I must note before beginning however that the article by Kamberelis cited
here (G. Kamberelis, 1994, "Genre as institutionally informed social
practice") is a really excellent (and LONG) discussion and review of genre
theory with lots of connections with and discussions of Bakhtin, Vygotsky,
Giddens, and others (including the work of Carol B. and Chuck B., I
believe, on this list).I think it's something that activity theorists
would get a lot out of if they don't already know it.
REGISTER
Halliday (1988, p.162) defines a register as "a cluster of associated
[linguistic] features having a greater-than-random...tendency to
co-occur." He then cites "scientific English" as an exemplar of the
concept because "any speaker of English for whom it falls within the
domain of experience knows it when he sees it or hears it" (1988, p. 62).
Other linguists, such as Biber (e.g., 1993, 1994; cf., Ferguson, 1983),
view registers as varieties of language which occur in speech situations
of varying contextual specificity. In the present study, I adopt a concept
of register which closely approximates Halliday's, while attempting to
relate it to the notions of genre and convention developed above.
The relationship between the concepts of register and genre is in
fact a complex and often confused one in the study of discourse. It may
even be the case that the primary difference is merely one of
disciplinary preference, and that the two words denote basically the
same phenomenon. Thus, Swales (1990) points out that the discomfort
re-cently caused in linguistics by the introduction of the concept of
genre may be due to the fact that register is a longer-established notion
in that field.
At least two perspectives on the relationship of genre and register
can be found in discourse linguistics. Couture (1986, p.82) details one
position:
genres can only be realized in completed texts or texts that can be
projected as complete, for a genre does more than specify kinds of codes
[i.e., registers] extant in a group of related texts; it specifies
conditions for beginning, continuing, and ending a text.
In contrast to genres, then, registers for Couture--as for Halliday
(1988)--are groupings of linguistic features into certain functional
communicative "codes" or styles (e.g., journalistic writing, "legalese").
Genres and registers are, in this view, theoretically indepen-dent, and
can be combined in various (although not all possible) permutations.
Martin (1985) describes a fundamentally different genre-register
relationship than the one proposed by Couture. For Martin, genre is
manifested in register, and register is manifested in language. That is,
genre is an abstract sociocultural system which underlies and determines
the permissible combinations of registral components (these components
being categorized within the systemic linguistic frame-work of field,
tenor, and mode), in a way similar to that by which register constrains
language.
While Couture and Martin differ in regard to the relationships they
posit between register and genre, they agree in clearly distinguishing the
two concepts from one another, and in placing the latter on a more
abstract plane. For the purposes of the present study, I will make a less
absolute distinction between register and genre, while retaining the basic
difference in level of abstraction; to a certain degree, my version of the
relationship synthesizes the views of Couture and Martin. Thus, whereas
genre represents the complex, conventionalized *rhetorical* form and
content of whole texts, register primarily represents the patterns of
co-occurring *linguistic* structures that comprise such texts. By
"rhetorical," however, I do not intend to exclude such linguistic
structures--properly speaking, the generic level of text will *include*
the registral level, as register is centrally involved in solving the
rhetorical problem(s) for which a genre is constituted. But register will
*not* typically include many of the generic components of text, since the
latter make sense only holistically. It is thus possible to describe a
disembodied "strip" of language as, for example, legal register, while
generic components would make little or no sense if similarly
decontextual-ized. In this view, then, register is a (theoretically)
detachable part of a larger organic concept--genre. In actual language
use, however, register will rarely if ever occur in such a detached mode.
Bakhtin (1986, p.64) places genre and register (which he calls
"functional style") in substantially the same relationship as I have:
[Functional] [s]tyle is inseparably linked to particular thematic
unities, and--what is especially important--to particular compositional
unities: to particular types of construction of the whole, types of its
completion, and types of relations between the speakers and other
participants in speech communication.... Style enters as one element into
the generic unity of the utterance. Of course, this does not mean that
language style cannot be the subject of its own independent study. Such a
study, that is, of language stylistics as an indepenent discipline, is
both feasible and necessary. But this study will be correct and productive
only if based on a constant awareness of the generic nature of language
styles...
The idea that register is a component of genre motivates the
analytical methodology employed in the present study. As described in
chapter 3 of the present work, language-oriented discourse analysis is
undertaken at two levels: the rhetorical (or generic), and the registral.
But whereas genre subsumes register--and registral elements (inasmuch as
they have rhetorical significance) are therefore considered at the former
level of analysis--at the latter level register is treated as primarily a
self-contained phenomenon. (From chap. 1 of "Scientific Discourse in
Sociohistorical Context: The Philosophical Transactions of the Royal
Society of London, 1675-1975. In Press from Erlbaum.
Best,
Dwight