soul w h e n it
falls together with body. This activity is called m e m o r y , or remembering.
Plato calls m e m o r y the "'preservation of sensation.'47 His m e a n i n g is now clear: m e m o r y
unifies the flux of physical affections into successive presentations. A second artist in the
soul unifies the "preservations" of m e m o r y into mental images. Hence, according to Plato
both sensation and " i m a g i n a t i o n " are activities o f mind conducted independently of body.
M e m o r y "preserves" sensation by operating prior to all physical occasions, and, by impli-
cation, images arise in the m i n d through an "artist" who operates prior to and i n d e p e n d e n t l y
of m e m o r y . This analysis completes the first step necessary for Plato's philosophical task: it
separates soul from body and shows ( l ) how soul, through m e m o r y , is fully responsible for
sensation, and (2) how mental images are a unified intuition o f " m e m o r i e s " at two removes
from body.

42 Phil. 39a5-7.
43 See Theaet 203a ft.; Phaedo 96a5-97c
44 Phil. 39b3-4; cf. Rep. 6, 510a-2, 5t0e-3; 7. 514 ft.
45 Phil. 39c--e; Hackforth refers this hne back to Phd. 32c, (p 75. n. 2). See above n. 24.
46 Important conclusions follow for Plato's posmon although they are beyond the purpose of the Philebus and
are not made exphcit here. While body provides the occasion for memory to "write down" opinions and assertions,
memory is able to do so only because it knows something opposite to bodily experience. Exactly the same relation
must follow in the formation of mental images by the "painter" m the soul. Memory supplies presentations which
occasion a pamtmg, t.e., a unified mental image, but the painter can compose such a umfied image only because he
has present to himself a model which is of another order from the presentations of memory. Thus, the objects which
serve as the origins of our mental images can never themselves be physical, but must be opposed to that which is
physical, in knowing, soul grasps the intelhgible, and only the lntelhgible can supply the content of mental images.
The intelligible serves as a model or norm toward which the soul must "look'" whenever it wJshes to know.
Insofar as the writing of memory can be umfied into an image, it can be so unified only m virtue of a norm which is
independent of body. Mental images are in no sense derived from body or sensation, but are nothing other than a
relation to norms which are prior to all "'sensation" (itself constituted by memory). The second artist paints
pictures, that is, renders intelligible, the partially mtelhgtble presentations of memory. In so doing, this artist
completes memory and allows it to be, just as memory allows sensation to be by rendering the affections of body
partially intelligible. Plato fails to identify the "painter" m the soul, but we mtght speculate that in some sense this
"painter" operates like recollection.
47 Phd. 34a10.
 A R I S T O T L E ON M E M O R Y 385

2. Let us now turn to Aristotle's view of memory and his critique of Plato. W e will
consider the first chapter o f the D e M e m o r i a . The text as a whole is quite complicated and
should be divided into two major arguments. Aristotle first argues that memory is limited to
past experience only and rests in sensation rather than in intelligence. He then argues that
memory relates a present image to an absent object of past experience.
Only objects o f p a s t sensation or contemplation are proper objects of memory. Memory
is not a spiritual activity constituting sensation. Rather, memory is posterior to sensa-
t i o n - i t s e l f constituted by sensible o b j e c t s - - a n d memory requires awareness of a time
lapse since the original experience took place. Awareness of the passage of time lies
essentially in the faculty of sense-perception as distinct from the faculty of intelligence;
therefore, memory too lies in sense-perception rather than in intelligence.
The genesis of memories and mistakes due to memory requires that objects be impressed
onto mind as a seal is impressed onto a wax tablet. After the original experience leaves an
impression, memory relates the retained mental image to an absent object, and this relation
is like that of a portrait to its model, Aristotle concludes that memory is a state of imagina-
tion relating as a copy to its original and that memory rests in sense-perception rather than in
intelligence. With his Platonic language, Aristotle reaches a thoroughly anti-Platonic con-
clusion.
Before explaining the operation of memory, Aristotle limits memory to objects of past
sensation or contemplation and locates it in the faculty of sense-perception. He notes that
mistakes are often committed on this point. 48 For example, Plato associates memory and
intelligence as well as relating memory to past, present, and future sensation. Aristotle
argues that memory requires awareness of a time lapse since the original experience. This
awareness requires an essential use of imagination and lies in the faculty of sense-
perception. Therefore, memory too essentially lies in the faculty of sense-perception and
relates only to objects of past experience.
Aristotle first considers the role of time in experience. It is easy to see that memory relates
only to the past. 49 The future is the object only of opinion or expectation; the present is the
object of sense-perception or contemplation. 5~

No one would say that he remembers the present, when it is present, e.g. a given white object at the
moment when he sees it; nor would one say that he remembers an object of scientific contemplation at
the moment when he is actually contemplating it, and has it full before his mind---of the former he
would say only that he perceives it, of the latter only that he knows it. 51

The senses and the mind relate to their respective objects in the same way. 52 Whenever an
object, sensible or noetic, actualizes sense or m i n d - - t h e m s e l v e s yearning for such actuali-
zation the activities of sensing or thinking necessarily ensue. 53 Through its a c t i v i t y - -

48De Memoria 1,449b10.

49 De Memoria 1,449b15.
5o De Memoria 1,449b10-13.
51 De Memoria 1,449b15-18.
52 See De Anima 3, 4,429a17: "Mind must be related to what is thinkable as sense is to what is sensible." All
translationsof the De Anima are those ofJ. A. Smith from the Oxford translationreprinted In McKeon, The Basic
Works of Aristotle. For a clear and concise treatment of this difficult passage in the De Anima see J. Owens, "A
note on Aristotle, De Anima 3.4,429b9," Phoenix 30 (1976): 107-18.
53 Perceptionand thinking occur whenever sensibleor noetlc capacities take on the form of an object withoutthe
individuated matter of that object. See De Anlma 1, 5. 410a25; 2, 4, 415a20, 415b24; 5, 417a3-5, 3, 4,
429a13-15; L. A. Kosman, "Perceiving That We Perceive On the Soul III, 2," Philosophical Review 85
(1975) : 506; T. J. Slakey, "Aristotle on Perception," Philosophical Review 70 ( 1961). 471-75.
386 HISTORY OF PHILOSOPHY

informing the senses or the mind---objects .constitute sensation or thinking and character-
ize experience as past, present, or future. Thus, in the present an object actually informs
the senses or the mind, and perceiving or knowing ensue. When actualization of a faculty
by an object formerly took place, but the object no longer actualizes its faculty, the
experience of the object is past and we "remember" it. 54
According to Plato, soul independer~tly of body.constitutes both sensation and contem-
plation; therefore, Plato separates soul from body and concludes that through memory
soul constitutes all sensation, past. present, and future. According to Aristotle, objects,
sensible or noetic, constitute all experience by actualizing the senses or the mind. Aris-
totle's account begins not in soul as prior to body but in objects and the way in which they
temporally determine both sensation and contemplation. 55 Therefore, Aristotle limits
memory to objects of past experience alone. Sensation is a spiritual activity dominated by
memory according to Plato; for Aristotle sensation is a sensible activity dominated by
sensible objects. Aristotle "corrects" mistakes concerning the objects of memory only by
presupposing his own account of sensation and locating the origins of sensation in sen-
sible objects.
"Memory is, therefore, neither perception nor conception, but a state or affection of
one of these conditioned by lapse of time.'56 Because memory relates only to objects of
past experience, it requires both a past actualization of the senses or mind by an object and
an awareness that the actualization is past. Aristotle returns later to the notion of memory
as a state or affection of perception or conception. 57 Here he deals with the relation of
memory to time.
Memory immediately involves awareness that a perception or thought took place for-
merly, for example, " 'I formerly heard . . . this,' or 'I formerly had this thought,"58
Since memory requires awareness of a time lapse, it is connected to the perception of
time. "Consequently only those animals which perceive time remember, and the organ
whereby they perceive time is also that whereby they remember. ''5~
The problem is to identify the "organ" through which animals perceive time. Aristotle
identifies it as the faculty of sense-perception rather than the faculty of intelligence. The
faculty of sense-perception and the faculty of intelligence differ in their use of images.
The former uses images essentially, while the latter uses then only accidentally. There-
fore, Aristotle introduces "the subject of imagination" in order to show that perception of
time, and consequently memory, uses images essentially and so must be located in the
faculty of sense-perception rather than intelligence. 6~
Aristotle recapitulates an earlier discussion of imagination. 6~ All thinking requires an
image produced by imagination; the content of images can be supplied only by
sensation. 62Therefore, all thinking in some sense originates in sensation. Hence, thinking

54 D e Memoria 1,449618-23.
55 Cf. Metaph. 1, 1.
56 De Memorta 1,449b24-25.
5~ De Memorta 1,450a26.
58 De Memoria 1,449b23.
59 De Memorta 1,449b28-30.
60 De M e m o n a 1,449b30; the Oxford translation reads "presentation" for phantasia throughout because of the
"psychological" connotations of "imagination," (see n. 2 on line 449b30.) I use "imagination" throughout in the
restricted sense of"that faculty m virtue of which an image can arise before us," De A nima 3, 3,427bl 5,428a. For
further d~scusslon of this difficulty m translation see Lycos, p. 496.
61 See De Aroma 3, 3,427b ff.
62 Throughout this discussion I assume Aristotle's view of sensation as a sensible actiwty, i.e , the actuahzation
of the senses by a sensible object. See Sorabj~, p. 6
 ARISTOTLE ON MEMORY 387

about sensible objects and abstract thinking do not differ by the presence or absence of body
and images. Both require images. They differ in their use of images.
Images represent all objects quantitatively. Because the content of images is supplied by
sensation, thinking about objects as sensible utilizes images quantitatively in virtue of their
direct relation to sensible objects. Therefore, thinking about sensible objects essentially
utilizes images.
Abstract thinking requires images only accidentally. Images serve as the content of
abstract thought, but this thought does not consider images in their essentially quantitative
nature. 63Objects represented by images as quantitatively determinate may be either quanti-
tatively indeterminate or not quantitative at all. 64 For example, a geometer must form a
quantitatively determinate image of a triangle in order to think about it, even though a
triangle may not be quantitative and the geometer does not consider quantitative features of
his image, such as the length of the line. 6~Therefore, even though an image supplies what
we think about, we think in abstraction from the quantitative image. Abstract thinking
involves images but does not use them.
We saw earlier that according to Plato memory preserves sensation by "writing" words in
the soul. Mental images arise through a "second artist, a painter" who, at two removes from
body, unifies the presentation of memory (sensation) into a whole (mental images). Thus,
Plato interprets sensation and mental images as progressively more untried spiritual activity
following upon the emancipation of soul from body. The content of mental images, like the
content of sensation, arises from soul knowing something which is independent of body.
Aristotle shifts the origins of sensation from soul to the sensible object insofar as it
informs our bodily senses. Body and soul work together in that all thinking, even the most
abstract, requires images whose contents come through the sense organs from sensible
objects. On this view mental images do not provide a unification of a sensible flux because
the images lie outside this flux, as Plato would have it: rather, the content of mental images
is immediately present in sensation, and imagination lies in the faculty of sense-perception
because it depends upon the actualization of the senses by the object. Because Plato
separates soul from body, he must locate in soul the explanation of all experience, sensible
as well as noetic; hence, different activities of soul require different "artists" present in the
soul. Aristotle's shift to a positive relation between body and soul enables sensible objects
to serve as the origin of sensation and to provide the content of mental images: different
kinds of thinking bear varying relations to the same sensible origins.
According to Aristotle, all thinking uses images, the content of which is supplied by
sensation. Thinking about extended objects uses images essentially and so essentially
originates in sensation; abstract thinking uses images accidentally and so accidentally
originates in sensation. The question is, does memory use images essentially, as thinking
about sensible objects does, or accidentally, as abstract thinking does'? Aristotle argues that
perception of time essentially utilizes images and so is located in the faculty of sense-
perception and not in the faculty of intelligence. He then concludes that since memory
requires the perception of time, memory too belongs essentially to the faculty of sense-
perception. He adds, however, that memory, like images, belongs accidentally to the
faculty of intelligence.

63 "'To the thinking soul ~mages serve as if they were contents of percepnon. 9 That Is why the soul never
thinks wahout an image" (De Aroma 3 . 7 . 431a15, cf Slakey, pp 483-84)
64 De Memoria 1 , 4 5 0 a - 4
65 De Memorta 1 , 4 5 0 a 4 - 7 .
388 HISTORY OF PHILOSOPHY

We are aware of the passage of time only when a perceptible movement takes place. 66
Observation of perceptible movement uses images essentially. Therefore, the cognitions of
magnitude, motion, and time all belong to the same faculty, the primary faculty of sense-
perception. 67Since memory belongs to that faculty which perceives the passage of time and
memory involves awareness of a time lapse, memory too must always essentially involve an
image even when an intellectual object is remembered. 68 Therefore, memory belongs
essentially to the faculty of sense-perception, although accidentally memory also belongs to
the faculty of intelligence.
Abstract thinking may depend upon memory to supply its content, that is, a quantitative
image; but once the image is supplied by memory, thinking considers it abstractly and not
quantitatively. Therefore, all objects which are properly capable of being imaged--that is,
sensible objects--are objects of memory, while those objects which are imaged only for the
sake of thinking about them in abstraction from that image belong to memory only acciden-
tally. 69
Therefore, memory belongs properly to the faculty of sense-perception and only acciden-
tally to the faculty of intelligence.7~Furthermore, Aristotle adds, if memory were a function
of intellect, then none of the lower animals would exhibit it. 71 This view is clearly false,
because all animals which are trainable possess some memory. Hence, memory belongs to
intellect only accidentally--it may supply the content for an operation which is itself
abstract and peculiar to man. Essentially memory belongs to sense-perception and so is
common to man and some lower animals.
This point completes the first argument in Aristotle's position. Memory is essentially
located in the faculty of sense-perception and properly bears upon sensible objects of past
experience, although objects of abstract thought also belong to memory insofar as they must
be quantitatively imaged in order to be remembered. The remainder of the first chapter of
De Memoria explains memory as an essential relation between an image and a sensible
object of past experience. Since memory is related to intelligence only accidentally, intelli-
gence and contemplation largely drop from the discussion of memory. 72
These conclusions are distinctly, perhaps systematically, anti-Platonic. Plato insists that
memory and "remembering" apply equally to past, present, and future sensation. 73 Since
memory constitutes all sensation, past, present, and future sensations rely equally upon
memory both to constitute them, insofar as is possible, and, simultaneously, to make them
intelligible, insofar as is possible. The view of memory as constituative of sensation
presupposes an essential disrelation of body (and sensible objects) from soul (and intelli-
gible o b j e c t s ) . TM Here we reach the crux of Plato's position: body is passive and helpless--it
can only occasion this or that sensation--while soul alone, because it knows something
opposite to body, actively construes sensation, thus caring for body and leading body to its

66 Phys. 4, 11,219a4-9. Also see C. H. Kahn, "Sensation and Consciousness m Aristotle's Psychology,"
Archiv fitr Geschtchte der Philosophie 48, heft 1 (1966):61-62.
~7 De Memoria 1,450a9-11; cf. D. W~ Hamlyn, "Koine Aisthesis," Monist 52 (1968): 195ff.; Sorabjl, pp
75-76.
68 De Memorza 1. 450a12-14.
69 De Memoria 1,450a22-25.
7o De Memorza l, 450a13-14.
1,450a16-18; cf. Metaph. 1, l, 980b29. Also cf, Hamlyn, p 196.
7t D e M e m o r l a
72 Recollection properly belongs to intelligence, and so Aristotle returns to mtelligence when he discusses
recollection in De Mernorla 2.
73 Phil. 39d-3; cf. above p. 384.
74 Theaet. 185e4-6.
 ARISTOTLE ON MEMORY 389

proper objects. Therefore, memory belongs to soul independently of body, is prior to

temporal distinctions, and is properly associated with intelligence.
Aristotle denies each of these points. He intends to separate memory from intelligence. In
order to do this, he first denies that memory applies equally to all experience past, present,
and future; memory is limited to past experience alone. Memory is so limited because
sensible or noetic objects originate experience by actualizing the senses of the mind, and so
these objects characterize all experience as past, present, or future. Because sensible or
noetic objects constitute all experience, all experience divides into the past presence,
presence, or future presence of objects. Memory describes our relation to the past presence
of objects. But the object is able to play such a dominant role for Aristotle only because of
the positive relation obtaining between body and soul.
Here we reach the crux of Aristotle's position. All thinking in some sense requires an
image, and images originate through sensation, that is, the actualization of the senses by a
sensible object. Sensation is not a spiritual activity conducted by soul alone, but a sensible
activity originating in the actualization of the senses by sensible objects, and such actualiza-
tion is required for all thinking. Mind does not operate insofar as it can be independent of
body; mind operates because it can consider abstractly images which originate in sensation
as a sensible activity.
Thus, at the end of the first argument in the D e M e m o r i a , Aristotle has redefined---contra
Plato---the soul-body relation, sensation, and the role of sensible objects in all thinking. In
so doing Aristotle "corrects" Plato. Memory does not bear upon all experience but is limited
to past experience only. Memory does not supply presentations needing to be unified into
images; memory essentially utilizes images, the content of which is supplied by sensible
objects through sensation. Memory is essentially located in the faculty of sense-perception
and is only accidentally located in the faculty of intelligence.
The second part of Aristotle's argument in the D e M e m o r i a concerns remembered
images--their genesis, sources of error concerning them, and their relation to objects.
Aristotle virtually quotes the T h e a e t e t u s and then explains the relation of memory to its
objects in strikingly Platonic language.
Aristotle now turns to memory as a state or affection of perception or conception. 75When
present to its faculty, sensible or noetic, an object constitutes perception or conception.
Memory relates to such objects when they are no longer present to the senses or mind. "One
might ask how it is possible that though the affection alone is present, and the fact [to
p r a g m a ] absent, the latter--that which is not present--is remembered. ''76 Aristotle first
explains the genesis of images through sensation and mistakes of memory connected with
this genesis. He then explains the relation of established images to objects which are no
longer present and how mistakes occur in this relation. This explanation establishes Aris-
totle's position and concludes the chapter.

It is clear that we must conceive that which is generated through sense-perception m the sentient soul,
and in the part of the body which is its seat--viz, that affection the state whereof we call memory--to
be some such thing as a picture. The process of movement involved in the act of perception stamps in,
as it were, a sort of impression of the percept, just as persons do who make an impression with a seal.77

75 De Memoria 1,449b26.
76De Memoria 1,450a26-28.
77De Memorta 1,450a28-32, cf. De Anima 2, 12 Cf. this sectionof the De Memoria withTheaet. 191c8-e; cf.
above pp. 380-81.
390 HISTORY OF PHILOSOPHY

The formation of images by imagination is not identical with sense-perception, but is "a
movement taking place as a result of actual sense-perception. ''7~ Thus, imagination,
which memory uses essentially, depends upon perception in two ways: for the content of
images and for the formation of images. 79The metaphor of an object making an imprint in
wax is fitting, for Aristotle, because the motion of the object makes the imprint while the
object itself, through this motion, supplies the content of the imprint.
Socrates rejects the metaphor of the wax tablet for several reasons: s~ (1) it fails to
explain mistakes which do not involve perception; (2) it incompletely emancipates soul
(and memory) from body: (3) it fails to distinguish between ideas and perceptions as
contributing to the content of mind, or soul. These difficulties are bypassed in the Phile-
bus because soul (and memory) are completely emancipated from body. In the Philebus,
we reach memory as a sort of scribe and a second artist who paints pictures from the
presentation of memory.
But in the De Memoria Aristotle has already corrected the arguments of the Philebus by
separating memory from intelligence, relocating memory in sense-perception and relating
memory to past experience only. Sensible objects, rather than soul, originate sense-
perception and mental images; consequently, mental images, some sort of pictures, are
located immediately in sensation, itself required by all thinking. Given this radical reinter-
pretation of memory and sensation, Socrates' objections to the metaphor of the wax tablet
are meaningless. (1) Since the content of images is supplied through perception, there are
no mistakes in the content of images which do not involve perception. (2) Memory is
directly located in sense-perception, and even during abstract thought soul cannot be
completely emancipated from body. (3) In respect to the presence and content of images
there is no distinction between ideas and perceptions. Aristotle's rehabilitation of sensa-
tion and memory makes not only the image of the wax available to him, but the entire
account of mistakes in the Theaetetus. s~
Actualization of the senses by sensible objects accounts not only for the origin of
mental images but also for bad memories in the genesis of mental images. Bad memory
occurs not because of body per se, but because of a defect of body. Young and old people
alike possess bad memories because the state of their bodies, either extreme growth or
extreme decay, inhibits good actualization. 82 "The former are too soft, the latter too hard"
to take and retain a good impression. 8~ Likewise with those who are too quick or too slow;
the former are too soft, and so while they take an impression easily, they do not retain it;

78 De Aroma 3, 3,428b30; much of this discussion seems to be &rected against Plato; see De Anima, Bks 2 and
3, trans D. W. Hamlyn (Oxford: Clarendon Press, 1968), pp 132-34: and Lycos, pp. 401-508.
79 DeAnima 2. 12,427a18; 3.3.
so Theaet. 196a-d
sJ See Theaet, 194c4, also 191c8-d. My account here naturally emphasLzes the difference between Plato and
Aristotle in order to show how Aristotle changes Plato's views for his (Aristotle's) own purposes. But we should
also note one ~mportant assumption which Plato and Aristotle share: neither provides a criteria for the correctness
of memory independently of the object remembered. According to both Plato and Aristotle, insofar as mental
images have eplstemic ~rnport they lmmedmtely refer to an extramental origin, e,ther mtelhglble in the case of
Plato or sensible in the case of Aristotle. Consequently. the entire question of the verification of mental ~mages
independently of their extramental origins never arises for either thinker.
s2 De Memoria 1. 450b ff.; Aristotle experiences no diff, culty ,n moving from perception to body, presumably
the sense organ, because "'there are not two things, but one organ which can be v,ewed either from the point ofwew
of ItS physical nature or from the point of view of Its function" (Hamlyn, De Aroma, p 114). Sense-perception is
nothing other than the operation of the sense organ (see De Anima 2, 12,424a17-27,425b27, 426a16. 427a3,
431a-b19. 432bl
s3 De Memorla 1,452b9-10
 ARISTOTLE ON MEMORY 391

the latter are too hard and so take no impression at all. 8+ The genesis of the image fails in
these people and a fortiori they cannot relate images to objects; hence, they possess bad
memories.
The remainder of the argument establishes Aristotle's constructive account of memory. It
explains how images relate to past objects of sense-perception and how mistakes occur in
this relation. "When one remembers, is it this impressed affection that he remembers, or is it
the objective thing from which this was derived? "'85 On the one hand, if we remember only
the impression retained in the mind, then we remember nothing that is absent. +6In this case
past objects of experience become unknowable. On the other hand, how can it be possible
through an image to remember that which is absent? s7 The answer to these questions lies in
the positive relation between mental images and objects of past experience.
A mental image can be regarded either in itself (kath' hauto) or as a likeness (eikon) of
something else. When considered in itself, an image is "'only on object of contemplation or
imagination."ss That is, we consider the image in isolation from its extramental original and
not as mnemonic token of an absent object.
But considered as a likeness, an image serves as a mnemonic token. In this case, we grasp
the image as related to the object which produced it. Because sensible objects account for
both the genesis and the content of mental images, generated images immediately resemble
their sensible originals. 89 Due to this resemblance the eikon immediately puts us into the
presence of the original sensible object, and in this way we remember an absent object
through a present image. This conclusion utilizes and completes Aristotle's views of sensa-
tion and images vis-a-vis the problem concerning the relation of images presently in the
mind to absent objects of past experience.
Aristotle develops his comparison of mental images to a picture. If we "contemplate" a
picture of Coriscos as a likeness of the man Coriscos, then the picture serves as a
"mnemonic token" of Coriscos because its content is wholly provided by Coriscos. Mnemo-
nic exercises then try to preserve one's memory by repeatedly contemplating something as a
likeness to an original. 9~ Thus the picture, or any image, used as in relation to an absent
object serves as a mnemonic token which through its content puts us into the presence of the
absent object.
Aristotle's account of memory here seems to place us squarely in Plato's metaphysics of
participation, According to Plato a copy (eikon) by definition resembles or participates in its
original and thereby immediately puts us into the presence of that original. +t In some sense
all nature is an image, intelligible only in relation to its formal original. Aristotle's ar-
gument that as relational a mental image is a copy which puts us into the presence of that

~4 De Memoria 1,450b7-13+
~s De Memoria l, 450b12-13.
~b De Memoria 1,450b14-15.
~7 De Memorta 1,450b15-17.
De Memorla 1,450627-29.
~9 See Sorabjl, pp. 2-3.
Anstolle adds a peculiar example here: ff one had never seen Coriscos, but hoped to recognize him on sLght,
one might look at his portrait so that the memory of the portrait v,ould reform the expectation and he would
recognize Conscus at first sight. At first glance this example seems to contradict the earher distinction between
memory as our relal~on to objects of future experience But the past object of experience here ts, of course, the
ptcture, and the future object is the man whom the ptcture represents. In this case memory reforms expectation only
because of the relation between the portrait, already seen, and the man, about to be ~een. ThLs example appears to
be tacked on to the main argument, perhaps as a reply to anticipated objections (see Phd. 39d-3 ~. Perhaps Aristotle
wishes to show that hts account handles tb~s relatmn at least as easily as does Plato's account.
9t F. M. Comford, Plato and Parmenides (New York. Bobbs-Memll. 1939, pp. 93-94.
392 HISTORY OF PHILOSOPHY

original which produced it is expressed in strongly Platonic language and clearly utilizes
Plato's notion of the relation of a copy to its original.
But in recognizing this Platonic origin of Aristotle's account, we must also see that
Aristotle adapts Plato's account to his own purposes. 92 According to Plato, the physical
realm is a copy--body and sensible objects contribute nothing to formal knowledge--while
the formal, nonphysical, realm is the original of physical copies and serves as the proper
object of knowledge. Aristotle rehabilitates the status of sensible objects so that these
objects serve as originals for ideas in the soul and thus are the lowest objects of knowledge.
Thus, the status of the physical world is entirely different for Plato and Aristotle: for Plato it
can be only a copy, while for Aristotle it can serve as an original.
According to Plato, copies must be seen as such, and failure to see them as nothing other
than a relation to the original which produces them results in mistakes at once metaphysical,
epistemological, and ethical. Plato constitutes all knowledge as a relation of soul, that is,
the knower immaterial and independent of body, to that which is immaterial and knowable.
Therefore, in order to explain knowledge and the various operations of soul, Plato requires
only a dynamic relation of soul to the knowable. Any notion of a mental image independent
of the knowable forms is by definition self-contradictory, just as is any notion of a copy
independent of its original
According to Aristotle all thinking requires an image which originates in sensation.
These images can be considered either in relation to their sensible origins, or, as in abstract
thinking, apart from these sensible origins. Therefore, for Aristotle, images, and all copies,
can be legitimately considered either in themselves o r in relation to their sensible origins.
Hence, mental images can stand independently of the original which produced them for
Aristotle, while such a notion involves a self-contradiction for Plato.
This independent status of mental images raises a problem for Aristotle which is wholly
absent in Plato. Aristotle must consider mistakes not only in the genesis of images but also
in reestablishing the relation of images as independent in the mind to the now absent
sensible objects which produced them. Since, for Plato, images necessarily involve an
immediate relation to their original, problems of reestablishing the relation of images to
their originals can never arise. But, for Aristotle, since mental images can be taken either in
themselves o r in relation to their original the problem of mistakes when a copy-to-original
relation is reestablished becomes critical. His explanation of the problem of these mistakes
concludes and completes his analysis of memory.
Even people with good memories make mistakes. Such mistakes lie not in the genesis of
the impression, but in the relation (or the lack of it) of an image to its object. Such relational
mistakes occur in two ways. ( 1) We can fail to see an image as relational--we just take it in
itself--and hence we fail to remember the real object which produced the image. 93 In some
cases, we do suddenly remember the object of past experience. Sudden remembering
results from a sudden shift in perspective from contemplating an image in itself to seeing it
in relation to its extramental sensible origin. As relational, the image serves as a mnemonic
token illuminating an absent object, and we suddenly remember. 94 (2) The second type of
mistake seems to be more pathological: believing that one remembers when in fact there has
been no past experience. In this case images always appear as r e l a t i o n a l when indeed no
relations obtain. If all images always appeared this way, it would become impossible to

92 On Aristotle "adopting" Plato's views, see Kahn, p. 66.

93 De Memorta 1 , 4 5 1 a 2 - 5 .
9~ De Memorta 1 , 4 5 1 a 5 - 8 .
 ARISTOTLE ON MEMORY 393

separate images as mnemonic tokens from images as required for abstract thinking or as
mere imaginings. Such mistakes occur in those suffering mental derangement. 95
This explanation of mistakes concludes Aristotle's discussion of memory and the first
chapter of De Memoria et Reminiscentia. Memory is essentially related to sense-perception
and only accidentally related to intelligence. Thus, the proper objects of memory are
sensible objects inherently capable of being imaged or intelligible objects, such as triangles,
insofar as they are imaged. Memory is constituted through a positive relation between body
and soul and is a state or affection of perception or conception, 96 Memory is the "state of
imagination related as a likeness to that of which it is an i m a g i n a t i o n . . . [and memory] is a
function of the primary faculty of sense-perception . . . . -97 Memory is sharply disrelated
from recollection and abstract thinking is performed by recollection. For this reason people
who are "slow," that is, do not make good use of their intelligence, often excel at memoriz-
ing; while intelligent people, being quick-witted and clever, excel at recollection.

3. We can now return to our original problem: the status of the Platonic language in De
Memoria 1. On the one hand, this language is Platonic and is rightly understood as indicat-
ing the presence of Platonic metaphysics in this text. On the other hand, Plato's doctrine of
participation has been radically transformed by Aristotle's reinterpretation of sensible
objects, sensation, and memory.
The Platonic language of the De Memoria et Reminiscentia cannot be ignored. By means
of this language, Aristotle invokes Plato's causal relation between originals and copies. But
Aristotle does so only after he systematically opposes his own doctrine of memory, sensa-
tion, and mental images to that of the Philebus and quotes Plato's rejected image of the wax
tablet as a model par excellence of mind and memory.
Here, then, is the key to the mysteries of the De Memoria. The importance of Platonic
language in Aristotle need not mean the importance of Plato in Aristotle. Rather, it indicates
the richness of Plato's metaphysics and language for Aristotle's own purposes and reinter-
pretation. In the De Memoria we possess clear evidence of this reinterpretation of Platonic
language. Aristotle first criticizes Plato and then on the basis of his criticisms asserts the
model of memory which Plato explicitly rejects. Thus, at the conclusion of the argument,
when Aristotle relies on Platonic language and concepts, the content of the argument can be
understood only as Aristotle's own. The Plato present in the De Memoria is a Plato
corrected: Aristotle's Plato.

Trinity College, Hartford, Connecticut

95De Memoria 1,451a8-9.

96De Memoria 1,449b24-25.
97De Memoria 1, 451a 16-18.