PREFACE

In recent years, burgeoning sociological and historical studies of sci­ence have considerably complicated the domain in which philosophy of science might be expected to clarify scientific practice.

Perhaps the root failure of the methodologists was to attempt to trace scientific knowledge to the epistemological activities of the in­dividual scientist. For reasons that will be evident in the main argu­ments of this book, this strategy must fail to locate a suitable sense of objectivity for scientific practice. From the classical methodologists, some echoes of Popper’s fallibilism will be the major residue. Recent books by Rellone, Bloor, Latour and Woolgar, and Ravetz have influ­enced what follows more than citation can make evident. These books have not always been treated kindly by reviewers, but have raised issues that seem to have decisively shifted the appropriate philosoph­ical perspective on science toward a more historical and sociological standpoint.

In Chapter 1, the important insight that scientists are created through acculturation is accepted largely in its Kuhnian version. The result of this acculturation process is that scientists are always confronted in scientific activity with both theories and complex data, and are at­tempting to fit the two together. The contemporary philosophy of science is not, therefore, obligated to solve the general epistemolog­ical problem of priority of theory or data, but should take as its subject matter the temporal accommodation of the two. Rationalism or em­piricism in the philosophy of science results when theory is thought primarily to guide data, or the reverse. Since both forms of guidance occur, rationalism and empiricism are true of episodes in scientific history, but neither can be a complete characterization of scientific progress. We must give up the idea that new theory destroys older theory, or that new theory must be an extension of older theory. The conclusion is thus that general methodologies based on empiricism or rationalism are inadequate to the dynamic interaction of theory and fact in science.

The dynamic interaction of theory and fact is then studied through other resources. The discussion introduces the view that scientific instruments break the connection between theory and observation, allowing the dialectic of theory and data to take place, and that the use of instruments establishes data domains, which are what theories adapt to. Instrumental splitting of theory and data means that data can be gathered independently of current theorizing, and used to constrain that theorizing. When new data domains are established by the use of new instruments, older data domains and the theories that are adapted to them do not disappear, but they can be split off as settled in principle.

Chapter 2 examines the social and cognitive norms present in sci­ence, using the results of recent work in the sociology of science. The differences between the scientific disciplines are also noted to the extent that sociological techniques can make them relevant. This in­vestigation shows the splitting of the larger domains to which theories adapt. The existence of controversy in science is examined to see how it can select superior views without such bitter acrimony that coop­erative research would be impossible. The social mechanisms of Chapter 2 are the means whereby a consensual fit of theory and data can be won from a welter of individual perspectives, a process that turns discovery into justified science.

The continual adjustment of theory to data means that successful science may not be recognized immediately, but may appear only after false starts and misleading development. Tinkering with theory and data means that scientific knowledge must emerge from this tin­kering in terms of our model, and that scientific knowledge isn’t (typ­ically) instantaneously recognizable. Science differs from other forms of human activity in that its intellectual constructs and instruments are the products of a special internal evolution, but the attempt to fit resultant theory and data is structurally much like trial and error else­where.

Chapter 4 is the presentation of the structure of science obtained from the synthesis of the ideas presented here with the valuable in­sights of the past philosophies of science. The interaction of theory and fact can result only when facts are seen to be negotiated data that have survived controversy, and not as a means of veridical epistemo­logical access to the real world. Facts appear only as data text requir­ing interpretation. This means that fact and theory may both give way in the dialectic of progress, facts becoming irrelevant to new domains, and theories becoming maladaptive to the data domains for which they were intended. The process of accommodation of theory and fact is then studied in a series of scientific examples. Objectivity is rescued as the relationship of best explanatory theories to their intended data domains. It is suggested that much of the formal work produced in empiricist philosophy of science can contribute to an understanding of the structure of the public domain of fitting negotiated fact with adaptive theory.

In the Appendix, some lessons are drawn about the human sciences from our discussion. The same structure of accommodation of theory to fact is to be expected, but there are some interesting differences. The data domain of human science theorizing is typically either the splinter domain of the rational, utility-maximizing economic agent or the entire complex domain of man in a historical social setting. These two approaches are obtained by different points of entry of under­standing into the human sciences, and can’t be resolved as they stand. Attempts at progress by proposing intermediate data domains have been too theoretical, and have been rejected by the disputants be­cause they cannot be tied down to an appropriate instrumentarium. These inconclusive debates are examined as confirmation of the anal­ysis of the relationship of the natural sciences and the human sciences that follows from a consideration of data domains.

The author would like to thank Donald T. Campbell, William C. Wimsatt, and Howard Darmstadter for admitting in public to having read his earlier book on the philosophy of science. Current efforts would not have been possible without encouragement from and con­versation with Robert Paul Wolff, Bill Heintzelmann, Bob Saltz, Christina Erneling, Alf Bang, Bob Yamashita, as well as the euphoria consequent to the victory of the transworld depraved philosophers.

Amherst, Massachusetts

April 19, 1983