Although this study answered many of my questions about instructing student collaborative writing groups in an electronic environment, obviously there is much left to research. As schools continue to acquire computer technology, I hope that more departments join efforts in developing new teaching and learning strategies for collaborative writing efforts as well as composing interdisciplinary research teams to pose and answer questions.

Linguistics and Composition Sara Kimball

It's very rare that you ever get a free ride from some other field. -- Noam Chomsky ( Olson and Faigley 34)

Linguistics has long informed teaching and research in composition. During the 1950s, insights into the systematic nature of language from American Structuralism, the dominant linguistic theory in American universities, helped bring about a significant reorientation in the perspective of composition teachers, allowing them to see themselves not as linguistic police enforcing prescriptive etiquette but as observers of student language. Later, the influence of sociolinguistics encouraged composition specialists to regard themselves as observers of the social and interpersonal contexts in which students write. These changes in orientation not only made composition pedagogy more humane, but without such fundamental shifts in attitude, composition research would not have been possible ( Nystrand et al. 273).

Since the 1960s, scholarship in second language acquisition and sociolinguistics has aided composition researchers trying to devise approaches that would meet the needs of speakers of nonstandard English and English as a second language. For example, the concepts of error analysis, first-language interference, and interlanguage, a learner's approximation of a target language that is systematic and a sign of learning rather than a random collection of errors, has informed the work of Mina Shaughnessy, David Bartholomae, and other researchers and teachers of basic writing ( Bartholomae68-70, 76-79, Montgomery (99-104). Sociolinguistic notions of dialect variation have contributed to an increased awareness that the students in our classes are not linguistically deficient but are competent users of varieties of language with systematic structures and histories of their own. As Lester Faigley notes, the NCTE Statement on Student's Right To Their Own Language of the 1970s is the product of an intellectual climate heavily influenced by sociolinguistics (81). Later research within sociolinguistics has helped researchers in composition to understand that difficulties experienced by nontraditional students are not simply linguistic but also have to be understood as rhetorical, cultural, and political ( Montgomery104-8). Work from the anthropological and ethnography of the speaking side of sociolinguistics, for example, Shirley Brice Heath Ways with Words, published in 1983, has influenced both research methodology in composition and fundamental ways of thinking about literacy. By the 1970s and 1980s, scholarship influenced by pragmatics, discourse analysis, and functionalist linguistics appeared with some frequency in composition journals and monographs ( Nystrand et al. 285-88, Faigley80, 89-91, Larson219-23).

AN UNWARRANTED PESSIMISM

By the late 1980s and early 1990s, however, linguistics had apparently become a less salient influence within mainstream composition scholarship. Lester Faigley, for example, notes that Stephen North ignores linguistics in his book The Making of Knowledge in Composition (80). In her survey of the impact of linguistics on composition between 1950 and 1980, Sharon Crowley concludes that while linguistic theory was attractive to teachers of composition in the 1950s because of the intellectual poverty of composition teaching at that time and the long association of composition with grammar (481), its influence on the whole has been disappointing and that linguistics "cannot provide teachers with the wider focus on composing that is necessary to develop a comprehensive theory of composition" because it offers "an extremely narrow, noncontextual view of what it means to be a user of language" (499). Even Frank Parker and Kim Sydow Campbell, who criticize Crowley's limited presentation of linguistics, are fairly circumspect in their claims about the possible benefits, restricting their illustration of the relation between theory, application, and practice to speech acts and pragmatics and dismissing Sledd's observation that the study of language structure, history, and the social functions of dialects provides useful intellectual grounding for writing teachers as special pleading from an expert in these fields (311 n. 5). To cite a more recent example, neither North in an assessment of possibilities for research in composition published in 1997, nor any other contributors to the volume in which his essay appears, mention linguistics as an influence on research.

At first glance, the pessimism seems justified. Institutional changes such as the formation of linguistics departments separate from language departments, a process that started in the late 1940s but accelerated in the 1960s and 1970s, in part because of the prestige of transformational-generative grammar in the academy, tended to isolate linguists from composition ( Faigley85). The increasingly abstract nature of theoretical linguistics has perhaps made it seem irrelevant too. The ethos of linguistics as a field that presents itself as scientific and empirical (and in the case of generative grammar, mathematical) has long proved uncongenial to the more humanistically oriented elements in English departments, ¹ but since the early 1980s the emphasis within Chomskyan linguistics has been on finding universal properties of language rather than on the constructions and rules of individual languages, a set of goals that seems to have little to offer either the researcher in composition or the teacher ( Faigley 83).  

But the picture of linguists isolated within linguistics departments pursuing syntactic esoterica is not entirely accurate, and much of the pessimism is unwarranted because it is based on limited and inaccurate views of linguistics. Linguists have always been part of English departments, and especially in the many universities that do not have separate departments of linguistics, English and other language departments often provide linguists with institutional homes. Many English department graduate and undergraduate programs require at least some coursework in English linguistics or the history of the language, and it seems a safe bet that most students and faculty in English departments and composition programs have had at least some exposure to linguists and their ideas.

Although transformational-generative grammar is the branch of linguistics most readily identifiable to outsiders, and Noam Chomsky is perhaps the only linguist well known to the educated public, not all linguists work within the Chomskyan tradition. It is perhaps safe to say that any linguist trained within the past thirty years is a Chomskyan in the sense that the gospel according to MIT provided part of our training and has helped shape our outlook, but many of us, including those in phonetics, some types of phonology and morphology, anthropological linguistics, sociolinguistics, educational linguistics, psycholinguistics, second language acquisition, discourse analysis, pragmatics, and historical linguistics have stuck to chugging along on Faigley's "slow, datagathering local" (84) and have prospered in our own ways. Chomsky's idealization of language as the theoretical competence of an idealized speakerhearer using an invariant idiolect was a productive abstraction on the theoretical level, but it has not gone unchallenged, ² and challenges to concepts from Chomskyan linguistics have provided productive avenues for research in empirically oriented fields within linguistics such as psycholinguistics and second language acquisition ( Newmeyer and Weinberger "Otogenesis," Cahill).

In some ways, theoretical linguistics and sociolinguistics have been such intimate parts of the formation of composition as an academic field that ideas from these fields are no longer easily recognizable as foreign. If we strip away some of the misappropriations and misunderstandings of the past forty years--for example, the confusion of "generation" or "derivation' with invention implicit in some of the early enthusiasm about generative rhetoric, the confusion of the syntactic notion of "transformation" with the writer's notion of stylistic option implicit in early work on sentence combining, and the use of "recursive," a term describing the mathematical properties of rules in early generative grammar as a synonym for "nonlinear" in process research--there is a solid core of intellectual influence that is so fundamental to modern composition thinking that its origins are no longer salient. ³ It seems unlikely, for example, that any scholar in composition would seriously question the notions that power relations are enacted through language or that speakers and writers signal various aspects of their social identity through their use of language, ideas that have been developed and examined in sociolinguistic research. Many ideas from sociolinguistics in particular are most familiar as sociolinguistics in the literature on basic writing, but the fact that so much of literary theory has been deeply influenced by linguistics, especially the European version of Structuralism ( Nystrand et al. 282-84; Faigley88) also contributes to the sense that concepts originally from linguistics (e.g., discourse community as an extension of speech community) are simply part of the intellectual landscape in composition.

Although some of the reluctance to acknowledge the influence of linguistics may stem from widespread attitudes toward scientific, empirical, or mathematical approaches to knowledge, some of the inability to recognize the influence of sociolinguistics in particular may come from a tendency within composition to regard knowledge as dialectic rather than cumulative, an attitude that is sometimes too ready to discount old knowledge in favor of novelty. The discouragement about linguistics is analogous to the fate of process within composition scholarship: the idea of writing as process was embraced as a virtual panacea, commodified by the textbook industry, and then widely rejected as a topic of research, both in reaction to initial excesses and because of its more or less accidental historical ties with expressivism ( Tobin7-9).

The embrace of novelty is no doubt in part a reflection of fickle academic fashion, but it is also a function of the fact that, as a field, composition is driven by a pedagogical imperative. A large part of our job, which is made more difficult by unrealistic demands from administrators and the general public, is to find effective ways of intervening in writing. Unlike linguists, we do not have the luxury of description without action. But the danger is that we can be moved by the pedagogical imperative into ill-informed action. We sometimes look for quick fixes--Chomksy's free ride from some other field--without necessarily thinking through the relation between theory and practice, forgetting that our job in its largest sense is to articulate theories appropriate to our own concerns, both as the backing for effective pedagogy, and as knowledge for its own sake.

And this is precisely where things have gone awry historically in the relation between linguistics and composition--when linguistic theory has been applied directly to practice, for example, in teaching transformational-generative grammar as stylistics or invention. After forty years of research, it should be clear that teaching students linguistics, or any form of formal grammar, will not improve their writing. ⁴ Expecting students to master style through the study of grammar, whether traditional or according to a particular linguistic theory, makes about as much sense intellectually as expecting pitchers to study physics in order to improve their pitching, even though a pitch can be described in the equations of physics.

LINGUISTICS AS INTELLECTUAL GROUNDING  

Crowley is right to criticize the quick-fix mentality in attitudes toward linguistics (498), but the solution is not rejecting linguistics as a field that can inform composition theory and practice. As Faigley notes, "a categorical dismissal of linguistics from rhetoric and composition may be premature" (83). The pitcher may not have to learn physics, but the people responsible for the conditions under which the game is played can contribute to better pitching by knowing enough about the physics of pitching to design better balls and practice activities that do not wreck pitcher's arms. Similarly, acquiring a background in the structure and history of the English language is one of the intellectual responsibilities of composition professionals in their roles as researchers and teachers because such a background can help teachers fashion classroom environments that allow writing to flourish, and it can provide productive directions for research into the nature of writing.

In the remainder of this chapter I would like to illustrate how linguistics can provide perspectives on writers and their work by looking at two areas in which the concerns of linguists and those of compositionists intersect. The first--in some ways easier because it is less obvious--is how definitions of language from formal linguistics can contribute to an understanding of writing. The second is perspectives from linguistics on the relationship between spoken and written language.

DEFINITIONS OF LANGUAGE

Writing involves using technology to make thought manifest in visible symbols with a relationship to language. ⁵ To intervene effectively in the work of writers, we need to separate out and understand the parts of writing that belong to language, thought, and technology. Linguistic definitions of language help separate the linguistic aspects of writing from the cognitive and technological.

All natural languages exhibit dual articulation: they consist of sounds and meanings linked by a system of rules called a grammar. A central tenet of linguistics since de Saussure is that, whatever the shape of the rules, the link between sound and meaning is arbitrary. No sequence of sounds uniquely, or ideally, conveys a particular meaning.

This is a formal description of language as an observable phenomenon. The generative tradition in its various manifestations has always had as its goal going beyond formal description to explain language as a biological faculty and to explain how it is acquired ( Olson and Faigley10-13). Current theories try to isolate Universal Grammar, those aspects of language that are hard-wired in the human brain at birth and that constitute a common human inheritance ( Chomsky and Lasnik 14 ). The working hypothesis is philosophically nativist; its premise is that human beings are born with a set of principles common to all languages already present in brain structure and with knowledge of how these principles may vary along a limited set of parameters. Language is acquired as the result of exposure to particular languages during the first years of life, and acquisition involves learning the values appropriate to each parameter--in other words, learning how the principles are embodied in the language the child hears. For example, a principle that seems to be universal is that syntactic rules are structure-dependent: they rely on structural relationships for their conditioning not on the linear order of words ( Cook2). The ordering of elements within constituents, however, differs from language to language. A child is born knowing in some sense both that rules depend on structure and knowing possible structures; exposure to language teaches her to select particular structures. Exposure to English, for example, teaches her to generalize (apparently quite rapidly) that the head of a noun phrase or verb phrase (e.g., the noun in cat in the hat or the verb in caught the mouse) comes to the left of the phrase. A child hearing Japanese, by contrast, learns that heads come to the right (roughly the equivalent of cat hat in and mouse caught). This process of acquisition is natural, unconscious, and biologically determined. Chomsky has repeatedly likened human linguistic capacity to a bodily organ, viewing acquisition as the natural growth and development of this organ ( Rules134-35).

At first glance, none of this seems directly relevant to teaching writing. But it is very directly relevant to one's sense of what one is doing in the classroom. First, it provides the warrant for regarding all of our students--no matter what their level of preparation or social background--as competent users of language. Universal Grammar and dual articulation provide intellectually sound arguments for viewing all languages, and all dialects of a language, as linguistically equal. If the relation between sound and meaning is arbitrary, then judgments about the superiority or inferiority of a language or dialect are social; they have no linguistic basis. If all human languages are at some fundamental level the same, and we are all born with the capacity to learn any human language, then all human languages are, in linguistic terms, equal. The equality of linguistic codes cannot be dismissed as a sentimental trope of the political left or a fond hope of optimistic but naive teachers; it is a consequence of human biology.

A biological theory of language also implies a theory of translatability. It is possible to express any thought capable of being held by a human being in any language. Labov, for example, in a paper that has been highly influential in composition and language education, demonstrates that it is possible to construct a syllogism in the most colloquial variety of Afro-American Vernacular English (12-15). A theory of translatability implies that whatever the relation between language and thought, it is not direct. Chomsky, for example, has long described language and thought as interrelated but fundamentally separate cognitive systems. Although there are challenges to this view from within cognitive science and linguistics ( Kempson), and it is possible that features of grammatical systems, such as marking of tense, aspect, and plurality may mediate thought, shaping and constraining speakers' worldviews ( Lakoff), a number of considerations should suggest that, ultimately, language is not thought. People without language (aphasics, deaf people who acquired neither spoken language nor sign language in infancy, and infants) are capable of various cognitive activities, for example, computation ( Pinker67-69). Anecdotal accounts of thought without language, (e.g., thinking in images) receive some empirical support from experiments ( Pinker70-73), and any writer who has ever struggled to find the proper words to express her thoughts is deeply, familiar with the idea that thought is possible--if not necessarily satisfying--in the absence of language. But more important, no human language is an adequate medium for the mental computation that underlies thought, because, as Stephen Pinker explains, language does not in and of itself offer the means for resolving coreference, deixis, or lexical ambiguity, and natural languages lack the logical explicitness necessary for mental computation (78-82). It takes more than linguistic knowledge to understand or produce a sentence like Norma cut a piece of pie, put it on the plate she'd taken down from the shelf, and gave it to her sister.

If language is biological and organized along universal lines, writing is artificial because it is always mediated by technology, and unlike spoken language, it is explicitly taught and learned. The distinction is a crucial one and in part what linguists mean when they speak of the primacy of speech. One of the benefits of recent research in computers and writing is that it foregrounds the technological aspects of writing, but it should not be forgotten that the simple writing tools we often take for granted--pencils, pens, and paper--are also technology. Because writing inevitably involves the use of tools, it ineluctably intertwines language and thought as cognitive systems with cognitive systems involved in the manipulation of tools.

Some of the difficulties writers face may be primarily technological rather than linguistic or cognitive, or they may reflect the complexities inherent in trying to integrate three different systems. This should not be an idea foreign to our experience as teachers and writers. Teachers of basic writing, for example, have long known that some of their students' problems stem from unfamiliarity with the tools used for writing ( Shaughnessy 14 - 16 ). If successful writing depends upon the fluent integration of thought, language, and tools, someone unaccustomed to manipulating a pencil or pen may in part be hobbled by the tool-using aspects of the task and unable to deploy language and thought with spontaneous fluency, a phenomenon that should not be alien to anyone who has struggled to learn a new word-processing program. But difficulties integrating technological, cognitive, and linguistic aspects of writing are probably not unique to basic writers and novice computer users. One cause of some types of writer's block may be interference among competing systems under conditions that force the writer to focus on one aspect at the expense of others.

One of our goals as teachers should be to arrange conditions to allow students opportunities to practice integrating thought, language, and tool-use. We can also reformulate some basic research questions by defining their domains more clearly. For example, one source of disillusionment with sentence combining is that no one has provided a satisfactory account of how it works ( Strong337; Crowley490-91). It is possible, however, that under the right circumstances sentence combining can increase syntactic flexibility not by tapping directly into linguistic competence, but by providing an opportunity for practice. As Larson, for example, notes, sentence combining seems to work best as a technique for revision that engages students in reshaping their own prose (218). Perhaps it succeeds when writers are actively engaged in the linguistic manipulation of intellectual content that they have created because it allows them to practice integrating the linguistic and tool using aspects of writing without undue interference from the cognitive aspects. In other words, it functions as exercises do in sports or music, as a way of isolating, practicing, and making automatic a small part of a larger, more complex whole in a context free of the stresses of real-world performance ( Strong; Elbow238).

LINGUISTIC PERSPECTIVES ON SPOKEN AND WRITTEN LANGUAGE

The idea that some of the features viewed with disfavor in student writing have their source in spoken language is not new to composition. It figures prominently, for example, in work on the influence of nonstandard dialects on the writing of so-called basic writers. The ideas of Walter Ong, who claims that differences in the syntax of written and spoken language reflect cognitive differences between "literate" and "oral" thinking have also been influential, if not uncontroversial, in composition, and the increasing use of real-time and delayed conferencing systems in networked composition classrooms has recently stimulated thinking on spoken and written linguistic strategies.

How might a perspective from linguistics inform thinking within composition on the relation between speech and writing? Over the past twenty years linguists have published a number of studies of spoken and written language based on corpora of various sizes that have the virtue of providing empirical evidence to confirm or challenge common-sense assumptions. For example, the vocabulary deployed by writers in a variety of genres tends, on the whole, to be more extensive than that available to speakers ( Chafe and Danielewicz87-89). There are also differences in the stock of vocabulary items typically used in writing and speaking, especially at the extremes of formal, highly informational writing on the one hand and casual conversation on the other. Writing, especially when it is highly informational in focus, tends to favor nominalizations ( Chafe and Danielewicz99; Biber104), and it typically shows greater lexical density, or a higher proportion of content words to function words (e.g., prepositions and articles) than speech ( Halliday59-62).

The differing relations of speakers and writers to their audience are also marked linguistically. Chafe and Danielewicz, for example, find that their written samples show a higher degree of referential explicitness than their spoken samples (90). They see spoken language as showing indications of speakers' involvement both with their audiences and with themselves, which they define as a concern for concrete aspects of linguistic interaction and concrete reality marked in linguistic features such as use of first and second person pronouns and in the use of spatial and temporal adverbials (105-7). The written language in their sample, by contrast, shows markers of detachment, or an interest in ideas that are abstract and not tied to specific people, or to real-world places, events, or times (108). Among these markers of detachment are use of the passive, which renders events abstract by focusing on action rather than agents, and use of generic adverbs (many of which act as sentential adverbs rather than verb phrase modifiers) such as usually, normally, or primarily (109). Biber finds general differences in markers of involvement and detachment and in explicitness of reference too, but his sample is much broader and more finely calibrated for distinctions of genre. Not surprisingly, he finds that personal letters and fiction show greater use of features of involvement and situation-dependent reference than other written genres (160-61).

On the syntactic level, there appear to be differences too, but there is less ostensible agreement among researchers as to their precise nature. Chafe and Danielewicz find that speakers in a variety of situations tend to rely on chaining together independent clauses to relate ideas, while writers in a variety of genres are more likely to use embedding (102-5). Halliday, by contrast, defines the language in his spoken sample as more grammatically intricate than the language in his written sample (63-68), but under his system of analysis hypotaxis (or chaining) is a more intricate operation than embedding, or parataxis. Biber, whose study is the most detailed and draws on the most extensive corpus, is able to make the finest distinctions, finding, for example, that embedded constructions such as nominal that-clauses and wh-clauses, and adverbial subordinators, which typically co-occur with markers of involvement such as first and second person pronouns, are characteristic of his spoken sample (229-30), while relative clauses formed with wh-pronouns (which, who), for example, are more common in written discourse.

Each of these studies has its limitations, especially those based on limited samples, and the linguists' conceptions of audiences for written discourse would probably strike many in composition as naive. The value for composition lies in the way these researchers regard their samples. It is significant, and perhaps a characteristically linguistic approach, that in accounting for differences between spoken and written language each researcher focuses on differences in production and processing. Simply by virtue of the anatomy of the human vocal tract, spoken language is temporally bound and organized linearly. Composition is performed on the fly, and revision is expressed in dysfluencies, hesitation, backtracking, and redundancy. Chafe and Danielewicz claim that some of the features associated with various types of spoken discourse, for example, restricted vocabulary, may reflect limitations in short-term memory. For Halliday, speech and writing can be considered forms of discourse associated (in our society) with extreme points on a continuum from most spontaneous to most self-monitored. Spoken language (at least casual conversation) is usually relatively unselfmonitored, while writing is typically highly self-monitored. He claims that hesitations, false starts, and anacolutha monitored use of language (68-71).

If we can return to the idea that writing represents an intersection between thought, language, and technology, focus on the production aspects of speech and writing suggests that some of the "spoken" features of problematic student writing, for example, problems with pronoun reference, immature syntax, or inappropriate register, represent a failure to exploit the technology of writing in ways that are normally considered appropriate to an academic context. Some of these problems may simply reflect discourse that has not received the elaboration possible from and expected of writing in our culture: it is too close to speech for comfort because it has not been ground finely enough in the technological mill.  

CONCLUSION

If I have seemed to lead the reader on an extended ramble through esoterica and then through a dull empirical wilderness to the mundane conclusion that student writers need to learn to revise and edit effectively, that is in some ways precisely my point about how linguistics can inform composition. It provides no revolutionary insights into the nature of invention, and although it provides terms for features of spoken and written discourse and empirical studies showing how these features occur in various genres, it provides neither a theory of style nor a theory of stylistic value; these are our jobs as composition researchers and teachers. What linguistics offers is warrants for productive attitudes toward writers and their work. It allows us to see our students (and, indeed, all writers) as competent users of language within particular social contexts, and it helps us to set goals for writers that are realistic. On the methodological level, if there is anything that linguistics has taught composition, it is that it is possible to temper the pedagogical imperative with reflective distance, to regard--at least temporarily--the product of writers as a subject of study rather than an object of intervention. Linguistics does not offer composition a free ride, but it may help us figure out where we are going.

NOTES