Continuing my progressive descent into vulgar materialism (I use the words "progressive" and "vulgar" in positive senses!), I’d like to develop the line of thinking of my previous post, "Reading under Neoliberalism." I will use the questions Joel Burges asks in a comment to guide my reflections here. His questions are too good to cosign to the comments section of my previous post. I will begin with a caveat: everything below is, as with my previous post, provisional and only vaguely sketched. Critical comments will do much to help me sharpen my primitive ideas.
Joel asks whether my approach to literary study, at least the approach I take when I discuss historical changes in reading practice, is marked by an "empirical turn," an "operative assumption that we will know more if we get more empirical -- not just materialist in the sense of assuming that economic conditions lead to cultural elaborations, but in which we turn ourselves into something like sociologists." The short version of my answer is simply yes. Indeed, there is some reason to believe that the academic study of literature more generally is swinging away from the era of theory toward an empirical orientation, if recent studies are any indication. We might recall new work in cognitive science and literature; the rise of evolutionary literary studies; "distant reading" research programs, spearheaded at Stanford by Franco Moretti, and other database-driven forms literary study; Bourdieu-inspired literary sociologies (McGurl, Casanova, Jim English come immediately to mind); the "postpositivist realist" epistemology of Satya Mohanty and, here at Arcade, of Paula Moya; the myriad anti-theoretical children of Walter Benn Michaels (one need merely look at the 20/21 series for excellent criticism in this vein); and so on.
The longer form of my answer comes with numerous necessary caveats and complications.
"Is an empirical turn in literary studies a turn away from theory, from, say, bridging textual analysis and conceptual thinking?"
This question assumes a stronger distinction between the empirical and theoretical that I am comfortable with. After all, isn’t the work of Bourdieu both thoroughly empirical and theoretical? Doesn’t Foucault make all sorts of empirical claims (ranging from claims about prison systems to claims about the history of science to claims about how discourse produces power relations)? Isn’t Lacan interested in correcting Freud’s fallacies, relocating psychic processes not in the minds of individuals but in relation to intersubjective processes of recognition and "within" structures of language? Do not Jameson, Žižek, Hardt, Negri, Laclau, Mouffe, and a range of theoretically sophisticated Marxists and post-Marxists all base their arguments, at least in part, on empirical claims about capitalist economies?
Likewise, all empirical studies are, I would argue, necessarily suffused with theoretical abstractions. Joel correctly identifies many of the abstractions I rely on to make my case: "literary market," "reading public," "sophistication," "literary culture," "postwar." There’s no way to study the world apart from our abstractions, theories, and interpretations, even if those interpretations are the translation of photons hitting our optic nerve into terms discernible by our cultivated mental capacities. The question is, What are our best theories? What theories should we reject?
The theory I reject is the notion that we should see in literary form an elaboration of material contexts on the model of homology. The theory I accept is that texts and contexts are dynamically linked together in a greater whole or totality, whose determinants do not necessarily operate according to a logic of homology. Causes do not necessarily look "like" effects. To the degree that “theory” in the academic humanities tends to refer to the former of these two intellectual frameworks, then I do reject theory, though in a partial and highly qualified way. I am more interested in "mechanical causality" than "expressive causality," to use Jameson’s terminology in The Political Unconscious.
"Is an empirical turn in literary studies a turn away from hermeneutics, from, say, textual analysis -- and what would we gain from that?"
I don’t see how we can avoid hermeneutic activities in the classroom as long as we ask our students to read individual texts -- I tend to teach individual texts in much the same way that they were taught to me -- nor do I think that there is some simple empirical practice apart from interpretive, cultural, and historically situated frameworks. That said, I think a lot of self-avowedly materialist criticism and theory today makes large empirical claims without doing the legwork to back up those claims. That’s what I take to be the source of Moretti’s frustration with literary study.
In our monographs and articles we have a habit of sliding between perfectly valid hermeneutic claims and large historical claims, often based on three or four close readings, often without explanation or with vague gestures toward some notion of discourse. This is the academic version of what the journalist Daniel Radosh calls “trend journalism” -- three examples of anything can be selected to argue for a historical trend. If we supplement textual analysis with an empirical orientation, we will possibly learn more about the material determinants of literary history and we will also learn what claims we should not be comfortable making with great confidence. Like Socrates, we will at least know what we don't know.
"Literature departments are... notoriously bad at making the normative and conventional ways in which their members read and write clear to students… So… shouldn't we also examine what knowledge we already transmit, and how we might do it better?"
Yes, I enthusiastically agree that we should study the normative and conventional ways we read and teach. We should understand how and to what effect we transmit knowledge to our students.
Indeed, my interest in empirically analyzing postwar literary culture is motivated by explicitly normative concerns. I begin from the premise that certain practices of reading are good and desirable. Reading long, complex novels is salubrious for human wellbeing. Cultivating the attention required to understand and appreciate poetry improves us. Literary reading gives scope and depth to life. These claims are normative -- and not strictly instrumental -- to the degree that they have no foundation. No empirical study will be able to prove to a persistent skeptic that literature matters. No data beyond self-reporting will explicate words like "wellbeing," "improvement," and "scope and depth."
My second assumption -- really, in a longer work, which I fantasize about someday writing, it would be my argument -- is that literary culture is unnatural, in the sense that it isn’t a spontaneous or inevitable development in human affairs and existence. We don't just decide to care about literature; and we don't automatically move from such caring to a society that enriches and supports what we care about. Our reading culture is, instead, the product of considerable investment, education, and political work. Humans may at all times have generated one sort of narrative art or another, but a society where all persons have the opportunity and capacity to appreciate literature requires hard work and years of institution-building.
If our empirical and critical work is grounded in the norm of producing such a "reading public," then we cannot help but self-reflectively understand our own teaching in relation to the broader project of the production of such a public. This doesn't mean that every critic would take or teach sociology and economics classes, but that every critic would understand that when they teach a course on Shakespeare, they are always whether they intend to or not linked to a larger public-producing machine, the University, which itself interlocks with other social spaces -- the book club, the marketplace, little magazines, and institutions of primary education.
Let me be the first to cheer you on, Lee, in this lucid and very brave exposition of empiricism for literary studies! I have been brewing up ideas like these myself, and I am thrilled to see your take on these essential issues for the future of the discipline. (Thanks to Joel's comment on your last, we've turned from the fate of literary culture to the fate of the discipline...I'm not going to address, for the moment, your ingenious way of connecting the two at the end of your post.) I completely agree that we will need much richer data, sociological (economic, anthropological, cognitive-scientific...) models and training if we are to pursue the kinds of questions that literary studies has been raising, and the kinds of answers we've been talking about in recent threads here. In pointing out the weakness of the "homology" model, you go straight to the heart of the issue. That model is incredibly convenient for professional textual interpreters--and it generated some of the most virtuosic New Historicist performances. But it's also obviously inadequate as an explanation of the relationships between texts and "contexts." Which suggests that, as Moretti has argued, the role of theory--not capital-T Theory, but theory worthy of the name--will lie in model-building and hypothesis-generation. (And, of course, that kind of theory is essential to empiricist scientific pursuits of all kinds.) Such theories will stand or fall on empirical work. Thus empiricism doesn't mean dispensing with abstractions but daring to--as they say over in the other divisions--"operationalize" and measure them as best we can.
So: sign me up; I can think of a few others who would sign up, too!
And now, since every silver lining has a dark cloud, a purely pragmatic hesitation. In Joel's comment I detect a troubled awareness that the common currency of literary studies ("the knowledge we already transmit") is in fact textual interpretation. And the criteria of merit are some mixture of originality, relevance to shared concerns of the discipline (politics, ethics, aesthetics?), virtuosity, and plausibility. These are, I'd say, the key ingredients when it comes to getting journal articles or book manuscripts accepted; but they seem to have little to do with the aim of the kind of work you're sketching here, where the tests would have to be something more like: Is the data-set adequate to bear out the explanation offered? Does this add to or revise existing explanatory paradigms? Can the model be generalized to further applications? Could the claims be replicated? --In other words, social-scientific criteria of value (not that I think those are cut-and-dried!). How would the institutional structure of the discipline have to change to support "empirical" work in literary studies? Or am I wrong--no changes are needed to get this going, and I should stop making excuses for myself?!
More to come, but in the meantime, to Andrew: yes, you more perfectly articulate my question, especially in the third paragraph above. I am increasingly an advocate of integrating these different models of explanation--with a particular interest right now in what David Bordwell calls "historical poetics," but also other models--with our more routine models. I have to say, I do continue to believe in textual interpretation and Theory, as long as the project's aims are to some degree self-consciously articulated within the parameters of those models. This is in part personal: one of my strengths is to blend these two models. At the same time, I think the claims the blend of those models enable can be a bit overwrought. On the pedagogical side of things, my point is that we don't in the classroom always make our models very clear: instead, we just interpret in front of our students, and assume they know how we do it--perhaps because close-reading is part of high school, but more likely because literary critics are often hostile to articulating in what way their forms of knowledge-production are not original and virtuoso, but procedural, the product of conventions and testable models. Indeed, even when the model is textual interpretation or Theory. Anyway, I will try to say more about this in another post when I have more time to elaborate at length.
Thanks Andrew and Joel for your spot-on questions. How will our institutions, practices, and theories (or Theories?) change in an empirical era? It's hard to say in advance -- and given that what gets called Theory is a hugely diverse undertaking, even harder to say systematically -- but the proto-empirical studies I cite in my post give some hints of what a new landscape of literary study would look like. In short it wouldn't look structurally too different from what currently exists, but it would feel more methodologically historical, stylistically demotic, and ontologically modest.
Among the studies listed above, one approach has been to learn about some scientific discipline (cognitive science, evolutionary biology, computer science) and find ways to "apply" insights from those determinate fields to existing literary questions, often transcoding terminology from the former field into the latter. I don't think anyone who is doing interesting work in this vein is having any unique trouble publishing that work, so little institutional transformation would be required if that's our future. Another approach, more in the Moretti vein, would change English departments into collective enterprises, on the model of the natural sciences. Graduate students would join research projects in graduate school, work with faculty advisers, and publish co-authored studies that are based on the work of multiple people. Dissertations would become sorts of afterthoughts to research. I doubt this approach will grow beyond a dedicated niche of researchers, but I might be wrong.
I think there is a third approach, between these two poles, which acknowledges the fact that what many of us are interested in isn't literature in the abstract -- the system of genres, the economics of literary production -- but particular literary works, movements, and periods. This third approach would require less an institutional revolution than an epistemological correction. We would displace a charismatic Theory with numerous small secular "theories" (indeed, Edward Said's "secular criticism" very much resembles what I have in mind). We would ask graduate students to do archival research when appropriate, to explicitly avow the models of causality they employ, to state the stakes of their projects more overtly, and to learn how to challenge large claims whose evidentiary bases are inadequate. This might lead to a more modest sort of literary study, where virtuoso readings remain one sort of thing we produce but where the inferences that we feel comfortable drawing from those readings are circumscribed.
Brilliant post, once again. I have long found it mystifying and vexing to see scholars invoking arguments clearly based on some kind of observation of the world in their own work, only to turn around and dismiss results from cognitive science and social science as irrelevant, "vulgar," etc. These are just the same questions, folks, only better put and asked of more people!
So hooray for the integration of cognitive-scientific results into literary study. Who knows, we might even solve some theoretical questions, instead of endlessly citing authorities at each other. Of course, it would be nice if we could stay away from fruitless data-mining (studies that show for example that the nobility in novels have, surprise surprise, houses in the country) and tendentious handlings of data (certain studies of Sherlock Holmes come to mind)... But the cog. sci. integration at least seems like it has a bright future, even if the data-trawling forms of empiricism may or may not.
What seems to be at stake across all the responses here, from Lee's and mine to Andrew's and Josh's, is the impulse to become more reflexive about what theory can do, our models of causality, and, indeed, our grammars of motives in implementing them. Indeed, I lay in bed this morning mapping out my own. But my work is not the subject here, even as this conversation has prompted me to think through my own model of causality.
1. I am totally on board about rejecting homologies as a causal model, unless, that is, the critic-scholar is self-conscious about his or her homologous model. My issue is when that self-consciousness vanishes: indicate what model you are using and the purpose it has for you, and no--demands of publishers and tenure committees aside--don't make unduly broad historical claims based on something that is unable to do that. If your purpose is Theoretical, then great: name your concept, and "use" the texts to explore it, but still give a reason for why those texts add up to a good way of exploring that concept. If your purpose is theoretical, then great: name your model--contemporary American literature is largely the product of the R&D university--and test it, seeing both how the specific claim and the more general model here works.
2. I'd advocate something like the same in the classroom. Explain your model, your approach; explicate the conventions--hermeneutic, empirical, poetic, Theoretical--that the discipline uses to think about fiction, or language, or whatever really. As a result, our students may better understand what we do in the humanities, especially literary studies, and appreciate it by its difference from what admittedly more objective fields do pursue. We need to name the model, or models, we are using, both their disciplinary applicability and their idiosyncratic status. Like it or not, procedural knowledge is something we do all the time; conventions are endemic even in a field that studies objects prized for their distinction and originality in some cases. We need to clarify this to our students. Indeed, I think that doing so--naming the model--is part of the struggle to preserve and to create a future for the humanities through our students.
3. The relationship between the discipline of literary studies and the broader reading public is vexed. My way of looking at, and even appreciating a film or show or text, is often alien to how a "non-critic" does the same thing; indeed, it is often tensely different (thus that adjective, "vexed"). This seems like a good thing to me, but both sides of the equation, though especially ours, could perhaps again benefit from naming the model of discipline versus reading public: the discipline likes explanation and interpretation, the public something closer to evaluation.
At any rate, thanks for the useful debate--which may in the end have nothing to do with empiricism, but with, to repeat the phrase yet again, naming the model.
And by the way, I include grad students among those who we could teach in much more explicit ways regarding models, conventions, and procedural knowledge, and not just through discussions of theorist X's idea of explanation, but quite literally too in courses that take up the rhetoric--understood quite robustly--of the discipline in order to explicate how to write in and challenge the models of that discipline.
Despite the fact that "formalism" has long been a term of art in literary studies, I don't think we've given nearly enough attention to the painstaking analysis and description of form. We invoke form for various reasons ("classically," as a way to make the text into a self-contained packet of meaning), but the analysis and explication of meaning has totally swamped our attention to form. I think the painstaking description of form is a worthy aim of literary study and would, by comparison, note that Darwin would have been nowhere without the previous three of four centuries of work devoted to the mere description of forms of flora and fauna in the natural world. We need to undertake a similar enterprise.
I've taken a first crack at making the case for such study in my Literary Morphology: Nine Propositions in a Naturalist Theory of Form. I've got a more personal approach to the subject in a recent post at The Valve where I discuss how I fell through the rabbit hole thirty years ago while undertaking a structuralist analysis of "Kubla Khan" (if nothing else, look at the three illustrations and ponder the progression from 1 to 2 to 3). That didn't work out as planned. Instead I discovered that the poem had quite an elaborate structure that no one had noticed, despite all the attention devoted to it. And some of that structure could have been discovered simply by giving serious attention to how Coleridge used commas, semi-colons, colons, and periods at line ends. We were asleep at the switch.
Here, courtesy of Nina Paley, is a rather more compact statement about the current state of affairs in literary studies.
* * * * *
A point of comparison: We would displace a charismatic Theory with numerous small secular "theories" (indeed, Edward Said's "secular criticism" very much resembles what I have in mind).
Midway through the last century the sociologist Robert Merton argued for theories of the "middle range" (not quite his term, but it will do), theories with less scope than grand Theories of Everything (e.g. Marx), but with more scope than individual case studies.