General Issues

In composition research, we have clashing epistemologies. They exist within the communities of researchers, and among the consumers of composition research. Most college writing teachers are trained in the humanities and look toward analysis, deductive reasoning, rationalist introspection, and other forms of theoretical speculation more often than they look to experimental, empirical research for answers. Secondary English teachers are trained in colleges of education and are also unlikely to develop sophisticated skills with experimental jargon in their undergraduate programs. Writing teachers deal more in complexities and uncertainties than in probabilities or statistical principles. They tend to prefer classroom narratives over experimental studies. As Stephen North has pointed out, positivist methods of research constitute a "method that seeks to approach certainty by reducing uncertainty." One can never quite apprehend truth with this model of research, but there is never any doubt that truth (in this case, a patterned, systematic theory of "human science") is out there. Many teachers know very well that "It varies" is a better answer than some of the predictive equations derived from statistical models.

So why do we continue to seek answers in quantitative studies?

Even though compositionists are not always trained in the methods of social science research, as Phelps points out (vi), and even though scientific methods may be inappropriate, composition specialists are often called upon to provide this kind of data to administrators who are accustomed to clear and concise presentations. When composition program directors seek funds for such things as new programs in "writing across the curriculum," reduced class sizes, or classrooms filled with expensive computer technology, they are often asked for "research" that demonstrates the "cost-effectiveness" of such expenditures. Administrators typically find numerical data, often pre-test/post-test measures, more compelling than theoretical arguments, naturalistic research, or case studies that provide potential analogies for the proposed activities. North's argument for organizing his book around the characteristics of researchers was that "we" constitute different research communities. Perhaps there is another more important point: the consumers of our research certainly differ from us. When "hard" scientists, social scientists, and leaders of professional programs make decisions about budgets in our institutions, we are pressured to use the "tools of the master" to get the resources we need for our writing programs. However, I find it hard to imagine many natural scientists or social scientists who would be impressed by some of the inappropriate aping of the social sciences that has gone on in composition research.

Granted, there are some areas where quantitative answers are extremely useful to all of us as research consumers. I can recall dozens of times when I have cited research on the scarcity of sustained writing in the nation's secondary schools, or some of my own quantitative research on revision or composing processes with computers. The answers they provide are useful for answering "how" questions: How much writing is happening? How often do students revise? How are writers using computers differently from typewriters or paper and pen?

How much change is there in sentence complexity from the fourth grade to graduate school? However, the more basic interpretive questions have to be dealt with in qualitative terms: Why did these things happen? Why did other things not happen? What are the underlying assumptions that account for x, y, and z? What kind of cultural blinders am I wearing when I interpret? Are there competing, coexisting explanations for our observations? And so on.

One of the contributors to Moran and Lunsford's book, John Briggs, invokes rhetoric as a cure for the "the swings between extremes of anxiety and euphoria" on the part of researchers with naive hopes for their endeavors. Echoing an essay by Douglas Park, Briggs observes that "an endemic desire among researchers to build theories of rhetoric and composition, apply them directly to pedagogy, and witness the revolutionary perfection of a discipline" has led to this malady. As one of many who had such naive hopes, I can attest that this was the spirit of the times in the 1970s and early 1980s. Our faith in the ultimate perfectability of our theories rested on good works. If we worked hard enough to discover rigorous methodologies, most often social scientific, and used them with rich multidisciplinary theories, our findings could lead us to a theory of composition from which sound pedagogy could be derived. Lloyd-Jones says of his early hopes for empirical research, "if it were just better done, we'd have great improvements in the teaching of writing" (202). Of his more recent views, he writes, "we were probably over-optimistic about what could be discovered by empirical methods" (202). Like Briggs, Lloyd-Jones argues that the theoretical home for scholarship in writing is still a humanistic edifice built of philosophy, philology, rhetoric, and literature. He writes, "we belong with the humanists, not with the social workers" (207). I would argue that we need both because writing occurs in the kind of world where social workers may be as important as humanists. The economic and political issues associated with writing within the larger culture lend themselves to methods borrowed from political science and from anthropology as often as from Lloyd-Jones' favorites.

Conclusion

Even with all the epistemological and methodological differences in the field, we have learned a great deal from composition research. We have learned that our ways of composing are not merely straightforward executions of the structure of the final form of a text. We have learned that writers construct complex cognitive representations of their texts in progress, of their audiences, and of their roles as writers. We know a great deal about how successful writers gather material or revise their drafts, and we are learning to pass these strategies on to novice writers. We have learned about the contexts for writing in composition classrooms, in classes "across the curriculum," in business and industry, and in the individual's own private world. We have learned that some things work as we teach writing (e.g., peer groups, revision) and that some things don't (e.g., extensive teaching of grammatical terminology in lieu of writing). We have also learned that writing can be empowering in some lives, and disabling in others, and a few of the reasons why. We have learned all these things from a body of research that is varied, with sometimes incompatible methodologies.

What we need in composition research may not be "better" methods or "the" appropriate research paradigm, but a way of synthesizing information from diverse sources. The ultimate paradigm shift within English compositionmay be the loss of belief in the term "paradigm" as a useful metaphor for what we are seeking. Nearly every empirical researcher I have studied over the past two decades, including myself in earlier studies, refers to the need for a model of the composing process, or, with a small concession to complexity, for a model of the composing processes. We have heard repeated calls for a more flexible, humanistic approach within our research ( Connors, Bizzell, Burton, Voss). Perhaps what we require is not a model at all.