Philosophy Of Science And Its Discontents

Steve Fuller

• Publisher: Taylor & Francis
• Year: 2019
• Print ISBN: 9780367282899
• eBook ISBN: 9781000303148
• Edition: 1
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

The most important and exciting recent development in the philosophy of science is its merging with the sociology of scientific knowledge. Here is the first text book to make this development available.

1
Mythical Naturalism and Anemic Normativism: A Look at the Status Quo

1. The Mythical Status of the Internal History of Science, or Why the Philosophy of Science Is Suffering an Identity Crisis

Can a discipline suffer an identity crisis? Presumably, if it either loses its subject matter or discovers that it has lacked one all along. Richard Rorty (1979) took the former position to criticize all of philosophy in Philosophy and the Mirror of Nature. To compensate for the restricted scope of my critique, I am resorting to the more radical latter position. I aim to show that, strictly speaking, philosophy of science has an imaginary locus of inquiry, known as the internal history of science. Its main historiographical assumption is that there is a natural trajectory to the development of science when it is regarded as a knowledge producing activity. Internalists divide over even this simple definition, especially over the identity of the relevant "regarders." Some, like Imre Lakatos (1981), believe that only the philosophically informed historian, by ridding the history of science of the social impurities that typically vitiate scientific judgment, can see the natural trajectory of knowledge production. This, in turn, gives the Lakatosian a privileged perspective from which to offer advice on the future conduct of science. Others, like Dudley Shapere (1987), believe that scientists themselves come gradually to perceive this trajectory over the course of history as they "learn to learn" about the world. Shapere's philosophically informed historian is then directed to identify the emerging patterns of knowledge production across the disciplines. Most internalists seem to fall somewhere between Lakatos and Shapere, though my later discussion of Larry Laudan's (1984, 1987) recent work will show that the middle ground is much more unstable than it first may appear.

The internal history of science is not just another private philosophical fantasy but one that has seduced the imaginations of other humanists and social scientists with an interest in the nature of science. These inquirers have, for the most part, uncritically accepted the burden of defending an "external" history of science, even though little more is conveyed by the expression than a general opposition to internalist history. In practice, however, an externalist is anyone who believes that science, like any other human activity, cannot be discussed intelligibly unless it is treated as something that is essentially embodied, where "embodied" means being bounded in space and time. For example, when psychologists appeal to human cognitive limitations or sociologists appeal to local interests in order to explain the course that the history of science has taken, they are treating science as essentially embodied. Despite the natural advantage of the externalist program from the standpoint of a unified science of human behavior, externalists have undercut this advantage by acquiescing to the image of history that internalism implies, an image borrowed from the principle of inertia in Newtonian physics, which postulates that bodies continue in a regular motion unless subject to interference. The "body" in this case is what philosophers call "rational methodology," i.e. reliable means for attaining some end, normally called "truth." Given this image, externalists seem to be concerned only with what draws the scientist away from the course that rational methodology projects, a decidedly second-class enterprise.

As we are beginning to see, an entire rhetoric accompanies the internal-external distinction and the debates that have since engulfed philosophers, historians, sociologists, and most recently, psychologists, A quick guide will prove useful for understanding what follows. Since the existence of an internal history of science looks more plausible with ever more abstract characterizations of the actual history, there is considerable ambiguity about the referential function of such key internalist terms as "science," "rationality," "truth," "autonomous," "epistemic," and "cognitive." But here are some rules of thumb. "Science" tends to be the name given to the activity to which the other five terms pertain to the greatest extent, usually as determined by today's philosophically informed historian. In practice, this means that physics is virtually everyone's paradigm case of a science. "Rationality" and "truth" are more often described in functional terms than in truly substantive ones. Thus, "rationality" is to "truth" as "means" is to "end." To say anything more substantive would be to open the Pandora's box of positions in the philosophy of science.

However, one implication of defining rationality and truth in purely functional terms is that the sorts of acts that might count as violations, or dysfunctions, of rationality or truth are always in the back of the internalist's mind. Indeed, this point serves to motivate the idea that science is a distinctive enterprise, one which cannot merely adapt to its ambient social setting, but which must often resist the call of the social. When scientists engage in this resistance successfully, they are 'autonomous," a term with unusually strong psychologists overtones by internalist standards. Externalists have something to study (so internalists claim) precisely because science is frequently dysfunctional, largely because scientists prove not to be autonomous. Finally, "epistemic" and "cognitive" are increasingly used as synonyms, though typically the former refers to science as a rationally organized body of knowledge (as in the German Wissen), whereas the latter refers to science as a rationally directed way of knowing (as in the German Erkennen). The two terms often appear as modifying "factors," when activities that make positive contributions to the production of knowledge are being discussed (for a history, cf. Graumann 1988). That they are used almost interchangeably reflects the extent to which internalists presuppose that there is an inherent link between the quality of knowledge processes (cf. cognitive) and the quality of knowledge products (cf. epistemic).

In addition to stacking the terms of the debate in her own favor, the internalist prevents externalist encroachment by deftly moving between historical modalities. An especially agile slide is from what did happen to what ought to have happened, via what would have happened-if certain ideal conditions had been historically realized. Indeed, an internalist might go so far as to blame, not the historical agents, but the historian herself for being unable to abstract the essential course of scientific progress from the surfeit of data that confronts her. However, this aggressive methodological posture conceals a remarkable weakness on the philosopher's part. The internal history of science is supposed to describe those episodes in which science has been done "by the rules," that is, in ways that philosophers have reason to believe will provide most direct access to truth. As it turns out, there are rather few of these episodes, which suspiciously correspond to the most famous incidents in the history of science (Newton, Lavoisier, Darwin, Einstein), which even more suspiciously are presumed to play a significant role in explaining the less exemplary episodes (e.g., the scientists in these episodes were merely imitating, anticipating, or completing the work of Newton, Lavoisier, Darwin, Einstein). We will see early in this chapter that such a view of the history of science cannot withstand much scrutiny.

Not only do internalists slide between value and causal significance, but they also presume a curious relation between normative and descriptive accounts of science. Since so few episodes conform to her preferred rules, the internalist is forced to say that most of the history ought to have been other than it was, which, as far as the entire history of science is concerned, means that she is offering a largely normative account. The complementarity suggested here between descriptive and normative adequacy is generally taken for granted in contemporary philosophy of science. For example, when Karl Popper (1972) began to think that falsification increased the reproductive capacity of the human species, he started to deny that falsification was "merely" a normative account of science but argued rather that it was part of the "situational logic" underlying all evolutionary developments. Of course, it is entirely possible for a philosopher to be blessed with most of the history of science conforming to her normative account. Larry Laudan (1977, ch. 5), for one, has explicitly aspired to be such a philosopher. Nevertheless, the normative aspect of the internal history of science tends to be asserted only once its descriptive inadequacy has become clear. Indeed, the greater the discrepancy between the internalist's norms and the historical record, the more likely that the discrepancy will be blamed on the scientists' failure to live up to the norms. In short, norms become the "higher ground" to which the philosopher's falsified account retreats!

This glimpse into how the philosopher manages not only to salvage but also to strengthen her normative account in the face of resistance from the history of science is just one of many rhetorical strategies that internalists have developed in order to keep the burden of proof firmly planted on the externalist's shoulders. I will simply comment on three of these strategies and leave it to the reader to spot them as she reads through the arguments addressed in this book and in many of the works that I cite. The reader will notice that a general feature of these strategies is that they involve an implicit double standard.

The first strategy is to circumscribe the sphere of the normative so that the philosopher presents herself as evaluating, but not actually directing, the development of science. This gives the philosophy of science its apolitical sense of critique, which is reminiscent of evaluation in aesthetics. The strategy can be seen at work in the preference that philosophers have shown for talking about what sort of research should have been pursued at a certain point in history (when nothing can now be done to change what happened) over what sort should be pursued in the future (when something can be done to change what would otherwise most likely happen). At face value, the past-to-present and present-to-future modes of science evaluation are susceptible to empirical treatment to roughly the same degree (e.g. both involve reasoning with counterfactuals [cf. Elster 1978, 1983]), yet since the futuristic mode invites explicit policy direction in a way that the historical mode obviously cannot, the futuristic mode has been shunned by philosophers.

But what, then, does the philosopher gain by restricting her normative focus to evaluating? By foreclosing the option of intervening in policy matters, the philosopher is also discouraged from engaging in a causal analysis of her own value judgments, which is to say, her appraisals of good and bad science are never subject to feasibility constraints. An internalist like Lakatos can declare that the history of science could have proceeded much more rationally by taking a course entirely different from the one it actually took, without his contemplating the costs and benefits that would have likely been incurred by the alternative trajectory. In short, an aversion to policy blinds the philosopher to the fact-laden character of values. What appears to be the optimally rational move from the standpoint of abstract philosophical criteria could turn out to be an implementation nightmare, which on any non-philosopher's theory would count decisively against the rationality of such a move. But, of course, to court facts in this way is to conjure up the image of science as an embodied activity, which would, in turn, offer the externalist too much dialectical leverage.

This brings us to the second rhetorical strategy by which the internalist advantage is maintained, namely, by manipulating the level of abstraction at which science is described. As I shall show below, whenever a concrete example of autonomous science is sought, social, political, economic, and psychological factors are immediately brought to the fore. However, philosophers nimbly deflect the full force of these moves by straddling the two extreme attitudes represented by Lakatos and Shapere. On the one hand, internalism can be defended (a la Lakatos) by arguing that a preferred model of scientific rationality can be instantiated by an indefinite number of sociologies, given that a methodology like falsificationism does not imply a specific way of organizing knowledge production. For example, it is open as to whether the "conjecturing" and "refuting" functions of falsification are to be seen as two cognitive processes operating in one scientist, in two distinct groups of scientists, in the same group of scientists, each of whose members is her own conjecturer and someone else's refuter-and the list of possible sociologies goes on. On the other hand, internalism can be defended (a la Shapere) by arguing that the preferred model of scientific rationality is simply the one that has in fact evolved during the history of science. The philosopher pitches her description so abstractly because she wants to concentrate exclusively on those features of the actual situation that have contributed to the growth of knowledge. In that case, to bring in the extra sociological detail would simply be redundant, if not potentially misleading to the naive historian.

What is missing between the Platonism of Lakatos and the Hegelianism of Shapere is the conceptual space that the externalist aspires to stake out, namely, the empirical preconditions, or "feasibility constraints," for the various models of scientific rationality. Contra Lakatos, not every social arrangement could (even in principle) support a preferred model of rationality; but contra Shapere, the current social arrangement is not necessary to the growth of knowledge and in fact could probably be improved upon. In its role as the normative wing of externalist history of science, social epistemology proposes to explore the implied intermediate possibilities (cf. Meja & Stehr 1988; Schmaus 1988). For even a move as scientifically basic as distinguishing "theory" from "observation" is not innocent of implicit sociological considerations. At the very least, such a distinction normally presupposes a division of labor between observers and theorists whereby the theorists ordinarily trust the reports of observers, whose skill was developed independently of any training in the theories judged by those reports; yet the theorist's trust in the observer's reports does not extend to determining what the theorist must do in light of such reports; and so forth.

The third, and final, rhetorical strategy is undoubtedly the subtlest, as it pertains to what counts as the "content" of science. I will shortly sort through the philosophical vexations surrounding this notion, but for now it will suffice to enumerate the sorts of things that the internalist wants to include and exclude by raising the specter of content. I imagine that a reader not already familiar with recent philosophy of science will find this enumeration somewhat undermotivated. For example, the private thoughts (what used to be called "the context of discovery") of the scientists are generally excluded, unless (as in the case of Faraday, Darwin, and other meticulous diarists) they are written so as to be tractable to methodological (or "cognitive") analysis. The public expression of science in journals and books is generally included, insofar as this reflects what the internalist can recognize as "the logic of justification." Any feature of the journals and books that seems to be driven by the "mere" pragmatics of scientific writing, such as article formats, length, and citation conventions is excluded.

Given that the transmission of knowledge depends on these pragmatic features, the internal history of science finds itself in the peculiar position of being a theory of scientific change that lacks any account of science as a process (cf. Hull 1988). Moreover, while internalists are not oblivious to the fact that what one says (in print) and what one does (in the lab) often diverge in significant ways, they are not thereby moved to examine the nature of these discrepancies. In fact, field studies of what scientists do in their work sites are generally regarded as centrally in the purview of externalist sociologists. Indeed, Maurice Finocchiaro (1973, chs. 16-17) has observed that the mark of the internalist is to ignore a basic canon of the historical method, a canon that would have utterances be treated as just more events to be explained, rather than as transparent accounts of other events. Thus, whereas an internal historian typically regards a scientist's statement of method as a description of her actual practice, it would be more historiographically sound to regard the scientist's statement as a retrospective rationalization of whatever she happened to have done, which would then need to be determined by other evidence. Why does the internalist blithely privilege what scientists say over what they do? A clue to the answer Hes in the internalist's residual Platonism, which interprets the surface of science's own record-keeping as deep because of the close resemblance that the trail of scientific journals and books bears to something that a Platonist would readily find significant, namely, the dialectical development of a system of propositions. Indeed, the extent to which the scientists' background deeds deviate from their up-front words perversely feeds the Platonic impulse by suggesting that the inability of scientists to act as they should does not prevent them from knowing, and hence saying, how they should act.

2. Dismantling This Myth, Step By Step

The philosophical significance of the internal history of science is established by an argument like the four-step one presented below. (After presenting this argument, I will systematically challenge each premise and short-circuit the inferential links between them.)

• (1) Most of the best cases of scientific reasoning have exhibited an independence, or "autonomy," from other sorts of deliberations. This autonomy may be defined psychologically or institutionally, but in either case the general idea is that the deliberators do not let non-epistemic matters intervene in the epistemic assessment of a scientific claim.
• (2) What makes these cases of scientific reasoning so good is their autonomy.
• (3) The role of autonomy in increasing the likelihood that the most rational, objective, and/or valid scientific claims will be chosen has been sufficiently great to make autonomy worth recommending as a general methodological strategy.
• (4) There is prima facie reason for believing that suboptimal cases of scientific reasoning-when less rational, less objective, or less valid claims were chosen-are traceable to the intervention of non-epistemic matters. From this tenet follows the division of cognitive labor canonized by Laudan (1977, ch. 7) as the arationality assumption: namely, that philosophers study scientific reasoning in its autonomous phases, whereas sociologists study it in its heteronomous ones.

Philosophers associated with "positivism" (Ayer 1959), "rationalism" (Hollis & Lukes 1982), and "scientific realism" (Leplin 1984a) normally differ on many things, but not on their allegiance to the above four tenets. In fact, virtually all these philosophers have written as if the tenets logically unfold in the order indicated, with (1) entailing (2), which then entails (3), which in turn entails (4). On the contrary, however, one could, as a matter of historical fact, accept (1) without accepting any of the other three tenets. Admittedly, this would not make for a very potent philosophy of science. Still, it is important to see that each successive inference could be blocked, if only to get an initial feel for the logical looseness of internalism and the problems that consequently lay ahead for it.

But let us begin on a positive note by granting both the intelligibility and truth of (1). The presence of autonomy in all the best cases of scientific reasoning must be supplemented by other inductive tests before the move from (1) to (2) can be licensed: Have an equally large number of suboptimal cases of scientific reasoning exhibited autonomy? And even in the best cases, might not different factors be responsible in each case for the reasoning's turning out to be as good as it was? After all, commonality of effect need not imply commonality of cause. If further empirical study produces affirmative answers to either of these questions, then the move from (1) to (2) is blocked, and the presence of autonomy will have been shown to be as epiphenomenal as the "fact" that all the best scientific reasoners have been males. However, assuming that these empirical hurdles are surmountable, there is a more trenchant conceptual barrier between (1) and (2).

The barrier involves getting a clear sense of what the content of a science might be. After all, if nothing else, "autonomy" implies the fixing of the will upon some object. In the case of autonomous science (or the autonomous scientist), the object seems to be knowledge, or epistemic content. But what exactly is that? The internalist tries to define content so as to separate it from anything obviously material which would make it susceptible to an externalist analysis. It is easy to think of this strategy as a safeguard against a form of weakness of the will that might befall the scientist who too closely associated the pursuit of knowledge with one of its material ...

Table of contents

• Cover
• Half Title
• Title
• Copyright
• Contents
• Acknowledgments
• Introduction: Philosophy of Science and Its Discontents
• 1 Mythical Naturalism and Anemic Normativism: A Look at the Status Quo
• 2 Reposing the Naturalistic Question: What Is Knowledge?
• 3 Reposing the Normative Question: What Ought Knowledge Be?
• Coda: Epistemic Autonomy as Institutionalized Self-Deception
• Bibliography
• Index