Presented at the AERA meetings, Montreal 1999, in session 17.26
Islands of Truth and Wisdom:
How Many Epistemologies Do We Need?
This symposium was organized and chaired by Andrea DiSessa (University of California, Berkeley). The participants were Paul Cobb (Vanderbilt University), Beverly Gordon (Ohio State University), and myself. The discussant was Allan Collins (Northwestern University)
Andy DiSessa described the plans for the symposium as follows:
This panel discussion seeks to engage a critical tension in contemporary educational research. On the one hand, education is a diverse enterprise with many foci of attention. In addition, science always grows from diversity; a static orthodoxy without ebb and flow of new ideas and trends cannot constitute progressing science.
On the other hand, the dangers of factionalism and of a splintered field cannot be ignored. A field of research with too many voices will find itself impotent in its appeals for influence and support from the wider society. More importantly over the long term, mature sciences have always solidified broad commitments to a set of core ideas. Indeed, they come to assume standards of evidence and arguments, even if these are inarticulate. Before broad base-level agreements are fashioned in scientific pursuit, the temptations of divisive, insider-outsider rhetoric, and parochial, self-serving "standards" of "truth" or intellectual value are obvious.
At what stage are we in the pursuit of field-wide agreement?most particularly with respect to basic epistemological assumptions about what constitute adequate argument and demonstration? What is the proper balance, at this stage, between the pursuit of field-wide coherence and the recognition that we are, as yet, well short of that goal? How should we--if indeed we should--promote higherand broader standards of scientific pursuit in the field?
After an introduction, the panel will proceed with introductory presentations by a group of educational researchers outlining their views on these issues. To promote productive interaction, presenters will be asked to address a common set of issues, such as the following.
1. Identify, briefly, the intellectual line that constitutes your "way of knowing" (methods and standards of evaluation, in the broadest sense). Provide some sense of your long-term aspirations in terms of wide-scale agreement: What would the success of your program be like?what would you "know"; whom would you like to convince and when would you expect to engage in convincing them; what would you be able to do, based on your (future) accomplishments?
2. Briefly position yourself with respect to others on the panel and the intellectual lines they represent, or with respect to other landmark points of view. Which of these do you see as "allies" and which as "competitors"?
3. What, if any, strategies do you have to offer for promoting convergence and high standards of scientific accountability, assuming you believe these to be reasonable goals?
A commentator will then draw out central agreements and disagreements and set the stage for an open discussion among panelists and the audience.
Comments on Interpretive Research, by Martin Packer
Well, there's so much to say, and only 10 minutes to say it in. And it is both challenging and very interesting to be discussing these issues with representatives of such a range of positions.
How many epistemologies? I call the kind of research I engage in interpretive--it is, if you want to be precise, hermeneutic phenomenology. Almost 25 years ago I looked around for ways to study the interactions I was videorecording between newborn babies and their mothers, and I found nothing I was satisfied with in the mainstream psychology of the time. I started to draw from a variety of sources to put together a workable interpretive approach, one that recognizes the special character of human action--and of humanity.
Since then I've worked to create and defend a space for this kind of work, in the face, at least initially, of incredulity that anything other than the standard empirical-analytic model research might be possible. So part of me wants to say, 'As many epistemologies as people wish!' Although Thomas Kuhn (1970a) described a 'normal' science, a mature science, as unified by a single paradigm, it's not clear that the social sciences need to operate in the same way. Certainly AERA has in recent years provided a home for a heterogeneity of research approaches, and I'm grateful for this. And in the natural sciences, researchers are now so aware of Kuhn's work that overthrowing paradigms has become the norm. Reflexive awareness transforms human practice and alters its character; Kuhn's description of science has changed the way scientists work.
So one response is, the more islands the merrier.
On the other hand, as I see it, the issues we are discussing today actually cut deeper than epistemology, and here I see both an argument for seeking some unity in this diversity, and a possible way to proceed.
Research programs can be defined by their techniques of choice, and by their views of knowledge, truth and validity; together these define a methodology--a logic of method. But research programs are also defined--and Kuhn recognized this with his notion of paradigm -- they are also defined by their ontological commitments: by their notions, albeit tacit and unarticulated, of what kinds of entity exist.
Epistemology is about knowledge; ontology is about existence; about being. About what kinds of stuff there is to be known. And epistemology is dependent on ontology. My slogan here could be 'No epistemology without ontology!' In fact, some ontological commitments lead to intractable difficulties at the levels of epistemology and methodology. Empirical-analytic research in the social sciences--the kind of research taught in traditional research methods and statistics courses--still, in my view, adopts a dualistic ontology, one that sees the mental and the material as distinct and incommensurate realms of being. The intractable problems this ontology gives rise to are well-documented: if knowledge is ideas in the mind, how do we ever ensure it corresponds to the way things in the world actually are? Even if, as the logical positivists proposed, attempting to solve this conundrum, knowledge is to be found instead in linguistic statements--in propositions, in assertions--how do we compare these with things as they really are, to assess the validity of our knowledge? Karl Popper (1963) claimed that Tarski's meta-semantics had solved this problem and rehabilitated this correspondence theory of truth; Kuhn (1970b, p. 265) pointed out the fallacy of this claim. And smart people like Samuel Messick (1989) have continued to struggle with these problems, without success.
A different ontology is needed, one that views humans not as distinct from, as set apart from and contemplating an inhuman, disenchanted, material world, but that views sees us as part of the world, as living together in a human world.
This is the first of three overheads I'll be showing you--and I won't have time to explain any of them to you satisfactorily! Each breaks all the rules about the amount of information to display in a visual graphic! But this is what I call my ontological blueprint: an attempt to show how we live and act and interact together in a social context that provides the background to our activity, and that exerts what Charles Taylor (just up the street) calls a constitutive causality (Taylor, 1971)--the being of both artifacts and humans--what we are--is not just regulated but constituted by , in combination, the social context of a community of practice, and the participation of a member of that community.
Such a non-dualistic ontology sets the stage for a different epistemology. We can know the world because we are of it, because we are shaped by it, and because we have shaped it. Physical laws are not social conventions--as Alan Sokal (1996) likes to say, in parody of postmodern arguments--but they do reflect human concerns. (Galileo's [1638/1954] Dialogues Concerning Two New Sciences begins with the topic of "the resistance which solid bodies offer to fracture"--including a very practical discussion of which animals can "fall with impunity from heights" and which cannot. Dogs, cats, grasshoppers, ants and children, can, but not elders or horses. That's to say, the "facts" that Galileo seeks to "demonstrate... by geometrical methods" are very practical facts; facts relevant to human practices.) And if that's true of the laws of physics, certainly sociological, psychological, educational knowledge--theory--is also shaped by human concerns. Our knowledge is never disinterested, neutral, detached and, in these senses, objective--human interests structure our theories, both physical and social, whether they are, to use Habermas' (1971) distinctions--instrumental interests, or communicative, or emancipatory interests.
It follows that knowledge, structured by human concern, is linked to human action too. Theory and practice are 2 sides of a coin; 2 arcs of a hermeneutic circle. The second overhead is an effort to illustrate how knowledge is always interpretation, arising from, an articulation of, the practical understanding gained through participation in a social practice.
And this means--well, many things. One thing it means is that rhetorical considerations can play a valid role in our choice of the form of representation in our theory. To be deliberately provocative: formalism has its rhetorical powers--but so, for example, does narrative. I worked for two years as a computer programmer, and loved it. Taught myself LISP... I've nothing against computers. But computer formalisms cannot handle ambiguity, and humans thrive on it. We enjoy it, we create it, we exploit it. I'm just completing a book on work in the Michigan public schools, as they struggled to cope with NSF's state systemic initiative, in combination with Republican Governor John Engler's market-place reforms--and I've employed some of the techniques of creative non-fiction writing (e.g., Gerard, 1996). Narrative has the power to convey a sense of place, of agency, of human striving, of contradiction, and of the interplay between the scope of human action and larger social forces- a power that formalisms are hard pressed to match. Formalisms may offer, as Andy has suggested (diSessa, 1991, p. 237) "analytic precision," but they fail to grasp the crucial plurivocity of human phenomena.
The third overhead shows a logic of inquiry, a methodology, a way techniques like interviewing, transcribing, ethnography, conversation analysis, situation analysis, can be woven together into a coherent research approach that respects the non-dualistic ontology and the epistemology of interpretation. Again I don't expect you to grasp the details, but I hope you get an overall sense of an investigation that moves in a circular fashion between the academy and the field, and between a focus on the researcher (preparing, reflecting) and a focus on the participants (grasping and articulating the terms of their participation in a community of practice).
Andy has asked us to declare our alignments with the other panelists, so let me attempt this briefly.
I imagine that first wave cognitive scientists might balk at much of what I've just said, but I think both Alan Collins and Paul Cobb are second wave cognitivists, to use the distinction Paul draws upon in his recent Educational Researcher article with Janet Bowers (Cobb & Bowers, 1999).
Paul Cobb has, I think, moved from a cognitive constructivism a la Piaget, which uneasily posited knowing individual child (epistemic subject) and known material world, towards exploring, for example, how a classroom task is both defined by a teacher and yet understood differently by students in different positions in the classroom community.
I've followed his work with interests. I would ask him to pay attention now to the kind of relationship children enter into in school: the impersonal relation of student and teacher (cf. Packer, 1999). And to the larger networks in which every classroom is embedded, networks of administration, financing and evaluation, that define what goes on within the classroom behind the backs, as it were, of the students and teachers.
Allan Collins, too, is now working in messy situations, aiming to characterize people's social interaction, with flexible revision of research design, even co-participant design and analysis.
As cognitive scientists move their investigations away from putative internal information processing, and towards the detailed description of practice and interaction, the dualistic ontology becomes harder to sustain. Action and interaction are human, yet in the world. They challenge the mental-material split. The "rich representations" Allan is now seeking will, I predict, draw on everyday natural language rather than artificial formalisms.
Turning now to Beverly Gordon, I must confess I'm a little uncomfortable, as a white male, not just Anglo but actually British, for heaven's sake, asserting solidarity with black feminist inquiry. Positional epistemology would surely assert that my location is a particularly murky cultural blind-spot.
But an interpretive, a hermeneutic approach to research demands that I take very seriously the way my position, and my choice of stance and attitude, influence what and how I know. Who I am, and how I am, make a big difference. And I've found that when, for example, I enter a school in the industrial Midwest, a school that is working-class, ethically mixed, rather to my surprise I can be the other, the stranger, expected not to understand, standing on the margins, someone to whom things must be explained. So, even as a white male I can, I believe, share a setting, a context, a site of struggle, with people whose background and position are very different from my own.
In these respects, I find a convergence of interests with all the other speakers.
Cobb, P., & Bowers, J. (1999). Cognitive and situated learning perspectives in theory and practice. Educational Researcher, 28(2), 4-15.
DiSessa, A. A. (1991, October 16-19). If we want to get ahead, we should get some theories. Paper presented at the North American Chapter of the International Group for the Psychology of Mathematics Education, Blackburg, Virginia, U.S.A.
Galileo, G. (1638/1954). Dialogues concerning two new sciences. (H. Crew & A. de Salvio, Trans.). New York: Dover.
Gerard, P. (1996). Creative nonfiction: Researching and crafting stories of real life. Cincinnati, Ohio: Story Press.
Habermas, J. (1971). Knowledge and human interests. (J. Shapiro, Trans.): Boston: Beacon Press.
Kuhn, T. S. (1970a). The structure of scientific revolutions. (2nd ed.). Chicago: University of Chicago Press.
Kuhn, T. S. (1970b). Reflections on my critics. In I. Lakatos & A. Musgrave (Eds.), Criticism and the growth of knowledge, : New York: Cambridge University Press.
Messick, S. (1989). Validity. In R. L. Linn (Ed.), Educational measurement, (3rd ed., pp. 13-34). New York: Macmillan.
Packer, M. (1999, April). The ontology of learning. Paper presented at the annual meetings of the American Educational Research Association, Montreal, Canada.
Popper, K. R. (1963). Conjectures and refutations: The growth of scientific knowledge. (4 ed.): New York: Harper & Row.
Sokal, A. (1996). A physicist experiments with cultural studies. Lingua Franca(May/June), 62-64.
Taylor, C. (1971/1985). Interpretation
and the sciences of man. [Originally published in The Review of Metaphysics,
25, 3-34, 45-51], Philosophy and the human sciences: Philosophical papers,
(Vol. 2, ). New York: Cambridge University Press.