Saturday, August 29, 2015

On the Poverty of Literary Cognitivism 4: The ‘Middle Way’, What It Can and Cannot Do

Literature is not made out of consciousness, it’s made out of words.

– J. Hillis Miller

No scholar should find humiliating the task of description. This is, on the contrary, the highest and rarest achievement.
— Bruno Latour

What do I mean by the ‘middle way’ and why is the term in scare quotes? The second question is easy: because I’m not quite serious about it, a little serious, but not all the way serious. As for what that middle way is, it includes the various approaches to literary criticism Alan Richardson had in mind in the essay review that prompted these posts, plus those approaches that favor evolutionary psychology. So this middle way is a sprawling mess of literary criticism and not internally unified.

In what sense are they in the middle? What’s on either side of them? Well, to one side you have the ‘hard core’ cognitive science that I embraced early in my career and that I discussed in my second post, What I Learned When I Walked off the Cliff of Cognitivism, while the formal description that I discuss in my third post, The key to the treasure IS the (form of the) treasure, is on the other wide. These middle way criticisms have little or no interest in literary form nor are they interested in computational mechanism (some of them even explicitly disavow computation as belonging to “first wave” cognitive science). Thus they work in between the kinds of things that have absorbed my professional attention.

Whether or not this middle way can serve as a bridge from one to the other, well – that’s an interesting question. A bridge does need to be built, but it is not going to be built by thinkers who don’t recognize that either explicit models or formal description are important intellectual activities.

The creation of explicit computational models can only be done by people with the appropriate technical expertise. Those people can be found in various disciplines, but few of them are literary critics. Is there anything literary critics have to offer those thinkers? Yes, formal analyses of literary texts. And the middle way people, do they have anything to offer? Well, I suspect they might object to the question. But that, after all, is not so different from the issue that Richardson posed in his review: Is there anything cognitive critics have that the psychologists and neuroscientists need?

My position is that they need, they can make good use of, those formal analyses, though they don’t know yet know it – in part because they don’t know that such things exist and so haven’t had a chance to examine them. It’s not at all clear to me that they can make much use of middle way criticism, though, as always, I could be wrong. But that’s what this post is about: to explain these things.

In the first section I revisit the idea of computation and argue that the “embedded” cognition of the so-called “second cognitive revolution” presupposes and would be nothing without the computational ideas of the “first” cognitive revolution. Then I use the metaphor of a building (such as a cathedral) and its materials to indicate why current cognitive criticism will necessarily fall short of a robust understanding of literary phenomena. What’s the answer to that problem? You guessed it, the study of literary form, to which I return in the third section, where I also argue the literary form is a way to link up with a Latourian view of social process. In the final section I argue that we free literary interpretation from the pretense of objective knowledge so that it become an openly ethical criticism in the sense that Wayne Booth has advocated.

Computing and Embodiment

The cognitive critics believe that they’ve left computation behind, as we can see in a passage that Alan Richardson and Francis Steen wrote in response to an essay by Hans Adler and Sabine Gross (152):
As Adler and Gross (ibid.: 197) themselves note, in fact, a “more comprehensive notion of human cognition” has over the past decade or so largely displaced the narrower, more exclusively computational, and effectively disembodied notion that the term cognitivism now conveys for many cognitive theorists and researchers (Varela et al. 1991: 71). More recent theories of cognition instead seek to acknowledge the bodily instantiation (if not basis) of mind, the emotive aspects of cognitive activity, and the social embeddedness of cognitive development and functioning.
This, I believe, is a reference to a so-called “second cognitive revolution” and which is, I believe, a bit misleading. I’m particularly skeptical about the effect that phrase “effectively disembodied” has in conjunction with computation, though that conjunction certainly is out there and has been and is influential. It is also superficial.

I supposed that “effectively” allows them to discount computing while still acknowledging that, yes, computers are physical systems and computing is a physical process. Whatever it is that computers do it is thereby in that sense embodied. But there is more to computational mindset than the mere fact of computing. It is a very explicit way of thinking about how mind-like systems are organized and constructed. That explicitness, that sense of design and construction, is somewhere between weak and missing in much of this second-revolution thinking.

Let me say it again: Computation was at the heart of the co-called cognitive revolution. No computation, no cognitive revolution. It was computation that allowed cognitive science to displace behaviorist psychology from the center of the academy. It made the mind real in a way it hadn’t been before.

Once those ideas had settled in it became professional respectable to think about the mind without having explicit computational models. And so we have this “second cognitive revolution,” but if you look closely you’ll see that computation still hangs around in the corners. Those ICMs (idealized cognitive models) that the cognitive linguists are fond of, many of them are computational models. That is one kind of embodiment and without it nothing else matters.

The more recent notion of embodiment emphasizes that cognition is grounded in perception and action and, in one version, emphasizes the role of metaphor mappings. Depending on the details of your conceptualization, these may be two different things or simply two faces of one notion. But that’s a detail that doesn’t matter very much. However, without computational embodiment in some form this second-revolution embodiment is “effectively disembodied”.

Furthermore I believe that the idea that literature is made of language is not one we can bracket or dismiss as a detail to which we will attend later. To dismiss computation as an aspect of language, as the cognitive critics are in effect doing, is shortsighted. Literary cognitivists are primarily concerned with what words mean, and so they gravitate to cognitive science that links ideas to perceptions and actions while ignoring cognitive science that treats words as physical phenomena arrayed in strings the we speak and listen, that we read and write.

Theory of minds, mirror neurons, these phenomena, much loved by cognitive critics, do not involve language directly. And, despite their origins in cognitive linguistics, cognitive metaphor nor conceptual blending are adopted as tools for investigating how cognitive and perceptual objects are organized among themselves rather than for how language is a tool for manipulating those objects. Cognitive criticism is a criticism that holds the text, the signifiers, at arms length. And the same is true of its kissing cousin, evolutionary criticism.

No doubt many of these researchers would deny that, and in a few cases the denial has some force. But for the most part the denial is empty. What makes this denial superficially plausible the flexible notion the profession has adopted about the text. Are they interested in the text? Of course they are, they’re literary critics.

But what IS the text? Well, sure, it’s the signifiers. But when most critics talk of the text they mean something beyond the signifiers. Just what that is, that will vary from critic to critic and occasion to occasion. The only way to clarify this situation is to let the text be signifiers, and to understand them as tokens on which a computational process operates. At present that is the only way I see of creating an understanding that actually and explicitly couples those (mere) signifiers with psychological and social processes.

Cathedrals, Stones, and Engineering

The problem with literary cognitivism as Richardson presents it, and as I have seen it develop over the past two decades or so is that it gathers a lot of interesting material together under the banner ”Let’s Understand Literature and the Literary Mind” and then doesn’t do anything particularly compelling with it. Place a lot of interesting cognitive materials into a bag called Literature does not magically turn them into a model of the literary mind. The interest remains with the discipline-specific pieces, not the whole. There is no whole, not in any deep sense.

It’s as though one undertook an inquiry into the nature of cathedrals, or, for that matter, cottages, steel mills, suspension bridges, whatever, by investigating the physical, chemical, and mechanical properties of stone, mortar, glass, wood, steel, and so forth. If you’ve got a taste for that sort of thing – and who doesn’t? – that can be a lot of fun, and you learn a lot. But just what it is that you learn about cathedrals as buildings, as sites of worship?

At some point in your inquiries you will learn, for example, why flying buttresses are necessary. And you will learn about the properties of the stone used to construct them. But why were cathedrals designed so as to necessitate the use of flying buttresses? Why have a large central room with a high ceiling and no supporting columns directly beneath that ceiling?

There is an answer to that question, but you are not going to find it by studying the properties of your building materials. And the people who’s primary interest is in the properties of materials have no reason to have any professional interest in why cathedrals are built with high ceilings as that reason tells them nothing of interest about the phenomena they study.

* * * * *

I am reminded of a remark that Dan Dennett made some years ago about biology. He inserted it into a parenthesis in a 1994 article [2] where he gives an overview of this thinking up to that time (p. 239):
It is only slowly dawning on philosophers of science that biology is not a science like physics, in which one should strive to find 'laws of nature', but a species of engineering: the analysis, by 'reverse engineering', of the found artifacts of nature - which are composed of thousands of deliciously complicated gadgets, yoked together opportunistically but elegantly into robust, self-protective systems.
I am entirely in agreement with his emphasis on engineering. Biological thinking is “a species of engineering.” And so is cognitive science and certainly the study of culture in general. But this sense of engineering seems weak in second-revolution cognitive science and certainly in cognitive criticism.

But it is not weak in computational cognitive science. Working with those models requires a strong sense of design and structure. “Architecture” is a term of art in computer science in general and “architects” are recognized as being responsible for high-level design of computational systems, whether hardware or software.

I don’t see any architects among the literary cognitivists or the literary evolutionists. One might object that architecting a literary mind is pretty much like architecting the mind in general and trying to do that is crazy. That is true on both counts, on the scope of project and on the intention and desire.

But one thing architects do is studying existing buildings. And one way you study them is by drawing pictures of them. Those pictures capture (some aspect of) a building’s form.

And that – you can see this coming, can’t you? – is why I study literary form, even drawing pictures of forms. The inner mechanisms of the literary mind are resistant to observation. We have lots of proposals for bits and pieces of mechanism, but little sense of how they work together. But we do know that they work together in literary texts, so why not study the forms of those texts, seeing them as traces of the workings of the inner mechanisms.

Form, the Mind, and the Group

And so once again I’m beating the drum I’ve been pounding for years: analyze, describe, and study form, not simply in bits and pieces, but whole texts, end-to-end. I believe that those descriptions, in addition to being basic to the development of our knowledge of literature itself, are also an important contribution to a trans-disciplinary understanding of the human mind in general.

The cognitive, neural, and evolutionary sciences are all built on specialized inquiry into narrow domains. That’s how science works, that’s how knowledge works – mastery of intricate piles of details. Somehow all those detailed and specific investigations and models must be assembled into an overall account of the mind. And to do that you would think we need to observe the overall mind, at work, day in and day out?

How can we do that? Can we stick a person in a portable fMRI machine and have them wear it as they go about their daily life? No, there is no such device. And if there were, the data it would pump out would exhaust the internet. Nor can we have a dozen drones following people, each with its own set of observational instruments sending data back to a team of investigators.

But, to a first approximation, writing a poem or telling a good story engages a wide range of human capabilities in an integrated fashion. Keith Oatley argues that literary texts are simulations [3], in the computer science sense of the term, of living life—and a half century ago Susan Langer talked of the arts as providing virtual (her word) experience [4]. Whatever the mind’s mechanisms are, a whole bunch of them are in use when making or comprehending art.

We – the human sciences – can use artistic activity as a proxy for the whole mind in integrated action. We CAN study novels and poems, paintings and sculptures, songs and symphonies, and so forth. Not only can we study the works themselves, we can study people comprehending them, we can even image brain activity while they’re doing. We can do, and are doing, all of that.

Now, a description of a poem certainly IS NOT a description of the mechanism that built or comprehends that poem. Not at all. But it does place constraints on proposals for such a mechanism, providing, of course, that the description is couched in terms that are commensurate with those of the “tinker-toys” from which one builds mechanisms.

At the same time proper descriptions also provide, in a way, an answer to the question: What does it do? What does it do? you ask. Why it builds things like THAT, where “that” indicates, not the poem directly, but the proper description of the poem. That description, by its nature as a description, reveals something about the poem’s design.

Those descriptions in my previous post, of “Kubla Khan”, of Heart of Darkness, and of President Obama’s eulogy for Clementa Pinckney, describe objects created by the whole mind and work, and intended to be taken up by whole minds. In the case of Obama’s eulogy roughly 5,000 of those minds were directly present when Obama delivered the eulogy and we can track the responses they made – the clapping, the shouts, standing up, the tears (we can see people crying in the video) – and correlate them with the eulogy itself.

While some literary critics will write as though texts exist for them (I have an example in the next section), they aren’t. Texts are written for ordinary readers. Before there was writing, stories were told to readers who were physical present, as were poems. They provided a vehicle through which people’s minds were coupled into a single dynamic interaction.

I have argued that explicitly in my book on music, Beethoven’s Anvil [5], and, less formally, I have made the same argument about literature in an open letter to Steven Pinker [6]. And it is the form of the text that guarantees the coupling [7]. Now consider this passage from Bruno Latour’s Reassembling the Social [8], intermediaries and mediators (p. 39):
An intermediary, in my vocabulary, is what transports meaning or force without transformation: defining its inputs is enough to define its outputs. […] Mediators, on the other hand […] transform, translate, distort, and modify meaning of the elements they are supposed to carry.
Given that distinction I suggest that we treat the ‘form’ of a text as an intermediary while it is the ‘content’ that is a mediator. We have now linked the notion of computational form with their existence as social objects.

THAT is why the notion of computation is central to literary study. It gives us clues about the operations of the literary mind, and gives us another way of thinking about how the literary mind is also a ‘collective’ mind. Far from being “effectively disembodied” it shows us how texts can bridge the gap from the privacy of individual minds and the public requirements of collective society. Now only are texts embodied in the brain through computational processes, but they embed individuals in the group.

I have no particular reason to believe that the other human sciences are ready to recognize the value of high quality descriptions of literary texts, and that’s not why I do it. But, as the more imaginative and investigators feel the need for a more comprehensive view of human psychology – was William James’ Principles of Psychology both the first and last comprehensive psychology we have? – they’ll look to the study of the arts, including the literary arts, as a way of exploring the interaction of our manifold psychological capacities. Well-constructed descriptions of texts will be essential to that work.

At the same time literary study can contribute, not only to the general study of the human mind, but to the study of computation as well. It is not as though the nature of computation is given and one can look it up in a handbook somewhere. Understanding literary texts as computational objects will surely enlarge our general understanding of computing.

Meaning and Ethical Criticism

What, then of meaning? What would this naturalist criticism, as I’ve come to call it, say about meaning? After all, it was meaning it was tracking down when I began my work on “Kubla Khan” and it was meaning I was modeling with cognitive networks, no?

Yes and No.

Yes, I was after meaning when I began work on “Kubla Khan”, but that’s not what I found. But I found form, and a deep relationship between form and meaning. For it is the poem’s form that places the final line of the first movement – “A sunny pleasure-dome with caves of ice” – into the middle of the second movement as “That sunny dome! Those caves of ice!” And it is the poem’s form that line playing the same role in the second movement that the pulsing fountain plays in the first movement.

And, while I may have been after meaning in my work on “The expense of spirit” what I found was semantics, which is different. A semantics dumb as rocks. As I emphasized in my previous discussion of that work, that cognitive network model is not an interpretation or reading of the poem. Those diagrams do not represent meaning. They represent an embodied neural structure which, when activated in the course of one’s life, reading texts included, gives us the experience of meaning.

Meaning is thus subjective, not in the usual sense of varying from one individual to another, often wildly and idiosyncratically so. It is subjective in the sense of pertaining to subjects, existing within a subject’s mind. Meaning thus cannot be examined from the outside. Meaning may also vary idiosyncratically from one individual to another but only at the margins of our collective life. For without substantial agreement on the proper usage of the signs and symbols that pervade cultural life, collective life would be impossible.

But it is precisely at the margins where criticism in the sense of aesthetic and ethical judgment is needed. For that is where new collective understandings emerged. Such things cannot be imposed from the outside through the tools of objective investigation; they must be negotiated within the group. Critical exploration of those subjective meanings can modify and extend collective understanding.

Consider my standard passage from Kenneth Burke’s essay, “Literature as Equipment for Living” from The Philosophy of Literary Form (1973, but originally collected in 1941 and published in the 30s). Using words and phrases from several definitions of the term “strategy” (in quotes in the following passage), he asserts that (p. 298):
[...] surely, the most highly alembicated and sophisticated work of art, arising in complex civilizations, could be considered as designed to organize and command the army of one’s thoughts and images, and to so organize them that one “imposes upon the enemy the time and place and conditions for fighting preferred by oneself.” One seeks to “direct the larger movements and operations” in one's campaign of living. One “maneuvers,” and the maneuvering is an “art.”
It’s at the cultural margins where people’s experience evades and exceeds current understandings. That’s were literary texts can help. Critics can, in the Socratic manner, be midwives in the birthing of new cultural forms.

Human ‘nature’ is open-ended in the way that board games such as chess are open-ended. Biology provides the game board, the pieces, and the basic rules. But the strategy and tactics are culturally elaborated. One must follow the biological rules, but there are many many games one can play and still keep true to those rules.

One can have objective knowledge about the nature of the game board, the pieces, the basic rules, and about how the game has been played in the past. One cannot, in general, have objective knowledge about how the game could be played in the future much less about how it should be played. The future offers many possibilities. There is no one best way that is inherent in the game. There may even be alternative ways that are contradictory. How does one choose?

The ethical critic’s job is to aid in the choice – I have this from Wayne Booth, The Company We Keep: An Ethics of Fiction (1988) – by exploring life’s possibilities through the explication of literary texts. If that’s what you want to do with a text, well a given text can support multiple and even contradictory explorations. The critic who wants to do this is welcome to gather support wherever she will, from whatever disciplines, hard, soft, or in-between. But the upshot will always represent an ethical choice, not an objective knowledge. Nothing is to be gained by confating the time.

It is time the profession made the distinction. Not once and forever, but rather in recognizing the necessity of the distinction, to provide the ground on which it can undergo perpetual renegotiation. In the course of that renegotiation new cultural forms will emerge.

Understood in this way, ethical criticism is a separate activity from cognitive criticism and from the formal analysis that I advoate. The ethical critic is free, of course, to call on whatever ideas (s)he finds congenial but is not obligated to any of them. I assume that ethical critics would continue the various existing lines of politically-inflected critical discouse, because those discourses ARE ethical criticism. And cognitive critics are free practice ethical criticism as well. But the claims they advance as ethical critics will be different in kind from the claims they advance as cognitive critics. Critics of either kind are free to draw on formal analysis and description.


[1] Alan Richardson. Reframing the Adjustment: A Response to Adler and Gross. Poetics Today 24:2 (Summer 2003), pp. 151-159.

[2] Dennett, Daniel C. "Self-Portrait." A Companion to the Philosophy of Mind, edited by S. Guttenplan. 236-244, Oxford: Blackwell Press, 1994.

[3] Oatley, Keith. Such stuff as dreams: The psychology of fiction. Oxford: Wiley Blackwell, 2011.

[4] Langer, Susan. Feeling amd Form. Charles Scribner and Sons, 1953.

[5] Benzon, William. Beethoven’s Anvil: Music in Mind and Culture. Basic Books 2001. Download final drafts of chapters 2 and 3, in which I explain coupling:

[6] Benzon, William. An Open Letter to Steven Pinker: The Importance of Stories and the Nature of Literary Criticism. (2015) 20 pp. Download:

[7] Literary Morphology: Nine Propositions in a Naturalist Theory of Form. PsyArt: An Online Journal for the Psychological Study of the Arts, August 2005, Article 060608. Download:

[8] Bruno Latour. Reassembling the Social: An Introduction to Actor-Network Theory. Oxford UP, 2005.

No comments:

Post a Comment