Kuhn, Historical Philosophy of Science, and Case-Based Reasoning

I. Introduction

History, if viewed as a repository for more than anecdote or chronology, could produce a decisive transformation in the image of science by which we are now possessed.

Thomas Kuhn

So reads the well-known opening sentence of Thomas Kuhn’s Structure of Scientific Revolutions (1962). ¹ Thirty-five years later, his prediction has certainly come true—in ways he helped initiate but could not have predicted, and even came to reject. Of course, we should add sociology, anthropology, psychology, economics, rhetoric, cultural studies, and feminist analyses of science and technology to the mix that has come to constitute science studies. The relevance of most of these disciplines was foreshadowed in Kuhn’s work.

As the most common reference point in science studies, Kuhn has become all things to all people. Whether you seek an account of science as a radical or a conservative enterprise, one that is progressive or cyclic, finely argued or irrationalist, modernist, postmodernist, or even premodernist—Kuhn is your man. Kuhn himself is partly responsible for being so hard to pin down. Under fire from philosophical critics, he retreated to some degree from most of his more exciting and radical theses. ² And after moving from Princeton to the Linguistics and Philosophy Department at MIT, he retreated further, into a kind of Quinean linguistic analysis of science—so that the man whom Stephen Toulmin and Alasdair MacIntyre had credited with historicizing that last bastion of the historically transcendent, the physical sciences, could say in 1991 that history of science was not so important to his project or to philosophy of science! ³

Nearly as puzzling as the retreat from history is Kuhn’s turn away from the cognitive psychological and sociological initiatives of Structure. After all, he himself had suggested that sociology was the preferred route to developing his concept of paradigm (in the large sense), and cognitive psychology the way to analyzing his concept of exemplar (“paradigm” in the small sense). But in the years that followed, his historical work remained strongly internalist, he firmly rejected the new sociology of science, and he showed little interest in the ongoing revolution in the cognitive sciences. ⁴

Whatever the merits of his later work, the early Kuhn is still worth mining for insights, particularly his account of learning by example. On this occasion my focus will be on the period from Structure to The Essential Tension (1977) and Black-Body Theory and the Quantum Discontinuity (1978), ⁵ and on cognitive science rather than sociology. Developments over the past decade in schema theory and case-based reasoning resonate with Kuhn’s claims about exemplars, direct modeling, and intuitive performance. In turn, they challenge traditional conceptions of methodology and boost the stock of methodologies of science based on historical and contemporary case studies. Ironically, however, as we shall see, instead of historicizing everything in the professional historian’s sense that Kuhn intended, “historical” methodologies render all appeals to historical (and contemporary) cases Whiggish or presentist, in the sense that they appeal to history in the service of present problems.

Section II of this paper locates Kuhn within both modern and premodern intellectual traditions, a “two-cultures” ambivalence that section III develops by asking about Kuhn’s target audience. We are then prepared to wonder, in section IV, whether Kuhn himself can be consistently anti-Whiggish, and why he would want to be. Sections V and VI show how much of Kuhn’s treatment of historical exemplars can now be captured by the aforementioned developments in case-based reasoning and schema theory. These developments reinforce my previous conclusion that, for Kuhn, good science requires bad history (and vice versa) and, so, further challenge the standard, anti-Whiggish interpretation of Kuhn’s opening statement. I draw some methodological consequences in the final section.

II. The Method of Doubt Versus the Method of Belief

Since the time of Descartes and Locke, epistemologists and methodologists of science, whether inductivists or deductivists, have followed what we may term “the method of doubt.” Descartes’s first rule of method, in the Discourse on Method (1637...