“Belike this show imports the argument of the play” – Some Thoughts On Moretti’s Paper

Oy! People — non-musicians — once thought that they understood music better once they’d discovered the mathematics in it. And of course, there was — and is — a lot of mathematics in it, music being what it is.

Then again, the best musicians, and the best listeners, have also known all along that music is prime. Any analysis of its meaning is necessarily derivative, and adds nothing of as much significance to our enjoyment or our understanding as the music itself does. This is less obvious to us when we analyze literature, which seems superficially to be made of substances — unlike music, or mathematics — which we all have in common. This is an illusion, as any comparison of Shakespeare to a corporate annual report or a novel by Tom Clancy should make perfectly clear.

It’s fun to do the analysis, but I’m not persuaded that we’re any the wiser for it.

literature, which seems superficially to be made of substances—unlike music, or mathematics—which we all have in common.

Everyone has numbers in common.

@dsquared: I’m not that familiar with the history of Wales, but what I do know about it tends to support the idea that having a dirge for their national anthem wouldn’t be particularly inappropriate.

Everyone has numbers in common.

…indeed, much more so than we have any particular cultural signifier, whose significance can be lost on people from a different time period of the “same” culture, let alone a foreign one.

This is an illusion, as any comparison of Shakespeare to a corporate annual report or a novel by Tom Clancy should make perfectly clear.
Are Tom Clancy and Shakespeare evidence against your view? By and large, Skahespeare is nowadays mostly popular among people who like to analyse their texts, Tom Clancy more popular in the rest of the world. That’s not what you would expect if analysis was a derivative thing that chages nothing to the enjoyment or understanding.

One day, after we add to these skeletons the layers of direction, weight, and semantics, those richer images will perhaps make us see different genres – tragedies and comedies; picaresque, gothic, Bildungsroman … – as different shapes; ideally, they may even make visible the micro-patterns out of which these larger network shapes emerge. But for this to happen, an enormous amount of empirical data must be first put together. Will we, as a discipline, be capable of sharing raw materials, evidence – facts – with each other? It remains to be seen. For science, Stephen Jay Gould once wrote, fruitful doing matters more than clever thinking. For us, not yet.

Doesn’t this sound a little … dreary? Faintly … corporate? Is this really likely to produce more interesting results than Johnson, Hazlitt, Lamb, Eliot, Empson, Vendler, or dozens of other critics who confronted Shakespeare’s work with nothing more than their individual sensibilities?

geo: Well, just because it sounds a bit dreary and corporate doesn’t mean it will be; some quite boring methods can lead to pretty interesting results.

In fact, I think Vendler, whom you mention, is a good example of this. It’s been a while since I read her book on the sonnets, but as I recall, it involved quite a lot of diagramming and counting — words repeated within a sonnet, words that appear in the couplet as well as the quatrains, or in more than one quatrain, etc. I wish I could remember the vocabulary she invented, but there was some, wasn’t there? “Key words” or something, which she looked for in each sonnet.

Formalism, chart-drawing, etc. don’t have to be opposed to “individual sensibilities.” At its best (as when Vendler does it), formal analysis can reinforce sensibility, and vice versa. And I find that the pure-sensibility, isn’t-this-beautiful type of criticism can be just as dull as pure, unenlightened word-counting.

Quite right, Aulus; pure sensibility can be dull. The question, though, is whether word counting and diagramming can be anything but dull. When you say that “some quite boring methods can lead to pretty interesting results,” what do you have in mind? (Myself, I find that aspect of Vendler on Shakespeare not very interesting.)

When you say that “some quite boring methods can lead to pretty interesting results,” what do you have in mind?

Euler characteristic? classification of surfaces? … …calculus?

Zamfir @8:
I think good analysis can add to understanding/enjoyment of a work, but OTOH I’m not sure Shakespeare is mainly popular with those who like to analyze texts. That Shakespeare is performed (and, to a lesser extent, read) so much means that a lot of people have some lines from Shakespeare in their heads even if they’ve never read much or any criticism.

It’s also true that there are many lines that will go past people if they only hear them onstage and are not already familiar with the play. I once saw a performance of 1 Henry IV and no one reacted at all to certain (arguably) amusing and (arguably) well-known lines (e.g., ‘Home without boots, and in foul weather too?/how ‘scapes he agues, in the Devil’s name?’ is one that comes to mind) because, except in the hands of an extraordinarily skilled actor, lines like this, delivered quickly, will not make much of an impression on people who don’t already know the play. (A bit off topic, sorry)

Fair enough, Aulus; I’ll grant that literary scholarship is necessary, and sometimes even interesting. Perhaps someone should write a book called What Are Scholars Good For?

Oh, dear.

geo: Wouldn’t you at least concede that the graphs are interesting in a “a picture is worth a thousand words” kind of way?

Salient’s comment isn’t totally off the wall but seems to me to miss the point in a way that is almost the perfect obverse of the reason I dislike papers like Moretti’s. At the level of math, the work is surely not yet ready for complete mathematization, and needs a close focus on the data, and close collaboration between study of the subject matter and examination of appropriate mathematical patterns. On the one hand there seems to be an almost total misunderstanding of what work in this area is like, a false assumption that papers that make a first attempt at conceptualization (horrors!) of a new problem are putting forward something like a quasi-metaphysical theory of the system of the universe. On the other hand there is almost total neglect of anything other than counting uncategorized “things,” and a faith in theory so strong that it’s content with those first attempts and never attempts to build on them.

The question, though, is whether word counting and diagramming can be anything but dull.

This seems awfully strong. Imagine that someone’s doing a bit of interesting literary analysis that, by the nature of the argument, requires a large amount of sifting through evidence — is there some point at which such an argument crosses over into “word counting and diagramming” and instantly becomes dull? (Of course not. It’s a continuum. But still: how exactly do we distinguish normal literary claims, backed up with evidence in the usual way, from claims that are just a bit too word-county or diagrammy to be interesting?)

I definitely agree that these sorts of techniques can be done badly (just like any technique), but to me they just seem like a particularly large-scale version of the evidence-sorting that literary critics already do. There’s nothing inherently dull or mechanical, for instance, about backing up your claims with selected quotes from the primary text — but just as a selected quotations are an appropriate type of evidence for certain claims, word-counts and diagrams are an appropriate type of evidence for certain other claims. If I want to make a point about how an author’s style changed over time, for instance, I could pick a quote from an earlier work and contrast it with a quote from a later work, but it might be more informative to do some word-counts. It’s a new kind of information that’s distinct from the sort of stylistic differences that are readily apparent from the experience of reading, and so it has a lot of potential to surprise. I dunno, do I not have the right to be intrigued by this kind of stuff?

Erratum: in 25, for “Oh what a rogue and peasant slave am I” read “Now I might do it, pat.”

Let me comment a bit on why I said what I said about Moretti and underlying mechanism.

After I made the statement John quoted, I quoted a passage from Sydney Lamb’s Pathways of the Brain that was similar in spirit to Moretti’s passage about seeing “the underlying structures of a complex object.” Lamb is a linguist of the Chomsky generation and, more importantly, a computational linguist, a man who’s built programs to parse sentences. Lamb really does have a handle on underlying mechanisms, but only of sentences and phrases, not Hamlet and the like. Many others have tilled those fields, of course, and they have a sense of the underlying mechanisms as well. Whether any of these folks have a start on the real mechanisms of language, that remains to be seen. But they do have serious proposals.

Do I think Moretti’s social networks are really underlying mechanisms of plays or novels? Not really (nor, in the end, does he). If that’s the case, then why did I say otherwise? Because it’s my guess that, by creating those graphs and thinking closely and carefully about them, Moretti has the barest sense of the possibility of underlying explanatory mechanism. Let me emphasize the fact that he had to create those graphs, by hand, and in the process of had to think closely about the texts. That close thought may not have been deep thought, but it was necessarily sustained, because it takes time to go through a text and create those things, and it takes time to draw them and manipulate them in even the most superficial of ways, cutting Hamlet out of one, and so forth. All that’s important.

Simply reading Moretti’s paper, even sympathetically, isn’t likely to give you the intuition that the creating of those graphs gave Moretti. Though, if you already have extensive experience working with such graphs – they’re used in many disciplines – you have a leg up on others, perhaps even on Moretti himself. You can’t learn math simply by reading the text book; you have to work the problems and do the proofs yourself. The same is true of the kind of exercise Moretti has undertaken.

And perhaps John somehow sees in Moretti’s work something of what Moretti himself sees in, while at the same time recognizing the Moretti doesn’t really have the “skeleton.” Still, it’s one thing to say that Moretti’s not looking at skeletons, it’s another thing to have something useful to say about just what one has to do to build an x-ray machine that’ll reveal the skeleton inside the text.

That’s a task I set myself some 35 years ago, under the tutelage of David Hays, one of the founders of the discipline of computational linguistics. Hays was in the Linguistics Department of SUNY Buffalo at the time, and I was a graduate student in English. He was using graphs to represent the semantic structure of language, not simply at the sentence level, but the discourse level. I learned something of the craft from him and managed a glimpse of the skeleton in Shakespeare’s Sonnet 129.

And then I hit a wall. The graphs were complicated and, at some point, arbitrary. It took a lot of work on a lot of graphs just to do a sonnet. I haven’t the foggiest ideal what it would take to do Hamlet. So I’ve not attempted to visualize the skeletons of other texts.

And I regard I regard that work as some of the most important that I’ve done. For it changed my conception of what it means to investigate literature. As for just how it changed my conception, well . . . I’ve blogged and blogged on that at The Valve and now at New Savanna, not to mention this that and the other paper. I’m tired of telling that story again and again.

But, who knows, maybe someone will pay attention to Moretti. I certainly hope so.

Salient: Perhaps the Stanford visualization people have something for Moretti. But I think he’d do better to chat with Martin Kay, one of the grand old men of computational linguistics. He needs to get a sense of what a real language mechanism is like, and what kinds of processes it has to execute. Kay knows that very well indeed.

Yes. On the other hand, some words are worth a thousand pictures.
Does that mean you are worried that something like Moretti’s approach might become so dominant that it starts replacing more traditional approaches?

@ holbo, craig:
Recently read 309-346 of Hegel’s Phenomenology on phrenology. This section, from “Observing Reason”, is actually not unconnected to the issues in the post. Anyway, Hegel makes me think the phrenologists themselves may have been much more interesting than we remember.

Zamfir: Nothing as serious as that. Mostly I just thought it sounded witty. But there’s a comment of Evelyn Waugh’s that somehow fits the case. He thought the film of Tender Is the Night was “a very good film of a rather poor book,” but that even so, “the enormously expensive apparatus of the film studio can produce nothing as valuable as can one half-tipsy Yank with a typewriter.” Exaggerated, as usual with Waugh — there are of course many great movies, as he would undoubtedly admit. But some words, with their intricately braided lattices of historical association and shimmering auras of half-conscious resonance, are more evocative than some pictures. That’s all.

@jim:

Act IV, Scene 3 has

Rosencrantz: Ho, Guildenstern! Bring in my lord!

Guildenstern is of course just off-stage at this moment, but Rosencrantz is plainly speaking to him. I think the definition of “speaks to” would have to be tortured out of all recognition in order to disqualify this line.