Ignorance in Educational Research:

Or, How Can You Not Know That?

Educational Researcher 22 (June-July, 1993): 15-23

Jon Wagner

Division of Education

University of California, Davis

In this paper I will argue that ignorance is a better starting place than truth for assessing the usefulness of educational research. This argument contrasts with popular parlance about research as the unadorned “pursuit of truth.” It also suggests that attention to truth alone in determining the value of educational research reflects a classic distortion of means over ends, a distortion that the sociologist Robert Merton (1938) called “ritualism” in his functional analysis of social deviance.

Truth and truthfulness are invoked commonly in judging particular aspects of the research process. However, some research projects are of little use to researchers or practitioners even though they reflect our highest ideals of truthfulness in data collection and analysis. Other projects can be very useful indeed,indeed, even though they fall short of such ideals (Davis, 1971). When we judge a research project solely on the apparent truthfulness of its parts, we neglect its larger purpose: generating new knowledge about education and schooling. To understand when research is likely to achieve this purpose, educational researchers must begin with ignorance, not truth.

Taking ignorance as a starting place in assessing the value of educational research has implications for how we think about educational research, how we teach it, and how we frame and support relationships between researchers and their subjects. As a point of departure for examining these implications, consider the potential research value of a high school yearbook. Clearly, the information and images it contains owe much more to previous yearbooks than they do to a framework of critical, educational analysis. Most photographs of students and faculty are carefully posed, all have been carefully selected, and some have been altered in the darkroom to remove skin blemishes, strands of flyaway hair, and so on. The content of these images is confined by and large to official school activities. No information is provided about the academic performance of students, the school’s curriculum, its financial resources, nor about the credentials and careers of teachers, administrators or support staff. In addition, as a rule enforced by administrators and faculty who supervise the students who work on it, the high school yearbook typically avoids references to disciplinary issues, drugs, drop-outs and other indicators of academic, personal or institutional failure.

All this suggests that a high school yearbook falls far short of what we would like to see as a truthful account of the school it memorializes. On these grounds alone -- even if I know nothing else about the school -- I should be suspicious of the yearbook’s veracity and validity. And, if I’ve actually attended or studied that school, it looks much worse, for I can then support my suspicions with specific examples of distortion and neglect, so many, perhaps, that my inclination will be to define the year book in terms of its deficits alone.

But let’s say I have attended this school, or studied it, and I want to compare it with another one, or to the same school at another point in time. Could the yearbook contribute to that effort? Perhaps it could. The yearbook pictures and names might help me determine the ethnic and linguistic heritage of the school’s students and teachers. Or, even if I knew about that from other sources -- though this kind of data has been collected by schools only recently -- the yearbook might be the only source of information about how these variables and gender correlate with participation in different school activities.

Or, let’s say I want to report about this school to people who know much less about it than I do through having attended or studied it. Would I consider using the yearbook to assist me in that reporting process? What if these people were colleagues of mine from another country, Brazil or China, for example? If I show them the yearbook they might notice that it was written in English. Is that a distortion of the school? They might also note the social organization of students into four classes, each identified by their graduation year, and note as well the in institutional celebration of the senior class. In yearbook photographs they might also find evidence of clothing, dancing and sporting activities that differ markedly from those with which they are most familiar. In helping me communicate all this to them, the yearbook would be a rich source of information indeed.

As these speculations suggest, we can assess the yearbook’s value to generating new knowledge with reference to either truth or ignorance. In the first instance, we ask How closely does it approximate a “truthful” picture of the school? Compared to a truthful account, in what ways is it flawed? In the second, we pose two somewhat different questions: How far beyond ignorance does this work take us? Compared to what we don’t know without it, in what ways can this work help us know more?

Both sets of questions may be worth asking, but I’d like to affirm an epistemology in which the second set comes first. That is, what makes a high school yearbook -- or a four-fold table, an index of statistical correlations, a narrative account of field research or any other document or representation of information -- useful in generating new knowledge owes something to the representation itself. But the kind of knowledge it generated and its value is framed by the ignorance of the person using it for that purpose. The implications of this generalization for the conduct of educational research become clearer when we look at two generic forms that this ignorance can take.

Blind Spots and Blank Spots

In constructing knowledge about education and schooling, educational researchers use a variety of different “materials.” These include data of various forms and types, direct experience, concepts and theories of their own or those developed by others, and so on. Some of these materials may help educational researchers answer questions that they have already posed. Others may stimulate them to ask questions they haven't asked before.

We can think of these two functions of the materials of educational research as responding to two kinds of ignorance. Materials relevant to questions already posed can be seen as filling in blank spots in emergent social theories and conceptions of knowledge. Materials that provoke scientists to ask new questions illuminate blind spots, areas in which existing theories, methods, and perceptions actually keep us from seeing phenomena as clearly as we might.

Educational researchers are not the only people with blind spots and blank spots. All scientists operate in a world defined by what they think and know to be true. What they don't know well enough to even ask about or care about are their blind spots. What they know enough to question but not answer are their blank spots. The same phenomenal categories are alive for non-scientists as well, and in some ways the particulars of these categories for scientists and non-scientists have much in common . This is in part what John Dewey had in mind when he wrote that, “Anything which can be called a study, whether arithmetic, history, geography, or one of the natural sciences, must be derived from materials which at the outset fall within the scope of ordinary life-experiences” (1963 [1938]: 73).

But one way of distinguishing between scientists and non-scientists and between scientists working in different disciplines or subdisciplines is by the characteristic configuration of their blind spots and blank spots, that is, by the structure of their collective ignorance. Following the work of Thomas Kuhn (1970), this structure can be represented for a particular tradition of inquiry (e.g., sociology, anthropology, psychology, history, etc.) by a “disciplinary matrix.” Rows of such a matrix can represent concepts or methods of investigation, columns phenomena that members of the discipline tend to examine. A matrix of this sort defines sets of related cells, each corresponding to the intersection of a particular concept or method and a particular object of investigation. Borrowing from different parts of a recent synthesis (Smelser, 1988), a small section of a matrix of this sort for sociology might look something like Figure 1.

Figure 1

One corner of a matrix of sociological inquiry

	Phenomena Under Investigation
Themes of Analysis	Jobs and Work	Sociology of Education	Sociology of Religion	Medical Sociology	Political Sociology
Social control
Social stratification
Status attainment
Bases of integration and differentiation: class, gender, age, race ethnicity, etc.
Social relationships: group, household, community, collectivity
Social change

The intellectual work of individual social scientists can be placed within a particular cell or set of cells. Some social scientists work within cells that adjoin along particular rows. For others the pattern involves adjoining cells within a particular column.

Materials useful to this work are defined by the rows and columns themselves. These materials allow social scientists to construct a more detailed matrix of rows and columns within a particular cell. In Figure 2, for example, I have taken one column from the matrix displayed above -- the column called “sociology of education” -- and elaborated it as a matrix in its own right.

Figure 2

Sub-matrix for sociology of education

	Phenomena Under Investigation
Themes of Analysis	Lessons	Classrooms	Schools	School Districts & Communities	The State
Social control
Social stratification
Status attainment
Bases of integration and differentiation: class, gender, age, race ethnicity, etc.
Social relationships: group, household, community, collectivity
Social change

Once again, blank spots represent cells in the sub-matrix that have not been investigated as adequately as we would like. Blind spots represent ways of organizing concepts or categorizing phenomena that are obscured by the sub-matrix itself. The simplified matrices presented above for sociology and for the sociology of education, for example, obscure attention to individual personalities, technology, the global economy, etc. A narrative example or two may illustrate these distinctions more clearly.

Blind Spots in Educational Research

During the 1970's, in response to increasing concerns about educational equity and the growth in the 1960's of policy interest in alternatives to conventional schools, Christopher Jencks and his associates conducted a series of ambitious studies about the outcomes of schooling (Jencks, Smith, Acland, Band, Cohen, Gintis, Heyns, and Michelson, 1972; and Jencks, Bartless, Corcoran, Crouse, Eaglesfield, Jackson, McCelland, Mueser, Olneck, Schwartz, Ward, and Williams, 1979). These studies used aggregate data analysis to investigate a “row” of variable characteristics of schools and a “column” of the effects of schooling for different populations of students. The effects in question were educational and occupational achievement subsequent to elementary and secondary schooling.

The blank spots upon which this work focused were defined by the scientists themselves, but the framework guiding this inquiry also obscured other ways of looking at similar and related phenomena. One set of phenomena obscured by these aggregate studies was the achievement and experience of individual students in individual classes. In their complementary research, Summers and Wolf (1975, 1977) took the careers of individual students -- not student populations -- as their point of reference and asked some of the same questions about individual classrooms that Jencks and his associates had asked about schools. In doing so, they came to quite different conclusions.

Jencks and his associates had focused on group effects -- the same approach taken by Coleman (1966) a decade earlier -- and found that organizational differences in schools made little difference to the outcome of schooling for different groups of students. But Summers and Wolf found that differences among classrooms within schools did generate significantly different outcomes for members of the same groups examined by Jencks. By aggregating students into groups and classrooms into schools, the Jencks studies collapsed these differences and made them invisible. By focusing on individual students and classrooms, however, the Summers and Wolf study could not describe the larger aggregation of “schooling” outcomes that the Jencks study described for particular populations. Each study was organized for the purpose of identifying materials that could help fill in a particular blank spot. However, even though it met conventional standards of research quality, each study also created by its very structure a distinctive blind spot.

We can also imagine more radical blind spots. At least Jencks and Summers and Wolfe were both working with similar kinds of data and somewhat similar questions. What happens when we compare these efforts to Paul Goodman’s analysis of compulsory education itself (Goodman, 1964). Given that lack of a real alternative, it is impossible to investigate the impact on student academic performance of compulsory attendance in school. However, it is also hard to take seriously studies of schooling that ignore it. As James Herndon (1972) has noted, when a student really likes class, what the teacher knows is not that it’s the best thing the student can imagine, only that it’s better than jail.

The same patterns found in educational research appear as well in social research applied to other problems of practice or policy. For example, educational research focusing on the work of individual teachers and research on the organization of schools frames the same configuration of blind spots that Van Maanen and Barley (1984) identified between research on formal organizations and ethnographic studies of workers.[1] Something similar occurs regarding research on all the ways schools change and research on particular efforts to change schools. A similar relationship applies among studies in education of “method,” studies of “philosophy and methodology,” studies of particular “variables and concepts,” and studies of particular “schools, settings, and organizations.” A recent illustration is provided by the exchange in Educational Researcher about “whole language” conceived on the one hand as a finite set of methods and assumptions about teaching language (McKenna, Robinson, and Miller, 1990a, 1990b) and on the other hand as a comprehensive social, pedagogic, and political paradigm for engagement and investigation in language development (Edelsky, 1990). Each approach momentarily frames for investigation a distinctive blank spot and tries to fill it in with detailed information. In doing so, however, the frame itself defines a host of blind spots.

The specialization of attention and effort required to generate new information within a particular blank spot leads to one of the great ironies of social research: Folk conceptions of the social order are frequently more complex and sophisticated than the matrices by which members of individual social science disciplines organize their work. In terms of the schooling studies mentioned above, for example it is a commonplace to note that different forms of instruction are more or less effective with different individuals. It is also common wisdom that the children of the wealthy and educated become wealthy and educated adults while children of the poor and uneducated grow up to become poor and uneducated. A vernacular critique of the Jencks study would emphasize contradictions between what I know from having learned more in some classes than in others and what these social scientists say about how schooling doesn't make any difference. A similar critique of the Summers and Wolf work would point to contradictions between how the the author’s description of how school classrooms make a difference and what I know about poor people staying poor and rich people staying rich.

One Discipline’s Row, Another’s Column

While both scientists and lay people have their blind spots and blank spots, scientists have organized their working lives around some of their blank spots. In doing so they stake both their reputations and livelihood on their ability to use materials, collect materials, and create materials that can fill in these particular blank spots with additional detail. Given this tall order -- and the distance between these blanks spots and those examined by their colleagues working in the “philosophy of science” matrix -- it is understandable that social researchers define the goal of their investigations in terms of truth. However, given the blind spots necessarily created through their dedication to particular blank spots, truth seems a peculiar and inappropriate point of reference for judging what they do.

Ignorance seems like a better bet, in two respects. First, ignorance provides a good criterion for assessing the value of particular lines of research. Does the research help us overcome the ignorance represented by either a blind spot or a blank spot, or does it not? Second, much more so than truth, ignorance encourages us to ask “whose” it is. As a matter of practice -- though the practice could change -- truth claims are framed to emphasize timelessness, anonymity, and independence from context and historical moment. Statements about ignorance, in contrast, are grounded in particular people, places, times and contexts.

These two considerations help describe one of the more prosaic ways in which ignorance is useful to social scientists in generating new knowledge: Adding a row or column from one discipline’s matrix to another generates a whole new set of questions. What is borrowed may be far from the cutting edge of the matrix it comes from and yet represent a powerful innovation for members working within the matrix to which it is added.

For example, during the last five to ten years there has developed among educational researchers an extraordinary interest in the social context of learning. Why and how has this occurred? Why is it news that the social contexts influence the conduct and outcomes of instruction? Would any lay person consider this to be news? Perhaps some, but not many. It certainly wasn’t news to the sociologist Willard Waller, writing in 1932. But it has become news to psychologists -- at least some of them -- and by far the largest number of educational researchers have psychology as their home discipline. Thus, what we have witnessed in professional journals and at annual meetings of the American Educational Research Association, is a collective process through which psychologists have added some rows and columns to their matrix of inquiry, rows and columns representing concepts, theories, and objects of inquiry from matrices associated with sociology, anthropology, and history. The news of “social context” is generated by the discovery among psychologists that the categories guiding their research kept them blind to important aspects of the phenomena they were trying to investigate.

This kind of disciplinary annexation is in no way limited to psychology. Discussions about re-incorporating the “individual” into anthropology (see Wolcott, 1991; Trueba, 1991; and Spindler and Spindler, 1991) or sociology point in just the opposite direction. Similar patterns appear among other social science disciplines and among the social sciences, natural sciences and humanities. In each case, members of a discipline become engaged by new questions (i.e. new blank spots) that were not even on their old matrix of inquiry. But these are not new questions on the face of the earth, just new to researchers doing the borrowing. They are old questions for those they have been borrowed from.

Do these annexations and extensions of disciplinary matrices lead researchers closer to the truth? I don’t think this description fits as well as another, that they move individual scientists farther away from ignorance. In this respect they are useful. A broader and more inclusive perspective, method or theory can help reduce ignorance about the limitations of the perspective, theory and method we were using in the first place. But there is no end to what we don't know about our perspectives, methods and theories. As a result, we can say whether or not a particular approach is useful to learning what we want to know, but there is no point at which we can say an approach is adequate or satisfactory for generating truth.

Researchers, Subjects and Practitioners

This way of thinking about quality and usefulness in research has some implications for what researchers do with each other, but probably not too many. Researchers work in institutional and social groups -- what Van Maanen and Barley (1984) call “occupational communities ” -- that have more or less constructed their own way of doing things and their own perspective on the issues I've presented here. These ways of doing things and perspectives are as rational as anybody else's, and they are familiar.

One implication this perspective might have, however, is that of encouraging researchers to be more generous and eclectic in defining what constitutes research and, as a result, to worry a bit less about what is and is not “real” research, focusing instead on the purpose of a line of inquiry and assessing how well that purpose is achieved by the particular study or approach in question. As Howe and Eisenhart (1990) argue, “The question of standards for qualitative research--indeed, for research of any kind-- is, then, a fluid one, and one that must be answered in terms of the successes and failures of inquiry. In turn, successes and failures can only be judged relative to given purposes” (p. 3, emphasis added).

Thinking of ignorance as a place to begin in assessing the utility of research may have greater implications for how researchers work with research subjects and for boundaries between research communities and communities of practitioners and policy makers. For example, if we replace the truth claims of researchers with more modest and appropriate claims of reducing ignorance, the latter beg to be specified in terms of the group for which ignorance will be reduced. This ties research closely to teaching, and that raises questions about how research is reported, to whom, and to what effect (questions I will turn to in a moment).

Defining research as a strategy for reducing ignorance may also make it more accessible to people who are otherwise intimidated by their vision of research as “pursuing truth.” In recent years that kind of access appears to be on the rise, at least for members of the so-called “professions.” Drawing on the concept of the “reflective practitioner,” professional associations and universities have given more attention to preparing professionals who can investigate the circumstances and outcomes of their work (Schon, 1983,1987; Agyris, Putnam, and Smith, 1985; Goodlad, 1990). One of the most visible of these efforts in education involves the increasing attention given to “teacher research” (Myers, 1985; Cochran-Smith and Lytle, 1990), an affirmation of teachers’ potential engagement as investigators in their own classrooms and schools.

These reflective-practitioner and practitioner-research efforts have met with a fair amount of criticism, or, at times, even scorn, by members of more traditional research communities. This is understandable, for the studies teachers conduct of their classrooms -- or those other practitioners conduct of the circumstances of their practice -- do not fit well within the inquiry matrix of a particular social science discipline. In many respect, criticisms of teacher research or practitioner research reflect failed efforts by researchers working in traditional disciplines or sub-disciplines to find a good match between their own rows or columns and those of practitioners. In this respect practitioner researchers create for traditional educational researchers some of the same problems in understanding that educational researchers create for researchers working within traditional academic disciplines.

However, if we keep the ends of research as well in mind as the means, we must acknowledge that blind spots and blank spots are at the core of all research endeavors. We can then apply to educational research, teacher research and other forms of practitioner research the same “ignorance-based” standards we apply to research writ large. In doing so we move away from asking whether or not the research generated truth, and ask instead whether it reduced ignorance, what kind, and for whom. Did the research help researchers, policy makers, participating teachers or others involved with it to fill in blank spots or to illuminate blind spots? If so, it was probably useful research.

Within this reorientation, teaching becomes a strategy for using research to reduce ignorance, and research a strategy for preparing to teach. While this articulates research closely with teaching, it also suggests how complicated teaching really is. That is, the people we want to teach have blind spots and blank spots defined already by categories, questions, and objects of their own concern and investigation.[2] To teach these groups effectively, we need to understand their own matrices of inquiry (e.g., Shavelson, 1988). Researchers invest a fair amount of time in doing just that for other researchers, a popular manifestation of which is the “literature review,” but schools and social life are not designed well for helping us do that for non-researchers. However, unless we understand the matrices of inquiry already in place for our students or publics we don’t know what blank spots exist for which we can provide new information, and we don’t know what rows or columns we can add to reveal blind spots.

Teaching and Cooperative Research

The connection noted above between research and teaching suggests the potential value of alternative strategies for conducting and organizing educational research, strategies, for example, in which research processes and results are shared with research subjects. Some of these strategies have been critically examined and advocated by Lather (1986), by Agyris, Putnam and Smith (1986) and by others. In terms of the analysis presented here, they represent one way to treat the ignorance of researchers and the ignorance of research subjects with mutual respect.

An intriguing illustration of this approach is provided by the work conducted by Howard Becker and reported in his book, Writing for Social Scientists (1986). The focus of Becker’s research and book are on how social scientists write and learn to write in and out of school. Reviewers of the book have been favorably disposed, describing it as a “valuable resource” (Mullins, 1987); as “humane, wry, reflective, gentle, and wise,” (Erikson, 1986); and as “delightful” and “sound advice” (Townsend, 1986). However, most reviews of the book have focused on the advice it offers writers while ignoring the research on which that advice is based. The book’s title and the publisher’s promotional materials may have contributed to this neglect, but it is also likely that the “cooperative” approach Becker took in conducting his research contributed as well.[3]

The sub-title of Becker’s book is, How to Start and Finish your Thesis, Book or Article. Chapter titles include the following: “Persona and Authority,” “One Right Way,” “Editing by Ear,” “Learning to Write as a Professional,” “Risk,” “Getting it out the Door,” etc. These titles reflect a central ambiguity: Is the book a description of how these concepts, issues and themes are manifest in the writing activities of social scientists, or is it a book on how to do them. This ambiguity recapitulates some of the abiding dualities around which educational researchers organize their work, including those of theory and practice, action and reflection, and research and reform (or, as reform appears in this case, the work of improving how individual social scientists write).

Within a larger enterprise of educational research informed by these dualities, Becker’s book is intriguing because of the match between its subject (i.e., writing social research) and its audience (i.e., social researchers who write). Because of this conjunction, the book exemplifies the prospect of looking at social research itself as a form of praxis (Lather, 1986). Within this prospect, the dualities noted above are integrated, not separated, in the working perspectives of educational researchers themselves.

In the introductory chapter, Becker describes how the book emerged within his own intellectual and professional life. Let me provide an unfortunately inelegant summary of this account: In response to the inordinate amount of time he spent discussing and editing the work of graduate students through individual consultations, Becker decided to teach a course on writing in which he and the students could work on these things collectively. As he stood in front of the class for the first meeting, he realized he didn’t know how they wrote. When he asked them to tell him, he began learning that they didn’t know how they wrote either. Through the class, Becker and his students worked on things they were writing and discussed the writing process they experienced as social scientists, and at the end of the class they talked about writing a paper about all this.

Becker wrote a description of the class and some of the ideas discussed in it. He distributed this to members of the class and to other colleagues, many of whom commented on the paper, and some of whom shared it with other social scientists and graduate students. Becker used these comments to revise the paper and eventually had it published in Sociological Quarterly (Becker, 1983) where others commented on it as well (Hummel and Foster, 1984). Additional readers -- probably all of of whom were also trying to write social science theses, books and articles -- wrote to Becker about the article, some of them sending long notes of their own about how these issues affected them

Becker revised some more, read some more, wrote some more, and talked some more. He began organizing this material as a book and showed it to a publisher. The publisher sent it to some readers, including social researchers and composition scholars. The composition scholars criticized the book for not taking into account all the work that composition researchers have conducted. But the the social researchers saw Becker’s work as a contribution to social science. The column Becker added to the social research matrix about writing was news to social scientists, even if not quite new to composition scholars. But that was enough, because the composition scholars didn’t write for social scientists and Becker did. It would be hard for them if they wanted to, because they don’t have the right matrix in mind. They don’t know the right blind and blank spots of social research.

Hearing from some composition scholars about the draft manuscript alerted Becker to their work in time to read it and incorporate it into his thinking and his manuscript. Even though composition scholars had not studied how social scientists write, they had studied how other groups of people write, and Becker was interested in potential parallels and contrasts. For Becker, these parallels and contrasts helped conceptualize the data he had collected about how social scientists write. In the book, he doesn’t describe this as data per se, but he does give a good account of his “method” in thinking through and writing the book.ook The account is good enough for us to construct a typology of his “data-sources” that might look something like Figure 3:

Figure 3

Categories of research subjects and data sources for

Writing for Social Scientists (Becker, 1986)

	Data Collection Activities
Research Subjects	Individual interviews	Group interviews	Analysis of written documents	Oral and written responses to research reports
Graduate Students in first class
Graduate students in subsequent classes
Readers responding to draft report of first class
Readers responding to article in Sociological Quarterly
Other social science colleagues
Readers responding to book

Displaying Becker’s data sources in this form helps clarify the “materials” he drew on in constructing his knowledge of how social scientists write. He doesn’t tell us how many research “subjects” fell within each cell of this matrix, but we know what the cells are. Some researchers might like to know much more about this; others might not, at least not for the kinds of knowledge Becker is constructing.

. Becker’s research and the book he wrote have also been involved in a lot of teaching, and the data source matrix helps describe that as well. For some people, the book itself represents the only teaching they encountered in conjunction with Becker’s project -- at least the book and/or its published reviews -- but that’s not the case for the graduate students and colleagues who exchanged written and spoken words with Becker while he was preparing it. For those people, the research itself helped reduce their ignorance, and Becker’s, and did so because it was conducted in collaboration with them and reported to them in an iterative manner over the course of several years.

As this account suggests, what Becker has conducted and reported on in Writing for Social Scientists is an extended project of practitioner research, or, as it happens, of teacher research. In this case, a teacher-initiated project of inquiry and investigation has been conducted in ways that are of value to the teacher himself, the work of his students, and the efforts of other teachers working in the same or similar fields. Becker’s work also represents a project of cooperative research, in which research findings, data-collection strategies and analysis were shared with research subjects, some of whom made their own contributions to the analysis appearing in the book (one chapter is a verbatim account of an analysis prepared by a reader of the first paper Becker wrote ). And, as at least some reputable scholars have noted (e.g., Cazden, 1987, and Erikson, 1986) the book reflects good social research as well.[4]

Is Becker’s account of social science writing truth? I couldn’t tell you. Does it help social scientists and those who teach social scientists reduce their own ignorance about writing? Yes indeed. Is it therefore good research? I think so, at least for social scientists, though I still wonder how well it sits with the composition scholars.

Knowledge about Ignorance

Distinctions between pursuing truth and reducing ignorance are not hard and fast. A case can be made that the same research methods, criteria for evaluating evidence, and logic of argumentation or exposition should apply within either model of science, or that pursuing truth in the face of ignorance is equivalent to reducing ignorance in response to truth.

However, distinctions that seem weak within the formal logic of research can still generate strong effects within the informal logic that guides the work of individual researchers. For example, the informal logic of most social researchers -- and of most natural and physical scientists -- focuses less on pursuing truth per se than on generating and reporting “new” truths. -- and This focus is made explicit in university criteria for evaluating the work of faculty members within merit and promotion reviews. These criteria laud contributions to “new” knowledge and denigrate research that “merely” confirms or synthesizes existing knowledge (Boyer, 1990). This distinction may be moot within the formal logic of social research, but it dominates the informal logic of conducting research in universities, including the informal logic of conducting educational research grounded in the social sciences. The distinction between pursuing truth and reducing ignorance is similar in its substance and effect.

One implication of these informal orientations to truth or ignorance is that they support different conceptions of scientific and public knowledge. In “pursuing truth” we assume that knowledge is both constituted and limited by empirical information. We also assume that we are reasonably well-informed about what we know and don’t know. By generating and organizing new information, scientists help people, including other scientists, learn more about what they want to know. “Truth” is “pursued” across an identified territory that has yet to be fully explored, and the science associated with this pursuit is dominated by blank spots.

In contrast to pursuing truth, a science based on reducing ignorance does not assume that what people know is constituted or limited by empirical information. It assumes instead that what people know expands to fill what they feel they need to know or want to know, and that “empirical information” is but one of many sources of knowledge. That is, in response to challenges of work, intimacy, political life, and other domains of activity in which they are engaged, individuals construct knowledge whether or not they have information. This knowledge -- some of which can take the form of self-fulfilling expectations -- both informs and reflects individual conceptions of gender, ethnicity, social class, community the state and schooling. People, including scientists, also construct ideas, concepts and stories of their own history and future, of groups of people they have never met, of economics, policy-making, government, and technology, even of times past and of territories, lands and events, for which their information may be limited to a single phrase or name. Researchers who recognize the ubiquity and power of these knowledge-making processes are not so sure that they know what they need to know, and blind spots are at least as important to their science as blank spots.

Drawing on these distinctions between different conceptions of knowledge and different perspectives towards science, let me offer four propositions related to the design, conduct and meaning of educational research.

First, conceptions of epistemology in educational research are closely connected to conceptions of pedagogy and method. Ideas about what knowledge looks like necessarily involve ideas about how it is acquired -- and can be acquired -- either by educational researchers or by other members of society. Research itself is a form of learning and research reporting a form of teaching. By helping to define what people don’t know and might learn next, ignorance is a central concern in both of these processes.

Second, and in part because of this, the theoretical problems that define educational research and the related social sciences can frequently be worked on firsthand in the social relationships between researchers, their subjects and their publics. This fact generates extraordinary ironies, tensions and contradictions within the social sciences themselves. For example, when researchers must invoke completely different pedagogies, metaphors and social theories to explain how students, parents, teachers, or researchers learn what they know, they undermine their credibility with these other groups and with researchers who care about the generalizability of their theories. However, when researchers invoke a common pedagogy and social theory to describe how they and members of these other groups learn, they undermine their credibility with other researchers and research subjects who look to them for expertise and authority. This tension can complicate greatly the lives of individual researchers, but it also contributes to the continuing vitality of educational research and the social science disciplines.

Third, with ignorance as a reference point, the potential utility of educational research by and for research subjects deserves more attention than most researchers have been willing to give it. Becker’s work illustrates the possibility of conducting educational research -- in this case research about learning to write -- in ways that are useful to subjects, other teachers and practitioners, and to professional researchers themselves. A growing number of other examples illustrate how similar objectives can be pursued in educational research involving teachers, students, parents, or administrators.[5] Of course, research by and for subjects is not only feasible, it is inevitable. It is what subjects do with researchers, whether the researchers want them to or not (Clark, 1991), but it might be done more gracefully, effectively and humanely if researchers acknowledged it directly.

Fourth, getting rid of truth as the goal of educational research does not necessarily mean abandoning efforts to be truthful in generating new knowledge. Rather than pursuing truth, however, researchers can explore different approaches to truthfulness as vehicles for reducing the ignorance of scientists and of non-scientists. In practice, that’s sort of how it works anyway. We try to chart a reasonable course between the foolishness of not caring about truthfulness at all and the distortions of life and work that arise when we care only about truth, and particular truths at that. As Sherwood Anderson observes through the narrator of Winesburg Ohio,

. . . it was the truths that made the people grotesques. . . The moment one of the people took one of the truths to himself, called it his truth, and tried to live his life by it, he became a grotesque and the truth he embraced became a falsehood. (1980: 23-43 [1919]).

In practice, we know much more about ignorance than we do about truth. That’s part of what makes truth so problematic as a criterion for assessing the usefulness of knowledge generated through educational research. But we do know enough about ignorance, or at least we can learn enough about it for particular people in particular situations, to use it in just that way.

Notes