I recently got a note from a fellow critic in which he talked about having had an Eminent Cognitivist visit his class. He said the fellow was bright, engaging, and utterly blind to literature as such. He offered interesting remarks about a poem, but treated it simply as a source of examples for his particular theory. Here’s the note I sent to my colleague (lightly edited). Consider it a place-holder for a piece to be more carefully done.
* * * * *
I know that mentality very well, because I've done it myself. I first took the cognitive plunge in graduate school. I was in English at SUNY Buffalo (back in the middle 1970s), but did a lot of work with Dave Hays in Linguistics. He was one of the founders of computational linguistics as a discipline. I joined his research group, learned his system, and submitted a long paper jointly to him and to Charlie Altieri, with whom I was taking a course in modern poetry. I took some passages from Patterson, Book V, and went to town on them with Hays's cognitive network model. It was a long paper, and a good one, too. But Altieri saw right away that I was using Williams as a source of examples for the cognitive model and didn’t really get to the poetry.
I did it up large for my dissertation, where I used Shakespeare's Sonnet 129 (Th' expense of spirit) as my text. Now, that was a great intellectual experience for me. I learned a lot. I've never again done that kind of full-dress cognitive analysis of a literary text. I suppose that, back then, I could have done a half dozen more sonnets and called it a book – though I'm not sure that book could have found a publisher – but I had no inclination to do such a thing. The possibility didn't even occur to me.
But, and here's the thing, if I could snap my fingers and change the world in one way, I'd run every literary critic through that experience. But why? Not to do THAT kind of cognitive analysis of a literary text. Or rather, to do it once, maybe twice, so you can learn from it. But how could that knowledge possibly be valuable to a literary critic if it doesn't give you something you can APPLY to texts in a series of readings?
The best answer I can give to that question is, alas, a vague one. First of all, you get intuition, a sense of things. You know how they used to justify the teaching of Latin and geometry because they're good mental exercise? That kind of thing.
It gives you new intuitions about language and thought. In order to do such models you have to think about language (and thought) in a more focused and detailed way than literary criticism, even good old fashioned close reading, requires. Psychoanalysis, Marxism, semiotics, phenomenology, deconstruction, in the end they don't operate very close to language, not to the nuts and bolts and gears and levers. That's what cognitive science is about, that clanking machinery.
And that, I think, is why you can't really APPLY it to the text like you can apply all these other disciplines we've been applying to our object texts. The application of those other disciplines is a loosy-goosey affair, which is fine. It leaves room for us to apply our intuitions about the texts themselves, in particular about form (as you know, my hobby horse).
But the cognitive science of language is mostly about words, phrases, and sentences, not so much about extended texts. Yes, there is some of that, but the texts aren't literary. It's about the blocks and mortar – to use my current favorite analogy – of the cathedral, not the cathedral itself. What interests us about the cathedral is its design, and you're not going to find that in an account of the materials from which it's made. I mean you can go through the cathedral and count and label every block and locate them on a (3D) map. You will now, in some sense, have described the cathedral. But you won't have captured its design.
Design and Description
That account of the design has to be formulated in different terms. The newer psychologies simply don't have the equipment for that. That equipment has to be built. How do we do that? I don't quite know.
That's why I insist on description. Yes, I know that description is not the end. And I also know that much of what I'm describing is just the blocks and mortar. But, for the most part, when I do the describing, I don't invoke the cognitive stuff. It doesn't much help. In my most recent work on "Kubla Khan" – I’ve used KK as my 'test bed', if you will, for years – I use a lot of diagrams (32 of them), and they're essential to the descriptive work. But they aren't derived from the diagramming conventions used in cognitive science for conceptual structures. They aren't the kind of diagrams I did for Patterson V or Sonnet 129. They're different, and don't require any specialized background in order to understand them. They are, for the most part, straightforward. And, alas, also a bit tedious, as there are a lot of them. But then, poems are complicated objects.
Glimmers of Form
Things get interesting when you start to compare texts – Lévi-Strauss on myth had a profound influence on me in this respect; I studied the Raw and the Cooked quite closely. Then you can begin to see glimmers of form emerging. That's what I'm after, glimmers of form.
And, there's a bonus. You don't have to make the comparisons on the basis of block-by-block descriptions. Shakespeare derived The Winter's Tale from Greene's Pandosto (see From Greene to Shakespeare). He followed it pretty closely, up to a point. But those little changes had a major effect on the overall story, allowing Shakespeare to have a happy ending where Greene did not. You can describe those differences without going block-by-block through both texts. And when you do, glimmerings of form start shimmering before your mind’s eye.
That's what I'm after, glimmerings of form.