Joshua Landy just posted a very interesting blog entry here, partly on how the fallacy of conversion tends to work in Derridean argument. I am no longer enchanted by Derrida -- I've been... deconverted? deprogrammed? -- but I was prompted to defend him in the comments to that post. That got me thinking a bit about what's effective, and admirable, about certain kinds of metaphors: the kinds that produce a sort of feedback loop between tenor and vehicle. Deconstruction, one might tweet, is almost entirely the notation and analysis of feedback loops in the texts it paraphrases through free indirect discourse. That is, it notes and makes very heavy weather of the hermeneutic circle. Since the feedback loop always occurs after the circuit is closed, those loops are more than the sum of their parts. If I say of A that it is B, B redounds upon A, and the original A I meant to indicate is changed once I say of it that it's (non-trivially) B.

Let me be opportunistic for a second, by quoting Landy discussing one of his examples of the fallacy of conversion:

You can have a lot of fun with this fallacy, ‘proving’ all kinds of delightful things:

3. All bananas are fruits.
therefore
4. All fruits are bananas.

Now it actually matters that statement 3 is false. Bananas are not fruits, but herbs. And any conditional whose premise is false is true. So it is true that if all bananas are fruits then all fruits are bananas.

Obviously that's not what's at issue here. But if you take a stronger (or weaker) sense of falsity here -- the kind of falsity that (as Donald Davidson has remarked) all metaphors exhibit -- then the claim that all bananas are fruits means that one or the other word is being used metaphorically. (This is an ordinary and non-vivid metaphor that tracks our lumping of everyday objects into naive kinds. It's a metaphor none the less.) Either bananas are fruits because a) the word fruit is being used metaphorically to indicate a sweet, pulpy, discrete object that we pluck and eat for desert, or they're fruits because b) the word banana is being used metaphorically to indicate a sweet, pulpy, discrete object that we pluck and eat for desert. (Clearly these kinds of indications are different, but one morphs into the other.)

If we take the metaphor as in a), then it would be clear that not all fruits are bananas. But if we take the metaphor as in b), then all fruits could be called bananas, that is things that we call fruits, as we did in 3, whether they are fruits or not, because they're sweet, etc. To make this plausible, someone might say, "Guess what: strawberries are not a fruit," and you might reply, "Sure they are -- they're, uh... they're bananas," by which you'd mean that since the whole world regards bananas as fruits it's just as reasonable to regard strawberries (and figs) as fruits too, as fullfilling the intensional fruit-criteria embodied in bananas.

Anyhow, to get back to the feedback loops in interesting metaphors, let me try out a couple of speculations, partly based on how it was that I once felt exhilarated and exalted by reading Derrida. How could the weird things Derrida (or any hard thinker or poet) said come to carry conviction for me? Put more narrowly, how could I accept his logically invalid conversions?

Accepting them requires the Quinean principle of interpretive charity, whereby we ought to (and tend to) understand something as meaning what it's likeliest to mean given what it says. Sometimes, when even likelihood is somewhat far-fetched, our charity can be stretched into new insight. The way we learn what we don't know is by having to work to understand -- and we do the work out of the principle of charity. The hard part is stretching the reader's charity without causing it to snap back into skepticism and disgust.

Derrida's use of "of course" in a sentence Landy quotes as an example of an unsupported conclusion ("the figure of the father, of course, is also that of the good") is actually an appeal to intuition, not a marker of logical consequence. All appeals to intuition require charitable interpretation: it's your intuition which does the work of seeing how this assertion might be thought of as patent when it isn't patent.

Let me pause to say that I regard the local defense of Derrida I'm mounting here as being also a critique of the method of deconstruction, which simply rejects the principle of charity wholesale. So I'm saying why deconstruction can yield insight, at its best, even while I have come reject its methodology and the general conclusions such a methodology almost always reaches.

I think one way that a metaphor can be good -- especially in literary language (like Derrida's) -- is to make this appeal to intuition. Derrida's example, in the passage Landy critiques, is part of a long analysis of Plato's calling a speaker of a an utterance the "father of his logos." Derrida's point is that our idea of fathership is challenged by this metaphor, not just our idea of the logos.

Whether or not this is true, it got me thinking about metaphors of the genitive form. While it may be that it's easy enough to see that the "blue waters of heaven" refers to the sky (in Pound's poem "Cino"), even there there's some feedback. And when Jesus calls himself "the light of the world" this affects how we think about both terms, "light" and "world." It's not the case that one indicates tenor and the other vehicle. If Babe Ruth is "the sultan of swat," his lazy power is indicated by the interaction of the two metaphorical words. When Swinburne calls the sea "mother and lover of men," he's doing a lot to our conception of both male adults (it matters that they're male) and of mothers and also of the singularity of men's ideas or memories or anticipation of the maternal. In all these cases, the metaphor cuts both ways, and that's what makes it powerful. (There's a type of ambiguity -- I have not my Empson about me -- which is very similar.) Lachrymae rerum says something about both tears and things.

One more example of this genetive metaphor (subjective genetive this time): Milton calls time "the subtle thief of youth." This is a double metaphor, or at least a book-ended one: time is a thief and what it steals is youth. Let's say the two major terms are those in apposition in the original line: time is the thief of youth. That's a relation you can't understand or parse without seeing how each term in it affects the others. How we think of time is determined by the fact that it is the thief of youth; but what makes something a thief of youth is that it has the characteristics of time, since youth is subjected to its depredations. You need to balance all three to get the meaning of the compound metaphor, and the cast of mind that can do that balancing to parse that meaning is the cast of mind appealed to by that arguments which suggest that feedback loops are in play.

Maybe one more example can clarify what I mean: when Stevens refers to "the sun, that brave man," the metaphor or predicate is one which affects our sense not only of the sun but of the man looking wistfully at the sun, and also of what bravery would be.

The sun is a brave man: would that mean (by conversion) that every brave man is the sun? Of course not. But it would mean that every man can be brave the way the sun is brave. Calling the sun brave, given the fact that the sun has no choice but to be what it is, therefore no choice but to be brave, is saying that all of us who have no choice but to be what we are can be brave like the sun. So there is an insight by conversion here: the sun is brave enough to be brave like a brave man, so if we are brave we can be brave like the sun, who is brave like a brave man.

[Edited to concede: bananas are fruits (non-metaphorically). See the comments. I mistakenly believed that most people were mistaken in thinking that they are fruits. But at least they weren't the fruit that Adam and Eve plucked, since bananas don't grow on trees.]