Peter Swirski, Of Literature and Knowledge: Explorations in Narrative Thought Experiments, Evolution, and Game Theory (London: Routledge, 2007), 196pp. £60 hb/£18.99 pb. ISBN 978-0415420594.
Fictions – novels in particular – embed so many incidental details in the process of telling their stories that they could act as repositories of information about particular places and times. Yet rather than simply a container for other types of knowledge, Peter Swirski wants to make a stronger case. The aim of Of Literature and Knowledge is to persuade us that “literature is a form of knowledge” (p. 6). But what sort of knowledge is it? He immediately gives us a clue when substitutes this claim for one he thinks “amounts to the same thing” – namely, that literature “can generate knowledge while coursing through the minds of its creators and/or consumers” (p. 6-7). (The extent to which you agree that these two formulations are equivalent will do much to determine how persuasive you find the rest of the book.) “That we learn from stories is a truism,” he says. The book sets out to explain what, and how.
Swirski’s central premise is that literature functions as thought experiments function, there is, he thinks, “a cognitive isomorphism between literary and philosophical counterfactuals” (p. 36). By arguing (contra John Norton) that thought experiments do contain new knowledge (rather than simply rearranging what we already know), so literature (as mode of thought experiment) must also contain knowledge.
But although thought experiments are central to his account, Swirski certainly isn’t a single-issue campaigner. Of Literature and Knowledge takes in modelling, thought experiments, evolutionary psychology, and game theory – exploring each topic in turn, and culminating with an extended analysis of Stanislaw Lem’s Memoirs Found in a Bathtub which puts into practice a game-theoretical analysis along the method he has laid out. The ultimate aim of Swirski’s (ongoing) larger project is to change the way we conceive of fictions, and adapt the study of literature accordingly.
In common with several other theorists calling for methodological revisions in literary study (Joseph Carroll, Franco Moretti) – Swirski believes criticism has taken some wrong turns (one assumes they wouldn’t be doing this otherwise). In particular, he is dismissive of the exegetical function of literature (p. 39), which he seems to view this as so much imposition (which it often is). But if he’s sceptical of criticism-as-cryptography, it’s not because he doesn’t believe the hidden meanings and mechanisms exist: he wants to make a distinction between the meanings imposed by exegetical critics, and the actual meanings – real mechanics which, of course, he thinks that his method might disclose.
Thought experiments are a particularly fruitful way to think about the type of knowledge literature embeds and performs. The scenarios sketched out by ethicists are certainly narratives, and sometimes, existing fiction is co-opted for the moral dilemmas it sets up. Swirski points to Styron’s Sophie’s Choice as a fictional case that is used to test ethical precepts. Be that as it may, that they are literary narratives is neither here nor there. The morally salient aspects of Sophie’s Choice do not require a whole novel, they can be sketched out in a one-line pitch: mother forced to choose which of her sons to save.
The potential reduction to plot points is a troubling consequence of this sort of approach: what makes a work literary is not just the sequence of events, but the precise words chosen to capture that sequence. Stripped to their narrative functions, the difference between Hamlet and The Lion King is all but invisible. It sounds like we might have gained some additional epistemic status for fictions at the cost of literary merit.
But (and here’s a clever move) Swirski believes reduction – to paraphrase, to a set of plot points – will do grievous damage to the knowledge literature contains. “Literary fictions are not mere conveyance belts,” Swirski asserts, “for propositions that could easily have been expressed without the ‘redundant’ encoding in narrative” (p. 28). That might sound silly at first: literature is nothing but a set of propositions. How’s the extra knowledge getting in? The special element here is the reader. It is when reading that the non-propositional knowledge is (implicitly) transacted through first-hand engagement of reader and text. Arranging propositions into narrative order adds value.
How much of this is added value is within the control of the author and how much a property of the text-reader dyad is unclear. Certainly, Swirski is eager to credit authors of fiction with profound insights into human nature which are only now beginning to be recognised by the sciences. Citing the actions of characters in a Poe story as evidence that “fiction tolerates and even embraces irrationality” (p. 119), Swirski thinks it “worth noting that Poe’s far-sighted fiction appeared a hundred years before the social sciences began in earnest to model human beings as less than monolithically rational calculators… we have much to learn to from the theorists who work in the medium of art” (p. 119).
It is true that cognitive scientists such as Herbert Simon and Gerd Gigerenzer have only recently begun to account for human irrationality, but less valid to claim fiction got there first. Swirski is able to credit authors with this precocious knowledge of human cognition because he has previously equated (behavioural) modelling in the social sciences with characterisation in fiction. This redescription of fiction as modelling is generous, but it ought to be clear that a successful description of effects (which is what Poe is doing) doesn’t imply an understanding of the causal (cognitive) processes that produce them (which is what Gigerenzer and Simon are trying to do). If Poe’s characters behaved with limited rationality this is because Poe is an accurate observer of human behaviour, not because he had anticipated by a century the cognitive scientists’ work on bounded rationality. There’s a significant difference between a description of symptoms and diagnosis.
This point is exemplary of several quite slippery moves that Swirski performs in order to get as much knowledge into literature as possible. Perhaps the most dubious of these is the manner in which he equates what is effectively tacit knowledge in the arts with explicit knowledge in the sciences. As anyone who had tried to express their “jist” knows, translating the implicit into the explicit is challenging. But collapsing the two licenses Swirksi to identify knowledge as being in an object rather than of it. “More than metaphorically,” he writes, “living beings are theories about nature” (p. 85). He credits psychology with admitting that biological adaptations are “no less than biological knowledge, and knowledge as we commonly understand it is but a special case of biological knowledge” (p. 82). We do, of course, talk of how nature “knows” how to solve a problem through adaptations, but this type of knowledge is only weakly analogous with the usual sense of the term.
Swirski seems to be riding on the notion that implicit knowledge holds epistemological parity with explicit knowledge, and helping himself to the consequent gains. By dissolving this distinction early in what is a well-sustained, book-length argument, he is able to argue forcefully for literature’s capacity to contain and transmit much more than incidental facts about history, society, and science.
These criticisms are well-intentioned. It is a testament to the cleverness of Swirski’s arguments that one is often suspicious of the conclusions, but unable to articulate precisely the objection or identify the presumed deviation from good sense.
Of Literature and Knowledge isn’t the definitive statement on the epistemology of fiction that its title suggests, but it is a complex and usefully provocative contribution to the field. Laudably ambitious and readable, it deserves to stimulate debate among those of us trying to work out the place of literary studies in a predominantly scientific age.
Jon Adams, London School of Economics.