Abstract
In this paper I explore a famous part of Plato’s Theaetetus where Socrates develops various models of the mind (picturing it first as a wax tablet and then as an aviary full of specimen birds). These are to solve some puzzles about how it is possible to make a mistake. On my interpretation, defended here, the discussion of mistakes is no digression, but is part of the refutation of Theaetetus’s thesis that knowledge is “true doxa”. It reveals that false doxa is possible only if there is a certain stock of abstract knowledge, conceptual knowledge, that is not awareness of the particular individual that is being described. The individual must be identified under some description, or seen as something of a certain kind. Error can only occur if the description applied misdescribes the situation, but then if it is to be applied falsely it must first have been known from somewhere else. So knowledge cannot be reduced to the application of descriptions to particulars, but is to be found in the prior possession of abstract descriptions that can be deployed in identifying particular individuals on the ground.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
I use the term “discernment” rather than “judgement” or “opinion”, though nothing much hangs on it. The term needs to comply with the grammar of Socrates' use of doxazein (to discern) which takes a direct object (I discern something). To discern truly is to discern what is really there and form a view of it that is true to how it is—rightly discerns what is there. What is discerned may be a quality or number or feature, but in the Theaetetus the examples are mainly things or people: individuals such as Theaetetus and Theodorus. The mistakes are typically misidentifications: seeing the person who is really Theaetetus and supposing that it is Theodorus. This is a “false discernment” or an error of mistaken doxa.
In the final part of the dialogue, Socrates considers the proposal that knowledge is doxa plus something else, namely “an account”, logos. This proposal also ostensibly fails (though some scholars have sought to excuse the failure, allowing Plato himself to approve of that definition or some version of it, which would survive the test). In my view that definition is also destined to fail because Plato thinks that doxa is the wrong place to look for knowledge, for reasons to do with the particularity of its objects. However, I shall not develop that part of the argument in this paper.
See, for instance, Ackrill (1966). Also (diagnosing a confusion about knowledge and knowing what something is) Runciman (1962), McDowell (1973, p. 196), Bostock (1988, p. 196) and Sedley (2004, pp. 121–123). All these think that Socrates is unable to solve the puzzle (though Sedley, at pp. 148–149, thinks only that Socrates is unable to solve it without Platonic metaphysics, whereas Plato has the resources). Contrast Fine (1979) and Gosling (1973, pp. 137–139) who think (as I do) that the problems are due to the TD thesis and are being diagnosed by Socrates to show up why that thesis needs to be supplemented or replaced. Similarly Barton (1999) suggests that Plato is exploring and rejecting a faulty account of thought which lacks the resources to deal with falsity. On Burnyeat, see below, note 50.
Theaetetus is a teenager. The point is not that his idea is naïve or childish, but that it reflects a certain stage of education (that is, training in geometry) at which attention to particulars is the normal route to knowledge. See note 17 and my forthcoming book, Catherine Rowett Knowledge and Truth in Plato, (henceforth Rowett K&T) chapter 7.
For Socrates as midwife, Tht. 149a-151d.
Meno 82b-c; Phaedo 75b. See also Phaedrus 249b for the idea that all humans have the capacity to think of things under classes or descriptions due to knowledge of the Forms.
Sedley (2004) takes this on board by suggesting that the Theaetetus delivers the Platonic brain child from a Socrates who lacks those resources. According to Sedley, although Socrates is portrayed as ignorant and unable to solve the problem, Plato is not thus bemused. On my reading, Socrates is not confused either, but is at least informally aware of the natural distinctions recognised by ordinary language and thought. But he is inviting Theaetetus to see what is missing, or at least to see that something is missing, by adopting the midwife approach. Theaetetus may be too far short of dialectic to get the point yet.
The reductive thesis is characteristic of the Megarians and other paradox-mongers, so Plato is at least implicitly addressing some existing puzzles in the philosophical air around and after the time of Socrates. It is also a reductive move that is still very common in young philosophers; it often manifests itself when they first encounter Plato's Theory of Forms: they wonder why there need be anything other than particulars, and “how Plato can be so sure that there are Forms”. Like Theaetetus they imagine that what we know are the particular things themselves and that we have no need of any intelligible objects besides the concrete tokens we encounter—a position which leads inevitably to the puzzles about “seeing as” and mistaken seeing as, that Theaetetus has run into.
The empiricist assumptions underlying this discussion are noted by Chappell (2004) (and see also Chappell's paper ‘Varieties of Knowledge’ in this volume). But he does not observe that the issue concerns concept-acquisition.
Treating it as a digression is common. E.g., Burnyeat (1980, p. 173). Attempts to make it relevant in some way despite its evident irrelevance to the proper topic are also common and can be found in Bostock (1988) and Sedley (2004, p. 119), Closest to my proposal (which takes this discussion to be integral to the refutation of TD and not a digression at all) are Fine (1979) and Chappell (2004, pp. 151–154).
See above, note 1.
I think “discern” is normally a success verb so I apologise for stretching the language a bit here.
Bostock (1988, p. 164) makes a case for thinking that Plato has distorted the grammar of doxazein and made it take a direct object where it would not normally do so. I agree. Plato seems deliberately to limit the discussion to discerning the identity, “what it is”, of something (i.e., a mere particular not a form—hence the term doxa and cognate verb). He is not concerned here, though he is elsewhere, with predications using the copula.
Those wedded to the idea that all such thoughts must be complex and propositional will want to say that the first part of the proposition is an indexical, as though the thought were “That is Meno”. But Theaetetus's proposal does not have room for that. He is trying to suggest that identifying an extensional item does not involve even one intensional concept, let alone two. Adding a second intensional idea, whether indexical or not, will not help here, since it is enough to discover that at least one is required. On the idea that Plato speaks of knowing things, not propositions see Bostock (1988, p. 199).
Again, discerning something is not knowing or believing a proposition, or applying a predicate to a subject. One thinks of just one thing: Theaetetus, say, or the number eleven. The thought is true or correct if that is what is there, and not if it is not. We do not need to find this satisfactory as an account of knowledge or knowing what something is (it is set up to fail).
A full defence would exceed the present task, but see Rowett K&T Chapter 9, where I appeal to issues concerning what “true” means in this passage, to explain the concern with identifying what something is rather than other truths about it. These issues are not materially relevant to the present argument except in that respect. It is worth observing that within the range of what Chappell is calling “objectual knowledge” (Chappell ‘Varieties of Knowledge’ in this volume), I am contrasting (a) the bare awareness of a particular object, (b) the awareness of a particular object as instantiating a conceptual type or description, and (c) the conceptual understanding that underpins the latter. Only the last of these is a good example of what Plato himself means by episteme.
Theaetetus is not good at following what part of his thesis has been defeated, and repeats ideas that have already been shown to have problems (first before the jury passage, when he repeats TD, and then in the third definition when he tries to add logos to a thesis that has already been discredited). Arguably this obsession with the idea that knowledge is fully achieved in the realm of doxa is part of his “geometrical” stage of education.
The TD thesis is a variant of the view of “knowing” advocated by thinkers in a tradition stretching from Zeno Vendler through Kenneth Sayre to Timothy Williamson and Trenton Merricks. These thinkers take knowledge to be of an extensional state of affairs (which is the equivalent, for propositional knowledge, of the extensional token or thing in the knowledge of things that Theaetetus is considering). See Merricks (2009), Sayre (1997), Vendler (1967) and Williamson (2000). Williamson would try to avoid the problem by omitting the word “true” from TD thesis: since knowledge cannot be false, it should not be described as “true” either he suggests. It would have to be mere discernment, and the contrast would be with failing to discern. But this leads to muddles about falsity exactly comparable to those raised by Socrates here (see the section on “false beliefs” in Rowett K&T Chapter 2).
An alternative way of thinking of this is not as reductive but inflationary: taking TD to be an attempt to inflate what is called doxa to provide all the conceptual resources that the full range of cognition requires This description of what is happening fits well for the parts of the discussion where Socrates attempts to supply some of the missing tools by deriving particular concepts and generic classes of ideas from the cognition of particular things. The inflationary reading can be found in Fine (1979). I think the reductive project is more what TD starts out to do, although it is adjusted as its shortcomings become clear. I agree with Fine that the puzzles about falsity follow from taking the TD thesis seriously, and that they reveal what problems it faces: they are the testing of the brainchild, not a digression before the Jury episode. Fine's view is rejected by Bostock (1988, p. 198) who is unable to understand how this would lead to the all-or-nothing premise (see below note 24).
E.g., an idea such as “Son of Euphronius of Sounion”, “Brightest of the Athenian youths in the geometry class” and so on (143e, 144c).
See above, Sect. 5.
It is not immediately obvious what “knowing” the thing means here. In the present passage it should be assuming the current definition of knowledge, which is under scrutiny. So “not knowing x” would be not having a true thought of that particular, and vice versa for “knowing x”. The issue is whether one can logically make a mistake about something that one truly thinks of, or about one that one does not truly think of. As we shall see, this is what justifies the all or nothing premise.
This premise has often been taken to be a mistake on Plato's part, and is attributed to his supposed “acquaintance” model of knowledge, as Fine (1979, pp. 70–71) observes, citing examples from Owen (1970), McDowell (1970) and Runciman (1962). But it is surely due to the constraints of TD, whereby all knowledge is direct factive reference to things as they are, leaving no room for gaps between sense and reference, or reference and (mis-)description. See below on Option E.
188a1-2 Οὐκοῦν τόδε γ' ἔσθ' ἡμĨν περὶ πάντα καὶ καθ' ἕκαστον, ἤτοι εἰδέναι ἢ μὴ εἰδέναι.
Theaetetus: Ἀλλὰ μήν, ὠ̃ Σώκρατες, ἄλλο γ' οὐδὲν λείπεται περὶ ἕκαστον πλὴν εἰδέναι ἢ μὴ εἰδέναι. 188a6-7. Note that οu̓δὲν λείπεται (there is nothing left, sc. for us), since other options are either excluded by the TD constraints, or irrelevant (as e.g., the learning or forgetting processes, see below). Similarly ἡμĨν at 188a1 (quoted in note 24) explains what is available to us, given the thesis we are investigating.
Burnyeat (1990, pp. 74–77) tries to distinguish (a) solutions to the faulty or misleading premise (which he calls J1, that if something is to be an object of one's judgement one must know it) that invoke tools of contemporary philosophy of language to render it harmless, and (b) solutions that show it to be false, because one can perceive a thing but not know it, using the wax block. But the wax block and other devices discussed here are really simple ways to supply the very tools that we know from philosophy of language. It is only if we think that Plato was struggling to find those tools that we should suppose he is in the grip of the harmful reading of J1 (or of the all-or-nothing premise).
For a similar recognition of the dialectical reason for the all-or-nothing premise, see Fine (1979, pp. 72, 77).
188a2-4.
Theaetetus 188b.
188b 7-10.
188d.
Burnyeat (1990, p. 78).
188e1. The change from doxazein to oiesthai is, presumably, merely substitution of close synonyms.
Note the definite article on τὸ ἕν at line 188e8. This could mean “the one particular thing we see” or it could mean “the idea of one”.
It seems likely that Socrates chooses to focus on the idea of “one” because he is playing on the idea that thinking of nothing is thinking of ouden (“not one thing”) while thinking of one is clearly thinking of hen, not of ouden. This might be fallacious, if he supposes that the mere etymology of ouden (“not one thing”) proves that thinking of one thing cannot be thinking of nothing, in a way that would not apply to any other abstract ideas or features of things besides “one”. But perhaps it is just a joke, and the proof lies rather in the idea that there must be some content to my false thought, even when I think something false. Unless I am thinking of nothing at all, the content of my thought must either be an extensional object (what I am thinking of) or some descriptive content (what I think about it). I cannot think nothing about it, and be wrong.
One might well quibble that the ambiguity of “being” is scarcely recognised in this argument, and not everything that “is not” in some sense is nothing. The discussion of falsity in the Sophist gives careful consideration to such cases. The absence of any concern about that issue here suggests that such cases, where something is F but is not Fness for instance, are simply not available here, because they presuppose a distinction between the extensional object of thought and the intensional description of it, a kind of conceptual content; which is just what is not available under TD in its basic form—although the (b) reading just described does break that constraint by hypothesising that we might think “one” about some one thing.
The allo-part of allodoxia is about the ‘‘other” or ‘wrong' thing: making a mistake is thinking not of the right one but something else. I translate it “wrong” to indicate that of two things one is right and one wrong, and the mistake is getting the wrong one. “Other-judging”, the translation favoured (over misjudging etc.) by McDowell, does not fully capture this.
See also Sophist 263e3-5.
The resemblance to Socratic dialogue is often observed. Sedley (2003, pp. 1–2) infers that the dialogues are Plato “thinking out loud”.
Theaetetus 190a.
‘It’ in the first sentence refers to ta pseudê doxazein, discerning falsehoods, at Tht 189d5. The noun dianoia (translated ‘thought’) seems to denote a part of the soul in which one thinks, or which is thinking, not the process of thinking (here expressed by the verb dianoeisthai, 189e2). By contrast at Sophist 263e3-5 the noun dianoia denotes the process of thinking.
Socrates checks that Theaetetus shares his understanding of the word “think”, suggesting that this is a technical mapping of the intellectual activities, not simple folk epistemology.
189c; 190b. Expressions such as “the beautiful” and “the cow” are ambiguous and could refer to particular cases. See Fine (1979, pp. 72–73).
Fine (1979, p. 74).
Barton (1999, pp. 176–177) makes a similar point against Fine, by imagining that Plato is discussing a kind of thinking that is like “grasping something”. As with my view, this is supposed to entail a direct relation to an object of thought, with no opacity. (I don't agree that “grasping” is the problem, but rather the directness of the grasp, and the lack of any tools with which to grasp it.)
Because this idea of knowledge of objects is inadequate, my discussion should be seen as supplementing and problematising the notion of objectual knowledge that is sketched by Chappell in ‘Varieties of Knowledge' in this volume.
See above, Option E.
194a-b.
On this point see Burnyeat (1990) (his section on “True and False Judgement in Retrospect”), pp. 119–123. He notes that the progress made by introducing the wax block would be just as great if the content in the wax block were true belief, not knowledge (since the crucial requirement for allowing falsity is the distinction between perception and another epistemic route). But this is just the point, he thinks: that nothing more than true belief has been added yet, and what is still missing is knowledge, as opposed to true belief. This is close to my reading, except in ignoring the difference between grasp of the type and mere token identification. There is still no way to turn recognition of a single item into conceptual understanding that knows how to go on for new cases. (Note that Burnyeat also sees that Plato's points are not fully captured with the tools of contemporary philosophy of language, but should be approached epistemologically, p. 121). I do not find Burnyeat's terminology (“true belief”) helpful here, but I share the idea that the content of memory is the content of previous cases of doxa.
The structure is comparable to Aristotle's description of phantasia, in De anima 3.3, and to the idea (common to Democritus, probably, and Aristotle) that one can derive an idea of the generic from repeated encounter of particulars. Plato writes (as does Aristotle) as though one first comes to think of particulars by way of the senses, rather than in thought. But in Aristotle too there are hints that thoughts can also serve. In Plato, the wax block certainly takes its imprints from thoughts as well as perceptions (191d).
Correct labelling is not at issue. We could omit that, though the label is useful in identifying the memory records, and explaining the mistake.
On the two alternative readings of Option E, see above, note 44.
I take it that if this complaint is to be relevant it must allude to a further set of cases where one can be wrong about “what it is”, given that the claim was that knowing is knowing, of something, “what it is” and that this is nothing more than doxa (discerning what it is). So far we have learned that this could suffice, if discerning involves matching the current thing with a stored image of the same thing. But if there are other cases (as the critic suggests) that will not fit that model, these now need to be addressed.
196a.
192a-193d.
See, for instance, the recollection argument in the Phaedo.
199d1-9.
This expression (as in “know Meno, who he is” at Plato's Meno 71b) is Plato's standard expression for having a certain item in one's repertoire of ideas.
See above 2.1.
See above 2.1.
194e.
199e.
If Theaetetus were right here, the whole enquiry into knowledge in this dialogue (which checks each proposed analysis of knowledge against the participants' existing concept of knowledge) would be suspect, since their notion of knowledge might be an ignorance bird, leading them to reject sound definitions and accept faulty ones. Neither Socrates nor Plato take this fear for real, although it is hinted at when Socrates ridicules the idea at 200a-b, and maybe also at 197d-e.
200a.
200b. There are two refutations here: one is Socrates' own (which shows that one would not know that one was wrong, so it doesn't solve the case we have in mind—so the fault is ignoratio elenchi); the other is a reductio by a logic-chopper (ἐλεγκτικός) who ridicules it by adding yet further mechanisms for knowing what you got wrong, which are iterations of the wax tablet and aviary, and vulnerable again to the same difficulties.
See the fuller treatment in Rowett K&T Chapter 9.
In associating his position with geometry, I am speculating beyond what is explicit in the dialogue, and borrowing some ideas from the Republic. For more on this see relevant chapters in Rowett K&T.
I would like to thank the many colleagues who have discussed these issues with me including Gail Fine, my colleagues at UEA (Faculty Forum), and those who read the paper in draft and sent comments, notably Oskari Kuusela and Anna Marmodoro. I thank the AHRC for financial support during 2011 and Merton College Oxford for a fruitful stay in Oxford, November 2011.
References
Ackrill J (1966) Plato on false belief: Theaetetus 187-200. Monist 50(3):383–402
Barton D (1999) The Theaetetus on how we think. Phronesis 44:163–180
Bostock D (1988) Plato’s Theaetetus. Clarendon Press, Oxford
Burnyeat M (1980) Socrates and the Jury: paradoxes in Plato’s distinction between knowledge and true belief. Proc Aristot Soc Suppl Vol 54:173–191
Burnyeat M (1990) The Theaetetus of Plato. Hackett, Indianapolis
Campbell L (ed) (1861) The Theaetetus of Plato. Oxford University Press, Oxford
Chappell T (2004) Reading Plato’s Theaetetus. Academia Verlag, Sankt Augustin
Fine G (1979) False belief in the Theaetetus. Phronesis 24:70–80
Gosling JCB (1973) Plato. Routledge and Kegan Paul, London
McDowell J (1970) Identity mistakes: plato and the logical atomists. Proc Aristot Soc 70:181–196
McDowell J (1973) Plato’s Theaetetus. Clarendon Press, Oxford
Merricks T (2009) Propositional attitudes? Proc Aristot Soc 109:207–232
Owen GEL (1970) Plato on Not-being. In: Vlastos G (ed) Plato I: metaphysics and epistemology. Doubleday, Garden City, pp 223–267
Runciman WG (1962) Plato’s later epistemology. Cambridge University Press, Cambridge
Sayre KM (1997) Belief and knowledge: mapping the cognitive landscape. Rowman and Littlefield Publishers, Lanham
Sedley D (2003) Plato’s Cratylus. Cambridge University Press, Cambridge
Sedley D (2004) The midwife of Platonism: text and subtext in Plato’s Theaetetus. Clarendon Press, Oxford
Vendler Z (1967) Linguistics in philosophy. Cornell University Press, Ithaca
Williamson T (2000) Knowledge and its limits. Oxford University Press, Oxford
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Rowett, C. On Making Mistakes in Plato: Theaetetus 187c-200d. Topoi 31, 151–166 (2012). https://doi.org/10.1007/s11245-012-9126-y
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11245-012-9126-y