Steve Fuller (1994) Cognitive Science of Science: the Wave. Psycoloquy: 5(70) Scientific Cognition (9)

Volume: 5 (next, prev) Issue: 70 (next, prev) Article: 9 (next prev first) Alternate versions: ASCII Summary
PSYCOLOQUY (ISSN 1055-0143) is sponsored by the American Psychological Association (APA).
Psycoloquy 5(70): Cognitive Science of Science: the Wave

Reply to Bookstein, Catania, Hardcastle, Roitblat,
Shafto, Slater & van Brakel on Scientific-Cognition

Steve Fuller
Department of Sociology and Social Policy
University of Durham
Durham DH1 3JT
United Kingdom


As someone who was identified (correctly) as skeptical of the project outlined in Giere (1992), I argue that a Cognitive Science of Science was desirable a half-century ago, when the logical positivists first proposed something like it, but is now an anachronism -- a point implicitly realized by some of the contributors and reviewers. I stress the need to examine the socio- historical circumstances that make grand synthetic projects like Giere's appear attractive.


Cognitive science, philosophy of science, cognitive models, artificial intelligence, computer science, cognitve neuroscience.
1. I will frame my response to the reviewers of Giere (1992) in the terms of Slater (1994), who bravely goes beyond questioning the feasibility of a "cognitive science of science" (CSS). She raises the embarrassing but necessary question of whether it is even desirable: Why would anyone want to get involved in such a project? Regardless of the obvious difficulties with making their research programs cohere together, Giere and most of my fellow contributors take the desirability of a unified account of science as pretty much self-evident, presumably out of some Newtonian impulse that says it's better to have one big theory than a lot of little ones to cover roughly the same phenomena.

2. Some of the other reviewers, however, have suggested other motivations -- at least for what they would LIKE to see a CSS do. Prominent among these are (1) helping scientists better understand, and perhaps improve, their day-to-day work, and (2) helping students learn scientific concepts more effectively. While (2) is an explicit interest of several of the volume's contributors, (1) surprisingly receives very little attention. Yet neither was a major interest of the movement from which Giere draws inspiration for his project, namely, the logical positivists.

3. This last point is of more than mere antiquarian interest because one way of getting at Slater's concern is to ask about the socio-historical circumstances which motivate scholars to propose grand synthetic projects. Projects that may have made sense fifty years ago may make no sense now, or at least they would have to be radically reconceptualized in order to make sense in today's world. Clearly, Giere and most of the contributors are banking on this not being the case, given their interest in retaining most of the traditional problem-set of the philosophy of science. But it remains an open question nonetheless.

4. As Hardcastle (1994) and van Brakel (1994) observed, much of the "cognitivist" work in the volume does little more than relocate the standard fare of positivist and post-positivist philosophy of science to more "naturalistic" settings, such as computers, mental models, and neural circuitry. In fact, it would be fair to say that most of these chapters offer empirical confirmation of notions that were cooked up at least thirty years ago.

5. Putting aside the "social epistemologists" Houts and Haddock, Gorman, and myself -- whom the reviewers realized were rather anti-CSS in spirit -- Churchland and Giere himself are the only two CSS adherers who openly challenged any of the key commitments of the received view in the philosophy of science. (Both oppose, for rather different reasons, the idea that theories are primarily linguistic entities.) In that case, instead of proclaiming a new paradigm in the philosophy of science, Giere perhaps should have said that finally the old paradigm has reached the phase of maturity characterized by "normal science" and "puzzle-solving."

6. But is this newfound maturity well timed, or is it just a case of philosophy of science having withered on the vine? Recently, historians have begun to clarify the context from which the logical positivists emerged (e.g., Galison, 1990). Telescoping much of this research, we can say that the positivist idea of "unified science" was meant to serve two political goals whose relevance to today's world is uncertain at best.

7. One goal was to blunt the rhetorical force of German academics who promoted a "two cultures" view of knowledge, in which Naturwissenschaft appeared responsible for Germany's loss in World War I (courtesy of the chemists and physicists who collaborated with the Kaiser) and Geisteswissenschaft was portrayed as the counter-rationalist response ready to bring Germany into a new age of greatness. This was not simply a matter of humanists seizing the political initiative from the natural scientists. In addition, they were assisted by engineers and other applied scientists who were more than happy to present themselves as purveyors of "craft knowledge," so as to distance themselves as much as possible from the chemists and physicists (Herf, 1984).

8. A second goal was to recover what the positivists believed, on historical grounds, to be the emancipatory roots of the Scientific Revolution. Instead of deferring to scholastic authorities, one could defend knowledge claims simply by making one's reasoning explicit and producing publicly observable phenomena. The positivists believed that all genuine scientific claims could be reduced to this patently democratic mode of presentation. Indeed, Otto Neurath went so far as to attempt to construct a universal iconic language that would eliminate the barriers of scholarly jargon that stand between lay people and scientific findings of relevance to their lives (Fuller, 1994b). We scarcely do service to this goal when we recall the positivists' interest in a "theory-neutral observation language."

9. Not surprisingly, outside of philosophy, logical positivism's legacy has been strongest as an Enlightenment rhetoric that can be brandished in fields where tendencies toward "counter-rationalism" are most evident (i.e., the social sciences) as well as in popular accounts of science designed to show that scientists are marching lockstep with democratic and liberal forces in society.

10. However, CSS is operating in a completely different cultural setting, one in which disciplines are no longer at odds in quite the same way (though a kind of "two cultures" problem persists: cf. Gross & Levitt, 1994). Our historical understanding of the Scientific Revolution has changed (cf. Shapin & Schaffer, 1985), and our sense of the relationship between science and democracy has become more complex (cf. Ravetz, 1971). Although I obviously cannot argue the point here (cf. Shadish & Fuller, 1994), I would maintain that a grand synthetic project that addressed these contemporary problems would look nothing like logical positivism, and hence nothing like CSS.

11. The reviewer who probably realized this point most clearly was Shafto (1994). He chides most of the volume's contributors for failing to attend to what distinguishes science from other cognitive processes. Being heirs to logical positivism, they found the conflation natural, as science would seem to be nothing more than a generalized problem-solving ability or cognition rendered self-conscious. While this orientation may have made sense earlier this century, science nowadays is principally identified with its distinctive institutional and technological forms, about which few of the contributors had anything of interest to say. This point helps explain the expectations that the reviewers brought to the book (see paragraph 2 above).

12. Since the launching of Sputnik in 1957, Americans have been inclined to take the level of science literacy in school children as a benchmark in the global balance of power. In the post-Cold War era, this metric has remained intact, while its meaning has shifted from concerns of "national security" to those of "economic competitiveness." In the positivist picture, children did not play such a strategic role. As is to be expected, then, several of the reviewers, especially Roitblat (1994), singled out Chi's and Carey's chapters for praise and thoughtful engagement, as their experimental work on conceptual change in children seemed redolent in larger implications. Nersessian, who likewise sees her historical work as relevant to science education, was generally well received. And those who remember Thagard (1988) will recall that his computer programs originated as automated logic tutorials.

13. Moreover, while it would have seemed preposterous fifty years ago for philosophers and psychologists to offer advice to NATURAL scientists on how to improve the conduct of their inquiries, the advent of "Big Science" has gradually eroded this resistance. The greater involvement of public monies in science funding and the increasing impact of scientific research on the larger society have understandably raised the level of scrutiny and, hence, the level of incompetence and fraud brought to light. Not surprisingly, then, there is a tone of desperation in Catania (1994) and Bookstein (1993) when they observe that this book won't offer scientists much assistance.

14. In addition, uncertain employment opportunities for new scientists as well as the routinized, instrument-driven character of much ongoing scientific research have created the impression that the heroic vision of science championed by the positivists is now an endangered species. Thomas Kuhn was only an early case of what has since become a steady stream of disaffected trained scientists who seek intellectual shelter in the history and philosophy of science, where they can memorialize the classical image of scientific research (Fuller, 1994a).

15. Based on the contributions to this volume, as well as other work in CSS, the locus of historical interest is certainly the mid-19th to early 20th century, the period normally regarded as the last when first-rate scientists could be expected to excel in a variety of activities -- such as theorizing, experimenting, and evaluating -- that are nowadays divided among a team of specialists in a laboratory (Buchwald, 1993).

16. Ironically, none of the reviewers noticed that Giere himself -- the man who brought us all together in October 1989 to discuss the possibility of a CSS -- ends the volume by disavowing any interest in having his philosophy of science "mingle" with cognitive science (Giere, 1992, p. 483). He professes the historically saner goal of integrating selected findings from cognitive science into Science & Technology Studies (or STS), a field that takes very seriously the cognitive implications of changes in the size and shape of science over the course of this century (Fuller, 1993).

17. So, maybe Giere's invocation of the positivists at the start of the volume was not to be taken so seriously, after all. Nevertheless, from an STS standpoint, we need to follow the fates of positivist artifacts in a decidedly non-positivist world. Thagard's computer programs and their various imitators are good starting points. As Glymour, Giere, Gorman -- and many others not included in the volume -- have repeatedly observed, the formal structure of ECHO is hardly more sophisticated than the paper-and-pencil syntactic systems of the positivists, whose Achilles' Heel is generally said to have been the inability of their formalisms to reproduce scientifically credible reasoning.

18. Thagard, however, enjoys the sociological advantage of having his formalisms embedded in computers, which are increasingly regarded as virtual agents in scientific inquiry. In other words, people are more likely to read the formalized outputs of a computer more charitably -- much as one would a fellow human being -- than they would a string of symbols in an open book, even if the cognitive complexity of the book and the computer was, strictly speaking, about the same. It also helps that a human being is usually attending to a computer while it is working. Thagard's continued professional success suggests that sociology may indeed be able to work miracles where logic can only dream!


Bookstein, F. (1993) Geometry as Cognition in the Natural Sciences. PSYCOLOQUY 4(65) scientific-cognition.2.bookstein.

Buchwald, J. (1993) Design for Experimenting, in P. Horwich (ed.), World Changes. (Cambridge: MIT Press) (pp. 169-206).

Catania, A.C. (1994) Information, Cognition and Behavior: Telling What They Want to Hear. PSYCOLOQUY 5(39) scientific-cognition.5.catania.

Fuller, S. (1993) Philosophy, Rhetoric, and the End of Knowledge: The Coming of Science & Technology Studies. Madison: University of Wisconsin Press.

Fuller, S. (1994a) Teaching Thomas Kuhn to Teach the Cold War Vision of Science. Contention, 4, 81-106.

Fuller, S. (1994b) Towards a Philosophy of Science Accounting: A Critical Rendering of Instrumental Rationality. Science in Context, in press.

Galison, P. (1990) Aufbau/Bauhaus: Logical Positivism and Architectural Modernism. Critical Inquiry, 16, 709-752.

Giere, R. (1992) Cognitive Models of Science. Minneapolis: University of Minnesota Press.

Giere, R. (1993) Precis of: Cognitive Models of Science. PSYCOLOQUY 4(56) scientific-cognition.1.giere.

Gross, P. & Levitt, N. (1994) Higher Superstition: The Academic Left and its Quarrels with Science. Baltimore: John Hopkins University Press.

Hardcastle, G. (1994) Why Don't We Yet Have a Cognitive Science of Science? PSYCOLOQUY 5(43) scientific-cognition.6.hardcastle.

Herf, J. (1984) Reactionary Modernism. Cambridge: Cambridge University Press.

Ravetz, J. (1971) Scientific Knowledge and its Social Problems. Oxford: Oxford University Press.

Roitblat, H. (1994) A Representational View of Science. PSYCOLOQUY 5(44) scientific-cognition.7.roitblat

Shadish, W. & Fuller, S. (1994) Social Psychology of Science. New York: Guilford Press.

Shafto, M. (1994) What Can Insiders Learn From Outsiders? PSYCOLOQUY 5(30) scientific-cognition.4.shafto.

Shapin, S. & Schaffer, S. (1985) Leviathan and the Air-Pump. Princeton: Princeton University Press.

Slater, C. (1994) Why We Should Not Rely on Cognitive Science to Unify Philosophy of Science. PSYCOLOQUY 5(45) scientific-cognition.8.slater.

Thagard, P. (1988) Computational Philosophy of Science. Cambridge: MIT Press.

van Brakel, J. (1994) Cognitive Scientism of Science. PSYCOLOQUY 5(07) scientific-cognition.3.vanbrakel.

Volume: 5 (next, prev) Issue: 70 (next, prev) Article: 9 (next prev first) Alternate versions: ASCII Summary