such a view of genre also takes social purposes at their word, thereby ignoring why certain social purposes exist in the first place as well as what institutional interests are most served through these purposes and their enactments. According to Terry Threadgold, genre theory is significant because of the relations it reveals between genres and institutions, power, the construction of subjectivity, as well as “the relations it permits/enables/constrains and refuses between readers and writers, textual producers, and receivers” (102). Threadgold’s critique hinges on SFL genre theory’s use of genres as a starting point for textual analysis while overlooking the “web of social, political, and historical realities” in which genres are enmeshed (106). As Threadgold elaborates:
What we need to know is how institutions and institutionalized power relationships and knowledges are both constructed by and impose constraints on (and restrict access to) possible situation types and genres. We need to know why certain genres are highly valued, and others marginalized. We need to understand the changing history of such valorizations. We need to know why some genres are possible, others impossible, ways of meaning at given points in history. We need to know how and why these factors construct identities for social agents . . . and how and why some social agents are able to/willing to resist and others to comply with existing situational and generic constraints. (106)
At the same time, Threadgold, following Derrida, also argues that because texts are always performances of genres, genres are less stable than SFL approaches imagine: “Genres and system cannot therefore have static, fixed values, and the extent to which they are predictive of choices in lexico-grammar is constantly subject to slippage and change . . .” (116). Indeed, as Brian Paltridge has demonstrated in his analysis of Environmental Studies research articles, genre identification depends more on contextual cues and interactional and conceptual frames than on structural and linguistic patterns (Genre, Frames and Writing 84-85).
While such findings raise questions about SFL approaches to genre pedagogy, the research and debates within SFL genre approaches have been crucial in establishing how genres systemically link social motives/purposes to social and linguistic actions. By arguing for genre as a centerpiece of literacy teaching, SFL genre scholars have debated the ways genres can be used to help students gain access to and select more effectively from the systems of choices available to language users for the realization of meaning in specific contexts. In the next section, we will discuss how scholarship in historical and corpus linguistics has also informed work in genre study.
Genre and Historical/Corpus Linguistics
Although work in genre within historical and corpus linguistics has not yet had a great impact on rhetorical genre theory and Rhetoric and Composition studies (with the notable exception of Amy Devitt’s work), it has much to contribute to research and teaching of genre by accounting for the nature of typology and for language change. In this section, we will examine debates over genre categorization within historical and corpus linguistics, and how such debates might help clarify confusion between genres and modes within Rhetoric and Composition studies. Then we will examine how corpus based studies of genres provide insight into how and why genres change.
According to Hans-Jürgen Diller, the field of historical linguistics became interested in text classification when it expanded its scope of study from sentences to texts (11). Within text linguistics, Diller describes two trajectories of classification: Deductive and Inductive text typologies, which parallel in some ways the difference between what Todorov describes as analytical (or theoretical) versus historical (or empirical) approaches to genre classification, described in Chapter 2. Deductive text typologies, which Diller represents through the work of Robert Longacre (The Grammar of Discourse), seek to create over-arching categories for genre and text classification in a way similar to how Northrop Frye sought to identify universal archetypes in order to classify and describe relations between literary texts. Longacre, for instance, “bases his typology of ‘Notional’ or ‘Deep Structure’ text types . . . on the ‘notional categories’ which in his view underlie human language” (Diller 12). The four “notional text types” (or modes) for Longacre are Narrative, Expository, Behavioral, and Procedural, and together they overarch and help categorize surface text types which Longacre calls genres (Diller 12-13). For example, the Narrative mode overarches genres such as fairy tales, novels, short stories, newspaper reporting; the Procedural mode includes such genres as food recipes, how-to books, etc.; the Behavioral mode includes essays and scientific articles; and the Expository mode includes sermons, pep-talks, speeches, etc. (Diller 13).
Rather than starting with apriori categories, inductive text typologies classify text types based on perceived textual patterns. Douglas Biber’s work in corpus linguistics has most influenced such an approach to genre classification. Corpus linguistics, using large scale electronic text databases or corpora, allow researchers to conduct systematic searches for linguistic features, patterns, and variations in spoken and written texts. In Variation Across Speech and Writing, for instance, Biber begins by identifying groups of linguistic features (what Biber calls “dimensions” such as “narrative versus non-narrative,” “non-impersonal versus impersonal style,” “situation dependent versus elaborated reference”) that co-occur with high frequency in texts. Then, applying these dimensions to a statistical analysis of a corpora of twenty-three genres, Biber examined the degree to which these dimensions appear within various texts in each genre. Based on such studies, Biber has been able to identify a great deal of linguistic variation within genres, suggesting that genres can be defined in terms of more or less complexity.7 (For more on Biber’s analysis of textual clusters on the basis of shared multi-dimensional, linguistic characteristics as well as his historical work mapping the rise and fall of genres, see Dimensions of Register Variation: A Cross-Linguistic Comparison; Discourse on the Move: Using Corpus Analysis to Describe Discourse Structure; and Sociolinguistic Perspectives on Register.)
This notion of “more or less” has played an important role in historical and corpus linguistic approaches to genre categorization. Based on Eleanor Rosch’s theory of prototypes, which takes a psychological (as opposed to a classical) view of human categorization, such a typology identifies membership within genre on the basis not of “either-or” but on the basis of “more-or-less, better and poorer” (Diller 21). As Brian Paltridge explains, prototype theory describes how people categorize objects according to a prototypical image they have conditioned in their minds by socio-cultural factors, while classical theories describe categorization based on shared, essential properties within objects that result in objective assessment of category membership (Genre, Frames and Writing 53). The famous example in this case is the way some birds, such as sparrows, are “birdier” birds than others, such as ostriches. The notion of prototypes, related to Wittgenstein’s idea of family resemblances, allows genre researches to define text membership within genres on the basis of how closely their structural and linguistic patterns relate to the genre prototype. Some texts, thus, are closer to their genre prototype while others function more on the periphery of prototypicality, or, more accurately, on the boundaries of different prototypicalities, as in the case of mixed genres. The important point here is that the relation between texts and genres is not simply based on features internal to both, but more powerfully is based on learned, conceptual relations between “memory, context and frames,” thus rendering “the notion of prototype as a principle of selection, organization and interpretation of genre frames” (Paltridge, Genre, Frames and Writing 62).
Prototype theory has important implications for genre study and teaching. Within SFL genre theory, J.R. Martin has used it to distinguish between typological and topological genre classifications: “For purposes of typological classification, we have to define just what percentage of causal relations is required for a text to qualify [as a member of the genre]. The topological perspective on the other hand allows us to position texts on a cline, as more or less prototypical . . .” (“Analysing” 15). The topological approach thus allows SFL approaches to genre teaching to use the teaching-learning cycle to move students towards more and more prototypical genres through sequenced assignments. At the same time, corpus linguistic-based analyses of genres have allowed researchers and teachers working in English for Specific Purposes (as we
