Readercon panel: "A Theory of Narrative Aesthetics Informed by Cognitive Science"

[kate_nepveu]

Description:

Eric M. Van (+M) with discussion from R. Scott Bakker, John Clute, John Crowley, Glenn Grant, John Langan, Charles Oberndorf

Talk/Discussion (60 min.). What brain mechanisms (and evolutionary rationales) underlie the fundamental narrative elements of plot, character, and style? Which narrative element seems to be more fundamental than generally recognized, when looked at from this perspective? How can the deep meaning of a narrative work be rigorously conceptualized in terms of information storage in the brain? On how many simultaneous levels do we process a story as we try to fully understand it? Van offers up the beginnings of a theory of narrative aesthetics informed by cognitive science, with feedback at every stage from the audience.

This was the middle panel that I took notes on, but it's the one I'm least sure I understood, so it seemed best to post it first before my notes got any more incomprehensible. (I went over my notes on the first panel shortly thereafter and so am a lot more confident in them.)

I should also say that I'm trying to do less work on the notes—I used to try and turn everything into nice coherent paragraphs, and the end result was re-writing the whole damn thing from scratch (yes, even with notes on the Palm), which took forever. I'm still trying to indicate connections and make the notes grammatical, but in many places this is much more dialogue-like than prior reports.

Notes on "A Theory of Narrative Aesthetics Informed by Cognitive Science"

First, I actually have to reconstruct who the panelists were, because they weren't introduced and the name cards weren't set up in correct order. (This was a deliberate omission and I disliked it strongly. It really would not have taken that much time out of the panel.)

From left to right:

1. John Clute, who I recognized from a prior panel
2. John Crowley, as identified by the person sitting next to me
3. someone whose nametag I wrote down as C-- C-----, who must be Charles Oberndorf
4. someone whose nametag I wrote down as J- H---, who must be John Langan
5. someone whose initials were GG, who must be Glenn Grant
6. and a man with longish blond hair, who by process of elimination must be R. Scott Bakker

My notes identified the non-Clute-or-Crowley speakers by position in the row, and while I think I have all the attributions right, I welcome corrections.

Setup

Van began by saying, this project started with the movies: what elements do we look at to see whether a movie is good or bad? E.g., this book/movie is beautiful but has a bad plot.

In neuroscience, this is the smallest number of concepts that you can fit the brain's activity into: 1) perception and sensation; 2) processing (cognition, including memory); 3) emotion; 4) action. Three of these correspond to elements in books: 1) style, 2) story & plot; 3) character (not an obvious mapping; wait for it).

1. Perception, sensation, and style

He addressed style first, because there's the least work on it. He defined style as actually consisting of two things: the meaning & appropriateness of word choice, and the sound of words and rhythm of sentences. In movies, we can separate these; he used the example of Eraserhead, which he thinks is about how beautiful images can be if you don't process their meaning. However, we don't make this separation as readers, and good stylists tend to do both well. He gave as a counter-example some of Philip K. Dick's "horribly written novels," where he would have the wrong words and great flow, or just the opposite. He thought that studies on music processing (of which there have been few) are probably going to lead to further insight on assonance, but that word choice is a challenge in terms of studying.

Crowley jumped in here to say that he thought Van was mistaken about what counts as style in fiction. It's not sound or the construction of a sentence, it's something else entirely: the hearing and reproduction and mingling of different voices. Rarely do you value a book because of one voice speaking, he said. That's why PKD can be good and bad at the same time.

Clute quoted Tolkien quoting someone else: "cellar door" is the most beautiful phrase in the language. That needs to be explained in a theory. On a similar note, Crowley cited someone reciting an alphabet in their native language, one unknown to the audience which thought it was gorgeous. These, he said, have nothing to do with fiction.

Van replied that he does think we respond to sound as readers, but didn't elaborate further, instead moving on to the next item.

2. Processing, story, and plot

Van: we are hardwired to turn information into stories. This is well established in cognitive science. Anything causal has to be stored as a story.

What's more interesting about our hardwiring is the way we comprehend story. He cited studies that gave the usual subjects (college sophomores) a story (Inuit, African?) that presupposed cultural knowledge that was unknown to the subjects; actually, the story couldn't be parsed without knowledge of what the culture believed to be true (regarding the afterlife, he thought it was about). The subjects went away for a week, and then were brought back and made to retell the story. They now told it in way makes sense to Western culture, because a story literally cannot be remembered without context.

[Ed: "Literally cannot be remembered" seems like a pretty strong statement to me; what if someone was told to memorize the story, not just given it to read? Anyone know more about this?]

Bakker: it's even more profound than this. When you want two brains to communicate, you automatically have the problem of encoding their neural processes to link up. Syntax in language is a way of building up meaning over the course of time; narrative is one more way we encode a type of holistic understanding in a lineal form. If consciousness is an artifact of the need to encode info to communicate—here my notes break off and say, basically, I think that Bakker was trying to suggest that consciousness might be co-extensive with language, but I'm not sure I understood him properly.

I think we have a sort of leftover bit of talk here. Here's what my notes say:

Van: moving to stories: it's easy to conceive of stories that are parsed differently in the what-happened (the "date rape" example), but that's a special case, so he was going to confine himself to stories where we can easily achieve consensus as to what-happened.

And then Van said "altruism" and that we were "jumping rails a bit." The panel came back to processing later, so I think it links up with 2a below.

3. Emotion, empathy, and character

We are hardwired for reciprocal altruism, toward strangers even, which gives us a very big evolutionary advantage. Apparently this works even down to vampire bats.

We've recently learned some of the neural basis for this. There's a population of cells in the brain called mirror neurons: they light up in the area of brain that does something, but also light up when you're just watching someone do something, like swing a golf club or feel emotion—provided the someone is a member of the "tribe" (in an us v. them sense)

An audience member asked if we have a neural understanding of how we classify tribe v. other? Van: No.

Clute: If I've never swung a golf club, will the same mirror neurons light up? Van: not sure, probably not. All the processes upstream from the mirror neurons are completely unstudied.

(Silliness: Crowley: my reaction would be, "what is that fool doing?" Van: in terms of golf, he's excluded from the tribe. Grant: and if someone who'd never seen golf before watched and had the mirror neurons in the golf-swinging area light up, then that shows that golf is genetically determined. Mutter from audience: "it's as good an explanation as any.")

Oberndorf: he's taught 7th grade boys for 22 years, and finds that very often the weak readers have low empathy; they can't do fiction because they can't go along with the character

Audience member: NPR is currently running a 2 part series on autism, language, and empathy.

Van: in the last 48 hours, he came up with this argument: humans are probably hardwired to expand the tribe, since the more people to be reciprocally altruistic with, the better.

The panel immediately made general noises of disagreement, saying that it depends on the resources. An audience member referenced game theory as saying that once the tribe is big enough to use about 1/2 the available resources, then the tribe will probably turn to countering outsiders.

Van: Yes, there are countertropisms, but without this hardwiring we'd probably all be dead. He thinks that when we meet someone new, we are hardwired to be inclusive unless there is pressure otherwise.

Audience member: this is very culturally-based. Cited economic studies about bargaining and cooperation. Van: this is not a good example because it's loaded with economic, market familiarity. Audience member: the point is that what's intuitively fair is culturally determined. Van: well, if that's right, then more openness to strangers = richer literature in a culture!

He continued, for a reader to refuse to identify with a character (label the character "not part of tribe"), the writer has really missed a trick, because there are so many ways that a writer can signal "member of tribe."

Bakker: in surveys of preliterature cultures, the rates of death-by-violence in males are very high; so this is really resource-dependent (I think "this" is "the ability to see others as tribe, if many of your friends are killed by others", but I'm not sure).

Van: okay, so I was very culturally bound when I came up with this idea that we're hardwired to increase the tribe, because obviously not. But cultural (my notes here are unclear, I think it's) willingness to empathize is relevant to literature.

Crowley: and we can extend empathy (not identification or sympathy) to villains, even, in books. Clute: though it's very difficult to keep empathy from sliding to sympathy. Van: good writers identify the point where slippage might occur, and have the person do something appropriately despicable. Example: Spike in Buffy.

Interlude: humor

Changing gears slightly, Van pointed out something that doesn't fit into this scheme at all: humor! Humor can't be subsumed into any of the four things above. What's the evolutionary basis? He thinks: if something violates expectations, it's likely to be worth remembering, so we evolve a positive association with it.

Crowley disagreed. Very often humor is the expected: show the banana peel, show the guy walking toward it, show him fall, show the peel again.

Clute said something that I didn't get all of; all my notes say is "most humor writers don't expect to be able do it again." Van responded with something I completely missed. Anyone?

An audience member argued that the original basis of humor was schadenfreude.

2a. Processing, story, and plot (revisited)

Van then went back to talk about how we figure out the what-happens in a story. (I'd move the notes on this section up, but it ties into the humor discussion.) He noted that he's still sorting out his terminology for this, but:

For each story element we encounter, we have to make sense of it on its face ("understanding"). The element can be fully understood; discreetly ambiguous; suggestive; incomprehensible; or insignificant. (For instance, in Dhalgren, a second sun appears, and it's incomprehensible. In another novel, it might be suggestive, but not there.)

When there are story elements that you've encountered previously that aren't fully understood, you are struggling with "comprehension": every new story element makes you go back and revisit the old, not-understood ones, giving them a "comprehension history."

Stories can be classified by what they do with that. Most stories have some reframing.

(Crowley interjected the major exception: in fairy tales, not only is every element supposed to be clear to you, but it's also explained by the teller.)

Van: the more distance you put between necessary pieces of information, the harder it is for the reader. However, it can be necessary for effects like suspense, drama, or irony.

Langan said that this ties into humor: how jokes work is that moment of insight at the punchline. Crowley: because the mind has to roll backward over the joke as you see that you got it all wrong. An audience member objected: mysteries have the same structure and they are not funny. Crowley: it's like the ugly surprise that you yell at, yelling & laughter are very close (example: carnival rides). Van: yes, and there's a social function to it being laughter rather than yelling. Langan: like management of aggression, when Clute embarrasses you, you laugh instead of punching him.

Back to levels of reading:

For every story element you encounter, you also have an idea of what the eventual significance will be. For instance, in a non-Dhalgren SF novel where a second sun appears in the sky, you expect an explanation because it's SF. The reader wants to be surprised at the paths along way, but does label items "this is important" when they're first encountered.

Van offered a "very out-there" theory for this: the original point of story was to avoid playing games with hidden clues etc.: because that's not how real life works! Story's function was to leave out the stuff that's inconsequential. Then we eventually de-tune this, citing Wolfe and Crowley as authors that layer and delay comprehension.

Crowley asked, how do you get the gene to look back through a story (and see how the layers were built, I think), since you can't do that for oral stories? Van: good question; he'd love to know the first book that plays games with comprehension. Crowley said it was Don Quixote, which didn't really get followed-up. An audience member responded to Crowley's original question: you have to wait for the bard to come back. Van: there are movies you have to watch twice, and they're very different on the second viewing. Clute: but the second or third hearing of Beowulf is not hugely different from the first.

Clute commented that Van was making it easier on himself by conflating story and plot. He was thinking, as he was listening to Van unpack all this, how little we actually know about story.

If I'm deciphering my notes properly here—and I might not be because there was a lot of overlapping—then there was a dispute between Clute and Crowley over what plot and story are. Crowley: story = series of events in time order; plot doesn't have to be in time order. I can't quite decipher what Clute's objection was to this, if anything. Crowley said that the neurological interest is not in the "endless tapeworm of story," but in the layers of comprehension: he would argue that's what ought to be meant by style.

Clute: the most interesting thing about story is that nobody can retell it in the way it was written.

Van: so it has to do with what's in the brain of the beholder before the story is received. It's very debatable, but he would argue that what we mean by calling a story profound or shallow, is that the story changed the pre-existing information in our brains, sometimes below the level of consciousness. Sometimes you can't articulate how you've been changed by a book you feel is profound, but that feeling is the representation of the change.

Clute: the first editor to understand a human story, will be the first alien we meet. I don't think we'll ever understand how story shapes and enables us. Crowley, concurring: we have story-shaped brains.

Van: fun thought experiment: upload all the information from the brains of the potential readership, extract the common elements, and see connections that can be made to a story: objective measure of the linkage from knowledge base. Or see that links contradict each other, or that there aren't many but they all go the same way.

Crowley: seem to suggest that the best would be the most complex under this model. Fairy tales, or other very simple stories, are profound because they're unresolvable: they set up a resonance, because they seem to be clear, then ambiguous, then back and forth: they radiate into all places you look (inside you) to find out what they mean. Hamlet, for instance, expands so far beyond natural limits of page. Clute: Shakespeare could do that because he thought of five meanings for every word.

Bakker said something about "narrative understanding" and "very theoretical" that I didn't hear. Anyone?

Crowley: the problem with evolutionary theory (studies? understanding? I missed the noun here) is that there always needs to be a practical reason, which is not going to get you very far. Van: yes, it's not sufficient, because we've exploited story for completely different ends, but the evolutionary theory needs to explain (or gives, I'm not sure) just a starting point. Clute: exactly! This is something he wishes sociobiology (or discussions about sociobiology) would acknowledge.

And that was the end of the panel.

Other notes

A lot of this, especially #2, seemed to me to tie into exposition (see the Noreascon 2004 panel report).

The most useful things I got out of this panel were actually the places where we lack information: Clute (who PNH, I believe, says has "a brain the size of a planet") saying that he's not sure we'll ever fully understand the influence of story on humans, which is going to comfort me when I fumble my way through the remaining panel reports; and the comment about the limits of evolutionary theory & sociobiology. But it was interesting to listen to.

Comments

[coffeeandink.livejournal.com]

This is fascinating. Thank you for writing it up!

[kate_nepveu]

You're welcome! I'm not sure the writeups are actually any quicker this way, but they do feel faster, at least, which means I'm more likely to keep doing it in the future.