Thomas Kuhn: A Philosophical History of Our Times
from Chapter Two:
The Last TimeScientists Struggled for the Soul of Science
A common assumption of the realism-instrumentalism debate as it is conducted today is that science is, in some significant sense, autonomous from the rest of society. At the very least, the success of science can be explained without referring to the societies that have supported it. However, once RID is fought in the arena of educational policy, such autonomy becomes difficult to maintain: How can science justify its autonomy, while at the same time meriting inclusion, if not privilege, in the key processes of social reproduction? From the Mach-Planck debate, two kinds of answers can be discerned that, over the course of this century, have proved as influential as they are unsatisfactory. The realist appeals to science as an exceptionally rational world picture whose adoption promises to add disciplined thought to any line of work. The instrumentalist portrays science as an economizing tool that can assist everyone in pursuing her ends without imposing any value orientation of its own. In practice, these alternative visions of “autonomy” have amounted to the dilemma: Use or be used.
When positions are developed dialectically, it is common for each side to scrupulously avoid the interlocutor’s shortcomings, yet to remain unwittingly blind to one’s own deficiencies. For Mach, the biggest threat posed by Planck’s realist educational policy was clearly indoctrination. This fear led Mach to conceptualize the cultural significance of the natural sciences in terms that closely conformed to the liberal doctrine of “academic freedom” as applied to both researcher and student. Unfortunately, this doctrine was originally designed with the humanities as the center of the educational system. By the first decade of the twentieth century, the range of applications of natural-scientific knowledge had become wider and potentially more dangerous than those of humanistic knowledge. Thus, to deny science its own value orientation was to license indirectly the appropriation of scientific knowledge for any purpose, including destructive ones, as World War I would ultimately demonstrate. For every Mach who resolutely refused to involve his scientific expertise in the war effort, there were plenty of Machians, especially among the chemists (including the “pacifist” Wilhelm Ostwald), who “freely” enrolled in the kaiser’s cause. 
Unlike Mach, Planck did not believe that the problem of scientific autonomy would be solved by a sharp separation of science and values, for that would only make science captive to those who have the power to impose their values on it. Science had to be socially recognized as its own
67. Johnson 1990, 180-83.
value orientation, alongside yet noncompetitive with the state, religion, and industry. An elite functionary for most of his career, Planck was alive to corporatist tendencies in the modern nation-state that eluded Mach’s democratic liberalism. Thus, Planck organized scientists in ways that enabled them to take collective control of the direction and application of their work, a strategy that included insinuating a distinctive natural-scientific perspective throughout the educational system. Nevertheless, as the religious rhetoric of the Mach-Planck debate brings out, the realist strategy placed science - especially an advanced science like physics - in an awkward political position of its own. For if the ends of science are not merely distinct, but increasingly divergent, from other societal ends, then students will need to be given early exposure to the scientific world-picture, in order to be attracted, or at least rendered sympathetic, to scientific careers. Thus, increased control of the curriculum would seem to be necessary for continued control of the research agenda.
As suggested earlier, a crucial factor in Mach’s long-term loss to Planck was his strict, perhaps even anachronistic, adherence to the old Socratic ideal of dialectical inquiry, admittedly the very one promoted by Wilhelm von Humboldt under the guise of “Enlightenment” when he reinvented the German university in Berlin at the dawn of the nineteenth century. Most closely associated with the classical humanistic mission of the university as citizen education, it was a resolutely praxis-oriented view in which the ends of science were regularly realized together by professor and student in the classroom. All students were, as the social psychologist Jean Lave now puts it, “legitimate peripheral participants” in the knowledge production process, active inquirers in their own right, not passive recipients of knowledge.  In Mach’s idiosyncratic version of this vision, the trac-
68. On learning as legitimate peripheral participation, see Lave and Wenger 1991. Lest Mach appear completely atavistic, I should say that the fall from Humboldtian grace had become obvious only in the generation prior to the Mach-Planck exchanges. Even in the late nineteenth century, students in the natural sciences were treated as researchers in that they were encouraged to try out experimental possibilities to see what worked. In that sense, there was no discrete periodization of “training” and “research.” See Olesko 1993, esp. 22. A brief history of the decline of the Humboldtian ideal in the German university system may be found in Schnaedelbach 1984, 12-32.
For German academics who lived through the final stages of the decline, such as the sociologist Max Weber, Friedrich Nietzsche (1844-1900) symbolized the “man of science” who lost his faith after the Humboldtian ideal had died. Nietzsche was a precocious philologist who fell into a permanent depression after his first book was severely criticized on “scholarly” grounds. (The rest is history.) His The Birth of Tragedy in the Spirit of Music (1872) invited readers to use the contemporary music of Richard Wagner to re-create in their own times the context that had enabled Greek tragedy to flourish over two thousand years earlier. Encapsulated in Nietzsche’s fate is a profound difference in sensibility between what “teaching” and “research” cultures in the academy have wished to conserve as the collective memory of their societies. Nietzsche represented the teachers, who stressed the recovery of the [contexts of the past that could continue to enliven the present, whereas the emerging tendency was to stress the recovery of past contents, even if that meant stressing the “otherness” of the Greek conceptual universe to our own. Implicit in the latter, now dominant, viewpoint is that the past is very much a “foreign country” best left as a preserve to specialists. The Priggishness of contemporary relativism discussed in sections 4-5 of the introduction should be seen as coming from this perspective, rather than the one Nietzsche represented, which would erase any clear boundary between the past and the present. On the trade-off between context and content as goals of translation more generally, see Fuller 1988, chap. 6; Fuller 1998b.]
HHC: [bracketed] displayed on page 131 of original.
tability of science to common modes of experience should constrain the development of science nearly as much as science should revise and discipline common modes of experience. Translation, not reduction, was his principle of scientific unification. Consequently, the research and education functions of the university were never sharply distinguished in his mind. Moreover, Mach imagined that the absorption of the natural sciences into the universities (and the Gymnasien) would serve to divest these institutions of their residual elitism.
However, ends need to be tied to institutions, in which case the desirability of the ends turns out to rest on the feasibility of the institutions. And here Mach’s vision came up short. Notwithstanding its Humboldtian provenance, the image of students in active inquiry with their professors was quickly molded by the institution of the seminar in the historical disciplines, which subsumed teaching under research, thereby converting the classroom into a market for recruiting the next generation of the discipline’s practitioners. This mentality was transferred to the laboratory, once the experimentally based sciences were brought into the universities, with psychology providing an especially instructive case in point, since most of the early labs were housed in philosophy departments, the core of the humanities.  Gradually, an alternative vision emerged, namely of a community of scientific researchers whose real work occurs in settings quite separate from the lecture halls where students are encountered en masse. Whatever remained of the old praxis-based ideal in what became Planck’s vision, it lay, as we shall see below, in the role that science plays in reproducing the social order, not in the conduct of science itself, which increasingly lost any sense of a realizable end. 
69. Danziger 1990, 17-67.
70. Given the current fashion of speaking of the work done in a modern scientific laboratory as consisting of “practices,” it is worth stressing that a citizen of the Athenian polis would find little in this usage that cohered with his understanding of praxis. For a thoroughgoing critique of the appeal to practices in social science, see Turner 1994, which was subject to a symposium review in Human Studies, July 1997, including responses by two of the leading practice theorists in contemporary sociology of science, Andrew Pickering and Michael Lynch. In what follows, I draw partly on the French Marxist anthropologist Godelier (1986, 130-37).
For the Athenian, a praxis was an activity that was done “for its own sake” in a special sense of that expression. The activity had to have a natural trajectory and ending that would [be recognized by the person engaged in the praxis. In addition, a praxis could define the trajectory and ending of other activities that did not have such self-defining qualities built into them. For example, before the introduction of slavery, agriculture was the paradigm case of a praxis. The cultivation and harvest of food were not interminable affairs, but ones pursued only as long as was needed to sustain a household. The emphasis was placed not on transforming nature but on participating in an activity pursued by many at once. Such praxis was the primary means by which a sense of civility was instilled, as each household tilled the soil in ways that enabled others to sustain themselves. However, after slaves started doing the agriculture, praxis was limited to free speech, that is, speaking one’s mind (and only one’s mind).
Two related notions need to be introduced here: techne and poiesis. The former designates any of a variety of client-oriented trades that require specialized (usually esoteric) training, while the latter refers to the characteristic products of such a trade. The client engages in praxis by directing the tradesman to make something to the client’s specifications. The tradesman’s job is to realize the client’s idea in the medium of his trade. The client might oversee the tradesman’s work, just to make sure all is going to plan. If the final product is judged good, then the client, not the tradesman, would receive credit for its successful execution. (The tradesman would receive payment for his labor, of course.) While such a merit system may seem odd by modern standards, consider the case of today’s teacher who typically receives credit for the excellent performance of her students on exams. In both the ancient and modern case, merit lies more in the design of realizable standards of performance than in the sheer ability to realize those standards in performance. If techne enables a tradesman to infuse matter with form, praxis enables the client to infuse the tradesman with direction.
Various complications can be added to this story. For example, what if some enterprising people try to turn speech itself into a techne? Would undergoing the relevant training enhance or diminish the status of speech as praxis? Such were the worries that animated Plato’s and Aristotle’s philosophical response to the Sophists. Nevertheless, our Athenian forebears would be in agreement that crucial features of today’s “scientific practice” disqualify it from counting as praxis. First, consider the claims made for the open-ended, indeed interminable, nature of scientific inquiry, as well as the stress placed on the essentially unintended (“serendipitous”) character of major scientific achievements. But perhaps even more telling is the image of a mounting body of scientific literature that seemingly diminishes in its ability to capture its objects of inquiry. More and more is written about less and less. In short, the sorts of reasons that have often been given for why the Greeks would never have countenanced the unbounded market mentality of capitalism can be used to show why they would refuse to dignify contemporary natural science with the title of praxis. Someone endlessly driven to make money maybe quite skilled, in the sense of possessing a techne, but the lack of purposeful closure to his activities would mark it as pathological.
At this point, Alasdair MacIntyre’s objection to this characterization of praxis is worth facing. His list of praxes includes scientific inquiry, as well as musical performance. See MacIntyre 1984, 187 ff. However, the difference between the two kinds of cases are important. A musical performance is properly regarded as a praxis because each moment of a competent performance recognizably contributes to an overall goal, the execution of a piece of music, during which both the performer and the client-audience experience pleasure. With the notable exception of improvisational performances, one’s musicianship typically depends on the ability to execute a piece of music that reaches a scripted end. There is no such clear connection between technical proficiency and success in scientific practice. Given the long hours of unproductive labor often involved in laboratory science, scientists are taught not to tie their efforts too closely to the likelihood of success. Rather, they are taught to see themselves as] [potential players in a story that will eventuate in a complete world picture: see Kuh 1970b, 38. For example, Max Planck was quite explicit about the vicarious character of the pleasure that most scientists would receive from their activities. And if Planck is right, that a mark of mature science is that increasing effort is needed to achieve comparable results, it would seem that we have on our hands less a praxis and more an addiction - with hallucinatory elements thrown in to capture the virtual status of any given scientist in his paradigm’s narrative of progress!]
HHC: [bracketed] displayed on page 132 of original;
[bracketed] displayed on page 133 of original.
It is beyond the scope of this chapter to explain exactly how Planck’s vision came to triumph, but the signs of its success are relatively easy to mark. In the first place, the unity of research and teaching of the classical humanist university underwent a subtle metamorphosis, one that can already be detected in the reactions to Max Weber’s famous “Science as a
Vocation” address of 1918.  Few academics any longer talked about students’ intellectually maturing as they participated in the professor’s scholarly interests. Rather, the professor herself is the one who now undergoes self-transformation through the process of scholarly discipline. Weber’s particular way of characterizing this process - which involved subsuming one’s ego to an endless and largely vicariously realized quest - was roundly criticized for its obvious reliance on the image of natural-scientific inquiry fostered by Planck and the physics community. Thus, humanists familiar with the classical pedigree of praxis, such as Ernst Robert Curtius, found Weber’s vision a monstrous perversion of the ends of inquiry.  However, in its place, Curtius could only invoke the scholar’s solitary participation in the thought of past minds, and the intuitive sense of private closure that it brings. Yet this too was certainly a far cry from the dialectical participation of professor and students in the Humboldtian classroom. It had become clear to all that academics were no longer in the business of bringing their students to intellectual maturity, let alone to personal empowerment. 
In short, Planck did not forsake Plato but shifted the Platonic imagery from Socrates to the philosopher-king. Planck sharply divided between the research and educational functions of the university. In practice, the state and the scientific community struck a deal. The state heavily invested in cutting-edge research as “bets” on the curriculum that will be needed to credential the next generation of citizens in jobs that will ensure the smooth operation of the social system.  In return, the scientific community offered its services in designing new principles of social stratification, ones based not on “status” or “class,” but on “general intelligence” and
71. See Weber 1958.
72. Curtius (1989) catches Weber’s implicit reliance on a physics-driven model of inquiry, but he neglects Weber’s explicit tendency to accept the convertibility of thought styles common to modern academic and corporate culture: entrepreneurship requires inspiration, and an efficiently organized scientific enterprise conforms to the same principles as govern a successful company. Interestingly, Weber sees this convertibility as not only inevitable but also of American provenance, when in fact the Americans refashioned it from the Germans. In chapter 4, section 3, this history comes to the fore, as Conant is largely responsible for importing the German model of academic research to the natural sciences after World War I. The model, in turn, was forged in the midst of heavy state and industrial involvement. See note 20 above. My thanks to Nigel Pleasants for helping me crystallize these connections.
73. Ringer 1979, 19.
74. Stinchcombe 1990, 312-13.
“problem-solving ability.” These qualities supposedly pertained to all kinds of jobs, but ultimately they were modeled on one’s facility with the kind of “work problems” found in classical mechanics textbooks. Indicative of this new insertion of science into the mechanisms of social reproduction and control was the attitude of Gestalt psychologist Wolfgang Koehler toward the widespread use of intelligence tests in schools.
Not only had Koehler originally studied physics with Planck in
The influence of this physics-centered view of thinking and intelligence was felt throughout Gestalt psychology. In their textbooks, Gestaltists were at pains to deny the Machian-popular view that physics was an inappropriate model for studying normal psychological processes because of its preoccupation with unobservable entities in artificial settings.  The earliest Gestalt-oriented experiments on problem solving were based on accounts of the electrical discoveries made by Benjamin Franklin and Michael Faraday.  These experiments, conducted by Otto Selz, highlighted the ability of subjects to transform their environments - essentially turning them into thought laboratories - and thereby remove the obstacles in the way of their problem-solving tasks. Moreover, while the subjects typically could not recount their thought patterns as a train of images, they nevertheless felt an
75. Koehler 1971, 112—13. On the vexed sociohistorical situation of the Gestaltists, see Ash 1991. The view that Koehler derived from Planck— that physical reasoning is ordinary reason≠ing rendered self-conscious—was popularized in late nineteenth century Berlin by Hermann von Helmholtz. See Hatfield 1990b, chap. 5.
76. Koehler 1971, 237=51.
77. Koffka 1935, 57.
78. See Humphrey 1951, 142-43. I say “Gestalt-oriented” because Selz, a member of the Wuerzburg school of psychology, held to a sharp split between the character of thought and perception, the former being “imageless.” The classical Gestalt psychologists typically held to a monistic view of the mind. On the significance of this difference for epistemology, see Berkson and Wetterstein 1984, 8-10, 106 ff.
unconscious “determining tendency” that kept them motivated until a solution was reached. The report of such experiences suggested that even ordinary people normally partook of some of the committed and slightly mysterious character of physical inquiry as Planck portrayed it. So perhaps Mach had overstated the case for the exotic nature of physics. At least, this was the verdict of the instigators of the “second cognitive revolution” of the late 1950s. 
Nevertheless, younger adherents to Mach’s critical approach to science, such as Karl Popper (1902-94), remained concerned. Popper’s doctoral dissertation was partly devoted to showing that these early cognitive psychology experiments did not clearly distinguish between thought that was simply driven to fit the facts into preexisting patterns - as is often connoted in the idea of a “mental set” - and a genuine scientific breakthrough, which addressed the facts in ways that substantially reconfigured preexisting patterns. We see here a psychological basis for Popper’s famous falsifiability criterion.  Given this lineage, it is hardly surprising that Kuhn found his studies of paradigm acquisition in the physical sciences moving him toward more general considerations of how children learn concepts.  The relevant developmental psychology experiments, which built upon the early Gestalt work, presuppose that children are inchoate physicists before they are anything else.
On the specific topic of intelligence tests, Koehler revealed his Planckian sensibilities most clearly.  He sharply separated the administrative role of the tests from their proper scientific merit. In fact, Koehler held that the primary function of the tests was to reduce uncertainty in how teachers classified students when tracking them through the educational system. Such bureaucratic efficiency was worth whatever interpretive confusions the tests’ coarse-grained measures might breed in the public - as in encouraging the idea that “intelligence” is a univocal substance that people have in varying amounts. Psychological researchers would not be so fooled, since (so Koehler believed) intelligence tests are clearly artifacts of the cul-
79. A good historical source on this revolution (and its learned despisers) is the series of interviews presented in Baars 1986, esp. 365-66 (interview with Herbert Simon on the influence of Otto Selz and the early Gestaltists on his own work). An indication of the continued closeness between psychology and physics is the story recounted by the early cognitive revolutionary Jerome Bruner concerning the controversial atomic physicist J. Robert Oppenheimer (in his later capacity as director of the Princeton Institute for Advanced Studies), who asked of Bruner’s own neo-Gestalt experiments: “Perception as you psychologists study it can’t, after all, be different from observation in physics, can it?” From Bruner 1983, 95-96.
80. A good source on Popper’s resonances with the history of psychology is Berkson and Wetterstein 1984.
81. Kuhn 1977a, 308-19.
82. Koehler 1971, 187-88.
tures that construct them and not the royal road to cognitive capacity. The ease with which Koehler could harbor such vastly different attitudes toward the educational value and the research value of these tests is reminiscent of Plato’s tolerance for “noble lies” that serve to stabilize the social order. While Planck and Mach would disagree on the import of this resemblance, both would agree that only practices that appear to have the warrant of the natural sciences are in a position to manufacture such myths in the twentieth century.
Planck’s view was effectively “naturalized” with the publication of The General Theory of Knowledge in 1925, written by Moritz Schlick, who acceded to Mach’s chair in
Such an explanatory narrative had been previously used by the psychologist Wilhelm Wundt in order to show that teleological and deontological ethics were earlier and later phases in the moral development of humanity. Schlick had now purported to show an evolution in the conception of economy, with Mach representing the earlier phase of “minimal effort” and Planck the later phase of “minimal principles,” the pursuit of which may actually involve much (pleasurable) effort. The crucial ambiguity in Schlick’s account, which carries over into contemporary philosophy of science and science policy, is whether he is talking about the development of the individual inquirer or that of a community of inquirers. The ambiguity matters when interpreting the idea that other life needs come to be subordinated to scientific ones: Is Schlick talking about the self-sacrificing scientist or the society that increasingly adapts its other functions to the needs of scientific institutions? The elision of these two interpretations clearly worked to promote Planck’s vision of science in society.
The social historian Fritz Ringer has examined the long-term effects of basing access to secondary and tertiary education on entry and exit exami-
83. See Schlick 1974, 94-101.
nations, the so-called merit-based system that has prevailed in
As the employment opportunities of academically credentialed people have increased over the last century, they have also displayed a distinct pattern, which Daniel Bell originally took to mean “the coming of post-industrial society.’  The pattern is one of intermediation, that is, the increasing need for academically credentialed people to survey, digest, and translate the work of other academically credentialed people for a third group of academically credentialed people. In some recent popular works, this labor process has been described as “symbolic analysis” and “knowledge brokerage.”  For science policy analysts, intermediation is the institutional correlate of cognitive complexity; for critical political theorists, it marks the corporatist sublimation of democratic impulses.  For still others, especially the German systems theorist Niklas Luhmann, these are two ways of talking about the same positive development, namely, an increase in social integration brought about by the continual redistribution of uncertainty across the various “intermediators.” 
In contrast to these diagnoses, intermediation can equally be seen as symptomatic of the growing disparity between the content and function of
84. Ringer 1979, 27-29.
86. See esp. Reich 1990. On the strength of his analysis, Reich became President Clinton’s first secretary of labor in 1993.
87. See, respectively, Pavitt 1991 and Held 1987, esp. 143-220.
88. Luhmann 1983. Academic intermediation is the intellectual wing of the state’s assumption of welfare and security functions that shelter the citizenry from the advances of capitalism. On how the need for such intermediation arises, see O’Connor 1973. Fuller 1999b, chap. 5, offers a severe critique of academic intermediation.
scientific knowledge, a tolerance for which we can already see in Planck and especially in his students Koehler and Schlick. If the ends of science are not merely distinct, but increasingly divergent, from other societal ends, then greater efforts need to be taken to render the two compatible with each other. The problem is exacerbated if the ultimate Weitbild seems to recede into the indefinite future and scientists are left with sheer “productivity” as their metric of progress, especially if the productivity of Big Science follows the familiar trajectory of product life cycles in Big Business. Later research is enabled by the rapid obsolescence of earlier research. Indeed, according to standard science indicators, the “harder” sciences are also the more “brittle,” as measured by the rate at which research is superseded. 
However, the plethora of fads and jargons associated with such volatility is quite unlike the stabilizing function that science continues to serve as the premier mechanism of social reproduction. Indeed, all the more reason - as we shall see James Bryant Conant argue in the next two chapters - for a segment of the scientific community to take a special interest in stabilizing background social conditions so that whatever changes do occur in the research agenda are the result of specifically scientific concerns. Yet the need to make the diverging ends of science and society meet increasingly require “intermediators” who rationalize bits of the difference to each other, but none who can make it all the way around from content to function, showing, say, how the continued administration of aptitude tests squares with our best theories of how human cognition develops. Given such a disparity, one might suspect that both science and society have lost their ends. 
To conclude, we still live with the mutual suspicions that made the original rounds of the Planck-Mach debate so memorably rancorous. What is to be feared more: a closed science that has been reduced to an off-the-shelf technology (Planck’s fear of Mach’s policy implications) or the mass indoctrination of a scientific theory that affords it an extrascientific significance that it would not otherwise deserve (Mach’s fear of Planck’s policy implications)? The Technological Menace or The Ideological Menace? This
89. De Mey 1982, 111-31.
90. One of the most celebrated expressions of the disparity between the content and function of science is the reflexive modernization thesis associated with the work of the sociologists Ulrich Beck and Anthony Giddens. See Beck 1992; Giddens 1990. Although usually attributed to the emergence of global environmental hazards, reflexive modernization may equally be the result of our increasing reliance on experts whose opinions change sufficiently often (perhaps because of their sensitivity to evidence) that the sheer publicity of these changes may be more responsible for the general perception of risk than any putatively “objective” source of risk.
polarization reflects the inherent instability of science as objective knowledge. Importantly, it is an instability that emerges more from the classroom than the laboratory. Contrary to Kuhn’s Planck-like view of the history of science as the circulation of research elites - the cutting-edge innovators versus the gatekeeping traditionalists - the Mach-Planck debate reorients us to the struggle within the educational community to represent those features of science that are worth reproducing in society at large.
On the one hand, if science is portrayed as the search for the most reliable means of adapting the material world to human ends, as Mach thought, then the success of science could be measured by the ease with which anyone can appropriate those means without having to rely on the presence of scientists. On the other hand, if science is portrayed as the search for the most comprehensive picture of reality, as Planck thought, then someone whose perspective is reduced or, in some sense, marginalized by science’s current trajectory will interpret it as that imperial form of cognitive relativism known as “hegemony,” the imposition of one special interest group’s view of the world on everyone else. On the one view, the scientists are invisible; on the other, they are overbearing. A closely related tension obtains between saying that science is both “universal” and “expert” knowledge at the same time. If it is universal, then presumably it is codifiable and perhaps even automatable, which would number the days of an elite scientific community. But if science is expert knowledge, then presumably it is esoteric and of limited inherent relevance unless scientific standards are attached to some mode of domination. 
If anyone epitomized Whewell’s vision of the mechanic remade as high humanist for the generation who came of age in the wake of the Mach-Planck debate, it was Michael Polanyi (1891-1976), autonomous science’s most eloquent twentieth-century champion. Although Polanyi had received his training in Mach’s favored field, chemistry, at the
91. The trade-off between universal and expert forms of knowledge is pursued in Fuller 1991, which is followed by several comments and a response by Fuller. We can also think about the choice between Planck and Mach in terms of the science education policy issues that helped motivate their debate: What best reveals that students have made progress in their knowledge of arithmetic - that they no longer need to memorize the multiplication tables because they have access to calculators (a Machian criterion) or that they have mastered the relevant parts of number theory needed to explain how multiplication works (a Planckian criterion)?
German science and raising its moral tone.  Polanyi ended the German phase of his professional career at the Kaiser Wilhelm Institute for Physical Chemistry under the directorship of Fritz Haber (1868-1934), the 1918 Nobel Prize winner (for his work on ammonia synthesis) who had developed poison gas warfare during World War I, while struggling to keep his institute’s military commitments separate from those relating to “pure research.”  However, Hitler’s rise to power in 1933 forced both Haber and Polanyi to emigrate, with Polanyi settling in the
Polanyi wrapped himself in the rhetoric of nineteenth-century humanism to offer a passionate defense of the culture of science in the face of barbarous bureaucrats and creeping socialists. But by virtue of his participation in the quotidian battles that science fought in the public arena, Polanyi could not manifest the sort of detached “scientific” rhetoric that pervades most of The Structure of Scientific Revolutions, a point in no small measure responsible for many of Polanyi’s original insights subsequently being attributed to Kuhn. Nevertheless, it is not hard to see that Kuhn owed more to Polanyi than the appreciative footnote to his magnum opus, Personal Knowledge, would suggest. Consider what Kuhn says here:
If authority alone, and particularly if non-professional authority, were the arbiter of paradigm debates, the outcome of those debates might still be revo-
92. Polanyi’s opinion is reported in Shils 1997, 253-54. The only other name Polanyi mentioned was Einstein’s.
Polanyi had first encountered Haber as
a student when the latter was a lecturer at the Karlsruhe College of
Technology. It is worth noting that
one possible source of Polanyi’s humanistic tendencies was his original
training as a medical doctor in his native
94. Polanyi 1962, which was followed by responses in the journal over the next five years, including one by Stephen Toulmin, which itself generated significant response (see chapter 6, section 8).
lution, but it would not be scientific revolution. The very existence of science depends upon vesting the power to choose between paradigms in the members of a special kind of community. Just how special that community must be if science is to Survive and grow may be indicated by the very tenuousness of humanity’s hold on the scientific enterprise. 
As Kuhn proceeds to discuss the special character of the scientific community, it becomes clear that he trades on the Polanyiesque trope of converting cognitive virtues to moral ones, qualities of understanding to ones of trust. Thus, in place of scientists holding beliefs on the basis of defeasible evidence, Kuhn finds scientists committed to a vision of reality on the basis of intuitive judgment. The sort of radical criticism that is the philosopher’s stock-in-trade appears, in the Kuhn-Polanyi lexicon, as a failure to respect the difference between one’s own social station and that of another inquirer. It challenges the soundness of trusting expert judgment during times of disagreement over the relevant standards of judgment.  Of course, Polanyi did not believe that all experts deserved to have their discretion so respected. In particular, he believed that social scientists could not be left to their own devices because of their tendency to extend some philosophical caricature of scientific practice - a “methodology” - to aspects of everyday life where it played no constructive role and, indeed, where it could be quite destructive if used to challenge established scientific judgment, the grounds for which may not be easily explained to the public. 
Polanyi’s dissatisfaction with social-scientific appeals to methodology
95. Kuhn 1970b, 167; Polanyi 1958, based on his Gifford lectures of 1951-52. For Kuhn’s acknowledgment of Polanyi, see Kuhn 1970b, 44 n. 1.
96. A leading sociology of scientific knowledge practitioner, Harry Collins, sides with Polanyi in a way that reveals the conservative implications that can be drawn from the day-today openness of scientific inquiry: “Even among the experts themselves, who have been trained to many levels above what can be expected of the public’s understanding, radically different opinions are to be found… It is dangerously misleading to pretend that the citizen can judge between the competing views of technical experts when even the experts cannot agree.” See Collins 1987,691. Ironically, then, while sociologists can step into the breach when philosophers cannot decide among themselves which methodology best explains a certain historical episode of scientific theory choice, the sociologists’ lack of scientific expertise prevents them from intervening in contemporary disputes among scientists trying to resolve their own theory choices! The underdetermination of theory choice by the evidence thus licenses, for the SSK practitioner, not the introduction of specifically sociological variables, but rather the discretionary judgments of the local experts.
97. Aside from the atomic bomb, the public image of the natural sciences was tarnished by the attempts of leaders of the scientific community to suppress without testing Immanuel Velikovsky’s best-selling Worlds in Collision (1950), which challenged the received wisdom of physics and chemistry by psychoanalyzing the creation myths of several cultures. When it turned out that some of Velikovsky’s astronomical predictions were correct, several social scientists took the opportunity to chastise natural scientists for failing to live up to their own avowed standards. This was the episode that specifically caused Polanyi to retrench his own position about the necessarily inexplicit character of true scientific expertise. See Polanyi 1967.
reflects an important but neglected point about the terms on which he advocated an “autonomous” scientific enterprise. Polanyi was no positivist. In fact, his low opinion of social science can be traced to its positivistically inspired doctrine of “value-free science,” which effectively established a division of labor between sociologist and policy maker: The policy maker dictates the ends she wishes to be pursued, and the sociologist designs the most efficient means for achieving them. This view, which Polanyi associated with Max Weber, enables the sociologist to live a Janus-faced existence as both policy instrument and pure inquirer.  The Weberian sociologist refuses to participate in deciding the external ends to which her expertise is put because her vocation dictates ends of its own. Thus, the sociologist appears to be the unholy hybrid of Cassandra and Cyclops, a seer who is at the same time professionally trained to turn a blind eye. While this position still haunts sociology today, not least sociology of science (as we shall see in chapter 7, sections 5-6), Polanyi would have recognized its origins in his own discipline, specifically as the stance adopted by his fellow chemists in the years leading up to the First World War. Arguably, this attitude was responsible for the wholesale recruitment of chemists to the kaiser’s cause, the ultimate failure of which eventuated in the massive public reaction against science that characterized the
At the turn of the last century, it was common for chemists to claim that they were engaged in “basic research” but in a sense that was intimately tied to practical ends. To today’s ears, this sounds contradictory, but an examination of the rhetoric of such leaders of the German chemistry community as Emil Fischer (1852-1919) and Wilhelm Ostwald - the respective winners of the 1902 and 1909 Nobel Prizes in chemistry - reveals that when they spoke of pursuing science as an “end in itself,” they meant to include the human intellect’s penetration into what had been traditionally regarded as “natural barriers,” especially the scarcity of land and raw materials that many Germans felt had given the Americans and Russians an advantage in the race to global domination.  Since the advent of polymer chemistry in
98. Polanyi 1974.
99. Johnson 1990, esp. 1, 9 n. 26, 73, 202.
the mid-nineteenth century, an important research strategy of the German chemistry community was to replace natural resources with human-made ones by analyzing naturally occurring materials into their essential elements and then introducing catalysts to rearrange the molecules into new materials according to their durability, resilience, and other humanly desirable qualities. Among the enduring fruits of this strategy were plastics and synthetic fibers.  Ostwald himself was famously converted to this ideology once Germany’s ill-fated support of the Dutch in the Boer War of 1899-1902 led to a restriction on German imports of fertilizer nitrates from South Africa. 
Most chemists espoused an “energeticist” philosophy that resembled Aristotelianism in many respects, especially its denial of any ultimate atomic constituents to matter coupled with an image of “nature” as a realm of pure potentiality that is given a specific form only through the application of human intelligence. The idea that nature might have an inherent character that scientific theory aims to represent was associated with the contemplative stance of the physics community, which (from the chemists’ standpoint) seemed to draw an artificial distinction between the “human” and the “natural.” Following Max Planck, physicists gave direction to their inquiries by sharing a common world picture in terms of which their particular research projects made sense, regardless of their remoteness from practical affairs. And as Kuhn would later do, most famously with his concept of “paradigm,” Polanyi appropriated this last aspect of the physicist’s world-view in order to prevent science from being further compromised in the public sphere.
However, considering Polanyi’s own scientific work in chemistry, the emphasis he placed on the “passion” and “commitment” of scientists to their object of inquiry must be understood in a special light that cuts against some recent attempts to make him out to be a latent feminist, if not a postmodernist more generally. Nowadays Polanyi is often read as having claimed that scientists need to empathize with what they study, that is, to adopt the object’s standpoint to the greatest possible extent - to “know it from the inside,” as it were.  However, it is not at all clear that Polanyi the chemist believed that objects were by nature so sharply defined as to have the sort of intrinsic qualities associated with a distinct “standpoint” or
100. Bernal 1971, 825-27.
101. Johnson 1990, 39.
102. A major case in point is the celebrated feminist biography of the Nobel Prize-winning geneticist Barbara McClintock, who spoke of her method of understanding maize chromosomes in terms of becoming one with them. For the biographer, Evelyn Fox Keller, McClintock’s manner of expression pointed to the characteristically holistic way in which women, by virtue of their early rearing, relate to the world. See Fox Keller 1983.
“subjectivity.” Rather, Polanyi’s assimilation of scientific inquiry to “personal knowledge” is better understood as a matter of scientists coming to realize the Promethean power their work contains and hence to take responsibility for defining the terms in which that work is used, that is, the contribution they make to materially constituting the objects of their inquiry. What can be too easily interpreted as the scientist’s receptiveness to nature is in fact the very opposite, a manifestation of the scientist’s will to power that is constrained only by a sense of moral responsibility, which in turn requires community reinforcement.
In short, Polanyi drew a very sharp ontological boundary between the human and nonhuman aspects of nature, but not between humanity and nature as such.  The problem with the “kaiser’s chemists” was that they believed their responsibilities as scientists ended with a demonstration of technical proficiency and usable results. Thus, regardless of the rigor of their training, they were not properly constituted as a moral community, a flaw that is all the more glaring precisely because the objects of inquiry by themselves did not exert sufficient constraint on the ends to which they may be put. In this respect, despite his Platonist sensibilities about the politics of science, Polanyi - like Ostwald before him - comes close to an Aristotelian metaphysical perspective that regards “artifice” positively as the mark of completion that humans impose on an otherwise unformed nature when they operate by the principle of intelligent design.
At the same time, the elitist undercurrent in Polanyi’s account of science is unmistakable, especially the unflattering analogy he frequently drew between, on the one hand, medieval self-governing guilds and indentured servants and, on the other, basic and applied researchers. (He seemed to think that all social scientists could be fitted into the latter category, perhaps in light of his Weberian interpretation of what they do.) This point has been typically missed by practitioners of science and technology studies who have attempted to contribute to the public understanding of science. When Polanyi alluded to science as a craft, he was envisaging it as an
103. This distinction can be understood in terms of the history of German idealism in the period between Kant and Hegel. Johann Gottlieb von Fichte (1762-1814) and Friedrich Wilhelm Schelling (1775-1854), two ideologues of the modern German university, represent the intended contrast. Schelling held that “Nature” has a will of its own that can be in either conflict or harmony with that of humanity. This view, integral to many an ecological sensibility, was not Polanyi’s, though it is presupposed in the sharp distinction between Geisteswissenschafl and Naturwissenschaft that had begun to dominate German academic life in the mid-nineteenth century. Rather, like Fichte, Polanyi stressed the primacy of the human will, with nature characterized in largely negative terms as resistance to the will’s strivings, but without any positive characterization of its own: pure passivity to the will’s pure activity. On this period of the history of German idealism, see Beiser 1987, 1992.
aristocratic leisure activity comparable to a sport or game in that its value lay in its intrinsic pursuit, not in some specifiable consequences. What Polanyi most certainly did not intend was that science was like more proletarian forms of labor, such as auto mechanics. 
However, Polanyi was not above mixing his metaphors, even when it made a mockery of the history of economics. He likened the mutual adaptability of basic researchers to a market in which the knowledge producers treat each other as the primary consumers of each other’s products, thereby yielding a “spontaneous coordination of individual initiatives.” Unfortunately, had the medieval guilds oriented themselves in such an exclusive fashion, it is unlikely that enough wealth would have been generated to enable the transition to capitalism.  Furthermore, Polanyi believed that the published record of scientists would function as the analogue of prices in the marketplace, especially insofar as the scientific literature indicated the researchers on whose shoulders their authors were standing. However, here Polanyi seemed not to distinguish between current and future prices, and hence failed to account for the fact that unproven products may appear more attractive to speculative investors than proven ones - and, more to the point, that these speculations may be right just often enough to motivate investors to seek a speculative big gain over an assured modest one. This phenomenon, amply in display in the boom-and-bust cycles of the world’s stock exchanges under advanced capitalism, was the raison d’Ítre for John Maynard Keynes’s General Theory of Employment, Interest, and Money and the subsequent financial safety nets introduced by governments
104. As portrayed in, e.g., Shapin 1992b, 1992c. This perspective was expanded in the first STS book explicitly aimed at a popular market: Collins and Pinch 1993. The book consists of some well-known STS case studies, shorn of their controversial philosophical implications, presented in the spirit of opening the door of the laboratory and letting the reader judge for herself what she sees. Collins and Pinch advised scientists that their own interests would be served by dropping inflated talk of “rationality,” “objectivity,” and “truth,” and by promoting instead the more ordinary image of science as “craft” described in the case studies. In that way, the public would learn to have more reasonable expectations of science, and scientists would not feel a need to promise what they cannot deliver. Despite the authors’ intent of providing friendly advice to scientists, the proletarian overtones of their craft rhetoric seriously backfired and only served to fuel the ongoing “Science Wars” discussed in chapter 7, section 5.
105. Polanyi 1962. A comprehensive critique of Polanyi’s pseudoeconomics of science is provided in Philip Mirowski, “On Playing the Economics Card in the Philosophy of Science: Why It Didn’t Work for Michael Polanyi” (paper delivered to the 1996 Philosophy of Science Association biennial meetings). The main reason it didn’t work, according to Mirowski, is that Polanyi’s “free” pursuit of science would have required coercing the rest of society to produce enough surplus to allow scientists to pursue potentially fruitless avenues of research with impunity. In other words, society would have to be adapted to the trajectory of science. For a similar diagnosis, see Fuller 1993b, 283 ff. We shall later see that Kuhn’s account of scientific change helped instill precisely this attitude in policy makers who defined ours as a “knowledge society.”
under Keynes’s name. This would seem to justify a science policy of protected risk seeking over that of enforced risk aversion. 
One plausible explanation for the metaphorical malfeasances in Polanyi’s economics of science is that ultimately his frame of reference was neither medieval guilds nor modern markets, but rather the fiduciary exchanges of nonliterate societies, especially two Sudanese tribes - the Nuer and the Azande - as described by Edward Evans-Pritchard (1902-73), professor of social anthropology at Oxford in the second half of Britain’s imperial period in Africa.  The spontaneously reciprocal exchanges of information and insight that enable the political economy of science to work so smoothly may be traced to the remarkable level of trust that scientists place in the integrity of those exchanges. In short, an unquestioning faith in the overall system allows for the identification and often tolerance of discrepant beliefs in particular cases. A precedent for this extraordinary behavior may be found in nonliterate societies, which typically lack objectified traces of their history, a fact that extends beyond the sheer lack of written records to the absence of any institution designed to preserve and interpret artifacts as vestiges of the tribe’s past.  Individual accountability is thus based on
106. The disanalogies between Polanyi’s normative political economy of science and advanced capitalism run even deeper, for Polanyi suggested that it is better to suppress novel results than allow foolish error to infect the body of scientific knowledge and perhaps provide a pretext for alarming the general public. Yet capital-intensive science has caused modern societies to endure exactly the opposite, such that novel but untested scientific results, when produced in a major research facility, are often not only publicized but also used as the basis for policy, so as to thereby instantly prove their worth. One would probably expect Polanyi to disapprove of these developments, were it not that the public’s belief in the efficacy of science rests largely on just this instant applicability to policy matters.
Interestingly, the contrary case - that research
should be restricted to that which can be communicated freely- has been
forcefully argued by the
Polanyi 1958, 287-294, draws explicitly on
Evans-Pritchard. A sympathetic
introduction to his work by one of his former students is
qualities that tribespersons manifest at the moment of questioning that reveal their affirmation of tribal values.
The viability of such a system obviously depends on the exercise of considerable control over the initial acculturation of tribal members, thereby alleviating any lingering doubts that the requisite virtue could be easily simulated in performance. Likewise, while publications are very much constitutive of the fabric of scientific knowledge, nevertheless they are not treated as a historian or lawyer would, namely as evidence of something that underlies and potentially undermines their surface pronouncements. Rather, they are treated as acts of “virtual witnessing,” in Steven Shapin’s terms, as if the reader were encountering the author face-to-face. It is this sense of epistemic transparency - rather than efficiency or precision - that should be seen as crucial to Polanyi’s analogy between the scientific publication and the price mechanism. The complexity of scientific and tribal exchanges may be unfathomable to all but the initiated, but it must be fathomable at least to the initiated. In both cases, the security attached to the knowledge that insiders possess is a measure of the security of their collective grasp on what lies outside themselves.
Polanyi was only among the first of a half-century’s worth of philosophical thinkers who have turned to Evans-Pritchard’s ethnographic fieldwork to test their intuitions concerning the possibility of cross-culturally valid standards of reasoning.  Nowadays the only vestiges of this late imperialist legacy are puzzles involving Nuer claims to have descended from birds and Zande adherence to oracles even in the face of substantial contradictory evidence. The work that generated these puzzles was conducted in the southern
This debate’s more celebrated participants
include Ernest Gellner, Peter Winch, Alasdair MacIntyre, Charles Taylor,
Steven Lukes, Martin Hollis, and Ian Jarvie. See
110. On Evans-Pritchard’s keenness to mobilize
the natives in
This meant operating within the tribal cosmologies. It was not enough to do as Evans-Pritchard’s teacher Bronislaw Malinowski (1884-1942) had done, to reveal the socially functional character of native beliefs and actions in Darwinian and Freudian terms. That would be to provide a sense of rationality that went above (and, in the case of Freudian explanations, “below”) the heads of the natives. Rather, Evans-Pritchard relied as much as possible on the natives’ own accounts of why they believed or acted as they did. Besides advancing the philosophical fortunes of cultural relativism, this strategy enabled the British to maintain the confidence of the natives by showing that they could deal in their terms. Indeed, Evans-Pritchard was unique among his contemporaries in insisting on the need for anthropologists to eschew interpreters and master the native languages themselves. 
From the standpoint of the history of the sociology of knowledge, Polanyi’s deployment of Evans-Pritchard to fuse the moral and cognitive orders in science marks Emile Durkheim’s revenge on Max Weber’s largely accepted account of rationalization as the mark of progress in complex societies. For Weber, rationalization incurs greater levels of potential scrutiny and accountability than ever before, as people have less firsthand knowledge of things that bear on their lives; hence, the increasing importance of such bureaucratic procedures as examinations and audits - not to mention recorded trials and experiments - where the adjudicative significance of written records presupposes a distrust of the inconstancies of personal observation, memory, and testimony. In contrast, Polanyi’s Durkheimian vision upgrades these cognitive deficits to moral virtues by inserting a curious form of what economists call time-discounting. 
Time-discounting is normally presented as a problem in rationality. People often favor short-term gain even if it means forgoing a much larger gain in the long term. In that sense, people discount the future in favor of the present. The processes of rationalization described by Weber arguably fall in this category, as the meticulous keeping of records ensures accountability only in the short run, since in the long run the number of stored records undermines the ease with which they can be accessed in future accounting exercises.  From this perspective, the Durkheimian discounts
111. On the need for anthropologists to master native language and customs, see Evans-Pritchard 1964, 79-80.
112. For an examination of time-discounting from an economic and psychological standpoint, see Price 1993; Ainslie 1992. I apply these ideas to the historiography of science in Fuller 1997d, 95-101.
113. Even when records can be translated into computer files, a cost is incurred, and indeed the specific mode of computerization may impede access after several generations of improvements in computer technology.
in reverse, preferring long-term goals to short-term advantage. In the long term, the members of a tribe or a scientific community must be able to live with each other, given the indefinite duration of their collective projects. Thus, a premium is placed on social mechanisms designed to avoid the introduction of friction into its members’ interactions, unless some clear collective gain is in sight. The burden of proof is shifted so that one trusts unless there is reason to doubt. In more individualistically competitive societies, the requisite “reason” may have just as much to do with personal as collective gain, but in both cases, the challenge to the status quo will succeed only if it serves the collective: i.e., I alone cannot stand to gain by doubting a colleague’s word, others must as well. It follows, in the scientific context, that the period of doubt is likely to lead to a still stronger commitment to the collective enterprise. This attitude is evidenced in not only Kuhn’s account of scientific change, but also the quotidian practices of scientific journal editors who refuse publication of negative results or theoretical critiques, unless the path ahead is as clear as the one being rejected. 
114. This position should be contrasted with the
provocative approach of the French anthropologist Dan Sperber, who has
promoted the concept of “quasi-beliefs” to capture native epistemic states.
Whereas Durkheim and Evans-Pritchard
(as well as Polanyi and Kuhn) assume that native knowledge claims imply a
deep, almost religious, commitment to the associated propositions, Sperber
argues that if alien cultures incorporate metaphor and irony the way we do,
then perceptions of radical difference may simply reflect a misplaced
literalism on our part, an interpretive failure born of an invalid
methodology. In other words,
relativism may be itself a quasi belief shared by Western anthropologists,
which the natives realize and use as
a basis for telling them what they want to hear and otherwise
playing inside jokes on earnest Western inquiries.
(The case of Margaret Mead amongst the
sexually active youth of