## Tuesday, May 27, 2014

### Patterns: Ramsay on Shakespeare, and Beyond

I was looking though the syllabus for one of Alan Liu’s courses, Literature + (New Media & Literary Interpretation: Close, Distant, and Other Reading) and came across an older, and fascinating, paper by Stephen Ramsay, In Praise of Patterns, TEXT Technology, Number 2, 2005, pp. 177-190 (with accompanying figures). It’s an interesting piece of work, both for what it says about Shakespeare and for what it says about methodology.

Methodologically, Ramsay tells us he began playing around with graphs of Shakespeare plays because, well, he was interested in graphs and wanted to see what would show up. After a fair amount of work, including a presentation to some mathematicians and collaboration with some data miners, something very interesting showed up. Alas, Ramsay doesn’t know quite what to make of it, but it’s still interesting.

And that’s just fine. That’s the kind of world we’re in. We have the capacity to find interesting things, but once found, explaining them is sometimes/often a problem. If we can’t figure out how to operate in that world (this is me speaking now) we’re not going to get very far with digital humanities. Whatever digital humanities is, it is not a positivistic haven of certain knowledge (I’ve now returned to Ramsay).

This post is going to be a long one, over 2K words. First I present Ramsay’s work, then I present some work I did on Shakespeare some time ago, work that speaks to genre by looking at a comedy (Much Ado About Nothing), a tragedy (Othello), and a romance (The Winter’s Tale). I conclude by suggesting that we’ve left the world defined by existing methods of hermeneutic analysis and exegesis.

Patterns of Loci and Scenes

Ramsay looked at how Shakespeare’s plays moved from place to place as they moved from one scene to the next. Here’s the graph he produced for The Comedy of Errors:

By standard mathematical convention such a diagram is called a graph; the ovals are called nodes and the arrows are called edges or arcs. Each node in one of Ramsay’s diagrams indicates a locus (my term) where a scene takes place and is labeled with a name or phrase designating that place. The edges indicate the transition from one scene to the following scene; the label on the edge indicates the following scene. It follows from the labeling convention that there will not be any edge labeled “1.1” as that is the first scene in any play and, as such, does not follow any other scene.

At the top of this diagram we see a locus called “A hall in Duke Solinus’s place.” It has one edge coming from it and directed to a locus called “The Mart.” That edge is labeled “1.2.” We now know that the first scene of the first act must take place in that hall in the palace. By following the edges in act-scene order from node to node we can trace the course of the play through space (the loci) and time (act-scene order).

Notice the locus at the lower right, “A street before a Priory.” As there is no edge pointing from it, it must be the locus of the last scene in the play. To the upper left of that locus we see a locus designated “Before the house of Antipholus of Ephesus.” At the right side of that node edge 3.2 loops back to the same node, indicating that two successive scenes (3.1 and 3.2) are set at that locus. In the middle of the diagram we can see that a number of scenes shift back and forth among loci. Those are the kinds of features that attracted Ramsay’s interest.

It turns out the graphs for the various plays look markedly different. The graph for Julius Caesar, for example, is very linear, while that for King Lear is linear for a while and then becomes more richly structured. Antony and Cleopatra is still more richly structured.

The question that Ramsay then asked is this: Are the graphs for the four traditional types – comedy, history, tragedy, and romance – much alike within the genre but different from those of other genres? It turns out that they are, though Ramsay did not determine this by visual inspection. First he characterized each play in terms of features of its associated locus-scene graph (p. 186):
1) the number of distinct loci,
2) the total number of scenes,
3) the number of “single-instance” scenes (I assume this means loci with only one scene),
4) the number of loops, where successive scenes occur at the same locus, and
5) the number of switches, where a string of scenes alternates between some one locus and one or more other loci.
With the help of Bei Yu, a graduate student at the University of Illinois who worked with the National Center for Supercomputing Applications, he used Baysian methods to classify the patterns of Shakespeare’s plays as indicated by the above five features.

It took a bit of work to figure out how to conduct the analysis, but they came up with something that worked (p. 188):
The comedies are, for the most part, clustering together, and the tragedies are, for the most part, clustering together. There are several curious anomalies, but the clear cases seem to have been adjudicated properly. Moreover, the curious anomalies appear to conform to some of the more famous critical statements about the plays. For example, one very inﬂuential critic has argued that both Othello and Romeo and Juliet resemble comedy, without making any mention of “low-level” structural features (such as scene loops and switches).
The remarkable thing is that this classification scheme worked at all. It’s not at all perfect, but it’s good enough that we have some serious thinking to do. Why should such simple and apparently purely formal features turn out to be reasonable indicators of genre, at least within Shakespeare’s corpus?

I don’t know, though I’ll have something to say in that direction in the next section of this post, and neither does Ramsay. But, for my money, he evades the issue, or rather, he doesn’t seem to know how to face it squarely.

Here’s the paragraph that follows the previously quoted passage:
One is tempted—almost behooved—to make sense of the entire arrangement. Indeed, we might almost convince ourselves that such matters as “number of scenes” and “number or loops” prove something essential about Shakespeare’s genres. But this, of course, is nonsense. These methods and visualizations prove nothing at all. Indeed, to assert that these extremely low-level features are somehow constitutive of genre would be to perpetrate a ham-fisted abuse of statistics and a grotesque parody of scientiﬁc method simultaneously. But what, then, does all of this do?
Here’s what bothers me: But this, of course, is nonsense. These methods and visualizations prove nothing at all. That denial is too strong, too defensive, and in misleading terms. It’s not a matter of proving anything or of being “constitutive of genre.” Ramsay comes closer to the mark at the end of the next paragraph:
It forces us to move our eyes over Shakespeare’s plays as Euler’s eye must once have moved over the bridges of Königsberg. I have never thought (at least in structural terms) of the ways in which The Tempest resembles comedy, the ways that All’s Well that Ends Well (one of the infamous “problem comedies”) resembles (of all things) history, or the ways in which Henry V resembles tragedy. The fact that I’m being led in such directions by low-level structural features—some of which barely register in one’s consciousness during the ordinary act of reading—raises an obvious question: How do the low-level matters of dramaturgy relate to the high-level matters of genre?
That’s the question we must ask ourselves. Ramsay has a bit more to say in this article, but he doesn’t propose an answer to that question. I don’t have an answer either, but I have done some work on Shakespeare that overlaps with some of Ramsay’s passing observations.

More Shakespearean Patterns, of a Different Kind

Ramsay’s graphs, and the five features he derives from them, are descriptions. One of the things Ramsay has gotten from those graphs is the realization that this or that play, which is normally classified as X, is also rather like Y. So, Othello is like a comedy, Henry V is like a tragedy, and so forth (and he lists more examples in the article than I’ve quoted above). At several points in his Anatomy of Criticism Northrup Frye has observed that, in effect, a tragedy is a comedy where the last act, the reconciliation, has gone missing (see my post Frye on Comedy and Tragedy). That is to say, and to put it crudely, there is a “deep play” that is expressed variously as a comedy or a tragedy or a history, whatever, depending on “circumstances.” That is, depending on those circumstances, the deep play will require either this or that pattern of locations and scenes to get from the opening state to the final state.

What are these deep plays and what are the circumstances that allow/force them to be expressed in one form or another?

It turns out that Shakespeare has undertaken a “natural experiment” that affords us some clues on this, one I’ve examined in At the Edge of the Modern, or Why is Prospero Shakespeare’s Greatest Creation? (Journal of Social and Evolutionary Systems 21(3): 259-279, 1998). In each of Much Ado About Nothing, a comedy, Othello, a tragedy, and The Winter's Tale, a romance, we find a protagonist who mistakenly believes the woman he loves to be unfaithful—in Much Ado I’m talking about Claudio-Hero plot. In the comedy the male protagonist makes the mistake during courtship; in the tragedy the mistake happens shortly after marriage; and in the romance, the mistake occurs well into the marriage. If we examine the relationships between the characters, we find that it gets closer as we move from one play to the next. And that's not all. There seem to be systematic differences among the configuration of characters in these plays. And that has led me to wonder whether or not those differences are related to the fact that we are dealing with three different genres, comedy, tragedy, and romance. Are these configurations merely incidental features of the plays or are they intrinsic to the different genres?

With this in mind, consider the following table, in which the first column names the function a given character takes in the play:
 Much Ado Othello Winter's Tale Protagonist Claudio Othello Leontes Mentor Don Pedro Deceiver Don John Iago Paramour Borachio Cassio Polixenes Beloved Hero Desdemona Hermione
Does this table depict something for which an explanation is necessary or does it depict a merely contingent set of relationships between these plays? If an explanation is necessary, what kind?

What that table suggests to me is that we have three different ways of “mapping” a single mind—Shakespeare’s, the reader’s, whatever–onto the multiple characters of a play. Claudio, Othello, and Leontes each has different capabilities; and so they draw on different capabilities within the reader.

Neither Othello nor Leontes has a mentor comparable to Claudio's Don Pedro. Don Pedro talked with Hero's father, Leonato, and arranged the marriage. We see that happen in the play. We must infer that Othello arranged his marriage to Desdemona, whose father didn't even know about the marriage. We know nothing about how Leontes managed his marriage to Hermione, but he doesn't have anyone associated with him who could be called his mentor.

Further, there is no deceiver in The Winter's Tale comparable to Don John or Iago. Leontes deceives himself. Iago, Othello's deceiver, is closer to Othello than Don John is to Claudio. Among the presumed paramours, Cassio is closer to Othello than Borachio is to Claudio. Polixenes and Leontes have known one another since boyhood; they are so closely identified that we can consider them doubles. Thus relationships between key characters and the protagonist become more intimate as we move from the comedy to the tragedy to the romance—and some characters, mentor and deceiver, seem to disappear.

Note that the protagonist also becomes more powerful as we move through the sequence of plays. Claudio is a youth just beginning to make his way in the world. Othello is a mature man, a seasoned general at the height of his career; but there are men who have authority over him. Leontes is king (and father); there is no mundane authority higher than his. Perhaps this increase in power is correlated with the apparent “absorption” of functions into the protagonist. The absorption of functions increases the behavioral range of the protagonist. And this increased range is symbolized by higher social status.

Finally, notice that as we move from the comedy, to the tragedy, to the romance, the deception moves further into a marriage sequence that begins with falling in love, moves to betrothal, then to the marriage ceremony, the subsequent consummation, and finally to settled married life. In the comedy the deception falls between the betrothal and the ceremony. It is between the ceremony and the consummation in the tragedy while it happens well into married life in the romance.

What have we got so far?

That table depicts a pattern, but one defined over relationships among plays rather than a pattern confined to a single play–a mode of analysis inspired by Lévi-Strauss’s work on myth. I’ve justified giving that pattern particular attention by the fact that the characters in each of the plays are involved in the same situation¬–a man is deceived about the actions of his beloved. Given how that common situation aligns the characters, other terms of comparison emerge: capabilities, position in the social system, and “distance” into the marriage sequence.

All these things seem bound together, but I still don’t see a causal model emerging from this pattern. It’s just a description, albeit an interesting one.

In that Franco Moretti has been interested in configurations of characters in a variety of texts, including Shakespeare’s plays, we can think of this pattern as existing in or at least being in touch with the territory he’s been investigating with his graphs (in the second pamphlet from the Literary Lab, Network Theory, Plot Analysis, 2011). This pattern thus establishes some kind of oblique connection between Moretti’s graphs and Ramsay’s giving us three related sets of patterns.

But Where’s an Explanation?

But how do we get an explanatory account out of these patterns? I don’t know. In my Shakespeare article I introduce quite a bit of psychology to deal with the deception plot: some psychoanalysis, some of what is now called evolutionary psychology (the term hadn’t been coined when I drafted the article), and some work on maturation during the life cycle. But it still doesn’t add up to a causal model for Shakespeare’s deployment of genre.

What it is, is speculation, albeit interesting and suggestive speculation. Nor do I see much hope of arriving at some kind of casual explanation without more speculation and more fishing expeditions in search of suggestive patterns.

Given what I’ve said about Much Ado About Nothing, Othello, and The Winter’s Tale, one obvious next step would be to examine Ramsay’s graphs for those three plays (which aren’t among the figures for his article) and see what turns up. Think of it as a fishing expedition of narrow scope. There’s no specific agenda, we’re not looking for anything in particular, but who knows, something interesting might show up–though we might have to do more than simply look for loops and count loci. We’ll probably want to look at what happens in each scene, and THAT’s likely to be a messy job.

Assume that something interesting does show up. What then? That’s only three plays out of over thirty. How do we extend those results to the rest of Shakespeare’s oeuvre?

The only thing I’m fairly sure about is that we’re not going to solve these problems by looking for models within the existing repertoire of hermeneutic strategies, for none of them were created to solve these kinds of problem. These aren’t hermeneutic problems. They aren’t about what the plays mean, they’re about the mechanisms and processes whereby they are constructed.

Is the profession ready to address such questions?

Addendum: Ramsay's graphs would be meso-scale graphs in the terms I set up in this post about conceptual topology, which became a section in this working paper: Toward a Computational Historicism: From Literary Networks to the Autonomous Aesthetic.