MAXINE HAIRSTON'S 1982 proclamation of a '.paradigm shift" claimed that the two allied disciplines motivating the new process paradigm were cognitive psychology and linguistics. By the end of the 1980s, one of these forces, linguistics, apparently had vanished. A noncontroversial aspect of Stephen North's controversial survey of writing research, The Making of Knowledge in Composition, is the omission of linguistics as an important disciplinary subfield.
North does not even include language or linguistics in the index. Another classification of writing theorists and researchers presented by Patricia Bizzell at the 1987 meeting of the Conference on College Composition and Communication (CCCC) also fails to mention linguistics as a major line of disciplinary inquiry ("Forming the Canon").
One could argue that North and Bizzell erred either by leaving out linguistics or by subsuming it under a broader heading, and in support of this argument one could assemble a formidable group of North American researchers of written language, who in the 1980s applied principles and analytical tools developed by linguists.(1) Nonetheless. one could also easily defend the position implied by North's and Bizzell’s classifications-that the influence of linguistics on the study and teaching of writing in North America has dwindled to such an extent that linguistics is no longer a major contributor of ideas. The demise of the influence of linguistics results not so much from the lack of substance in recent work on written language as it does from the lack of a dominant approach within linguistics that is applicable to the study of writing. Researchers of written language do not share common goals and methodologies, nor use the same terms, nor recognize common research issues, nor even agree about the nature of language. In spite of the brilliance of certain individual studies, the whole adds up to considerably less than the sum of the parts.
The situation was much different in earlier decades of CCCC. In the 1950s
[86]
linguists were leaders in the organization and published articles frequently in its journal. The major conflict within CCCC was characterized as "correctness” versus "usage." with linguistics contributing research on language variation to counter absolute judgments of "good” and “bad" English. Donald Lloyd’s call for a composition course built around linguistics in 1952 reflected the newly found tolerance for usage among the liberal wing of CCCC.
Correctness again became an issue following the publication of Webster’s Third New International Dictionary in September 1961, which set off a storm of public controversy over its practice of basing descriptions of usage and pronunciation on what people actually said rather than what experts assumed to be correct. The much vilified editor of Webster’s Third, Phillip Cove was a featured speaker at the 1962 meeting of CCCC. More important, however, was work in sociolinguistics arguing that dialects considered prestigious or “standard" gain their status by being identified with the wealthiest and most powerful groups in a society and not from their inherent superiority. Sociolinguists denied assumptions that speakers of "nonstandard" dialects are somehow deprived or suffer from a cognitive deficit by demonstrating that nonstandard dialects are as inherently logical as standard ones. A measure of the authority given to sociolinguistics during this time came with the 1974 Students’ Right to Their Own Language statement discussed in the previous chapter, which lists 129 entries on dialects and the teaching of writing in an attached annotated bibliography.
The work on usage and dialects was only a part of the influence of linguistics in the 1960s and 1970s. When rhetoric and composition blossomed as a discipline in the mid-1960s advances in rhetorical theory represented by the work of Wayne Booth, James Kinneavy, and James Moffett were paralleled by new directions in language study. By 1965, English teachers’ awareness of linguistics was viewed as the most important development in the first two decades of the CCCC (Correll). and linguistics was proposed as the basis of a modern theory of rhetoric (Young and Becker). The most important work on written language during this period was Francis Christensen's theory of generative rhetoric, set out in a series of four articles in College English and College Composition and Communication in 1963 and 1965, later collected with two earlier articles and published as Notes Toward a New Rhetoric in 1967.
Christensen's stylistic analyses demonstrated that textbook advice favoring the periodic sentence ran counter to the practice of published writers who frequently use right-branching sentences often referred to as “loose" sentences. Christensen renamed these right-branching sentences “cumulative” sentences and made them the basis of his pedagogy He assigned semantic levels to the modifiers of a main clause and devised a scheme to represent what he called "levels of generality." Christensen quoted this sentence from Sinclair
[82]
Lewis as an example of a semantically subordinate sequence of modifiers, each modifying the level above.
2He ~inP~~ ~~~"dS in the bichlea'q~t~~cuk ~;I~~I~IIUS In fne bichloride solution and shook them.
3 fingersdown.
4 like the Fingers of a pianist above the keys.
Using such examples as models, Christensen designed exercises that teach students to observe carefully and to describe accurately; he felt that practice in using nonrestrictive modifiers could generate the supporting detail that is characteristically absent from much student writing. In “Generative Rhetoric of the Paragraph,” he used the cumulative sentence as a way of analyzing units of discourse larger than the sentence. Christensen believed that the paragraph is structurally a macrosentence, and he found intuitive evidence that the paragraph is structurally similar to the cumulative sentence from the fact that many of his cumulative sentence examples could easily be translated into paragraphs if the nonrestrictive modifiers were made into complete sentences.
In successive issues of College Composition and Communication immediately following the initial publication of "Generative Rhetoric of the Paragraph” in 1965, Alton Becker and Paul Rodgers presented alternative views of paragraph structure. Becker proposed a semantic slot conception based on tagmemics, another variety of structuralism. He criticized Christensen’s model for its lack of a semantic theory adequate to explain in formal terms the relationships Christensen perceived. Reeker's criticisms were on the mark because no semantic theory existed at that time (nor does one exist today) that can account for these relationships. But by the same standard Becker’s tagmemic alternative also lacked an elaborated semantic theory and it had less intuitive appeal because it presented only two basic paragraph patterns. Rodgers’s criticisms of Christensen’s analytical model of the paragraph were more substantial. He argued that the paragraph is not so much a semantic unit as it is an orthographic unit. Instead of the paragraph as the basic unit of discourse. Rodgers advanced semantic units that he called stadia of discourse (conceptual chunks of discourse, not often coincident with paragraph divisions).
Considerable work followed in the 1970s from lines of research established in the 1960s, ineluding the continuation of the discussion of how coherence is achieved by Ross Winterowd (“Grammar”) and extensions of Christensen's ideas to the essay as a whole by Frank D’Angelo. Michael craciv. And Will Pitkin. But as these efforts grew in scope, their shortcomings became immediately apparent. Without an elaborated semantic theory, the structural classifications seemed too idiosyncratic and arbitrary, as well as too vague, to be the basis of pedagogy.
[83]
In the 1970s the energy in the language camp within rhetoric and composition passed to those interested in sentence combining, which grew out of the work of Kellogg Hunt on syntactic development in the 1960s. Like others at the time, Hunt was inspired by Chomsky's theory of generative grammar. His suggestion that sentence-combining practice would enhance the syntactic maturity of developing writers was demonstrated in several studies with junior-high age children, most notably by John Mellon, in Transformational Sentence-Combining, and Frank O'Hare. These findings were extended to a college population in a major study at Miami (Ohio) University conducted by Donald Daiker, Andrew Kerek. and Max Morenberg, which concluded that a sentence-combining curriculum could increase the syntactic maturity and overall writing quality of first-year college students. At first the researchers assumed that the gains were related (Daiker), but when they analyzed the extent to which syntactic factors influenced readers' judgments, they found that the measures were almost unrelated to the assessments of quality (Kerek).
The failure of researchers to associate linguistic variables with perceptions of writing quality cooled the interest in linguistics raised by pedagogical studies of sentence combining. After 1980. linguistics was no longer seen as a panacea for improving student writing, and applications of structural linguistics. Chomsky's generative linguistics, and sociolinguistics faded from composition journals. For a short time. European work on text linguistics promised to provide the comprehensive theory of discourse structure that the earlier efforts of Christensen had lacked, Teun van Dijk's theory of a semantic macrostructure underlying texts, most fully elaborated in Macrostructures (1980). was especially attractive for the study of structure in written discourse because it promised to resolve the ambiguity of Christensen’s levels of generality.
But van Dijk's ambitious project failed to provide adequately for the contexts in which language is used, and after several attempts to augment his semantic theory with pragmatic theory, van Dijk turned his attention elsewhere, For those who continued to work in text linguistics, the scope of their inquiry became more and more restricted. Even though there has been much research on written language in the 1980s from both inside and outside composition studies, no work has inspired the enthusiasm raised by generative
rhetoric and sentence combining, nor have their been any large-scale movements within the discipline based on linguistic research.
Given the abstract, theoretical direction of most North American universities' linguistics departments and the virtual absence of alternative language study within English departments. linguistics appears to have little chance of reemerging as a major disciplinary influence in composition studies. Nonetheless, a categorical dismissal of linguistics from rhetoric and composition may be premature. Questions concerning subject positions in discourse were
[84]
discussed in linguistics outside of North America in the 1980s, and there is much work now being done on language and politics.
Before I begin to use postmodern theory to investigate how subjectivity is understood in composition studies, I will consider in this chapter how subjectivity has been conceived in linguistics. In particular. I examine the critical linguistics movement, which, had it been better known, might have influenced composition studies during the 1980s and led to some of its current concerns by a different path.
The Limits of Formal Linguistics
Before moving to how critical linguistics might have influenced the study of writing, I would like to reflect on why language study within composition declined so quickly. The obvious answer is that the influence of linguistics was swept away by the movement toward understanding and teaching writing as a process, but the process movement alone does not explain such a quick demise. For underlying reasons we must look to the discipline of linguistics itself. If we ask what happened within linguistics, again there is an easy answer: Noam Chomsky. Chomsky's theory of transformational-generative grammar influenced the study of language in North America as no other theory had in the past. Shortly after the publication of Chomsky's second maior book in 1965, Aspects of the Theory of Syntax, linguists were either on board the fast Chomsky theoretical express or hopelessly behind on the slow, data-gathering local.
For those in other disciplines interested in questions concerning language and discourse, generative grammar at First appeared to be a methodological breakthrough, a way of describing the messy data of language with orderly rules that could obtain universally. These researchers, however, soon encountered the limitations that Chomsky had been careful to anticipate. Language could be orderly only if it were idealized: if actual language was used as data, the orderliness of language predicted by generative grammar soon disintegrated. Chomsky insisted that language be viewed as abstract, formal, and accessible through intuition. His goal for a theory of language was describing a human being's innate capacity for language, not how people actually use language. When asked what relevance the study of linguistics had for education, Chomsky has consistently answered: absolutely none.(2)
Gradually, those interested in studying discourse came to heed Chomsky's warnings. No matter how hard they tried, researchers could find no fruitful way of applying advances in formal linguistics for their own research programs beyond early language acquisition. Research in sentence combining as a
[85]
method of teaching writing is a good case in point. Early sentence-combining experiments developed from the concept of the "kernel" sentence in Chomsky's initial presentation of transformational grammar in Syntactic Structures in 1957. Students were given two or more short kernel sentences and asked to combine them into one, using a particular transformation signaled in the exercise. But by the time John Mellon published Transformational Sentence-Combining: A Method for Enhancing the Development of Syntactic: Fluency in English Composition in 1969, the first report of a major sentence-combining study, Chomsky had abandoned the notion of kernel sentences. Soon “transformational” was dropped
as an adjective modifying "sentence combining." and research in sentence combining proceeded independently of later developments in syntactic analysis.
Had Chomsky's influence been restricted to language theory, linguists in North America might have remained more active in the study of writing. In some ways generative grammar turned out to be like other radical theories that for a time gain many enthusiastic supporters but quickly lose their impetus when the supporters begin to diverge into warring camps.' But in another way the life of the generative grammar movement was different from that of other radical theories. Generative grammar altered the disciplinary map.
Before the rise of generative grammar, linguists were scattered in departments of anthropology, English, and other modern language departments. These linguists tended to share some of the interests of members of those disciplines, and the arrangement fostered interdisciplinary cooperation. The excitement that accompanied Chomsky's theory accelerated the formation and growth of separate linguistics departments committed to the theoretical study of structure in language. The methodology of generative grammar with elaborate sets of formal rules emphasized the difference of its project from those of other humanities and social science disciplines, as well as from other schools of linguistics. Theoretical linguists dismissed the questions of other disciplines such as those of language education as applied and "uninteresting." In their view, the only truly interesting questions in the study of language concern abstract universals underlying language.
To blame Chomsky. however, for the decline of linguistics within composition studies is not merely simplistic: it is wrong. The limitations of generative grammar were demonstrated when stylistic studies aimed at analyzing the "deep structure” of style failed to produce results beyond what could be observed from surface features (for example, Ohmann, “Generative Grammars”). Nevertheless, researchers of language in written discourse did not cease working when they realized that generative grammar was not useful for their purposes. Rather, they encountered again and again a fundamental difficulty met by earlier linguists beginning with Zellig Harris, who in a boundary-breaking article in 1952 had ventured beyond the sentence.
[86]
When these linguists analyzed stretches of language larger than a sentence, they attempted to apply those criteria they had used for analyzing phonemes, morphemes, and clauses. In the tradition of American structural linguistics, they assumed that a continuum of formal correspondences exists between smaller and larger units. What many researchers in written language did not consider was that the basis of text structure might be radically different from that of sentence structure, that no one kind of structural description might be adequate to characterize text structure. Models of text structure based on a few patterns, such as those of paragraph theorists Christensen and Becker, at first were attractive but inevitably failed to account for a variety of distinctions that readers perceived among different texts. These models were confounded when readers encountered a paragraph where a topic sentence could not be readily identified, or when readers with different levels of familiarity with the subject matter of a paragraph could assign differing interpretations of what is important.
Efforts to describe text structure have all been frustrated because texts--unlike phonemes, morphemes, and clauses--are semantic rather than structural units. Semantics has been the least developed area in American linguistics, as opposed to European linguistics, partly as a result of different readings of Saussure. American structural linguistics derived from Saussure a methodology that was well suited for describing the phonology of unstudied and often quickly disappearing languages. Because U.S. linguists grew up within the dominant ideology of behaviorism in the social sciences, they continued to make empiricist assumptions about language and ignored Saussure.s discussions of meaning.
European structuralists, on the other hand, explored Saussure’s proposal that since language is a self-contained system, the boundaries for meaning, like those for meaningful sounds or phonemes, are largely arbitrary. Following Roman Jakobson's analysis of how differences in sound are grouped as distinct phonemes according to shared articulatory features (Selected Writings), several European linguists, including Louis Hjelmslev and A J. Greimas, proposed a systematic analysis of semantics based on shared aspects of meaning. The broadening of European structuralism to questions of semantics led to an even more expansive use of structuralism to study culture as a system
of signs--a pursuit also known as semiology.
Most American linguists avoided such extensions. Instead, they took the advice of Leonard Bloomfield, who argued in 1933 that language can be studied scientifically only to the extent that meaning is ignored. The Chomsky revolution did not overturn this bias from structuralism. Indeed, Chomsky has frequently argued (most recently in The Generative Enterprise) that linguists