Monday, October 9, 2017

How I discovered the structure of “Kubla Khan” & came to realize the importance of description

I say discover because I regard the poem’s structure as something existing objectively in the world, prior to and independent of my work, or anyone else’s for that matter. However, I regard that discovery as tentative because it really hasn’t been confirmed. The profession – academic literary criticism – isn’t like that.

But I’m getting ahead of myself with this talk of a discovery that hasn’t been confirmed. Let’s set that aside for the moment. We’ll return to the issue at the end – though that’ll take awhile (you might want to get a cup of coffee, or some scotch, whatever’s appropriate).

This is about description, and how I came to realize the importance of description. That happened through my work on “Kubla Khan” – other texts as well, but that was the Rubicon. I’m going through this once again now because the profession seems to be in the process of discovering, or rediscovering description – e.g. the special issue of Representations (Summer 2016) devoted to it – and is still quite tentative about it. By my reading – and here I’m being polemical – they don’t know what they’re up to.

But then I didn’t know what I was doing either, not back then. That’s why I’m writing this, to emphasize how very difficult it can be to understand what you’re doing. It was only two decades or more after the fact that I came to understand what I had done in the early 1970s.

A structuralist analysis of “Kubla Khan”

In my senior year at Johns Hopkins I enrolled in a two-semester course on Romantic Literature, taught by (the legendary) Earl Wasserman. Keats, Shelley, and Sir Walter Scott in the Fall (1968); Wordsworth, Coleridge, and Jane Austen in the Spring (1969). That’s when I became hooked on “Kubla Khan”, the spring of 1969. I had this idea that the poem attained completeness by asserting its own incompleteness, a formulation that Wasserman loved, and that would have been well suited to the emerging deconstructive dispensation, though I didn’t frame it that way.

I stayed at Hopkins for a master’s degree in Humanities and wrote my thesis on “Kubla Khan”. I had become interested in structuralism and wanted to do a structuralist analysis – I refuse to call it a reading – of the poem, which was dripping with the binary oppositions so central to structuralist thought. Oh, I had other thinkers I wanted to bring to the poem, but I centered on Lévi-Strauss.

Therein lies the problem. It stalked and ambushed me and laid me low. When he began his work on myth Lévi-Strauss pursued the idea that a myth began with a binary opposition – life vs. death, nature vs. culture, that sort of thing – expressed in extreme form and proceeded by substituting successively less extreme forms of the opposition until, with one final substitution, it all but eliminated the opposition. Thus he had an intriguing scheme that explained why myths took the form they did. That’s what attracted me, the possibility of an actual explanation.

But he’d abandoned that scheme when he began Mythologiques. Oh well, never mind. Anyhow, I wasn’t dealing with a myth or group of myths, I was dealing with a poem. One poem: “Kubla Khan”.

Finding binary oppositions was easy – Kubla vs. the wailing woman, the wailing woman vs. the Abyssinia maid, Kubla vs. the maid, the dome vs. the caves, wailing vs. song, maid vs. auditors, Kubla vs. poet. They were all over the place, those oppositions. But I couldn’t see any order to them.

Here’s a fragment from one of my worksheets:


I’d typed the poem out in triple-space and then marked it up. I must have done this half-a-dozen times. Sometimes double spaced, sometimes triple. And I used different colors of felt-tip pens to indicate different aspects of structure.

It just went on and on. And then I gave up.

Of course there was more on my mind than those oppositions. For one thing I had been reading everything I could find on “Kubla Khan” in the Hopkins library. Along with bits and pieces of that scholarship I was also weaving my other core thinkers into the mix, Wittgenstein, Piaget, Merleau-Ponty, and Nietzsche. But it wasn’t working. In the spring of 1970 I stopped typing on page 142. Some time after that, I don’t recall when, I had the idea of treating line-end punctuation as a means of dividing the text into hierarchically ordered units in the same way that parentheses, brackets and braces are used in grouping elements of mathematical expressions. Once I had done that I could see that the punctuation structure of the poem matched the units I that had slowly emerged in my examination of binary oppositions.

This is what I ended up with for the first movement of the poem (36 lines):


The second movement (18 lines) is similar.

The fact that it was mere punctuation that brought things into line is important. Back in those days the issue that hung over all other issues was this: is the meaning a critic explicates something they find in the poem or is it something the place in the poem? Well, I didn’t place those punctuation marks in the poem; they are clearly there in the text. They are objectively there.

There was and is much more to the analysis than what I’ve indicated here, but I have no intention of trying to summarize those findings. If you’re interested you can read the papers [1]. My point is that I felt that I had my hands on something tangible, something real.

But – and this is just as important – I didn’t have an interpretation, not as those things were and are understood in the profession, nor did I have the kind of explanation I had found in Lévi-Strauss’s early work on myth. I didn’t know what I had. Let me repeat: I didn’t know what I had.

Something for sure. And it was real. But what kind of real thing was it?

I now know that it was a description. But I wasn’t thinking in those terms in the early 1970s. No one told me that’s what I had and I couldn’t figure it out for myself. It was just this strange lump of THE REAL.

And that – not knowing what I had – is why it took me two years or so to complete that master’s thesis. I took a tab of LSD in January of 1972 to get me off the mark and writing the damn thesis [2]. No, there was no mystical revelation, no marvelous lights, nothing like that. The conviction that I had to write the thesis was not an immediate consequence of the acid trip. But some weeks later, that’s what I did, started over from scratch and wrote the thesis.

I was both pleased and uneasy, disappointed. Pleased at having found this strange hunk of the real. Disappointed because it wasn’t what I’d been looking for. I’d been looking for something that made sense in terms of the concepts I’d brought to the work. Instead, it refused to cooperate with them.

Buffalo and cognitive science

Look back over that diagram of the first part of “Kubla Khan”. Notice how those 36 lines are divided into three parts, the middle of those in turn has three parts, and then one more time, the middle has three parts. That looks like a dol1 (ll. 17-24), within a doll (ll. 12-30), within a doll (ll. 1-36). That’s a nested construction. I read about such things in linguistics and had even played around with them a bit in a course I’d taken in computer programming in my junior year at Hopkins.

It seemed clear to me that if I wanted to explain the structure I’d found in “Kubla Khan” I’d have to learn more about such structures. That meant the emerging cognitive sciences. So I went off to SUNY Buffalo in the fall of 1973 to get my Ph. D. in English. And somehow I’d do more work on “Kubla Khan”.

There were a lot of brilliant and interesting thinkers in Buffalo’s English Department back in those days [3]. But I couldn’t find anyone who could help me with “Kubla Khan”. A fellow graduate student, Ralph Henry Reese, had been working with David Hays in Linguistics. I made an appointment, showed him my thesis, and became his student. Hays didn’t know what to make of that structure, but he knew there was something there, somewhere.

Hays had been a first-generation researcher in machine translation – I’m told that he coined the phrase “computational linguistics” [4]. I joined his research group with the intention of working on “Kubla Khan”. I didn’t make it that far, though I did some very satisfying work with Shakespeare’s Sonnet 129 [5]. By that time I was as much a cognitive scientist as a literary critic.

Whatever it was that I was doing, it certainly wasn’t interpretation. It was something else. The work on the Shakespeare sonnet had something of the feel of an explanation, but didn’t go far enough and I didn’t know what it would take to push it further along – 40 years later I still don’t know.

I finally began rethink that early work in the mid-1990s. While cruising the web I more or less accidentally discovered that, at long last, there were literary critics interested in applying the cognitive sciences to literature. Alas, they didn’t know about my early work and what they were doing was quite unlike it. They certainly weren’t thinking about computation.

It was in the process of thinking through that mismatch that I finally figured out that 1) I had been interested in form all along, and 2) what I wanted to do with form was to describe it. That thinking-through took a long time, years, probably more than a decade.

All along I had thought of myself as a theorist. That’s what had interested me back at Hopkins in the 1960s and 1970s, theory, but theory before it had become capital-T Theory. That’s what I’d done at Buffalo, work with David Hays in developing a theory of natural language semantics. It was that theory – or was it a model? - that I’d applied to the semantics of Shakespeare’s Sonnet 129. And it was that model that I couldn’t figure out how to apply to “Kubla Khan”.

Description, the modern ‘rithmatics, and diagrams

Description, however, is a bit different. One thing I realized is that there had been a strong and sharply descriptive aspect to my practical criticism all along. As I’ve said, my work on “Kubla Khan” was primarily descriptive. But there was a strong descriptive element to my work on Sir Gawain and the Green Knight [6] and my article on Shakespeare’s life cycle was grounded in a table depicting crucial relationships between Much Ado About Nothing, Othello, and The Winter’s Tale [7]. The same is true of “The Evolution of Narrative and the Self” [8]. In each case there is a descriptive core on which I built whatever theorizing I had on offer.

The thing about description is that, as far as I can tell, it doesn’t require an elaborate intellectual superstructure. You don’t need to know a whole lot of theory, or even Theory, to do it. All it requires is experienced observation [9]. That’s pretty much what I’ve been relying in my on-going work on ring-composition [10]. I don’t have a theory about ring-composition. I started with the examples Mary Douglas had worked out [11] and went looking for similar structures in different texts.

Once the discipline has amassed a collection of high-quality descriptive work, then individual researchers can set about building theories on that, theories we cannot even imagine now because we don’t have a descriptive base from which to work. Darwin did his theorizing over several centuries of accumulated descriptive work on the forms and lifeways of plants and animals. That’s what the literary Darwinians should be doing. Instead of cranking out half-baked theories and interpretations [12] they should be doing what biologists have been doing for centuries, getting descriptive control over the phenomena.

Alas, no one is going to do descriptive work if the profession is not prepared to recognize it. And it isn’t, not that I can see. The work of these new descriptivists seems so tentative. The introductory essay in the Representations issue is entitled “Building a Better Description” [13]. It surveys the issue and ends with six suggestions for building better descriptions. Where’s the essay telling us “What We’ve Learned by Crafting Better Descriptions”? That’s what’s missing.

What’s also missing is linguistics. Literature, after all, is made of language, and linguists study language. But there wasn’t an essay in the issue – which was derived from a conference held at Columbia in 2015 – by a linguist. That’s a major lacuna.

In Aspects of the Theory of Syntax (1965) Chomsky, for example, clearly distinguished between describing the grammar of a language and explaining it (pp. 18-27). Of course, when Chomsky talks about describing a language, he’s talking about creating a grammar for that language and that grammar takes the form of formal rules that are often expressed in the form of tree diagrams. The new descriptivists, however, seem to think that description is primarily verbal. To be sure, there is some acknowledgment that description need not be exclusively verbal [13, p. 2]:
An elusive object that travels by many names, and sometimes by no name at all, description’s dictionary definitions include representation, drawing, report, portrayal, and account. Description can take many forms, including lists, case studies, sequences, taxonomies, typologies, genealogies, and prevalence studies, and it involves many actions, including observing, measuring, comparing, particularizing, generalizing, and classifying, using words, images, and numbers.
The possibility of using diagrams in describing literary texts is not, so far as I can tell, even considered.

Diagrams and tables were central to the work Lévi-Strauss did on myth. They constituted a logic and a rhetoric that allowed him to investigate and display relationships between items within individual myths and between myths. Would these devices not be useful in describing literary texts? I think so; I know so [14]. They are indispensible to my practice [15].

Sydney Lamb, a linguist of Chomsky’s generation, though having a very different intellectual style, has remarked on the importance of visual notation [16]:
... it is precisely because we are talking about ordinary language that we need to adopt a notation as different from ordinary language as possible, to keep us from getting lost in confusion between the object of description and the means of description.
Literary language may not be “ordinary” – a matter of considerable debate from to time – but Lamb’s point still applies.

More recently Franco Moretti has made a similar observation [17]:
Third consequence of this approach: once you make a network of a play, you stop working on the play proper, and work on a model instead: you reduce the text to characters and interactions, abstract them from everything else, and this process of reduction and abstraction makes the model obviously much less than the original object – just think of this: I am discussing Hamlet, and saying nothing about Shakespeare’s words – but also, in another sense, much more than it, because a model allows you to see the underlying structures of a complex object.
Literary critics, however, had explicitly rejected this separation between analytic text object text back in the mid-1970s.

Consider title essay of Geoffrey Hartman’s 1975 collection, The Fate of Reading, as an exemplary text. Hartman is grappling with the fact that, no matter how intensely critics are oriented toward the texts they study, the act of analysis and writing requires distance. One cannot write about the text if and while one is immersed in reading it. Complaining that contemporary theorists — mostly French or under French influence — have come to privilege such writing over reading, Hartman asks (p. 272): “To what can we turn now to restore reading, or that conscious and scrupulous form of it we call literary criticism?” He then observes: “modern ‘rithmatics’—semiotics, linguistics, and technical structuralism—are not the solution. They widen, if anything, the rift between reading and writing.”

He is invoking the professional trope that elides the difference between reading, in the common sense of the word, and reading, used as a term of art for analytic examination and explication. Critical reading is merely, you know, reading. Same difference.

More recently, Tony Jackson writes [18]:
That is, a literary interpretation, if we are allowed to distinguish it as a distinct kind of interpretation, joins in with the literariness of the text. Literary interpretation is a peculiar and, I would say, unique conjunction of argument and literature, analytic approach and art form being analyzed.
Reading is writing is reading – I think.

Where are the new descriptivists on this issue? I think they belong to the side of Hartman and Jackson by default. They are either unaware of or they simply reject ‘rithmatic disciplines. They are working within an intellectual lineage that rejected them four decades ago. They’ll be crafting their better descriptions within that cleansed tradition.

In contrast, four decades ago I had opted for the those ‘rithmatics and produced an account of “Kubla Khan” that was unlike anything in the extant literature. It was primarily descriptive in nature, though it took me two to three decades to conceptualize it and value it as description.

These descriptivists are looking for descriptions but aren’t sure of how to construct them. I’d constructed one but didn’t know that that’s what I’d done.

I have no idea where they’re going from here. Myself, I intend to continue my practical descriptive work. I think of it as exploration, looking for new things. What new things? Structure in texts.

Description: A Mode of Discovery?

That brings us back to the beginning, my original work on “Kubla Khan”. I’d opened by asserting that I’d discovered the poem’s structure but that that discovery is only tentative because it hadn’t been confirmed.

To be sure, my basic work was accepted as a Master’s Thesis at Johns Hopkins in 1972: THE ARTICULATED VISION: Coleridge's “Kubla Khan”. Some years later (1985) I published a somewhat abbreviated version of that work in Language and Style and then, more recently (2003) I published a rather longer and considerably more sophisticated version in PsyArt: A Hyperlink Journal for the Psychological Study of the Arts [1]. But those publications don’t mean anything like confirmation, which would imply that that description becomes the way the profession thinks about the poem. All those publications mean is that a handful of scholars think that work is competent and worthy of consideration by other scholars. No more, no less.

What would it mean for the profession to regard the poem in that way? Here’s an analogy:
In biology: Centipedes are of the class Chilopoda and have between 30 and 354 legs, but always an odd number of pairs [19]. And so forth.

In literature: “Kubla Khan” is a 54 line lyric in two movements, each of which has a triple-embedded ternary structure. The last line of the first movement is repeated at the structural center of the second movement. And so forth.
The profession doesn’t have a mechanism to confirm discoveries of that nature. Previously unknown manuscripts and editions, yes. But structures of texts? The profession doesn’t recognize that there is anything there to be discovered.

For me, description implies that there is something there to be discovered. For these descriptivists ... I don’t know. I don’t know. Do they?


[1] This first paper is a highly edited and reframed version of the thesis I’d written in 1972. William Benzon. Articulate Vision: A Structuralist Reading of “Kubla Khan.” Language and Style 18: 3 - 29, 1985. URL:

The second is more sophisticated and detailed and has speculations about underlying psychological and neural mechanisms. William Benzon. “Kubla Khan” and the Embodied Mind, PsyArt: A Hyperlink Journal for the Psychological Study of the Arts, November 29, 2003. URL:

[2] For what it’s worth, I’ve told this story in considerably more detail: Touchstones • Strange Encounters • Strange Poems • the beginning of an intellectual life (2015, originally published in Paunch in 1975), URL:

[3] Bruce Jackson has written a wonderful short account of that department: Buffalo English: Literary Glory Days at UB, Buffalo Beat, 26 February 1999, online,

[4] David G. Hays, Wikipedia, accessed October 9, 2017,

[5] William Benzon, Cognitive Networks and Literary Semantics, MLN 91: 1976, 952-982. URL:

William Benzon, Lust in Action: An Abstraction, Language and Style 14, 1981, 251-270. URL:

[6] William Benzon, Sir Gawain and the Green Knight and the Semiotics of Ontology, Semiotica, 3/4, 1977, 267-293.

[7] William Benzon, 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.

[8] William Benzon, The Evolution of Narrative and the Self, Journal of Social and Evolutionary Systems, 16(2): 129-155, 1993.

[9] See a recent blog post in which I draw on the work of John Wilkins, a philosopher of biology, Experienced observation (& another brief for description as the way forward in literary studies), accessed October 9, 2017,

[10] Most recently I’ve been investigating King Kong. Those posts should be at the top of the list (as of Oct. 8. 2017) in you follow this link, which gathers all my ring-composition posts:

I’ve gathered some of that work into Working Papers at,

[11] Mary Douglas, Thinking in Circles: An Essay on Ring Composition, Yale University Press, 2007.

[12] For critical remarks about literary Darwinism, see William Benzon, On the Poverty of Literary Darwinism, September 2015, 45 pp.,

[13] Sharon Marcus, Heather Love, Stephen Best, 2016, Building a Better Description. Representations 135, Summer 2016, pp. 1-21. Ungated download,

[14] William Benzon, Description 3: The Primacy of Visualization, Working Paper, October 2015, 48 pp,

[15] My most recent paper on “Kubla Khan” is the most complex and sophisticated example (see note 1, above), but I’d also emphasize my ongoing use of tables in. e.g. my ring-composition work (note 10).

[16] Sydney Lamb, Pathways of the Brain: The Neurocognitive Basis of Language, John Benjamins 1999, p. 274.

[17] Franco Moretti, Literary Lab, Pamphlet 2, Network Theory, Plot Analysis, May 1, 2011, p. 4,

[18] Tony Jackson, “Literary Interpretation” and Cognitive Literary Studies, Poetics Today, 24 (2), 2003, p. 202.

[20] Centipede, Wikipedia, accessed October 9, 2017,

No comments:

Post a Comment