THEODOR EBERT

University o] Erlangen

P L A T O ' S T H E O R Y OF R E C O L L E C T I O N R E C O N S I D E R E D
AN I N T E R P R E T A T I O N OF M E N O 8 0 a - 8 6 c

In this article1, I should like to put forward an interpretation of a passage in

the Meno which is probably the earliest reference in Plato's writings to what
has generally been known as the "Theory of Recollection." Together with the
"Theory of Forms," to which it is systematically subordinate, the "Theory of
Recollection" is one of the doctrines most central to Plato's philosophy. The
upshot of this theory, as it has generally been understood by Plato's interpreters, might be stated briefly as follows : the human soul has had a vision
of the eternal Forms before birth; this knowledge of the Forms is, however,
lost when the soul enters the world, and has to be recovered by the process
of recollection. Since all true knowledge is knowledge of Forms, knowledge
can only be achieved through recotiection.
It would appear to be obvious that this theory, in comparison, for example,
with the logical achievements of Aristotle, or of Plato himself in his later
dialogues, is of little genuine philosophical interest; we may be interested in
it for historical reasons, but it does not provide an answer to a philosophical
problem which might still be accepted as valid today. On the contrary, it
would appear to be an obstacle to a solution of the problem it pretends to
solve, in that it poses the problem in somewhat inappropriate terms.

Although the above statement of the "Theory of Recollection" accords with

the interpretation widely accepted among scholars, it does not harmonize
with the texts as well as it should. Although the difficulties which arise in
any attempt to give a consistent interpretation of the texts relevant to recollection have been discussed, so far as I can see no satisfactory solution of
them has been forthcoming. I should like to draw attention to the two
principal difficulties posed by the Mena. Firstly, although this dialogue
contains the most detailed statement of the "Theory of Recollection," there
is, insofar as this theory is concerned, no allusion to the "Theory o.f Forms."
163

THEODOR

EBERT

In spite of H. Cherniss' claims against Klara Buchmann 2, I can find no

suggestion whatever in the Meno that recollection is meant to be a recollection of Forms. In introducing his puzzling statement that all learning is
recollection, Socrates is appealing to the religious doctrines of the Orphics
and/or Pythagoreans about reincarnation and metempsychosis.
The second difficulty lies in the scepticism apparent at the end of
Socrates' discussion of recollection (Men. 86b6-c2; cp. Phaedr. 257a).
Socrates' concluding remarks in the Meno look in fact as if they were a sort
of revocation of what has just been proved in a somewhat dogmatic way.
N. Gulley thinks that this retraction is Plato's way of conveying his doubts
about the Theory of Recollection to his readersL This suggestion would imply however, that within the space of a few pages, Plato is affirming and
questioning one and the same doctrine without giving his readers a hint as
to what the solution of this riddle might be. Gulley's interpretation seems to
me0 therefore, to be somewhat implausible.
Besides this "traditional" account of the theory of recollection, there is
another approach, which has found favour especially among nen-Kantian
scholars*. According-to the explanation which this approach has given rise
to, the philosophical rationale of the Platonic anamnesis is the apriori. What
is hidden behind the metaphysical mythology of a prenatal vision of Forms
is in reality nothing other than the modern philosophical insight that there
are concepts and truths which can never be formulated merely on the basis
of sense-perception, but which must be knowable independently of it, and
which are logically prior to it; those of pure mathematics for example. And
is it not a mathematical truth that is discovered, "recollected" by the slaveboy in the Meno ?
This account, which has been widely accepted by continental philosophers
since Natorp and N. Hartmann put it forward was not actually initiated by
the neo-Kantians. It was Leibniz who first insisted, in the course of his metacritique of Locke's criticism of innate ideas, that this was the true meaning
of Plato's recollection, and who first tried to force this meaning from the
obfuscations of the Platonist tradition ~. But neither Leibniz nor the neoKantians were able to give a convincing explanation of the Hatonic texts.
Although Leibniz contended that it was the Platonists only who gave the
theory of recollection its mythical form, he did not explain how this view
was to be reconciled with the fact that in the Meno as well as in the Phaedo
recollection is dealt with in terms of a prenatal life. Natorp is more critical
164

PLATO'S THEORY OF RECOLLECTION RECONSIDERED

of this point, and blames Plato for what he calls "a psychological, and what
is more, a metaphysical bias"" of the a prio.ri.
Thus neither the "traditional" nor the "aprioristic" interpretation of the
theory of recollection gives rise to a cogent account of the Platonic texts. In
order to overcome the difficulties they involve, we should, I think, bear in
mind the following principles :
(1) We should not interpret Plato's dialogues by making presuppositions
about what he might have wanted to convey to his readers, by assuming
either a "Metaphysica Platunis" or looking for anticipations of the achievements of modern philosophy. Even if Plato might in some respects have
anticipated Kant or Leibniz, looking at him through Kantian or Leibnizian
spectacles might prevent us from seeing other more important aspects of his
philosophy. We should attempt to explain what Plato himself is trying to say.
(2) We should take the literary form, that is to say the Platonic dialogue,
as the starting point of our interpretation. Plato's (written !) doubts about
the insufficiency of the written word in matters of philosophy would be
rather pointless if his dialogues were nothing but dramatized treatises. It
seems to me that we come nearer to the truth if we regard his dialogues as
an attempt to solve the problem of overcoming the danger of dogmatism
and misunderstanding inherent in written logoi without leaving philosophy
exclusively to transient oral conversations. In other words, we should regard
the dialogue-form as an answer to the problematic situation of the philosophical author. Plato is the philosophical pupil of a man who left nothing
written. The Platonic dialogue, unlike the dialogues of the Enlightenment
for example, is not the dressing of philosophical doctrines in the convenient
form of a conversation. The most striking difference between the Platonic
dialogue and that of later times seems to consist in the different function of
question and answer. To put it briefly : in Plato's dialogues questions do not
function as a mere staging of the (correct) answer; the respondent may be
wrong and although he may be wrong, this is brought out not by direct
criticism, but in the rather indirect approach of an etenctic questioning.
Socrates tries to show what is wrong with his interlocutor's thesis by means
of a questioning technique which is designed to deduce the absurd consequences of a wrong answer. It is this technique of elenctic questioning which
seems to be the specific feature of the Platonic dialogue, and which gives it
its character of a dialogue between author and readerL

165

THEODOR EBERT
II
It is a sound exegetic principle to start an interpretation of a difficult and
puzzling text with what seems to be clear in it; in the case of the passage
under discussion, this qualification clearly applies to the famous Socratic
geometry lesson (82b9-85b7) s. We can even rely on the text itself for this
assumption : the geometry lesson is meant to be an explication of Socrates'
statement that all learning is recollection (cp. 81d4-5). It is quite evident
from this lesson that the slave-boy is learning something, and in what sense
he is doing so. _After the problem of duplicating a square has been explained
to him, his first answer is erroneous, assuming that one can construct the
double square from the side of double length (82e2-3). He is then brought
to see the error of his first and second proposals, and finally confesses that
he has no answer to the problem at all (84al-2). From that point he is led
to the discovery of the solution : the square of double size can be constructed
from the diagonal of the original square.
It is, however, in no way evident that Socrates' young pupil is recollecting
something in this process of finding an elementary geometrical truth. In
order to give a rational explanation of this discovery one would not have to.
rely on the farfetched and implausible hypothesis of prenatal learning. The
boy corrects his mistakes and makes his discovery by way of trial and error
combined with a method of testing the proposed answers: by counting
segments of equal size in the squares to be compared. It is all the more
paradoxical that Socrates should interrupt his lesson on two occasions (82
and 84a-d) in order to tell Meno, on the first occasion, to pay attention to
the slave-boy's recollection; and, on the second, to ask him what point in
the process of recollection has been reached by the slave.
The promised exemplification of Socrates' puzzling statement turns out to
be rather unsatisfactory. In view of the difficulty involved in making sense
of the geometry lesson as an explanation of the "learning-is-recollection"
statement, if we take this statement at its face value, we would appear to be
justified in trying another approach. Let us suppose that recollection is used
here in a metaphorical way, and that what Socrates wants to draw attention
to is an analogy between learning and recollecting. We should bear in mind
however, that this supposition will shed its hypothetical character only if it
not only provides a plausible interpretation o.f the geometry lesson but can
also be shown to be in accordance with Socrates' speech in 81aS-dS, where
166

PLATO'S THEORY OF RECOLLECTION RECONSIDERED

the argumentation seems to be based wholly on religious and mythical
grounds.
In taking recollection to be a metaphor, we should however avoid connecting it with those theorems of modern philosophy which belong to a systematically and historically different context. In order to do so I shall first try
to give an analysis of what could be called the "essential" or "constitutive"
feature of recollection itself, quite apart from its function within a philosophical theory or within our beliefs about this theory.
As is shown by a passage in the Philebus (34b6-c2), the Greek
'dv6t~v~]o~'has two meanings. It may either designate the process of mere
reminiscence i.e. the bringing to mind of something one knows but had not
actually thought of, or the recollecting of something one had forgotten. Nosy
I take it for granted that ~dvdktvqo~c_ ~ is used in the second sense in the
passages relevant to the theory of recollection. This supposition finds
support in the many references to forgetfulness in these passages (Phaed.
73el-4; 75d7-76a7; Phaedr. 248c7), which make it obvious that ' dvd#vF
o ~ ' is understood as recollection of something that had been forgotten.
The fact that a process of recollection involves as its starting point a state
of forgetfulness is relevant to the structural nature of recollection itself. In
order to bring out this structure we have first to give an analysis of forgetfulness.
Forgetfulness is not a mere lack of knowledge but is characterized by a
specific reflectiveness : having forgotten something does not only mean not
knowing something, but implies a no.t being aware of this lack of knowledge.
It is, of course, true that when somebody says "I have forgotten so-and-so"
he is not in this state of total forgetfulness. But this only means that he has
overcome this state of total forgetfulness and has taken what might turn out
to be the first step in a process of recollection. We can refer to such a state
of total forgetfulness only in the case of talking about a third person, not in
the case of talking about ourselves. In the case of statements introduced by
"I have forgotten" the awareness (of forgetfulness) expressed has been
preceded by a lack of awareness about one's having forgotten. Thus, forgetfulness is like the blind spot in our visual field : there, too, we not only do
not see anything, but we are not even aware of our not seeing anything.
The reflexive structure of forgetfulness now gives recollection its crucial
feature : recollection consists of two cognitive acts : first, one has to realize
lthal one has forgotten something, and then one can recollect what one has
167

THEODOR EBERT
forgotten. These two steps may coincide in borderline cases, when, for
example, I am suddenly reminded by something of something I had forgotten. But even in these cases the awareness of something's being forgotten
is a logical presupposition for any recollection of it. Only if we recognize
what we have recollected as something we have known earlier but which we
do not actually know, are we entitled to speak of recollection; otherwise it
would not be distinguishable from any new information.
It is easy to see that this gradual succession in the process of recollecting
has its formal counterpart in the passage from error (fictitious knowledge)
to true knowledge : both forgetfulness and error can be characterized as an
ignorance about ignorance; we are lacking knowledge about a lack of knowledge in both cases. And in both cases the way to knowledge leads through
the awareness of one's ignorance.
The first point of my argument is, then, that the geometry lesson points to
this analogy between the process of recollecting and the process of learning
in the sense of "coming to know from an error." We have seen above that
the process of learning Socrates' young interlocutor is undergoing leads from
fictitious knowledge (the answer in 82e2-3), to the predicament in 84al-2,
and from this point to the discovery of the solution (85b). That Plato
wants to lay emphasis upon these steps in the process of learning is shown
by the dramatic staging of the geometry lesson: the questioning of the
slave-boy is divided - - by Socrates' questions directed to Meno in 82e and
8 4 a - d - into three sections (82b9-82e3; 82e14-84a2; 84d3-85b7). These
sections lead to error, predicament, and the discovery of the solution respectively. The fact that the dramatic arrangement of the mathematical lesson
stresses these steps in the process of learning is not, however, sufficient
proof that Plato/Socrates is actually pointing to an analogy between the
processes of learning and recollecting. To prove this we need further evidence. We find it, I should like to suggest, in the questions and comments
Socrates addresses to Meno in the two interruptions of the geometry lesson.
The first interruption (82e4-13) is a comment upon the slave-boy's first
erroneous reply. Socrates calls attention to the fictitious knowledge expressed
in this answer (cp. o[~zal d3~vat 82e5, o~'ezat el0). He then goes on
to urge Meno to "pay attention how he [i.e. the slave] will recollect
step by step (gc;~q;) as one ought to do in recollection" (82e12-13).
This remark implies that someone who is undergoing a process of recollection
will go through a gradual process in a necessary sequence. If we adopt the
168

PLATO'S THEORY OF RECOLLECTION RECONSIDERED

"traditional" account of the theory o.f recollection, the meaning of this
remark remains mysterious. Bluck, for example, has no satisfactory explanation for it~ The statement becomes understandable, however, if we interpret
it in the light of our analysis of recollection : Socrates is alluding here to the
formal characteristics the process of recollecting shares with the process of
learning. And he can do so at this point because the slave has been led into
an error, i.e., into a state corresponding to total forgetfulness. We may say
that Socrates is merely translating his previous comments upon the mere
belief, the fictitious knowledge of his interlocutor, into the metaphorical
language of recollection.
The second part of the geometric epideixis (82e14-84a2) brings the
slave to an awareness and confession of his ignorance. He now knows that
he has no solution to the geometric problem (84al-2). He has reached the
predicament, the aporia.
At this turning-point in the geometry lesson, Socrates interrupts his
mathematical questioning for the second time and asks Meno: "Do you
recognize, Meno, which point he has now got to in the process of recollection ?" (84a3-4). There is no answer to this question in the same metaphorical terminology, i.e., in terms of recollecting. Socrates does not state
explicitly that the slave is now in the state of - - speaking in terms of recollection - - recognizing that he has forgotten. This answer is left to Meno
and/or to the reader. Socrates, however, gives an explanation of the slaveboy's state in terms of knowing and not-knowing (84a4-bl). The boy now
does not know the solution any more than he did befo.re (84a4-6), but he
has taken a step within his state of ignorance. He has taken the step from
fictitious k n o w l e d g e ( dfezo ... ,6ze egS&al a6) to the awareness of
his lack of knowledge (vf;v ~... o~:x o~&v, o~'o~ezat e~gvat a7-bl).
It is, however, worth notidng that Socrates can give this explanation instead
of a direct answer. Socrates states where the slave has arrived in his process
of learning, that is he gives, in the non-metaphorical language of learning,
the answer to a question that was put in the metaphorical terms of recollection. The key to the riddle of the Platonic recollection lies in this shift from
Socrates' metaphorical question to his non-metaphorical answer. To be able
to read Socrates' remarks about the predicament o,f the slave-boy as an
answer to the preceding question presupposes, however, a recognition of the
structural analogy between the processes of recollection and of learning.
Socrates obviously cannot point directly to this analogy : this would imply a
169

THEODOR

EBERT

teaching of Meno (and of the reader) and not a mere guiding by means of
questions.

III
In spite of the plausibility of the above interpretation of the geometry lesson,
this leaves us with a problem. If the geometry lesson is designed to illustrate
the analogy between the processes of recollecting and learning (in the sense
of "coming to know from an error"), then any correction of any error will
be a case of "recollection." Is this the whole outcome of the Platonic recollection ? Does Plato merely want to make the somewhat trivial point that
any correction of an error (a false belief) logically presupposes the awareness
of that error ? Or does the Platonic recollection have a more specific significance in the Meno and in the whole of Plato's philosophy ? I think it does.
In order to see this significance, we shall have to consider the function of
the geometry lesson in the wider context of Socrates' discussion with Meno.
In the questioning of the slave we witness nothing but the correction of an
error concerning a particular geometric construction. In the context of the
dialogue, however, the geometry lesson has a much more distinctive function :
it is as it were, the play in Hamlet, a "play within a play," and just like this
performance, it has the object of displaying the state of the persons for
whom it is staged.
We shall see what this lesson is intended to tell us and is intended to tell
Meno about Meno if we turn first to an analysis of Meno's predicament
concerning the nature of virtue expressed in his speech at the beginning of
our passage (79e7-80b7). Meno is fair enough to admit that he has no
answer to Socrates' repeated question (79e5-6) as to the nature of virtue.
All his pretended definitions have been proved to be wrong. The confession
of his predicament, however, is brought out only in 80bl-2, after quite a
lengthy statement concerning what has happened to him in the course of
his discussion with Socrates. The way in which he describes his own aporetic
situation makes it clear that he does not know what his predicament means
to him. He is referring to himself as being "bewitched" and "deluded" and
"enchanted" by Socrates (80a2-3; cp. 80b6). These metaphors as well as
the famous and much quoted comparison of Socrates with the torpedo
paralyzing everybody who gets into contact with him (80a5-bl) are designed
to bring out a negative fact : the loss of the capacity to use one's own mind
170

PLATO'S THEORY OF RECOLLECTION RECONSIDERED

and one's own limbs. Meno sees enchantment where in fact he has been
released from the spell of fictitious knowledge. He believes himself to have
been paralyzed by Socrates, the "torpedo," whereas the paralysis of error has
in fact been taken away from him. So his remark : "And yet on countless
occasions I have made abundant speeches on virtue to various people - - and
very good speeches they were, so I thought" (80b2-3 - Lamb's translation)
does not imply a self-criticism and the insight that his speeches about virtue
have been based on mere belief; it is rather intended to remind his listeners
of how unnatural the situation is into which he has been brought by
Socrates' tricky questions.
For Meno, the predicament is a sort of defeat. He has been "caught" by
Socrates. But even in the confession of what he thinks is a defeat, he still
shows his conviction that he is superior to Socrates thanks to his rhetorical
education : Meno is a friend and pupil of Gorgias (cp. 71c-d, 73c, 76b), and
the somewhat artificial metaphors he uses to describe his aporetic situation
are made up in the style of his teacher. So is the comparison of the torpedo :
this conceit even uses the typically Gorgian stylistic figure of the homoioteleuton : za~z~ T~ ~,~aze~a vd~x~ z~ Oa2)~az[a (80a5-6).
One of Socrates' intentions in staging the slave's aporia is to correct
Meno's misunderstanding of predicament. This is to be seen most clearly in
Socrates' questions about the predicament of the slave-boy in the geometry
lesson (84a3-c9). After having characterized the step from error to the
awareness of ignorance as a progress brought about through the predicament
(84a4-bl), Socrates gets Meno to admit that their young pupil is now, i.e.,
after having been brought to see his ignorance, in a much better state (84b3b5). Then he goes on to point out a direct connection between this cognition
and Meno's conception of predicament : in describing the slave-boy's state
in terms of the torpedo metaphor, he is able to give this metaphor a sense
contrary to its original meaning in Meno's speech at 80a-b. Meno has to
concede that the slave has suffered no harm by having been brought to his
predicament and by having been paralyzed as by a torpedo (84b6-8).
Socrates' next question; by its ironical overstatement, makes it quite clear
that what he is pointing at is in reality not the slave's but Meno's predicament : he is alluding to Meno's remark from 80b2-3, when he says that the
slave (who for the first time - - cp. 85d-e - - has been confronted with a
geometrical problem), before coming to realize his predicament, had been
"only too ready to suppose he was right in saying, before any number of
171

THEODOR EBERT
people any number of times, that the double space must have a line of
double length for its side" (84b11-cl. - Lamb's translation). Meno is to
realize that what still appeared to him in 80a-b as a sort of knowledge which
had merely been rendered ineffective by the aporetic incantation, is in fact
a fictitious knowledge, is error. Therefore, the following question insists on
the difference between error and predicament : he who is labouring under a
misapprehension does not even try to find the truth, is not even longing for
knowledge (84c4-6). Socrates concludes his questioning of Meno with
another allusion to the torpedo metaphor : "Then the torpedo's shock was
of advantage to him ?" (84c8. - Lamb's translation). Here, as in 84b6-7,
Socrates' question implies a critique of the negative meaning in Meno's
previous use of the torpedo metaphor.
If Meno, in admitting his predicament, is at the same time displaying his
misunderstanding of predicament - - and this has been confirmed by the
corrections Socrates tries to elicit - - then there must be some point in looking
for a reason for this. Why does Meno misunderstand the function of the
predicament ? We shall find an answer to this question if we analyze Meno's
famous eristical argument (80d5-8). This eristical objection is brought
forward in opposition to Socrates' proposal to look for a common solution
to the common predicament (80d3-4).
Meno's argument consists of three questions. The first (80d5-6) consists
of contesting in a general way the po,ssibility of searching for something
when one does not know what it is ( zo~zo b~r162ogo~ga zd) :zaod:zav 6vt
do,Iv ) and this, it is understood, is the case with Socrates' question
regarding the nature of virtue. Meno's second question (80d6-7). is meant
to support or prove (),d o d6) the argument suggested by the first. It isto
establish why, in the case of a What-question, there can be no method of
searching, no z#,~xor ~'~r#a8o3~" : because the possibility of searching for
something presupposes that one has an idea (cp. :zOo~9~l~,~vor d7) of what
one is looking for. This condition can, of course, not be fulfilled if one
does not know what one is looking for. The last question (d7-8) brings
forward quite a subtle objection. Even if we pass over the difficulties
implied in the notion of looking for something one does not know, even if
we restrict our consideration to the case in which someone stumbles by
chance upon what he is looking for, it must be the case that such a person
would still not have found it; for finding something involves more than
merely stumbling upon it ( ~,zvyzdetv d7), finding implies an identifica172

PLATO'S THEORY OF RECOLLECTION RECONSIDERED

tion of what one has stumbled upon with what one is looking for. This
identification is however impossible if one does not know what one is
looking for.
It should be noticed that Meno's three questions show the influence of
the technique of eristical argumentation taught by the sophists : Meno is, as
we have already noticed, a pupil of Gorgias. This influence is most evident
in the way in which the objections are mustered, a device reflected in the
"even if" ( d xat 80d7) of the last question: after one difficulty has
been overcome, the nert is already waiting. Gorgias in IIeo~ zo~e~ &,zo~
uses the same pattern of argumentation 1~
The reason that Socrates' thesis about learning and recollection is provoked
by Meno's eristical objection is to suggest the necessity o.f a much more
careful examination of Meno's argument and its underlying premisses than
is usually given to it 11. It is easy to see that Meno's objection owes its
apparent plausibility to its implicit orientation toward the search for perceptible things. If we are looking for something that is an object of senseperception, we must know what it is that we are looking for : we must have
perceived it at an earlier time, or we must possess a sufficient description of
it. It is indeed quite absurd to look for something and not to, know it, if
what one is looking for is an object of sense-perception, i.e., if it can be
known in the sense of knowledge by acquaintance (connaitre, kennen,
7~,'&va,). The plausibility of Meno's argument disappears, however, if what
one is looking for is something that can be known as a matter of propositional
knowledge (savoir, wissen, egO&at ). When we are looking for the solution
to a riddle or a mathematical problem, we have no idea of what might be the
solution to the riddle or the result of the calculation. W e simply know which
conditions a proposed solution is to fulfill; we are looking for a term within
a relation.
Meno's argument - - and this is my second point - - rests o.n the implicit
presupposition that all knowledge has the logical structure of knowledge by
acquaintance. The misconception of knowledge implied in this presupposition
is not, however, merely a contingent error or a misunderstanding quite natural
to the philosophically untrained mind.
In order to see the full implications of this misunderstanding, we have to
examine the difference between knowledge by acquaintance and propositional
knowledge. For brevity's sake, Jet us put knowing (a) for knowing by
acquaintance and knowing (p) for propositional knowledge. Now the most
173

THEODOR EBERT
basic feature in the case of knowing (a) is that this sort of knowledge is
not articulated in propositional form, although 'u,,hat is known (i.e., the
known object) may have the form of a proposition : I can know (a) people's
opinions and judgments and poems. On the other hand, knowledge (p) has
always a propositional character; "knowing (p)" - - if used in the affirmative
can always be constructed with a that-clause.
Now this difference between "knowing (a)" and "knowing (p)" has
consequences for the meaning of the negative use of "to know" (not
knowing) in the two cases. Since what I know (p) is necessarily a fact,
something to be expressed in a true proposition, what I think I know may
be not the case, may be expressed in a false proposition. Thus one meaning
of "not knowing (p)" is "being mistaken." "Not knowing (p)" in this
sense implies "taking a proposition R, to, be true whereas Not-R is true."
But obviously this is not the only meaning of "not knowing (p)." Not
knowing the fact expressed in the true proposition R ought not to imply
taking Not-R to be true. It may as well mean "having no, idea about the fact
expressed by R," or just "not having yet made up one's mind as to whether
R or Not-R is true" (grammatically this latter case is indicated by the use of
an interrogative clause). "Not knowing (p)" in this second sense means
"being ignorant about." Thus, we have to distinguish two senses of "not
knowing (p)" : error and mere ignorance.
In the case of "knowing (a)" however, there is only one opposite meaning:
"Not knowing (a) B" is "not being acquainted with B" and acquaintance,
because of its non-propositional character, does not allow of error : thus there
is no, meaning of "not knowing (a)" which corresponds to, the first sense of
"not knowing (p)." I am acquainted with something if and only if I have
become acquainted with it (which usually means that I have perceived it),
and its idea has not escaped my memory.
Now it will, at first sight, seem rather unlikely that any supposed knowledge (a) should be free from error. Obviously there are cases in which we
are mistaken about an object we are acquainted with. I can, for example,
believe wrongly that I know someone's brother, if the person who has been
introduced to me as such is in fact somebody else. And the same is true of
all cases in which the description of the object consists of referring to
relational properties. (Proper names can also be interpreted as belonging to
this class of descriptions : to know Peter is to know someone whose name
is "Peter." There is, however, a crucial difference between these errors in the
-

174

PLATO'S THEORY OF RECOLLECTION RECONSIDERED

field of knowledge by acquaintance and a mistaken knowing (p). When I
am mistaken in believing I know (p) something to be the case, I do in
fact not know anything. After the discovery of the error, the pretended
knowledge (p) disappears completely. When, however, I am mistaken about
an object of knowledge (a), it is not the case that I do not know anything
- - on the contrary, I do know (a) this person or thing; I can, for example,
recognize him/it or give a description of him/it. What I do not know is who
or what he/it is. The error in these cases belongs to a pretended knowing (p)
that is parasitic on the genuine knowing (a). Thus we can in fact say that
there is only a simple dichotomy between "knowing (a)" and "not knowing
(a)"; this dichotomy co.rresponds to. the opposition of having in mind and
not having in mind.
IV
This examination of "knowing (a)" and "knowing (p)" now enables us
better to understand the implicit premisses of Meno's eristical argument. He
is arguing on the implicit assumption that all knowledge can be considered
to have the logical structure of knowledge by acquaintance. This assumption
implies that knowing (p) and its negation form a dichotomy, as does "being
acquainted with" and "not being acquainted with." The assumption of this
difference .eclipses the basic difference between the two meanings of "not
knowing (p)," between error and mere ignorance. For Meno., knowledge is
the presence in one's mind or memory of something known, lack of knowledge the absence of a content of knowledge. This misconception of knowledge obfuscates the difference between error and mere ignorance, as well as
the progress implied in the step. from error to predicament. Thus we now get
an answer to our question as to the reason for Meno's misunderstanding of
predicament : it is rooted in his misconception of knowledge.
This is confirmed by Socrates' first reply to Meno,'s eristical objection. It
seems to be significant that his first answer is not a rejection of Meno's
argument as a sophistical inference by analogy, but an explication of what has
been implicitly presupposed by Meno's questions : the simple dichotomy of
knowing and not-knowing; either, so runs the interpretation of Socrates, you
know something, and then you need not look for what it is, or you do not
know it, and then yon do not know what to look for (80e3-5) x2.
If Socrates characterizes Meno's objection as an eristical argument (80e2),
175

THEODOR

EBERT

this ought not to imply that Meno himself is aware of the sophism implied in
his questions. On the contrary, the naivet6 of his question in 81al-2 strongly
suggests that he is unaware of the falsehood of his underlying premisses. He
is in fact labouring under a crucial misapprehension : he believes he knows
what knowledge is, but does not.
In order to correct this error of Meno's, Socrates chooses a digression, the
subtlety of which is, however, not easily recognized by the modern reader.
Socrates assumes the role of an epideictic orator in the style of Gorgias.
Gorgias has been present behind Meno all the time in the dialogue (cp. the
allusions to Gorgias' teaching 70b3, 71c5-d8, 73c7, 76bl, 76c4, 79e6). For
his eristical questions in Gorgias' manner Meno gets in return a "theatrical
reply," a zoayx~} d~:z~xO~o,~r similar to the one in 76d3-5 (cp. 76e3).
Socrates' following short speech in 81a5-d5y with its conclusion that all
learning is recollecting, is, I want to. argue, a parody of a Gorgian epideictic
logos. It has to be read aloud in order to catch the abundant stylistic devices typical of Gorgias' rhetoric. Typical of the Gorgian character of Socrates'
speech is the breaking up of the sentences into small cola, accentuated by
homoioteleuta and alliterations2 * I shall quote a few examples of this:
d~x~]xoa ?d~ dwcS~c?ovze xa't ~/vva~z~v oow6v
(81a5)--twocolaofequal
length immediately at the beginning, with homoioteleuta and a twofold
rhyme within the second. The abundance of rhyming syllables in : ... z~v
[ ~ o o v ~e xa~ zCov [e~etc~v 8aot~ ... :ze~" c~v ... o~t'oe~ ~'dvat c~tcSdvat
(81al0-bl), or in : ~22ot :~o22o~ ~c~v~ot~z~v &sot zSgoc.. (81bl-2) ; the
alliteration in : xa~ zoz~ #ev ze,~evzSv - 8&'7 d~oOv~oxetv xa2ogoot (81b45). The elaborate alternation of ~, z, fl and the figura etymologica in:
c)e'iv c)~ &d zaaza d)~ 6otc6~a~a &afitGovat zbv fllou (81b6-7); the elegant consonance in the antithesis : xal zd &~dc)s xa't ,a g:v ~'Atc)ov (81c6).
The quotation from Pindarus serves to accentuate a prose which embodies
the rules or poetry :~5 rhythm and euphony are overriding grammatical
simplicity: witness the hyperbata (e.g. 81all-b1), and the sonorous,
rhyming genitive cases instead of the nominative (81a10, bl). Besides the
stylistic devices, both the composition and the solemnity of this speech provide further evidence that Socrates is speaking in the r61e of sophistical
orator. It starts off with a reference to authorities, but the pathetic attributes
Socrates is using (cp. aoclo&v 81a5, z$s~a a6, z$s~ot b2) conceal the vagueness
of what they are attributes of. Only Pindarus is mentioned by name, the
"many other poets" are rhetorical dumb actors. The solemnity evoked by the
176

PLATO'S THEORY OF RECOLLECTION RECONSIDERED

vocabulary is accentuated moreover by the distinction drawn between the
diction used in the speech and the language of ordinary people (81b4-5;
d2-3).
Only after Socrates, by referring to priests and poets, has created the
nimbus of a "higher" truth fo,r what these authorities are saying, is the
argument itself brought forward. Meno's impatient interruption serves to
bring out the artificialty of this procedure : Meno wants to know the argument first, and only then those who are using it (81a7-9). Moreover, the
main thesis of the speech is only an inference that cannot be based directly
on the authority of "many divine poets" (bl-2). This fact, however, is
skillfully concealed by the emphatic reference to Pindarus and by the long
quotation which is in fact a digression, as has been noticed by Bluck. 16
It is worth noticing however, that the doctrine of reincarnation is part of
the philosophical theory of Empedocles (cp. fragments 8, 11, 15, 1i7, 118
Diels/Kranz). And as is shown by Socrates' definition of colour in 76c-d,
Meno can be supposed to be well acquainted with the philosophy of Empedocles. The line in 81b4-5 3c5r7 a=o4vFa~e~,za2o~o, seems to echo the
verse in fragment 15 (Diels/Kranz) of Empedocles :z6 c}r7 flgozov xa)~{ovol.
We know that Socrates mocked the affected speeches of Greek orators
(cp. Menex. 234c-235c), we know his view that in a dialogue the answers
should be as short as possible (cp. Gorg. 449b-c; Prof. 334c-335a). If we
assume that Socrates' speech in the Meno is not meant to be ironical, we have
to explain the contrast between what he says and what he does, because his
speech is quite evidently constructed in the manner of a rhetorical epideictic
logos, and is delivered in lieu of an answer to a short and meaningful
question. It seems to be much more reasonable to assume that Socrates is
deliberately acting the r61e of a sophistical orator - - as he does in his two
speeches in the Phaedrus and in the great speech in the Menexev~us. That
this is not stated explicitly in the case of our dialogue has an obvious
reason : Meno is to be taken by surprise by a Socrates who suddenly shows
himself to be well acquainted with the tricks of sophistical rhetoric. Ir

V
The literary character of the passage 81a5-d5 has been misunderstood
because this speech has been taken to. be the enundation of a dogma of
Platonic metaphysics : the Theory of Recollection. Indeed, such a dogma
s

177

THEODOR

EBERT

seemed to require the imposing solemnity of this speech. However, reading

it as a parody of a Gorgian speech provides us with a key as to the meaning
of the concluding statement that learning is recollection. This statement, if
taken literally, would indeed be inconsistent with the subtle analogy between
recollecting and coming to know from an error which we have taken to be
the meaning of the geometry lesson. Socrates' puzzling statement has to be
read as an allegorical conceit, a stylistic device usual in Gorgias as well as
i~ Empedodes (cp. Gorg. fr. 5a; Empedocles ft. 55, 66, 99 Diels/Kranz).
The reader had to guess the meaning of these metaphorical riddles.
The sequel to the speech of Socrates confirms the view that the learning-isrecollection statement is a riddle in the form of a metaphor : Meno's first
reaction is not a doubt as to, the truth of this paradoxical statement, but a
question as to its meaning (81e3-4). What is more, the dialogue with the
slave is staged not in order to. give an answer to= Meno's question, but to
enable him to find one. This, however, implies the recognition of the
legitimacy of Meno's question as to the meaning of this statement.
The geometry lesson which is arranged for Meno's sake (~o,3 Z~exa 82a8)
aims at correcting Meno's misunderstanding of predicament by correcting
his misconception of knowledge. The geometrical problem is to find a
square, or the side of a square, which stands in a certain ratio to a given
square, or the side of the given square. The error of the slave consists in
assuming a false proportion, in taking two different ratios to be the same;
he believes that 22 : 42 = 1 : 2. The recognition of his error means, in
modern mathematical terms, that the magnitude looked for becomes an unknown, an x within an equation (22 : x 2 = 1 : 2). And the discovery of the
solution consists in seeing that, and of course why, the diagonal of the
original square is the unknown magnitude looked for.
The twist in this discovery of the solution now seems to lie in the fact
that the diagonal is not something unknown in the sense of not having been
seen as a line in the figure. Socrates had drawn the diagonals at the beginning
of his lesson in order to define the square (80C2-3), not the lines parallel to
the sides running through the centre of the square, as Thompson, Lamb,
and Bluck have argued. '18 The slave was acquainted with the diagonal before
he started to look for the solution, and his final discovery of the solution
implies seeing the diagonal as the side of the double square.
This fact now makes it clear that the structure of togos, in its double
sense of proportion and proposition, not only allows for error, predicament,
178

PLATO'S THEORY OF RECOLLECTION RECONSIDERED

and knowledge, but in certain cases provides the possibility of going through
these steps without acquiring any new information from without. It is a
characteristic feature of logoi (proportions as well as propositions) that they
combine different elements without blending them : their elements remain
recognizable and can be disconnected and combined in a new way and with
other elements. This is also the reason for Socrates' being right when he
claims that he is not teaching but only questioning his interlo.cutor : he is
merely giving maieutic assistance in such a process of combining and disconnecting logoi.
The only one to see this "didactic" strategy of the geometry lesson,
although rather incidentally, has been Leibniz. In the first book of the Nouveau Essais, he makes the point that mathematical truths are innate to our
mind in an implicit manner, "en sorte qu'on les y peut trouver en consid6rant
attentivement et rangeant ce qu'on a dSj~ dans l'esprit ... comme Platon l'a
montr8 dans un dialogue, o~ il introduit Socrate menant un enfant /t des
v~rit~s abstruses par les seules interrogations sans lui rien apprendre" (italics
are mine. - Th. E.).19
The geometry lesson is staged in order to correct Meno's misconception of
knowledge. Does Meno, learn what this lesson is designed to make him
admit, and if not, why not ? We get an answer to this question, if we
consider the questioning o,f Meno, by Socrates about the slave-boy's geometrical discovery (85b8-86b4). At first sight, in this part of our passage the
attempt is made to deduce that the boy must have learned his mathematical
knowledge in an earlier life. What is decisive, however, is that Socrates in
this passage is merely asking, not making statements, just as he did in the
geometry lesson. It is Meno. who concedes the premisses from which the
mythological interpretation of the slave's learning is deduced. In particular,
it seems to be worth noticing that the decisive question as to whether the slave
had his opinions beforehand, is put twice (85c4, eT) and both times as an
alternative question. Meno's replies are always in the affirmative. Socrates,
however, does not take Meno's conclusions at their face value, as is shown
by his reply to Meno in 8666-c2 : he merely wants to. make the point that it
is possible to, look for something one does not know. He has now reached an
agreement on this point with Meno (cp. 86c4-5) who, however, has been
brought to concede this from quite different, i.e., mythologica! premisses.
It seems to be plain that Meno does not realize the import of the geometry
lesson and therefore of Socrates' statement that learning is recollection. But
179

THEODOR EBERT
once again this misunderstanding ought not to be interpreted as merely a
transitory mistake, as a mere missing of the point of Socrates' comparison.
It seems rather to be a consequence of his misconception of knowledge.
Since Meno in the geometry lesson did not go through a process of "recollection" as did the slave, since he has not come to, see and correct his error
about the nature of knowledge, the consequence of a mythological interpretation is quite natural. In the case of knowing by acquaintance, a progress that
leads from a lack of knowledge to knowledge and is not based on outside
information, is possible only in the case o,f recollection, through which we
can come to know something we did not actually know merely by ourselves.
It is Meno's misconception of knowledge that causes him to persist with a
mythological interpretation of the slave's learning and of the "Theory of
Recollection." We should no longer, I think, follow him in this.

NOTES
1 The following article is, in its main argument, part of my doctoral dissertation which is to
he published under the title: Meinung und Wissen in der Philosiphie Platons. I am much
indebted to Mr. Michael Petry for many improvements in the English text.
2 See Cherniss' review of Klara Buchmann : "Die Stellung des Menon in der platonischen
Philosophie," i n : A]Ph 58 (I937) 498; also P. Friedl~inder, Platon vol. II, Berlin I964, 343.
The core of Cherniss" argument is based on a petitio principii : "Yet obviously it is while
disembodied that the soul got its knowledge so that what it 'saw' could be only nonsensible;
apd, since in the Meno 'to know' is admittedly to know the ~(~Og*, the ~ [ ( ~ that the
soul has known must be nonsensible." (Cherniss foe. cit.) There is no statement in the Meno
that would support Cherniss" claim that "in the Meno 'to know' is admittedly to know the
~Og
." C. Huber who, like Cherniss and Priedl/inder, takes recollection in the Meno to
he a recollection of Forms, concedes that "hier yon dem Ideen nicht ausdriicklich die Rede ist'"
(C. Huber, Anamnesis bei Platon,, Miinchen 1964, p. 516; cp. also N. GulIey, Plato's Theory
of Knowledge, London ~96a, p. z9).
See N. Gulley, Plato's Theory of Knowledge, p. 22.
4 See, for example, P. Natorp, Plates Ideenlehre, 5rd edition, Da,rmstadt I96I, p. ~xf., 142-144;
A. Stewart, Plato's Doctrine of Ideas, and edition, New York 1964, p. 26; R.E. Alien,
"Anamnesis in Plato's Meno and Phaedo/" in : Rev. of Met. z3 (:t959/6o) I7o; N. Hartmann,
"Das Problem des Apriorismus in der platonischen Philosophie," in : N. Hartmann, Kleinere
Sehriften, Berlin I957, PP. 48-85 (first published in 1935).
5 See G.W. Leibniz, "'Nouveaux Essais,'" i n : Leibniz, Die Philosophischen Schriften (ed. by
Gerhardt) vol. V, p. 74f- See also Diseours de M~taphysique, vol. IV (Gerhardt), p. 45z f.
6 Natorp, Ioc. cit. p. ~45.
The problem connected with the literary character of Plato's dialogues has been discussed
since Schleiermacher and C.F. Hermann revived interest in it. For recent arguments concerning
this topic see Leo Strauss, " O n Plato's Republic," i n : The City and Man, Chicago ~964;

180

PLATO'S

THEORY

OF

RECOLLECTION

RECONSIDERED

Jacob Klein, A Commentary on Plato's Meno, Chapel Hill 1965. The problem is discussed
extensively in the "Introductory Remarks" in Klein's book, which contain a critique of R.
Schaerer, La Question Platonicienne, NeuchStel 1938. See also the "Introduction" in Stanley
Rosen, P~ato's Symposium, New Haven/London 1968.
8 For a discussion of the mathematical background of this lesson see the illuminating article
by M.S. Brown, "'Plalo Disapproves of the Slave-boy's Answer," in : Rev. of Met. 21 (i967)
57-93. However, Brown does not examine the connection between the geometry lesson and
Socrates' speech in 8aa5-d5.
9 See R.S. Bluck, Plato's Meno, Cambridge I964, pp. ~6 and 297.
10 See Gorgias fr. 5 (Diels/Kranz).
11 Bluck in his commentary ad Ioc. says that "it is impossible to say who originated this

~ t G ~ l G O ~ ~()~0~... One thinks of the Megarians, and in particular of Enbulides." J.

Moline, "Meno's Paradox ?," in : Phronesis I4 (~969) 155-I6~ argues tha,t Meno's questions are
an argument ad h ominem to which Socrates replies with a philosophical paradox. Cp. too
B. Phillips, " T h e Significance of Meno's Paradox," i n : Class. Weekly 42 (1948/49) 87-9~.
12 Cp. to this passage P.M. Cornford, Principium Sapientiae, Cambridge I952, p. 52 : "The
dilemma assumed that the only choice is between complete knowledge and blank ignorance."
18 The 'speech' itself is finished at 8Id5. The following lines are a comment upon it.
1,4 Cp. Eduard Norden, Antike Kunstprosa, Leipzig, ~898, vol. I, p. 64.
15 Cp. Norden, op. cit. p. 75 ft.
16 Cp. Bluck op. cir. ad Ioc.
17 It should be noticed that in the foregoing discussions Meno himself has made two short
epideictic }d)~Ot in the manner of Gorgias : at 71ei-Tza5 (a passage put among the Gorgias
fragments in the Diels/Kranz edition of the Presocratics) and in 79e7-8ob7.
18 See E.S. Thompson, The Meno of PlatG London ~9Ol ad Ioc.; Lamb's footnote in his
translation of the Meno, in vol. IV of the Plato edition in Loeb's Classical Library; Bluck
op. cir. ad Ioc. The text in 82c2-5 allows for both interpretations. What is decis4ve, however,
is the mathematical context. Socrates, in 82b9-c5, is giving a step-by-step definiton of the
square, He starts from the notion of the quadrangle. (~E~:Q(~O)~OV) in 82b9-Io, proceeds
to the notion of the rhombus (quadrangle with four equal sides) in 82bl0-c2, and by
further specification arrives at the geometrical notion of the square (rhombus with two
equal diagonals) in 82c2-5. If we take, with the commentators, the lines referred to in 82c2-3 to
be the middle parallels (the transversals joining the midpoints of opposite sides), the
signifi6ance of this question becomes obscure and ununderstandable. First of all, Socrates"
question would be trivial: in any rhombus the lines in question do have the same
length. Secondly, and this is more important, Socrates would not h~ve given a definition
of the geometric figure dealt with in the following lesson. Up until then, he has defined
a rhombus, not a square - - as Bluck has rightly noticed (op. cir. ad Ioc.). Furthermore,
the slave would not be acquainted with what is most essential for the final discovery of the
solution, namely the fact that in a square (as in any rectangular quadrangle) the diagonals
must be of equal length. The figure drawn in 84e-85a is a square because its four sides are
formed by the diagonals of four squares, each of which has the size of the original square
and because its diagonals have each the double length of the side of the original square.
I do not think that Bluck is right in arguing (op. cir. p. 294) that the lines in question
cannot be the diagonals because "diagonals, (~ta/AF,~QOt are mentioned for the first time
at 85b4.'" There they only get their scientific label.
1.9 G.W. Leibniz, "Nouveaux Essais," i n : Leibniz, Die Philosophischen Schriften (ed. by
Gerhardt) vol. V, p. 74.

181