Volume 10 Number 2 1997
Stephen Cole presents this book as an attempt to resolve the twenty year old dispute between "Mertonian" sociologists of science and their (social) constructivist opponents. Cole, a leading representative of the Mertonian tradition characterises the dispute in simple terms. For the constructivists such as Harry Collins, Karen Knorr-Cetina and Bruno Latour the content of science is social. Constructivism, according to Cole, posits that the discovery of matters of scientific fact arise not from the application of rational criteria for evaluating evidence nor as a result of the determining force of `nature', but from "chance occurrences, the availability of particular equipment, the availability by chance of particular substances, and social negotiation among people inside the laboratory" (p.12).
By contrast, according to Cole, the Mertonian tradition refuses to accept this excessive relativism. To be sure, the rate and institutional location of scientific progress may be determined by social and cultural factors, but the specific content of scientific ideas cannot be accounted for simply in terms of the local and chance circumstances of its production. Scientific knowledge, according to Cole, has an irreducibly cognitive dimension. Although Cole accepts what he takes to be "the constructivists'" argument that "the evaluation of (claims to) knowledge is inherently subjective (this) does not mean that the cognitive characteristics of a new contribution (the evidence, theories, models and logic in a new paper) have no influence on the formation of these subjective opinions" (pp.230-231, my emphasis).
If the author finds a clear distinction between his view of science and what he takes to be the views of the constructivists, he nonetheless sees some possibilities for a resolution of this long-standing controversy. For a start, Cole agrees that the constructivists have got some things right. The traditional positivist view of science, he believes, implied that there would generally be a fair degree of consensus amongst scientists and that science would progress in an ordered and predictable fashion. Scientists, according to the traditional view, would make judgements based entirely on the the objective evidence of observation and experiment. On the basis of his own quantitative evidence of scientists' evaluations of National Science Foundation (NSF) grant proposals, Cole rejects this view. Instead, he argues that the empirical evidence supports the constructivists' claim that there is very little consensus amongst scientists working at the "frontiers" of knowledge, whether they are physicists or sociologists (p.135).
Despite this recognition of the value of the constructivists' qualitative and ethnographic studies of scientific work, Cole argues that empirical evidence of the constructivists' studies does not provide any support for their more general arguments. Although the constructivists give us detailed accounts of the research process, he suggests, they have no convincing way of explaining "why some ideas win out in the competition for communal acceptance" and some do not (p.59). This can only be done by looking at the "cognitive content" of the ideas and the "empirical evidence" upon which they are based.
The author's faith in the determining force of empirical evidence extends to his suggestions as to how the dispute between the `Mertonians' and the `constructivists' might be resolved. "I wonder", he asks with apparent naivety, "whether constructivists would be willing to participate in a joint research effort in which it would be agreed prior to data collection what the significance of various outcomes would be?" (p.234). Unfortunately the book ends shortly after this proposal is made, and the author provides us with few clues as to what this joint research effort might consist of.
Far from offering the reader any new way of thinking about the old debate between Mertonianism and constructivism much of Cole's argument simply reproduces many of its original terms. Indeed, in some respects this book is a particularly weak contribution to the debate, and makes a whole series of misreadings of the literature. Cole confuses, for example, the idea of constructivism (the validity of a claim to knowledge depends upon the conditions within which the claim is made) with the sceptical idea that there can never be any conditions for preferring one claim to knowledge over another. He also makes the bizarre assumption that there is a necessary relation between a commitment to relativism or constructivism, and a belief in the importance of the subjective views of individual scientists in determining the course of scientific change. Not only is there no necessary relation of this kind, but there is a certain irony to this view: for constructivist sociologists of science have, somewhat cautiously, preferred to examine the more directly observable and recordable features of scientific work, rather than engage in any analysis of the `motivations' or `subjectivity' of the scientist. In this respect at least, the constructivists share with Cole a tendency towards empiricism.
Making Science is clearly intended to elucidate fundamental issues in the sociology of scientific knowledge, but it has also another dimension. Whereas the first three chapters and the last chapter are devoted entirely to refuting the claims of the constructivists, the remaining chapters articulate a rather different intellectual and political agenda. Here the author's interlocutors are not the predominantly European constructivists, but his fellow sociologists of science of the American school. In particular, Cole examines Robert Merton's claim that scientists should be guided by the ethical principle that any evaluation of the quality or validity of scientific work should be made on the basis of universal and objective criteria. Against Merton's `universalism', Cole suggests that there are some situations - such as the evaluation of research grant proposals - where it is helpful for the scientist to be guided by her own subjective opinion. A total commitment on the part of scientists to the Mertonian ethos of universalism does not, he argues, enable science to progress as rapidly as it might.
Although Cole's attempts to refine some of the arguments of Mertonian sociology of science contribute very little to his broader dispute with the constructivists, they nonetheless betray a further unease on the part of the author with the constructivists' work. The problem for Cole is not just that constructivism is incoherent, but that it has forced the sociology of science to be "too preoccupied by epistemological problems" (p.237) and has, thus, displaced the earlier concerns of the field with, for example, ethical questions. If Cole's sense of the restricted terms of recent debates in the sociology of science is well-placed, Making Science provides no new ways of thinking beyond them. Teachers and students may wish to use this clearly written book as a representative example of the limitations of the Mertonian tradition in the sociology of science, but I doubt that they will use it as the resource for anything more enterprising.
Andrew Barry