Academic journal article
By Shulman, Lee S.
Journal of Teacher Education , Vol. 53, No. 3
A university professor was on her hands and knees in the darkest part of a tunnel when a sewer inspector stumbled over her.
"What are you doing here?" he asked. "I'm looking for the truth, or at least useful knowledge," was the reply. "I'm pretty sure someone must have left it here." "But it's so dark, muddy, and toxic here that I fear you will never find anything of value. Why don't you go look where there is a bit more light?" "Because the muddier it is, the better the chance that I will find something truly interesting. It's so unlikely that anything of value will be hidden in plain sight."
What kind of research on teaching is of most worth? To what extent should researchers in this field be conducting highly functional investigations that attempt to identify the key elements of accomplished teaching or the most important components of teacher preparation programs or experiences? Should we be asking whether teacher education programs significantly improve the likelihood that someone will teach effectively? Should we instead be conducting inquiries that explore the rich complexities of teaching, learning, schooling, and development and the contexts that support them? What genres of research are worth undertaking?
The tacit dialogue between the present articles by Wilson, Floden, and Ferrini-Mundy (2002 [this issue]) and by Florio-Ruane (2002 [this issue]) is nostalgically familiar. We designed the Institute for Research on Teaching (IRT) in 1975 on the basis of our critique of the then-prevailing prototype of
process-product research on teaching. We considered process-product research on teaching behavioristic, simplistic, and unduly dependent on standardized achievement tests as indicators of product. Indeed, the leaders of process-product research, such as Nate Gage (1978) and Barak Rosenshine (Rosenshine & Stevens, 1986), complained that their critics were unnecessarily "complexifying" the phenomenon of teaching, whereas the hallmark of scientific progress was increased simplification, not complication. Moreover, if research on teaching were to have the desired impact on policy makers, it needed to be both simple and clearly connected to easily understood indicators of student achievement. Finally, there was a moral message in the process-product tradition. Our bottom-line obligation as teachers was to the students and their learning. To study teaching without reference to students was unethical self-indulgence.
These two articles stimulated me to reflect on my history of work as an active scholar on teaching and teacher education. I thought about the nearly four decades of research in which I had been actively involved. And I began to wonder how, if at all, it added up.
I concluded that we may be asking the wrong questions and focusing on the wrong units of analysis. That is, individual studies rarely can be adjudged as valuable or trivial per se. Instead, we need to think about extended programs of scholarship, in which a variety of types of research are pursued, to maximize the value to be gained from studies of teaching. I want to tell a story of more than 30 years of research, of a series of research programs that cumulated into a meaningful knowledge base, an enduring policy initiative, and the spinning off of a number of significant lines of research.
I begin with my work on medical problem solving in the 1960s and 1970s, followed by the research on teaching as information processing that characterized the IRT programs. A set of studies on the development of teacher knowledge, with special reference to pedagogical content knowledge, followed that work, which transitioned into the Teacher Assessment Project conducted on behalf of the then-infant National Board for Professional Teaching Standards. The board's own validation studies of board certification were conducted in the year 2000, and a new program of Carnegie Foundation studies of teacher education is currently underway. …