Response to Ridgeway, Dunston, & Qian: Authoring the Questions: Research as an Ethical Enterprise

T imes of flux and fomentation always seem richly productive times to take stock. In literacy research, as we see in Ridgeway, Dunston, and Qian's study reported in this issue of RRQ and the Alvermann and Moore (1991) work that precedes it, the meta-analytic impulse emerges: What do we know about research in secondary reading, and where do we go from here? Often the motivation is largely an archival one (Alvermann, 1992); occasionally, the meta-analysis is motivated by the need to push a particular agenda (Hillocks, 1986). But typically, as with the Ridgeway et al. study, the motivation is to determine what we know and where we are (and, as the authors wonder, whether-given the methodological inconsistencies of the research-we know as much as we think we do). Before looking at a study about studies, however, it is worthwhile to examine the belief system that fuels the meta-analytic drive, especially in experimental research. Two images come to mind when I think of metaanalysis. The first image, the story circle icebreaker, is familiar: The first person whispers a story to the second, who recounts it to the third person, and so on, until the circle is complete and the story retold aloud. Invariably, the final version is far from the original. This image portrays what Jacques Derrida claims occurs in all discursive contexts anyway: Language is in an endless process of deferral-the term he uses is differance-in which there are no fixed signifieds (concepts) shared among people; what anything means depends upon the location of the text-the speaker/writer, the intention, the historical and political context. Derrida critiques rationalist theories of language, particularly Western logocentric tradition which assumes that the meaning of concepts is fixed before we articulate them in language. We can draw the comparison with the process of meta-analysis. Our confidence in the process-regardless of the sensitivity or accuracy of the meta-analytic instrument-must depend upon our belief in that logocentric perspective at several levels: that we can adequately represent what was studied originally (and that the original study re-presented phenomena accurately in the first place), that differences in the contextual locations of the original studies count for so little that an objective instrument can wash them away, and that the emerging meta-analysis itself, at the end of the spiral of endless deferral of meaning, can, in fact, produce reliable direction for further study. Perhaps, as readers inquiring about reading, we have extrapolated our confidence in sound-symbol correspondence at the word level further than we ought to, into the realm of research itself: Just because a letter sounds a certain way, predictably and in most contexts, it does not necessarily follow that a word, an idea, a study, or any phenomena that involve people and intention will mean the same, predictably and in most contexts. While I agree with Annie Dillard's claim that some things can be known and understood (albeit partially and differently), certain knowing may not be measurable, nor is it necessarily helpful in order to teach our children. Or, in the words of Ridgeway et al.: Research