Lillian Bridwell-Bowles UNIVERSITY OF MINNESOTA

Within the composition community we have some disagreements about research that we can trace back to Plato and Aristotle, with aggravations added by Descartes, Bacon, Locke, and others ever since. They center around reality and ways of knowing. Does objective reality (i.e., "truth") exist? And if so, how can it be apprehended? If it exists, does it exist apart from the language used to describe it? In composition studies, we welcome a range of answers to these questions and a range of theoretical and methodological positions, as our journals attest. Nevertheless, we have a long and clear history of favoring social scientific methods for research and a more recent and cloudy history of welcoming into the research community more philosophical, naturalistic, or deconstructivist inquiries. In this chapter, I hope to provide an overview of this history and some of the explanations for our biases. I shall also outline the objects of inquiry that our research investigates, some of the major epistemological differences that have coexisted during the period from the 1960s through the 1980s within the composition research community, the major methodological approaches used, and some issues that remain to be addressed as composition research continues to evolve.

Sources for the Beginning Composition Researcher

Within the space of one essay, I cannot hope to provide a comprehensive picture of composition research, but I will cite a number of major books ( Beach and Bridwell; Cooper and Odell; Lauer and Asher; North; Phelps; Hillocks; Freedman, Dyson, Flower, and Chafe; Moran and Lunsford; Flood, Jensen, Lapp, and Squire) on the topic to supplement this chapter. Taken collectively, these sources can provide a reasonably broad view of research in our field.

Given the label Stephen North has applied to Research in the Teaching of English (RTE), "the leading Researcher journal in Composition" (135), this journal is a useful one for the beginning composition specialist to examine as a case study of the modern history of composition research. Anne Herrington's article on twenty years of RTE provides just such a comprehensive overview from the perspective of composition studies. Even though RTE is not exclusively a composition journal, more composition submissions were reviewed by its editors than any other kind over the years from 1987 through 1989, thus demonstrating the impact of composition research on the larger community concerned with teaching all of the English language arts. Other journals listed at the end of this article provide the primary homes for most composition research.

The Object of Inquiry; or, What Are We Studying?

Any description of research in the field depends, of course, on where we mark its beginning. Despite the obvious existence of earlier research on composition ( Lyman, Hoyt), most writers mark the beginning of modern composition research with the publication of Braddock, Lloyd- Jones , and Schoer book, Research in Written Composition, published in 1963. Setting this date helps us to distinguish a modern period of research, with an emphasis on empirically observable data, from other periods in which the more typical kinds of scholarship were non-empirical, intuitive or introspective, and seated in rhetoric or philosophy, rather than in social science.

Prior to this modern period, and on into the 1950s and 1960s, the sparse empirical research that existed was focused on narrow issues or segments of language. Moran and Lunsford Research in Composition and Rhetoric is constructed around some of the topics that this early research frequently investigated, including separate chapters on such topics as the sentence, the paragraph, spelling, usage, and textbooks.

Various methodologies for teaching composition were a popular focus around mid-century, as revealed in the many comparisons of methodologies summarized in Braddock, Lloyd-Jones, and Schoer.

During the 1970s, many critics questioned these methodological studies, including Cooper and Odell, who characterized experimental research on teaching methodologies as "corn field studies" because of their reliance on statistical designs derived from agronomy. They and others such as Janet Emig and Martha King called for more basic research into questions about composing and the nature of texts.

Studying static products was unpopular throughout the expansionary years for composition research during the 1970s and 1980s. Instead of analyzing texts and their features, researchers began to study what Louise Phelps calls the "human science" or "psychology" of composition. Researchers wanted to know how human beings write, both behaviorally and cognitively. As I argued in 1984, "the dramatic shift came when [composition researchers] began analyzing what writers really do when they write, not what they 'ought to do' based on a priori logical assumptions" (3). In defining what she calls an "objectivist conception of text," Phelps lays out the history of the famed process/product opposition in composition. Among other things, the shift in priorities succeeded in "reconstituting the field as a research discipline" (132).

A myriad of process studies in the 1970s and 1980s responded to the calls for "basic research" on composing processes. Janet Emig is credited with being the first to conduct such research, and she was followed by Sommers, Perl, Bridwell, and others. Such studies examined various parts of composing processes, including invention, drafting, revising, and rereading during composing. Studies of the pedagogical technique called sentence-combining, which attempted to increase syntactic fluency without formal instruction in grammar, were also common during these decades (see Kleine for a review and critique).

Perhaps the most significant developments during the 1970s and 1980s were studies of the "cognitive" processes involved in writing introduced by Flower and Hayes. Using a technique called "protocol analysis" which was borrowed from cognitive psychology, researchers asked writers to talk aloud as they composed to gain insight into the mind at work "at the point of utterance." (See Steinberg and Dobrin for commentaries on this method.) Others used videotapes (Matsuhashi) or computer programs ( Bridwell-Bowles, Johnson and Brehe; Bridwell, Sirc, and Brooke) to capture behavioral records of composing processes in minute detail. With the advent of computers into writers' workplaces came research on the effects of word processing on composing and on the effects of computer-assisted instruction (see Bridwell-Bowles, "Designing Research," for a summary of much of this work). Many of these researchers attempted to build models of the cognitive processes involved in writing as they interpreted their data.

In recent theoretical work, we have begun to recognize new research questions on the product side of the process/product dichotomy. For too long we believed that the more intriguing problems in research were on the process side. With new insights from discourse analysis (see Coulthard and van Dijk for introductions) and poststructuralist theory (see Culler for a historical introduction), it is now clear that our conceptions of text demand just as many questions. White provides an accessible summary of the effects of new theories about reading texts and how they account for our responses to writing.

Given the complex relationships and interactions among writer(s), reader(s), text(s), and culture(s), many possible readings and interpretations of a "text" are possible. As Phelps asks, "What concrete fact of experience, then, are people pointing to when they 'refer' to the structure of a given text?" (138). This question is of particular significance in writing assessment studies where theorists are beginning to question judgments of writing quality made without adequate attention to context. In fact, if we question our judgments of writing quality, we question the results of much composition research that uses quality as a variable (but more on this later).