Plato's Theory of Knowledge (Routledge Revivals)

Norman Gulley

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781136200595
• Edition: 1
• Topic: Filosofia
• Subtopic: Filosofia antica e classica

First published in 1962, this book provides a systematic account of the development of Plato's theory of knowledge. Beginning with a consideration of the Socratic and other influences which determined the form in which the problem of knowledge first presented itself to Plato, the author then works through the dialogues from the Meno to the Laws and examines in detail Plato's progressive attempts to solve the problem.

Chapter III

Knowledge and Belief

1. Recollection and the New Method of Dialectic

The first formal exposition of the methods of collection and division is found in the Phaedrus. It is now generally agreed by scholars that this dialogue is later than the Republic,¹ It is also generally agreed that in several important respects its doctrine links it with the group of late dialogues which begins with the Sophist.² As for the question of whether it precedes or follows the Parmenides and the Theaetetus, I have already attempted to show how Plato’s new method of dialectic is designed to meet problems raised, but not solved, in the latter half of the Theaetetus, and this, in my opinion, is good reason for assuming that the Phaedrus is later than the Theaetetus.³

Together with the new method of dialectic the Phaedrus re-introduces the theory of recollection. The close link between method and theory is established at 249b–c, where Plato first describes the method of collection (synagôgê) in its application to sensible particulars, a process of generalisation and abstraction culminating in the recognition of a single common Form. This recognition is then said to be the recollection of that ‘true reality’ which the soul once knew. And though recollection is mentioned only with this first description of collection it is clear that this first description is closely related to the later description of it in 265d ff., where it is linked with the method of division; and this passage is closely related in turn to descriptions of the method in the Sophist (253d) and the Politicus (285a–b). It is to be noted too that, although it is to ‘collection’ as a process of direct abstraction from sensibles that recollection is explicitly related, Plato indicates that this process of abstraction is simply a first step in recollection, to be followed by a further stage of methodical analysis (249c–d). The obvious implication, once we take into account the relation between the earlier and later descriptions of ‘collection’, is that this further stage in the process of recollection has as its instruments the methods of collection and division. And the descriptions of these methods, in the Phaedrus and in later dialogues, suggest that fundamental in the process of recollection will be the recognition of relevant resemblances and differences between Forms in attaining knowledge of any particular Form. It is extremely important, I think, to see that there is this close association between the new conception of dialectic and the theory of recollection. Many scholars have argued that the theory of recollection is abandoned in the late dialogues. And it is essential to any examination of Plato’s theory of knowledge in the late dialogues to consider first the evidence for and against this view. This will lead to a discussion of Plato’s evaluation of perception within his new conception of dialectic. And this discussion in turn, by bringing to light the problems involved in Plato’s retention of the distinction between knowledge and belief, will lead to an examination of what Plato has to say about the distinction in the late dialogues.

We have already seen that the descriptions of the new method suggest that fundamental to the successful practice of the method in attaining knowledge of Forms is the recognition of relevant resemblances and differences between Forms. This concept of likeness’ or ‘resemblance’ (homoiotês) has an important place in the dialectic of the late dialogues. It is used not only with reference to sensible particulars which have derivative resemblances to one another as instances of the same Form, but also, in conjunction with its opposite ‘dissimilarity’ (anomoiotês), with reference to species of the same genus; and the genus itself is describable as a likeness’ in that it constitutes the common point of resemblance between its different species.⁴ And from the Phaedrus onwards this terminology is associated, within the new methods of dialectic, with the terminology of ‘kind’ (genos) and ‘part’ (meros) as descriptive of Forms.⁵ Together with these new methods and the new terminology which is used to describe them there is a significant change in Plato’s assessment of the cognitive value of perception. This change is apparent in the first place in the fact that the severe disparagement of the senses which was characteristic of the middle dialogues virtually disappears in the late dialogues. As we shall see in detail later Plato gives in the late dialogues a much more consistent and much more favourable assessment of perception and its objects. Thus in the Timaeus his evaluation of sensible’ images’ in relation both to Forms and to Space is so pointedly at variance with what had been put forward in the Theaetetus as a major criticism of perception that it is difficult not to conclude that it is designed as a correction of the views of the Theaetetus.⁶ And it is in the Timaeus that Plato explains the efficient cause of the ability of sensible particulars to function as images or likenesses of Forms, and most emphatically asserts the value of perception in suggesting the way to knowledge of the structure of the universe and of the laws which govern its movements.⁷ In other dialogues too Plato’s discussion of Visible likenesses’, taken in conjunction with his descriptions of the methods of collection and division, suggests that he is now making a much more consistent assessment than in the middle dialogues of the claims of perception.⁸ And a point of special significance which seems to be implied by this evidence is that the new methods of collection and division, in so far as they assume the complexity of the Form to be defined and assume also that the Form has perceptible instances, thereby assume the complexity of the perceptible instances of the Form. Thus these instances mirror in their complexity the structure of the Form and for all characteristics relevant to knowledge have a complete, though imperfect, resemblance to the Form. And this immediately points to the possibility of a systematic use of perception in acquiring knowledge. At the same time Plato feels that he is able to maintain the superiority of the Form to sensible particulars. The sensible particular remains indefinable and unknowable, its individual peculiarities falling below the limit of the systematic ordering which the new methods afford.⁹ It is the Form or ‘Kind’ which is definable and knowable.

But does Plato still maintain his doctrine of recollection? The doctrine is not mentioned after the Phaedrus, and some scholars argue that Plato’s new conception of dialectic and his reassessment of the claims of perception lead in the late dialogues to a modification of the theory of Forms which makes the doctrine superfluous as part of his theory of knowledge. Their arguments make a radical distinction between recollection, as an earlier technique of coming to know Forms, and division, as a later ‘technique of relation’ or of ‘specification’¹⁰ which supersedes recollection. Recollection is associated with the view that the relation between sensible and intelligible worlds is one of radical separation and that this permits no systematically ordered means of ‘mediation’, division with the view that no such radical separation exists and that the structure of ‘reality’ can be revealed through an exact analysis of class-concepts and their relations, with no further need ‘either for a transcendent dialectic or for reminiscence, which is the necessary means to it’.¹¹ Thus, on the one side, recollection is envisaged as a sort of ‘intuition’ of a simple object, effecting an immediate transition from the sensible to the intelligible world, and representing the only possible means of bridging the gap between the two; on the other side division is envisaged as a systematic means of ‘mediation’ between worlds no longer radically separate, and designed to serve the needs of a theory of knowledge which is now giving serious attention to perception as an important source of knowledge.¹²

Now as a preliminary to the discussion of this thesis it is worth recalling what sort of question the theory of recollection was designed to answer. It was the question of how one is to know that a proposition is true, and thus save an inquiry from being inconclusive (Meno 80d–81a). Similarly, when the theory is associated with the theory of Forms in the Phaedoy it sets out to show what assumptions are necessary for the Forms to be recognised. Thus in the middle dialogues the possibility of knowledge is dependent, for Plato, on the assumptions of the recollection theory.

If, then, the theory is abandoned in the later dialogues, on what alternative assumptions does the possibility of knowledge now depend? One apparent assumption, according to those who claim that recollection is abandoned, is that the method of division, provided that it satisfies in practice the formal requirements which necessarily belong to it as a particular method of analysis, itself affords a guarantee of the correctness of its determination of ‘reality’, and hence of knowledge. There is certainly no doubt that Plato was impressed by the apparently demonstrative force of the method. But there is no ground for thinking that this entails that Plato thought that by ‘mere analysis’, by a process of descending by stages of ‘pure thought’ independently of any criterion for knowing external to the method itself, it was possible to attain knowledge of the Forms.¹³ In fact Plato illustrates abundantly in practice and acknowledges in theory that the method is not, in itself, infallible.¹⁴ Thus it may distinguish, within a genus, a ‘part’ for which there is no Form, in which case the division is not a ‘natural’ one.¹⁵ Admittedly he does not elucidate further here what sorts of’parts’ have Forms, and what sorts have not, but he does at least make clear that merely to satisfy the formal requirements of the method is not sufficient for the determination of’reality’. Hence ‘the formal method alone may lead to any number of definitions of the same thing unless one has the additional power of recognising the essential nature that is being sought (Sophist 231c–232a). In short, diairesis appears to be only an aid to reminiscence of the idea.’¹⁶ One might add that it is essential to have ‘the additional power’, at every stage in the practice of the methods of collection and division, of recognising and selecting ‘real’ resemblances and differences, those which are resemblances and differences between Forms (Politicus 285b). The possibility of knowledge cannot, then, be made to depend on diairesis as a self-sufficient criterion.

And it is certainly not the case that perception now becomes the criterion for knowing. Nor do those who claim that the recollection theory is abandoned pretend that it does, though they certainly assume that the fact that perception and its objects are now valued more highly in relation to knowledge and the Forms has an important relevance to the question whether the recollection theory is abandoned. But what is this relevance? As far as the metaphysical status of sensibles in relation to Forms is concerned the thesis of the late dialogues is distinguished from that of the middle dialogues rather by its greater consistency than by any important change in doctrine. Thus the theory of recollection in the Phaedo had assumed that Forms have sensible instances which are ‘copies’ resembling their archetypes, and on this assumption had granted to perception the cognitive value of facilitating the recognition of Forms. And since Plato thus acknowledges the existence of determinate sensible characteristics he quite properly describes them as having ‘being’ without in any way contradicting the assumptions of his distinction between Being (the perfect or ‘complete’ Being belonging to Forms – to pantelôs on) and Becoming (the most appropriate comprehensive term with which to characterise the changing sensible world in distinction from the world of Being in the above sense). Any possible confusion is avoided once we distinguish, as Plato distinguished, (i) the distinction between Being and Becoming, as above, (ii) the distinction, within Becoming, between the derivative ‘being’ of determinate sensible characteristics and the ‘becoming’ which is the process of attaining this determinate state of ‘being’. This point we discussed in detail in the last chapter in examining Plato’s criticism of perception in the Theaetetus. Now these distinctions are preserved in the late dialogues.¹⁷ They do not make their appearance there for the first time as an indication of some important development in doctrine. When Plato assumes there the ‘being’ of determinate sensible characteristics he is assuming only what he had earlier assumed in the middle dialogues as part of his theory of Forms. Unfortunately his criticism of perception in the middle dialogues had led him into inconsistencies. It had led him at times to deny any cognitive value to perception and to deny too the existence of determinate sensible characteristics. It is failure to recognise this serious inconsistency in the middle dialogues which has allowed scholars to maintain that in the late dialogues there is an important change in Plato’s assessment of the metaphysical status of sensibles.¹⁸ For they fail to recognise that what Plato says in severe condemnation of perception and its objects in the Phaedo, the Republic, and the Theaetetus is inconsistent not only with whathe says of perception and its objects in the late dialogues but also with what he says about them in other places in the middle dialogues and, most particularly, with the assumptions of his theory of recollection, as presented in the Phaedo. Finally, by associating the distinction between Being and Becoming (distinction (i), above) with the middle dialogues only, as a distinction marking a severe disparagement of sensibles, and the distinction between ‘being’ and ‘becoming’ (distinction (ii), above) with the ‘maturer doctrine’ of the late dialogues, they wrongly assume an incompatibility between distinctions (i) and (ii). Once these confusions in their thesis have been noted, it is possible to see that all the positive evidence which they adduce for the view that in the late dialogues Plato assumes the ‘reality’ of determinate sensible characteristics is simply evidence that in the late dialogues Plato is underlining more emphatically and maintaining more consistently than in the middle dialogues what was assumed from the start in his theory of recollection.¹⁹ Hence it is evidence irrelevant in itself to the question of whether or not recollection is abandoned in the late dialogues.

There is one passage, however, which demands special consideration, since the manner of its presentation of the function of the method of division and its subsequent analysis of ‘being’ in terms of the Pythagorean concepts of Limit and the Unlimited have made it the passage most commonly appealed to as evidence for the advance in Plato’s theory which removes the doctrine of recollection from the late dialogues. For it is argued that it is most clearly in this passage that the method of division can be seen to be a method which ‘bridges the gap’ between sensibles and intelligibles, and here too, in the analysis of the generation of ‘being’, that the problems raised by the ‘separation’ between the two worlds can be seen to have been removed. The passage is in the Philehus (15a ff.). The method of division is here presented as one which supplies ‘intermediaries’ between the genus and the infinitely numerous particulars. The problem raised at the beginning of the passage appears to be that of the consistency of postulating the Form as a One, and yet subjecting it to division and making it a Many.²⁰ But Plato goes on to state that the problems of postulating the Form as a One are: (i) do these Ones or monads really exist? (ii) can they retain their unity and permanent identity if they ‘come to be’ in an infinite number of particulars?²¹ The method of division is then introduced as a method which will illustrate the ‘one-many’ distinction in a form which will prevent the ‘one-many’ problem from confusing the proposed discussion of the nature of the genus pleasure and its varieties.²² And it is introduced here as a method of classification, not of definition in terms of genus and specific differences. What the method does, says Plato, is to show precisely what and how many species there are of a given genus. The initial division of the genus into species is followed by subdivision of these species, and the process is continued until no further subdivision is possible, the ‘total number’ of species within the genus having now been discerned. This number is finite, and the species making up this finite number are described as ‘intermediary’ between the One genus and the infinitely numerous particulars.²³ Plato, making use here of the Pythagorean concepts of Limit and the Unlimited, associates Limit with the finite number of species of the genus, the Unlimited with the infinite number and variety of particulars. And in saying that only when the total finite number of species has been determined is it permissible to ‘let each of these species pass away into the unlimited’, he clearly implies – and this is confirmed by his subsequent illustrations – that with the limits of exact specification there has been reached also the limits of the intelligibility of the particular, the ‘infinity’ of their number and of their individual differences being alike irrelevant to their determinate nature. In this analysis it is clear that the element of ‘unlimited’ is located in the realm of particulars, and not at all in the realm of Forms. At the same time it is clear that particulars are characterised by ‘limit’ to the extent that they are amenable to the systematic ordering made possible through an analysis of Forms completely determinate in structure. Thus it is possible to say that, in so far as it is assumed that the method of division makes it possible to explicate precisely the structure of a complex Form, it is correspondingly assumed that the determin...