Notes on G.E.L. Owen's 'Notes on Ryle's Plato'

First of all, Gilbert Ryle is the 'Ryle' in the title. He was a very influential philosopher of the "ordinary language" school of philosophy: some of his greatest contributions had to do with philosophy of mind. He also wrote quite a lot about Plato. In other words, he was not primarily a scholar of the history of philosophy. He was primarily a philosopher, and he found a good deal of inspiration and took a great interest in Plato.

The question Owen is addressing might be put simply as follows: "Why did Plato write the second half of the Parmenides, and what did he accomplish in it?"

The answer Owen and Ryle offer is that the Parmenides is not concerned with proving any particular conclusions. Rather, it is concerned with laying out a thick nest of problematic arguments which inform much of the 'late' Plato. That nest of problems has to do with the problems of the Theory of Forms, but more generally it has to do with problems for any theory of types in more modern terms. Russell too took it as such. The central problem here, for Russell, is that of self-reference:

Russell seems to have found his way out of the thicket by saying that sets are determined by their members, not by specifying conditions. Another way out he seems to have liked was to say that self-referring statements are meaningless. But why?

Luckily, we don't have to talk about that important matter in all its modern glory today here in this class: it's awfully thorny.

Unfortunately, we are left with an even thornier set of things to talk about: the second half of the Parmenides.

The subject of discussion throughout is "unity" or "one" or "The One." By that it seems he means something like whatever it is that is just one. He wants to try to think about one by itself with no other qualities or aspects or parts or properties. Just oneness in its purest form. Now problems arise immediately, because oneness must surely be itself one, or so it seems. So self-predication enters the scene. And oneness must be in order to "be" one, and so it has being. So it seems that oneness has two aspects: being and oneness. Each of those must be and be one, and so on and so on and so on. The infinite regresses of parts or aspects of oneness is regrettable, because it seems to show that oneness is not one, but many, not one being, but many beings, not one one, but many ones.

That's not a very satisfactory situation.

Ryle apparently thought that one result of the discussion in the Parmenides was that in the sentences, " ____ exists" and " ___ does not exist," you cannot fill the gap with "unity." To do so is to commit a mistake, because "unity" is not the sort of thing that can be a 'specific' or 'proper' concept and fill that gap. I'm not sure what sense to make of Ryle's claim, whether it is a legitimate one, but it does seem to stop up the regress if it is legitimate. I know that he is not talking about grammar or the fact that I can simply type "Unity exists." He is claiming that it is a logical mistake to say "Unity exists." What sense do you make of that?

Plato develops his discussion in four sections, each of which is split into two parts.
  1. Hypothesis: One is
    Subject of the antinomies: One
    1. 137c-142b
    2. 142b-155e
    3. Appendix on paradoxes of instantaneous change: 155e-157b
  2. Hypothesis: One is not
    Subject of the antinomies: everything other than one
    1. 157b-159b
    2. 159b-160b
  3. Hypothesis: One is
    Subject of the antinomies: One
    1. 160b-163b
    2. 163b-164b
  4. Hypothesis: One is not
    Subject of the antinomies: everything other than one
    1. 164b-165e
    2. 165e-

More detail

  1. Hypothesis: One is
    Subject of the antinomies (antinomy is a name for these opposed arguments): One
    1. 137c-142b
      1. The one is one and it is not many (137c).
        Therefore, The one cannot have parts or members or be a whole.
        Owen thinks 1Ai confuses the 'is' of identity (example: Jacques is Prof. Bailly) and the 'is' of predication (example: Jacques is confusing OR Jacques is a man). That may be one source of the self-predication that besets the theory of forms.
        This thesis conflicts with 1Bvi.
      2. The limits or extremities of anything are parts of that thing (137d).
        Together with 1Ai, this leads to the conclusion that the one cannot have limits, shape, or position.
      3. Having more than one character or attribute makes a thing a plurality (139e-140a and perhaps 138c, 139c)
        Perhaps this thesis together with 1ai proves that:
        1. One cannot change in quality or
        2. be identical with...
        3. differ from...
        4. be like...
        5. be unlike ...
        • ... itself or anything else,
        • because 1Aiiia-e all attribute an additional character/attribute to one, and so involve one in having >1 character or attribute
        But perhaps 1Aiiia stems directly from the Identity/Predication confusion in 1Ai. From 1Aiiia-e, it is argued that the One is not equal or unequal to anything and that it has no temporal attributes.
      4. If the statement that S is P differs in truth-value from the statement that S is Q, then P is different from Q (139d, 142b-c (with small variant: substitute "sense" for "truth-value").
      5. Changes (strictly speaking, movements to a place) take time: to describe S as becoming P is to describe something temporally intermediate between an initial and a final state (138d)
        From 1Aii, the One cannot move in one place.
        From 1Ai and this thesis (and probably 1Aiv as well), the One cannot change place either.
        This thesis conflicts with the argument at 1Cii.
      6. If (when) X is becoming different from Y, it cannot be the case that Y is different from X: otherwise X would already be different from Y and not merely be becoming so (141b).
        Hence, if X is becoming older than itself, it is also becoming younger than itself.
        But from 1Aiii and 1Aiv, the Once cannot have temporal properties nor can it even remain the same age as itself. Therefore, it does not exist in time.
        "If S is becoming P it cannot yet be P" is the general premise here.
      7. What exists exists in time (141d-e).
        From this and 1Avi, the One is not in any way and is not anything: i.e. the One does not exist.
      8. If S is P, then S is or exists (141e)
        From this and 1Avii, it follows that the One is not one (or not the One).
      9. What does not exist can have nothing related to it.
        From this and 1Avii's conclusion, it follows that the One cannot be named and that there can be neither speech nor knowledge, perception, nor idea of it at all.
        This thesis conflicts with 3A.
    2. 142b-155e
      This movement uses some of the steps from the last movement (1A), but it also challenges some of the same steps as it uses (namely 1Ai and 1Avi).
      • By 1Aiv, if the One exists, its existence is not the same as its unity.
      • Hence, by 1Aiii, unity and existence are pluralizing parts of the One (this conflicts with 1Ai).
      • These pluralizing parts have pluralizing parts (oneness and existence).
      • These pluralizing parts have pluralizing parts . . .
      • Therefore the One is a whole with infinite parts and unity and existence are infinitely distributed.
      • What is more, if the One and its existence are differentiable, then they exhibit difference, which is distinct from either of them (143a-144a).
      1. What is not one is nothing at all (144c).
        From this it is argued that any part of a plurality is one, that anything divisible is divisible into some number of parts, and generally that any number is a number of units.
        This points out something interesting about unity and existence: we cannot abstract them away from a subject as we can things like 'largeness' or 'redness.' The reason is that if you abstract existence from a thing, you are left with something that nonetheless must exist. So you have to immediately reimport existence. The same holds for unity. (2Ai will try to challenge this).
      2. A whole contains and so limits its parts (144e-145a).
        This amounts to a claim that the limit or limiting factor is external to, not part of, what it limits.
        This conflicts with 1Aii and also 1Biv!
      3. If X is limited, X has spatial extremities (145b).
        Hence the One has shape.
        See 1Bv.
      4. If X has parts, X is identical with the aggregate of those parts.
        This says, in effect, that X is nothing more than its parts. 1Bii said that the whole is something external to its parts. In the Theaetetus 203e-205a, these two positions are given as an exhaustive disjunction. Here they are a conjunction. It is hard to see how they both could be the case. Sophist 252e-253c has an argument that supersedes this analysis.
      5. What exists must exist somewhere (145e).
        In Greek philosophy, to be somewhere is to be in something.
        From this and 1Bii and 1Biv, it is argued that the one must be both in itself and in another. Insofar as it is in itself, it is always 'in the same,' but insofar as it is in another, it is always 'in something different.' From that, the argument goes, it follows that the One is static and in motion and that it is different from itself. That argument is a fallacy. The problem is that 'the same' and 'different' above do not relate to one thing, but rather to distinct things.
        In the Sophist 259c-d, Plato chastises those who make that mistake by failing to complete predicates such as 'same' and 'different.'
      6. Anything is related to anything in one of three ways (146b):
        1. by identity
        2. by difference
        3. as part to whole or whole to part
        From the treatment of part-whole relationships in what follows, this distinction seems to say that 'partaking in unity' or 'being in a way one' is a part-whole relationship rather than one of identity or predication. That might serve to clear up the confusion in 1Ai (and elsewhere).
      7. What is not one is related to one by b and not by a or c (147a)
        Therefore things that are not one have no unity at all of any sort. And therefore, they have no number at all. They are neither identical nor predicated by one or any number.
        That supposedly conflicts with 1Bi.
      8. Words are names, and to repeat a word is to name the same thing twice (147d).
        Thus when one claims that the one is different from everything else and everything else is different from the one, one attributes to the one and everything else the same property, namely difference. So they share one property at least. And the One is like the others too! (this is all fallacious, and obviously so: why does Plato include it?)
      9. Smallness is small, largeness is large (150a-d).
        This is apparently used to claim that nothing but smallness can exhibit smallness, nothing but largeness can exhibit largeness (150a-b).
        This is a blatant claim of self-predication. By this point in the dialogue, Plato surely recognizes the problems with self-predication. Why does he nonetheless include it in his schema?
      10. If X is becoming Y, then at any 'present time' in the process X is (and not 'is becoming') Y (152c-d).
        Hence, if the one is growing older, at any present time it is older.
        Also the one is becoming younger and neither older nor younger than the others.
        Owen points out that this step is plausible if you use something like 'older' or 'younger,' but not plausible with things like 'an octogenarian' or 'six feet tall.' Step 1Avi, however, is plausible with things like 'an octogenarian' or 'six feet tall.'
        That is important, because it kills the assumption of the Theory of Forms that there are only two alternatives: either becoming or being. It depends on the question "Becoming what?"
        This conflicts with 1Avi.
      11. What exists has (or can have) other things related to it (155d).
        This is the converse of 1Aix.
        By this step, the one can be named and spoken of and there can be knowledge, perception, and thought of it.
    3. Appendix on paradoxes of instantaneous change: 155e-157b
Now imagine filling in the same level of detail for the rest:
  1. Hypothesis: One is not
    Subject of the antinomies: everything other than one
    1. 157b-159b
    2. 159b-160b
  2. Hypothesis: One is
    Subject of the antinomies: One
    1. 160b-163b
    2. 163b-164b
  3. Hypothesis: One is not
    Subject of the antinomies: everything other than one
    1. 164b-165e
    2. 165e-