September 3, 2008
This book pretends to be historically informed philosophy of science but it is nothing but a disgrace to its field. It is a parade of stupidity and error interspersed with long plateaus of verbose banalities.
Laudan's model of science is based on the concept of a "research tradition," which is a set of metatheoretic commitments (methodological, ontological, metaphysical, etc.). Theories are of minor importance; they are merely "designed to particularize" (p. 81) the research tradition in which they were conceived. These little particulars don't matter much; it is the research tradition that scientists care about: "a scientist's cognitive loyalties are based primarily in the research tradition rather than in any of its specific theories," which means that "he generally has no rational vested interest in hanging onto those individual theories" (p. 96).
History shows that in reality the priorities are the reverse, so Laudan is forced to falsify history to fit his model. For example, he claims that Leibniz and others "were seriously prepared to dismiss Newtonian physics because its ontology was incompatible with the accepted metaphysics of the day." (p 64). Leibniz was certainly uncomfortable with action at a distance (as was Newton), but to say that he was prepared to "dismiss" Newton's theory is ridiculous. Indeed, Laudan provides no evidence for his interpretation and gives no reference to support it. (This is not the only time Laudan provides an imaginative interpretation of history without giving any reference to support it; see, e.g., p. 88.)
The idea that theories are "designed to particularize" metatheoretic commitments also requires tinkering with history. Newtonianism is a prototypical research tradition, so obviously Newton must have drawn up this blueprint first and then developed his physics "to particularize it." Indeed Laudan claims that "the ontological and methodological frameworks for Cartesian [and] Newtonian physics ... were explicit even from their inception" (p. 75). But this is nonsense. "Hypotheses non fingo," for example, is a typical commitment of the Newtonian research tradition, but it was not "explicit from the inception," but rather added in the second edition of the Principia, whereas there were many hypotheses plainly labelled as such in the first edition.
Ok, so maybe Laudan was wrong about the inception part but the Newtonian research tradition, once established, surely showed its importance by the many scientists who took it as their highest loyalty, toiling relentlessly to "particularize" it? Not really. Instead scientists immediately started violating its assumptions. "In eighteenth-century natural philosophy, many scientists were simultaneously Newtonians and subtle fluid theorists. ... Boscovic ... set out deliberately to develop an new 'system of nature,' by picking and choosing from among the assumptions of two incompatible research traditions, Newtonianism and Leibnizianism. Maupetius attempted something similar. The work of their contemporary, Daniel Bernoulli, illustrates an analogous attempt to forge a compromise between the research traditions of Cartesian and Newtonian physics." (p. 104-105).
Indeed, with all this "picking and choosing" going on, Laudan admits that "if one looks at the great research traditions in the history of scientific thought ... one can see immediately that there is scarcely any interesting set of doctrines which characterizes any one of these research traditions throughout the whole of their history." (p. 97). In other words: a research tradition is a post hoc construction by its creator, from which its followers "pick and choose" as they please, about which "scarcely anything interesting" can be said. This Laudan chooses as his main tool for analysing the history of science.
Laudan wants progress to mean problem-solving progress. But if two theories solve partially overlapping sets of problems then "the problem of setting gains against losses in insoluble" (Collingwood, p. 147). Not to worry, though, because Laudan has thought long and hard about this and figured out how to avoid "the alleged Collingwoodian 'insolubility'" (p. 150): we simply determine which of the problems are more important. Wow! Why didn't anyone think of this before? I suppose it takes the sharp intellect of Laudan to come up with such a penetrating insight. This flash of genius leads to the following "appraisal measure": "The overall problem-solving effectiveness of a theory is determined by assessing the number and importance of the empirical problems which the theory solves and deducting therefrom the number and importance of the anomalies and conceptual problems which the theory generates." (p. 68). Laudan's later criteria (chapters 3 and 4) for evaluating acceptability and progressiveness of "research traditions" all fall back on this "appraisal measure." Since there are no rigorous rules for the weighting of problems, this "appraisal measure" is nothing but a pseudo-objective piece of fluff which in no way solves the original problem. This does not bother Laudan, however, who goes on to "apply" this "appraisal measure." Suppose, for example, that "a research immunologist must prescribe medication" (p. 108). As usual Laudan cannot resist prefacing his point with some dishonest slander: "Kuhn would insist that no rational choice can be made" (p. 109). As for Laudan himself: "My own reply to the question, of course, would be, 'choose the theory (or research tradition) with the highest problem-solving adequacy.'" (p. 109). As we saw, this in effect amounts to saying: choose the best theory. And this Laudan pompously declares to be a great advance in the "theory of scientific rationality" (p. 106).
Laudan hails it as a discovery of his that not all anomalies count against a theory; indeed, in his usual scornful tone, he laughs at the "irony" that other philosophers have "refused to recognise" this fact (p. 29). Laudan's great discovery is that "a problem can only count as anomalous for one theory if it has been solved by another" (p. 30). But this criterion is vacuous since any fact can be accounted for by some other theory, for example a theory dealing exclusively with this fact or an ad hoc modification of the original theory. It is not part of the criterion that the competitor theory must be very good; and even if this was part of the criterion this would still not exclude the ad hoc option since Laudan considers adhocness "a cognitive virtue rather than a vice" (p. 115) and maintains that "an ad hoc theory is preferable to its non-ad hoc predecessor (which was confronted with anomalies)" (p. 116). See pp. 118-119 for a discussion which becomes absurd in light of this inconsistency.
Finally, a partial list of further instances where Laudan is flagrantly misrepresenting the views of others, each of which would be enough to fail a sophomore essay. "[Kuhn] proposes essentially that it is the accumulation of a large number of anomalies which finally induces scientists to abandon a theory. ... Kuhn's account cannot be squared with the historical fact that scientists have often abandoned a theory in the face of only a few anomalies and have other times retained a theory in the face of an ocean of empirical refutations." (p. 37). No reference is given to support these claims (other than an utterly useless footnote which says only "Cf. especially Kuhn (1962)" with no page reference), and understandably so, since Kuhn never said those things. "Kuhn has been misled by his discovery that some empirical problems are not jointly shared between different traditions or paradigms (which is certainly true) into believing that no problems are identical" (p. 145). Again there is of course no reference since Kuhn never said anything of the sort. "Kuhn never really solves the crucial question of the relationship between a paradigm and its constituent theories." (p. 74) Of course not, since paradigms don't have constituent theories. "Kuhn's paradigms have a rigidity of structure which precludes them from evolving through the course of time in response to the weaknesses and anomalies which they generate. ... there can be no corrective relationship between the paradigm and the data." (p. 75). Nonsense; such improvements are a crucial part of what normal science is all about. "Both Kuhn and Lakatos seem to believe that the decision about which elements of a maxi-theory fall into this privileged class [of assumptions that are not questioned] is arbitrary and not governed by rational considerations: on their account, it simply 'happens'." (p. 100). Laudan goes on to spell out criteria for choosing which assumptions should be in this class which Kuhn and Lakatos obviously refrained from mentioning since they are moronically trivial. One last slanderous outburst: Kuhn's "mature-immature dichotomy" is "methodologically suspect because it effectively renders [Kuhn's model] immune from empirical criticism" since "any actual scientific examples which fail to fit [his model] can be explained away as proto- or pseudo-scientific" (p. 151). This is nonsense since, in Kuhn's model, once a science has established its first paradigm it remains a science and cannot go back to being a proto-science.