In this article I offer some ways of integrating theoretical linguistics into a literature department. I write as a theoretical linguist who also teaches literature.The article is in three parts. First, I explain why I think linguistics has an place in literary studies; second, I show how some of the major differences between linguistics and literary studies are resolved in Strathclyde's 'Ways of Reading' class; and third I talk in general about ways of bringing linguistics into the literature curriculum.
It is easy to explain why some knowledge of language should be useful in literary studies. Knowing what a preposition is, or a gerund, or a partitive, makes the analyst more sensitive to the medium from which literary texts are made, and provides a vocabulary for talking about them. Being able to isolate elements of language enables you to see things which otherwise are below the threshold of attention: the first two lines of Milton's Paradise Lost begin with 'of', and thereafter a higher than chance proportion of lines in the poem begin with 'of'. When we now look at Tintern Abbey we find that 23 of the 60 instances of 'of' in the poem are line-initial, and we have something previously unexpected to say about the 'Miltonic style' of this poem. None of this requires linguistics; just standard descriptive English grammar will do, perhaps supplemented with some newer terminology, and certainly stripped of its more oppressive prescriptive characteristics. In contrast, theoretical linguistics is the scientific study of language, which seeks to establish what forms language takes and why it takes these forms. What can a theoretical linguist do for literature which cannot be accomplished by an a theoretical grammarian? To answer this question we must acknowledge two of the major discoveries made by linguists in the past fifty years.
The first major discovery is that a single utterance or inscription (e.g. a sentence, part of a text) is always mentally represented in several distinct ways. In their mental representations, a surface representation of the text may be quite different from an underlying representation of the same text; this is the distinction known to many by Chomsky's terms 'deep structure' and 'surface structure' and while the theory has moved on, the discovery that there are different representations of the same sentence remains valid: Chomsky's current 'minimalist investigations' are centred on the problem of why this distinction exists – why for example the phonetic form and the logical form of a sentence should differ. One example of the relevance of this for literary studies is poetic metre: the underlying metrical form and the surface rhythm of a text can be very different. Not only does the metre underdetermine the rhythm (hence allowing great rhythmic variety within a single metre) but it is also possible for rhythm to communicate a surface metre which differs from the underlying metre: hence English iambic pentameter can rhythmically mimic a Sapphic, or Christina Rossetti can write a strict metrical poem which mimics a ballad (1), or Auden a loose metrical poem which mimics various strict metres. The fact that an instance of text can have multiple representations demonstrates the systemic complexity of language as a source of formal and interpretive richness, and linguistics offers a way of understanding how this richness comes about.
The second major discovery is that there are different kinds of meaning: there is no unified 'semantics', but rather a collection of different kinds of semantics. This has been suspected and reported throughout the twentieth century, but the explanatory breakthrough was in the work of the linguistic philosophers, Austen, Grice and Searle, and in linguistics particularly in the 'relevance theory' of Sperber and Wilson (2). Most importantly, we now have an ontologically parsimonious and psychologically realistic theory of metaphor and of irony (and other forms of metarepresentation), which are explanatory not just for everyday language but also for literature. I have argued elsewhere that many elements of literary form (most obviously genre) can be explained under a theory of pragmatics, and that in fact these 'forms' are actually meanings communicated by the text, things the text tells us about itself (3). In these ways, formalism finds a new way to return to literary studies, via contemporary pragmatics.
Both of these major discoveries offer ways in which literary studies can understand how language makes possible the richness of literary texts. Neither of these discoveries has been as influential as it deserves to be in literary studies. Perhaps linguistics feels too grounded, while in contrast literary studies is free to roam, and thus appears to liberate students in ways not possible for linguistics. I will shortly suggest that linguistics has its liberatory side as well. In the next section of this article I will look at two of the major differences between linguistics and literary studies and show how they are practically resolved in a class at Strathclyde University.
In the late 1980s my colleagues and I created a class called 'Ways of Reading' (its title a homage to John Berger's influential book Ways of Seeing). At Strathclyde it is the only class from that period still taught, and its organizatory principles have come to influence our whole undergraduate English curriculum; it was also the basis of a Routledge textbook which has sold 28,000 copies and is about to go into third edition (4). I think the class and book are successful in part because they temporarily overcome two of the major differences between linguistics and literary studies. The first major difference is that linguistics is concerned with generalization, and literary studies with specificity. Linguists do not care about any instance of language in itself, but only in how that instance is evidence of underlying regularities, while in contrast literary scholars value specific texts. The second major difference is that linguistics proceeds by problem-finding and problem-solving, notions which are somewhat alien to many ways of teaching literature.
The ‘Ways of Reading’ class is organized around topics, and thus around generalizations: a class on metaphor, a class on parallelism, for example. Each week we have one lecture and one workshop, with the students following detailed instructions on a worksheet. We begin with a class about how to ask questions about a literary text, and this sets the tone for the class. I'll give an example of an actual class which has been used to illustrate how it works; this is a class on narrative where our goal is to teach the students about Propp's notion that narratives might be segmented into types of event (drawn as an ordered subset from an ordered set). The workshop broke into about 25 parts a summary of the film Salmonberries by Percy Adlon, and presented them in randomized order. The task was first to find the most 'realist' or conventional way of ordering the events; and then to produce an antirealist or postmodern narrative from the same sequence of events. The point here is not the text itself, which is derivative from a film which the students haven't actually seen, but the way in which this text can provide material through which students can learn about our expectations about event order, and how these expectations can be manipulated. The linguistics here comes partly via the explicitly formalist thinking behind the class, but more through the idea that by inventing and manipulating verbal material we can make discoveries about underlying principles. Some classes are a long way from linguistics in their content. In one session we look at the book as a physical object and the artist's book, and in the workshop give (cheap, used) books to students and ask them to alter them to produce a new object/text from them by any means except fire. Here again the point is experimentation, a procedure derived from the teaching of linguistics.There will be a new session in autumn 2005 on the aesthetics of 'going for a walk': the flâneur and the Arcades, the dérive and Debord's 'walk home', the Artist's Walk, skateboarding, and so on. There is no actual linguistics here, but even though linguistics is not the subject matter of the class, it is always a linguistics influenced class in the sense that it is always about generalization and the finding and solving of problems. Our choice of texts for each session is left to some extent to chance; this has to be the case, because our underlying assumption is that no textual practice is unique. Salmonberries works nicely for the 'narrative/Propp' workshop because its narrative presents some surprises as regards family relations (the bedrock of Propp's theory), but there are thousands of other narratives which would also have been good.The educational roots of the class come in part from Richard Kohl's 'open classroom' with its radical pedagogy (5), and in part from the problemsolving methodology used in linguistics departments. (6) While theoretical linguistics might be committed to idealization and the discovery of universals, and hence operates in some ways by excluding, closing down, focusing, and regulating, it nevertheless has paradoxically been taught in ways which can be liberatory for students, and can be adapted to literary studies.
In the final section of this article I will suggest some ways of putting linguistics into the literature curriculum. Sometimes at Strathclyde we do this by having an option class whose subject matter is linguistics or English language, but because our curriculum is so heavily option-based, we run the risk of having a class which few students will choose to do. Some students originally chose to do literature precisely in order to get away from the technical kind of thinking required for linguistics; others seek instead the fun they will get from reading literary texts; and others, driven by considerations of relevance, find linguistics too far from their everyday concerns.The traditional way in which linguists have nevertheless sought to satisfy literature students is to make the linguistics as relevant as possible. The traditional British 'stylistics' way of doing this, promoted for example by Simpson (7), is to seek always to show how linguistics gets you somewhere with a literary text. Thus I have taught journalism students Searle on speech acts by showing how newspaper headlines 'fake' speech acts for specific purposes; and have asked students studying creative writing to write and then comment on their own fiction exemplifying Goffman's notion of facework.There is no question that these exercises engage the students, though I think there is an intellectual cost incurred by too intensively seeking to make theory relevant for practice. There is however another way of making linguistics relevant. New media, which students might know better than we do, can raise interesting linguistic problems; one of our students is about to write a dissertation on the syntax of text-messaging, and another who will be working on the sociolinguistics of blogging. And at Strathclyde we have discovered a characteristic interest in the linguistic analysis of non-standard dialect. Perhaps Scotland may have a special status here; strong dialect loyalty struggles with rigid proscription of dialect, such that multi-dialectalism is very widespread, and language is very visibly tied to issues of nation and social class. Thus we find that dissertations on topics relating to Scottish English are quite common, even on the basis of small amounts of class-based instruction, and that students are willing to learn substantial amounts of linguistic theory in pursuit of topics such as 'The difference between aye and yes' or 'The language of (the soap) River City'.
Though it is always possible to find some students who want to study linguistics as such, it is not resource-effective always to offer a linguistics or English language option. Because of this, we are open to letting linguistics appear in any class.The possibility of using theoretical linguistics in this way exploits two characteristics of the discipline as originally formulated by Chomsky: it is modular, and it is rationalist rather than empiricist. The modularity is manifested by a willingness to separate out a problem and deal with it independently of some other problem. One of my favourite tasks with students at any level is to get them to see how far they can get in asking questions and formulating hypotheses by looking just at the first and the last sentence of a text, and for the moment to ignore the rest. Chomsky's rationalism tells us that there is no pre-theoretical arrangement of the data, and that we never know in advance what a theory will be able to explain, or that any particular theoretical model is guaranteed to be right. Instead, the theory must take a risk and work out for itself what data it is seeking to explain. Both of these characteristics make linguistics opportunistic and thus available for intervention at any point in a literature class. In one of the first lectures of our first year literature class, the issue of 'embedded' narratives arises when we discuss Margaret Elphinstone's complexly folded novel The Sea Road; this is the right moment to explain 'recursion' as a characteristic of linguistic systems, and to point the students to Chomsky's argument about recursion and the evolution of language. Chomsky created generative linguistics in opposition to earlier linguists' adherence to a methodology or pre-theoretical assumption about what a theory should be; while we are hardly Feyerabendians, this opposition to method is another of the liberatory aspects of linguistics. Opportunistic and 'modular' uses of linguistics carry over to the use of any language in the class. Linguists are used to working on any language, whether they know it or not, and focusing just on the problem at hand. In our first year class we use New Penguin Book of Scottish Verse (8) because it is so multilingual, with its Latin, French, Old English, Icelandic, Gaelic, Scots, English and even Welsh (the Gododdin) texts. In one session we expect our students – few of whom know the language – to discover the metres of the early modern Gaelic poems. This is an easy task even for those who cannot pronounce the words (just looking at the written vowels will do the job), but it does require the willingness to be nonholistic, to separate off a problem of form from every other aspect of the poem.
I have written about ways in which linguistics can influence literary studies, but I want to end by acknowledging that the linguistics is being undertaken as part of a literature degree. A popular research topic for our students is 'The language of Irving Welsh' (or Jim Kelman, or some other Scottish writer who represents non-standard dialect in an interesting way). I remind the students who do this work that though they are working as linguists, the novel's author is not; Trainspotting is not a work of sociolinguistics. The language of the novel has an indirect relation to the Edinburgh dialect it represents; Welsh is undertaking linguistics as fiction, rather than linguistics as a scientific enterprise, and the student's linguistic work on Welsh's language achieves complexity and richness when it can combine the opposing demands of linguistics with the literary. Linguistics is the search for simplicity, literary studies is the search for complexity, and in a literature degree the opposition between these is true friendship. In this article, I have suggested that linguistics can open up new kinds of complexity in the teaching of literature.
Acknowledgement
This article arises from an event I organized in Autumn 2004 for the Subject Centre for Languages, Linguistics and Area Studies. The event was called 'New methods in literary linguistics and their relevance for linguistics', and was administered by Jane Copeland; it included presentations by myself, Barbara MacMahon and Sylvia Adamson.
References
1 Nigel Fabb and Morris Halle, 'Metrical complexity in Christina Rossetti's verse', College Literature (in press).
2 Dan Sperber and Deirdre Wilson, Relevance: Communication and Cognition, second edition (Oxford: Blackwell, 1995).
3 Nigel Fabb, Language and Lliterary Structure: The Linguistic Analysis of Form in Verse and Narrative (Cambridge University Press, 2002).
4 Martin Montgomery, Alan Durant, Nigel Fabb,Tom Furniss and Sara Mills, Ways of Reading: Advanced Reading Skills for Students of English Literature, second edition (London: Routledge, 2000).
5 Herbert R. Kohl, The Open Classroom: A Practical Guide to a New Way of Teaching (New York: New York Review, 1969).
6 Maya Honda and Wayne O'Neil, 'Triggering Science-Forming Capacity through Linguistic Inquiry', in K. Hale and S. J. Keyser (eds.), The View from Building 20: Essays in Linguistics in Honor of Sylvain Bromberger (Cambridge MA: MIT Press, 1993), pp. 229-256.
7 Paul Simpson, Language through Literature: An Introduction (London: Routledge, 1997).
8 The New Penguin Book of Scottish Verse ed. by Mick Imlah and Robert Crawford (London: Penguin, 2001).
