01 September 2005

Eristic, or Deep Nonsense?

Christopher Shields has a review in NDPR of Gail Fine’s collected essays, Plato on Knowledge and the Forms (see the full review here), which focuses mainly on the paradox about inquiry in the Meno. Shields poses the question: if the paradox involves fallacies, and Plato (as it seems) is aware that it does, then why does Plato bring in the doctrine of recollection to answer it?

Here is Shields’ representation of the paradox:

(1) For all x, either S knows x or not;
(2) If S knows x, S cannot inquire into x;
(3) If S does not know x, then S cannot inquire into x; hence,
(4) For all x, S cannot inquire into x.
If S is any arbitrary inquirer, then for any x unknown by S, S should not waste her time with x.

And here is Shields’ attempt to show that it involves fallacies:

…if (2) has any chance of being true then 'know' must mean something like 'knows all about x', whereas if (3) has a prayer, 'does not know' must mean not 'does not know all about x' but rather 'does not know anything about x' with the result that either (2) is true and (3) is false, or (3) is true and (2) is false, or, if we give both (2) and (3) true readings, then (1) is false and not the instance of the excluded middle it may have first seemed to be.

To simplify: Shields claims that the paradox commits the fallacy of false alternative. It is false that we either know a thing or we fail to know it. Things have parts, and thus we can know them in part and reasonably wish to inquire about the rest.

But this doesn’t really get around the problem, does it? For instance, suppose one holds (as it seems Plato does) that true objects of knowledge cannot have parts?

Or, even if we granted that they did, the problem would break out all over again, as regards each part. You either know that part, or you do not. Assume that you know it. Then you know all about that part. But then you know, in particular, the whole of which it is a part and to which it is connected—and then, once again, there would no need for inquiry. (And if you retort that you knew only a part of that part, an infinite regress results.)

So perhaps Plato reasonably thought that he needed something like the doctrine of recollection to resolve the paradox? Of course, the paradox would still, for all that, be a piece of eristic, insofar as it was proposed with the intent merely to dazzle or stun.


Thornton said...

At the risk of sounding naive, Meno's paradox about inquiry is an eristic "debater's thing" which Meno the character utters, and which Socrates can show is wrong without any appeal to a full-blown metaphysical doctrine of recollection, viz. simply by having a slave boy learn something. Why in heaven's name claim that Plato endorses this so-called fallacy of false alternatives because of the appearence of the doctrine in the Meno? It's like saying that because Hamlet was suicidal, therefore Shakespeare must have been too.

Sam Rickless said...

I'm sure this is not original, but I represent the paradox as follows:

1. Either S knows what she's looking for or she doesn't.
2. If S knows what she's looking for, then she can't inquire into it.
3. If S doesn't know what she's looking for, then she can't inquire into it.
4. So, S can't inquire into what she's looking for.

The fallacy is one of equivocation. To say (P) that S knows what she's looking for is to say one of two things: either (a1) that for all X, if S is looking for X, then S knows X, or (b1) that S knows the answer to the question: "What are you looking for?" Correspondingly, to say (Q) that S doesn't know what she's looking for is to say one of two things: either (a2) that there is something, X, such that S is looking for X and S doesn't know X, or (b2) that S doesn't know the answer to the question: 'What are you looking for?" In order for (1) to be true, (P, Q) has to be read either as (a1, a2) or as (b1, b2). For if (P, Q) is read as (a1, b2) or as (a2, b1), then (1) is false. But if (P, Q) is read as (a1, a2), then (2) is true but (3) is false. And if (P, Q) is read as (b1, b2), then (3) is true but (2) is false. Ultimately, then, (1), (2), and (3) can't all be true together.

The examination of the slave boy then does two things: first, it proves the falsity of (3), when (Q) is read as (a2), thereby contributing to a dissolution of Meno's paradox; second, it provides part of an argument for the Doctrine of Recollection.

Anonymous said...


I was probably unclear. Shields (if I read him correctly) doesn't say that Plato 'endorses' the fallacy; he says that Plato surely recognized that the paradox was a fallacy and wonders, then, why Plato dealt with it as though it needed a full-blown metaphysical response.

It's true, as you say, that the slave-boy shows that  the paradox reaches a false conclusion. But any instance of inquiry (and there are lots of them) would show the same. One wants to know why it reaches a false conclusion. Shields suggests that it would have been enough to deal with it as one deals with any fallacy--i.e. expose the fallacy. And yet Plato hauls in a metaphysical theory in response.

But I think in calling it a 'debater's thing', you are perhaps making another point: that we shouldn't put too much weight on the paradox. Why take it as a 'problem' which recollection is meant to answer, instead of (say) simply a provocative device that allows Plato to introduce, rather elegantly, a doctrine that he accepted on other grounds, and which he would have defended even apart from the paradox?



Posted by Michael Pakaluk

Michael Pakaluk said...


The difference between (a1, a2) and (b1, b2) seems to involve extensionality versus intentionality. If so, I don't think that you've represented the argument as a 'fallacy of equivocation' but rather as a worry about this distinction. To me, this seems right.

It seems that one needs to use definite descriptions in giving a logical paraphrase of 1: what's at issue is whether when S is inquiring, S knows "the thing S is looking for" or not. (Doesn't Quine deal with these kinds of paraphrases in W&O?)

To wit: it's not correct to construe "S doesn't know what she is looking for" as "There is something, X, such that S is looking for X and S doesn't know X", that is, as "There is a least one instance where S is looking for something and she does not know it." What the speaker of the paradox means is that it never happens that S knows the thing she is looking for. Thus, your (a2) is too weak; it can't be that the slave example is meant to show that (a2) is false.


David said...


I think it may be more accurate to say that Soc assesses Meno's mental abilities, then introduces both the paradox and the "myth of recollection" as a response. Meno is hardly a self-reflective chap and seems not to even know that he doesn't know so Soc has to try to get him to recognize his own ignorance. Unfortunately, Meno really isn't the type of person who can sustain a careful discussion regarding our soul's characteristics so Soc gives him a bit of a fiction and leaves it at that. The argument that Soc somehow endorses the myth stretches credulity since it assumes that Soc couldn't see the problems with it. If Soc is as intelligent as we claim his is, then we have to believe that he knew what he was saying and where it goes wrong. The myth is really just an ad hoc construction in response to a poor student's questions.

Michael Pakaluk said...


Your comment that Meno "seems not to even know that he doesn't know" is helpful. It's useful to think about the Meno paradox as placed in the group of concerns involving 'knowing that you don't know', in the Apology, Symposium, Lysis, etc.


Anonymous said...

Two unrelated points:

1. I wonder if it isn't right to expand "S knows X" to "S knows what X is"? Then, given an idea which Socrates introduces earlier on and which Meno allows, namely that if you don't know what something is then you can't know anything else about it, everything looks OK.

2. We might take Michael's post as suggesting that we supply an implicit assumption to the effect that the objects of knowledge are simple. Another context in which such an assumption might be implicit is the handling of the problem of false belief in the Theaetetus, where it is also taken for granted that you can't know and not know the same thing.