09 April 2005

Off to NYC

I'm off to NYC this morning for the NYC Colloquium in Ancient Philosophy this afternoon. But here's a brief comment on Annas' essay in the Leiter volume.

Her strategy seems to be: identify an objectionable way of doing ancient philosophy, popular in the '60s (she uses the literature on the Third Man, which Vlastos originated, as her stalking horse), and then take the reaction against that approach as determining the direction of the field for the next century.

Some initial questions, already posed by commentators:
1. Why identify the career of ancient philosophy within analytic departments with the career of ancient philosophy simpliciter?
2. Is this 'new direction' taken within analytic departments a discovery so much as a (rather late) recognition of the soundness of the approach that all along was favored by others?(Friedlander and Ross have been mentioned; one might add the Louvain school.)

But what I want to ask now is: What was so bad about the approach of the 60s? (Please hold your eggs and rotten tomatoes!) No, what I mean is: we need to identify accurately what was wrong with it, and I'm not sure Annas does this, although she does acknowledge the need: "In recent years we have seen a rejection of the plain ahistorical approach. This of course forces us to be more specific in defining what is meant by 'ahistorical' here"(29).

What does she say?

Annas first observes, sensibly, that rarely do cultural or historial circumstances bear on the interpretation of a philosophical text. (I don't suppose the Peloponnesian War might affect the soundness of the One Over Many.) Then, after summarizing the Third Man objection and the literature following Vlastos' famous article, she begins to set out her criticism:

All of this presumes that the 'theory of forms' is a defining part of Plato's thought, and that the issue of whether it can be held despite the 'Third Man' argument is one the answer of which is going to determine the central issues about Plato's philosophy (31).

But why isn't that a reasonable presumption?
Yet the 'Third Man' does not remain as a pressing unsolved problem. Rather, we can now see that a great deal of the time and fuss was off the point, and that the argument yields far better philosophical understanding if instead of plucking it out of its context and treating it as though it were straightforwardly a modern argument attacking a modern kind of theory, we try to put it into its broader philosophical context (31).

But I don't understand: aren't arguments such as the One Over Many, and the Third Man as directed at this, meant to have weight on their own? (No need to call them 'theories', just arguments.) And doesn't Aristotle, not a 'modern' philosopher, deal with them in the way that Vlastos did? Why is a concern with the argument a 'modern' approach that comes from our distance from Plato? (How could 'modern' or 'ancient' affect whether these are arguments, or whether if so they are sound or not? Would the passage of time render De Int 9 a non-argument? Or Word and Object chapter 2?)

She then says that acknowledging the context of the Third Man "involves...recognizing that the Parmenides is a dialogue, not a philosophical treatise". (Hard to tell from the second half of the 'dialogue'!) From this, the lesson to be drawn is that:

Plato presents us with arguments in a way designed to encourage us to pursue them as arguments. Only if we think about the conclusions, and the way they are reached, for ourselves are we making headway towards philosophical understanding (32).

But one might have thought that that was what Vlastos was doing!

Annas then gives an interpretation of the Parmenides which she presents as if it comes with the endorsement of this new 'contextualizing' approach, when in fact it is on the same level as every other interpretation of the dialogue so far put forward:
In the Parmenides it is Socrates who is the overconfident purveyor of a position which Parmenides' questions show that he doesn't really understand. ...Once we take this point seriously, we can see why the result in the dialogue is not that Parmenides tells Socrates that his forms are no good, but rather that he tells him that his commitment to the idea has been premature: he must train himself further in various kinds of positive and negative abstract argument before he can claim to have understanding of it (32).

Interesting interpretation, worth thinking about. But even if it is true, then, after we've done our dialectical homework... Do we say that the Forms exist or not? It looks like we eventually need to consider the soundness of that objection.
The fate of the copious literature on the 'Third Man' gives us a cautionary tale as to what happens when an ancient text is seized upon too unreflectively and treated as though we had a confrontation of theory with counter-argument like those in analytic philosophy journals (33).
The 'fate' of the literature, to which Annas refers, is that grad students and professors don't get impassioned about the force of the Third Man any more: it 'does not remain as a pressing unsolved problem' for us. But then why isn't this a 'cautionary tale,' rather, as to how our concern for the soundness of arguments and philosophical truth is not as great as that of the generation that preceded us? Why aren't they admirable, and we lamentable?

I don't see the argument, or reasons, for the lesson that Annas draws. It looks like an inference from influence to merit. Might makes right, for philosophy.


David said...

It isn't a reasonable presumption because there isn't a theory of forms. Even someone like Grube acknowledged that the "theory" is never articulated in the same way twice and that the lacunae regarding the relationship of the gods and eros to the forms is opaque. If Plato was a dullard we might allow that he never saw these problems but, obviously, he wasn't a dullard so we have to take these dark spaces seriously. We might also quibble with the utilization of the word "theory" when that didn't mean the same thing in Greek that it does in English for, when we say "theory", we seem to be referring to something more in line with a "paradeigma" or, maybe, a "logos."

Michael Pakaluk said...
This comment has been removed by a blog administrator.
Michael Pakaluk said...

That a view is not articulated always in exactly the same way doesn't count against its being a 'theory' in some reasonably broad sense of that term. (I guess Hume had no theory of cause and effect, because he states it differently in the Treatise and first Enquiry, and Kant had no transcendental deduction.) I'm not sure I would agree that the matters you mention are 'opaque', but even if they were, I don't think that Plato would be obliged in dialogues to say everything he holds about Forms. It seems to me that in the dialogues we find exactly what we would expect of someone who held a 'theory', and who was also (as we known) energetically creative: partial and changing statements of that view across many dialogues.

David said...

Hi Michael,

If something isn't articulated in the same way twice, how can one call it a "theory?" (NB. I don't think that either Kant or Hume held "theories" about matters in the way that we think of them today). How many elements need to be similar for us to be able to pronounce the various utterances a "theory?" When are we prohibited from saying such things as, "Plato believed blah-blah about the Forms?" In one place, Soc calls the whole logos of "forms" a dream (Cratylus 439c7). Are dreams theories? Are you discounting the language because it's inconvenient to the conventional construction of a theory? I don't ask this to be flip but how are you able to ignore the heterodox elements (eg. the very presentation of the "theory" in various dialogues) in favor of the "theory?" More importantly, why are you able to ignore them? Why are you allowed to ignore the context of various statements in favor of the "theory?" To return to the previous example, why is the "theory" called a dream? Did Plato "progress" in his understanding? And, if we are going to go for the jugular (aka the truth) in this matter, how do you square the fact that the gods seem to be trumped by the ideai with the conventional assertion in Platonic dialogues that they are divine? If something transcends them (eg. the gods), are they still divine or do they simply occupy a lower plan of divinity--albeit one not articulated by Soc or any other Platonic protagonist--than the ideai? Why didn't Plato flesh this out fully? Was he just blinkered? Again, I'm not asking this to be polemical or flip but, if someone as conventional as Grube can admit that there isn't a consistent articulation of a "theory of forms", how can we assert that Plato ever held such a thing? Because, at certain points in certain dialogues, the logoi appear similar? What do we do when the logoi don't appear similar? Are we guilty of priviledging Aristotle's account because we don't want to bother with doing the work of tackling the dialogues themselves? Are men such as Charles Kahn and Kenneth Sayre just wrong when they claim that there isn't such a thing as a "theory of forms?"

Sam Rickless said...

Reply to David: If Kahn and Sayre say that there is no theory of forms, then they are mistaken. In the middle dialogues, we see a method of investigation (the method of hypothesis) which is well designed to yield a theory in the modern sense (with axioms and theorems and auxiliary hypotheses). You start with a hypothesis that seems plausible (e.g., that it is by virtue of partaking in Beauty that beautiful things are beautiful), and then consider whether this hypothesis yields unacceptable results. If it does, then it's disconfirmed. If it doesn't, then it's (provisionally) confirmed. If the hypothesis itself follows from other hypotheses that seem plausible, then it receives further confirmation. The end result of this process is the creation of a system of logically interrelated propositions, i.e., a theory.

Reply to Michael: I agree that Annas draws the wrong conclusions from the fact that the huge TMA discussion that was started by Vlastos seems to have petered out somewhat. As she sees it, the fact that graduate students don't struggle with the TMA in the way their 60's mentors did suggests that a different approach to the TMA might be more fruitful, presumably one that does not involve as much in the way of logical reconstruction. (As Annas says, we should think of the Parmenides as a *dialogue* rather than as a *treatise*.) I understand Annas's frustration at the fact that the discussion seems not to have yielded any firm results. But I think the prescription (if it is in fact different from what Vlastos et al. were doing--and, as you say, it's not clear that it is) is worse than the problem it's supposed to solve. In my view, the TMA literature got stuck for two different reasons: first, because it did not extend the method of logical reconstruction to the other arguments of the dialogue as much as it should have; and second, because it was actually based on a misreading of the TMA passage itself (particularly the last sentence, which was clearly mistranslated in Vlastos's first 1954 paper). If we remedy these problems by logically reconstructing the other arguments of the dialogue in a way that respects the text as written, the mystery of the TMA (in fact, the mystery of the whole dialogue) disappears.

David said...


The classification of certain dialogues as "middle" is buttressed by the fallacious argument that we can date the dialogues based upon the internal evidence. Different teachings to different interlocutors isn't an indication that Plato changed his mind. It also ignores the very structure and dramatic content of the dialogue. For example, we are allegedly given a "theory of forms" in the Phaedo but the person telling the tale, Phaedo, is so forgetful that he doesn't even remember that Plato was present. Why can we priviledge his account of the "forms" over the Cratylus' statement that a logos of such a thing is a dream? And the Philebus speaks of
such things as a road. Is a dream a theory? Is a road a theory? Why are we allowed to priviledge the Republic and the Phaedo over the Phaedrus, Symposium, Philebus, and Cratylus?

You mentioned a "method" employed by Soc. What, precisely, is that "method?" All I see is a man asking questions and telling stories to his interlocutors.

Sam Rickless said...


1. What buttresses the claim that we can identify the Phaedo and the Republic (II-X) as belonging together is not only the internal evidence (that the protagonist, Socrates--not Phaedo, offers similar views for similar reasons), but also the testimony of Aristotle.

2. Different teachings to different interlocutors is at least an indication, though it is of course not unassailable proof, that Plato changed his mind. The reason is that, because of Aristotle's testimony, it is reasonable to suppose that Plato's protagonists speak for Plato himself. (Aristotle attributes to Plato doctrines that show up in the mouth of his middle period protagonist, Socrates.) And some of what is said by the protagonist of one dialogue is logically inconsistent with some of what is said by the protagonist of another (when the dialogues are identified as belonging to different periods).

3. Scholars who claim that the Phaedo and the Republic (II-X) belong together do not *ignore* the structure and dramatic content of these dialogues. What they say (or should say) is that the interpretation of dramatic and literary tropes should depend on the interpretation of the relevant philosophical arguments, rather than vice-versa. I think it's dramatically important, for example, that Parmenides and Zeno exchange a knowing smile after Socrates' speech in the *Parmenides*. But I don't think we are in any position to understand why they exchange the smile unless we understand the philosophical content of the dialogue.

4. Maybe I'm missing something, but I believe that Socrates, not Phaedo, provides us with the method of hypothesis and some of the propositions that form part of the theory of forms in the *Phaedo*.

5. I'm not sure what part of the Cratylus you're referring to, but as I see it, what Socrates says in the Cratylus, particularly at the end, is consistent with Socrates' position in the middle dialogues.

6. The *Philebus* is usually understood to be one of Plato's latest dialogues, and, along with the *Sophist* and *Statesman*, doesn't belong in the middle group. The *Phaedrus*, which, like the *Sophist* and *Statesman*, endorses the method of division and collection, probably belongs in the late group as well.

7. The *method* is described at Phaedo 100a and 100d-e. Asking questions and telling stories is but the tip of a very large iceberg.

David said...


You have cherry picked what you want from those dialogues which go into some detail about the "theory of forms" but you have said nothing about all of those dialogues which don't conform with the two or three upon which you are relying. Why does Soc call it a "dream" and a "road" in other places? Why does he couch it in a myth in the Phaedrus? Why doesn't he just come out and say, "these things are forms and they have these characteristics?" And, if I may be so bold, what about the context of those more technical accounts? Bk 6 of the Republic is part of a series of books which are called a digression. Is that irrelevant in our consideration of Soc's discussion? Why? If we treated Shakespeare's plays the way we are currently treating Plato's dialogues we would end up publishing papers with conclusions like, "Shakespeare thinks that horses are more important than kingdoms in Richard 3rd but he changes his mind in Troilus and Cressida." Why don't you think that it's important that Phaedo tells us the tale of Soc's last hours? Clearly Plato thought it was important enough to put those words in Phaedo's mouth. Why do you think it's unimportant when analyzing the dialogue?

The dating of dialogues is only uncontroversial to those who automatically accept the premisses of the argument. Deborah Nails's
review of Ledger in BMCR points out that stylometric analyses of Plato's work engage in circular argumentation at their very core. Strauss and Heidegger are quick to point out that what a philosopher omits in a given work is as important as what he includes. Word counting only tells us that a particular word hasn't been used in a certain dialogue. It cannot tell us why it isn't used or if its omission is important. Therefore, stylometric analysis is useless for understanding Plato.

As Cherniss demonstrated ad nauseum, Aristotle's testimony about Plato's teachings is suspect. I wouldn't rely upon him any more than I would rely upon Hegel to give me the best understanding of Plato. Why take Aristotle's words over Plato's works?

Different teachings to different interlocutors is no indication that one has changed one's mind. I tell my kids that lying is bad but I tell "white lies" to adults frequently. Have I changed my mind or am I tailoring my message to my audience?
Why aren't you paying attention to the rhetorical elements of speech?

The only thing the Phaedrus endorses is "cutting at the joints." How is that a method? That's like me telling someone to "step on the gas" and having someone call it my "theory of driving."

I'm speaking of Cratylus 439c7. Are dreams consistent with fully articulated theories?