Introduction

ix

Introduction

Our objective in this book is to state precisely and c1early where and why sociological analysis is ~;J!~s_ary in the understanding of scientific knowledge. Our main method is to present historical case studies. We then show how sociological analysis applies in these cases, and how it is an essential complement to even the most insightful interpretations derived from other perspectives. In establishing this last point we introduce important work on science drawn from psychology and philosophy. We try to present this non-sociological material in the most favourable light so that, rather than overestimating the power of the sociological approach, readers will appreciate that sociology makes a necessary contribution to a larger project of obtaining an understanding of science. We seethe sociology of scientific knowledge as part of the project of science itself, an attempt to understand science in the idiom of science. Other sociologists have attempted to develop perspectives on science using approaches which are uncharacteristic of science, and which do not accept or rely upon its methodological conventions or its accepted cosmology. We ourselves honour science by imitation: in our study of science we try to emulate its own matter-of-fact, non-evaluative approach. Ironically, some scientists and philosophers have assumed that since we neither praise nor defend science our objective must be to subvert it. They have failed to understand that for a social scientist to seek to justify science would be to deviate from its own non-evaluative precepts. One cannot simultaneously adopt a scientific approach and celebrate it. We begin our study with an examination of the reliance placed by science on observation and experience. A number of psychologists and philosophers have argued that observation is theory dependent. They say our minds actively create part of what we perceive, and do so in a manner that expresses the theoretical presuppositions of the perceiver. This view would seem to have c1ear advantages for a sociological approach to the acquisition of knowledge. Drawing upon

recent empirical and theoretical work, however, we argue in Chapter 1 rhat there is no need for sociologists of knowledge to take this view. It is perfectly possible and empirically plausible to accept the c1aim that perception is to a large extent 'modular' - that is, isolated from other components of our cognition, and only influenced by them in a limited way. This in no way rules out a sociological analysis. The reason is that the raw material.of OUT experiences..musr . .be tran_~fQrrn~into observation reports before it can. begin 10 enter into m~mi.nc..~n9wledge. In the second chapter, therefore, we look at ways in which a consensus is reached as to the correct interpretation of these experiences. We proceed by giving a sociological analysis of R.A. Millikan's experimental attempts to measure the charge on the electron. Basing our account on Gerald Holton' s detailed examination of Millikan's experiments and the records of them left in laboratory notebooks, we argue that Millikan 's interpretation of his results depended upon an available tradition in his subculture. This has profound implications for OUT understanding of the interaction between nature and culture. In Chapter 3 we look at the character of c1assification and try to formulate some general conc1usions about how words relate to the world. We begin by considering how c1assification would proceed if it were based wholly on assertions of resemblance and identity between ostensively given examples of a c1ass of thing. This is a simplification, and later discussion introduces the relevant complexities, but it serves to introduce the view of meaning called 'sociological finitism'. The essential point is that OUT c1assifications are always underdetermined by the promptings of experience or by previous acts of c1assification. Each new application of a term is sociologically problematic. A finitist account thus emphasizes the conventional character and sociological interest of c1assificatory activity, such as the fact that every act of c1assification has the form of a judgement, every act changes the basis for the next act, every act is defeasible and revisable, and every act involves reference, not just to the 'meaning ' of the term applied, but also to the 'meaning ' of all the other terms currently accepted for use in the context. Scientific c1assification does not, however, treat things as conventional c1usters of diverse particulars: it treats things as falling into natural kinds whose members can be thought of as essentially identical to each other, or as identical in certain important respects, for example as electrons, as a particular chemical element, or as members of a biological species.

Scicntifu: Knowlcdgc
and elaboration
01' the image

lntroductiou
built up over the course

xi
01' two centuries

Objects which are mernbers of natural kinds are things about which true or false staternents can be made, and to which fully general scientific laws can be applied. Chapter 4 deals with the insistently realistic idiom in which scientists treat the objects referred to in their laws and theories. Some sociologists see realism as a style or assumption that is ro be opposcd. Our view is that it can be understood and illuminated by the finitist account already developed. First, scientists' talk of essential identity is clearly not a simple response to ernpirical prompting. Ir must be thought of as a preferred strategy of a group of scientists. Second, the 'theoretical discourse' of scientists cannot be seen as deriving from a fixed set of theoretical statements with determinate meanings from which significance for particular cases emerges. It is all too easy to talk of Daltorr's theory, or Newtorr's theory or Mendel's theory, and conceive of them as sets of statements with fixed meanings. Once these theories are addressed as historical phenomena this approach faIls aparto As would be predicted by finitism, it is extremely difficult lo specify what a theory is as an historically situated entity, and it is quite impossible to identify it as a ser of statements. It is better to think of a scientific theory as an evolving institution. Illustrarion is provided by appeal to historieal studies of 'Mendel's theory '. The first fOUT chapters therefore show the role of tradition, con vention, consensus, and the social processes whereby these are upheld or dismantled in establishing and sustaining knowledge. Chaptcr 5 shifts the focus and seeks to show in a more synoptic way what sociological study can add to our underslanding of scientific knowledge and scientific practice. One concem here is to show that scientific research is conducted with reference to goals and inreresrs, and that both research and the evaluation of research must be understood as intrinsically goal-oriented activities. Two major examples are discussed: the role of chemical arguments in the legal disputes conceming a patent protecting the manufacture of red dyes derived from aniline, and the interpretation of experiments in which fermentation was first held to take place outside a living cell. Chapter 6 is concemed with how scientists defend the intellectual territory over which they wish to exert control, with how they demarcate science from non-science in order to present a credible image of specialist expertise and intellectual authority, A consideration of the origins of the 'experimental method' shows that our current image of science has been carefully forged scientists for specific practical purposes, by natural philosophers and and that further development

or more continues at the present time. Our tinal chapter looks at mathernatics and mathernatical reasoning. iI is particularly important that we inelude this topic in our rernit since mathematics has an important role in the natural sciences, and it is often claimed that mathematical knowledge defies analysis in sociological terms. We show that, on the contrary, sociological considerations have an essential role in understanding the status 01' the simple but nonetheless fundamental proposition that 2 + 2 = 4. We go on to provide a general discussion of the character of proofs in mathematics, and to indicate what we take to be the praper sociological response to such achievements. The text is the responsibility of the authors coIlectively, and the outcome of their working together over a considerable periodo But we do not believe in writing by committee, and have drafted the various chapters individually, taking account of OUTdifferent interests and competences. lf there is some diversity of style from chapter to chapter this wilI no doubt be its source. If there is less than complete "nmiql'fl(,; argumenr thi~ (,()\llri wel] have occnrred for the same
()f

', ,. ,
J t
f

reason,

but our intuition

is that any such inconsistencies form derives trom its origin

are unlikely

10 be major

ones. Thc books particular

as an introductory

,
t
t
f

course for graduate students. We hope that it will be of some value to others, from their final undergraduate year onwards, seeking a text in the sociology of scicntiric knowledge. Since there is no barrier in terrns of vocabulary or presumed background knowledge of sociology, and since the text concentrates on basic issues, anyone with an interest in history or philosophy of science, or cognate fields, might also find the book of value as a way of relating sociological approaches to their own. Conversely, the case study materials do not discuss scientific knowledge in a way which dernands any special training, so the book should be accessible to those with an interest in the sociology of knowledge and culture as traditionally studied. It should be remembered, however, that few of the historical case studies we cite have been written in order to advance the elaims of sociologists of scientific knowledge. A number o' the historians whose work we have borrowed may weIl hold a more rcalist or a more positivist view of scientific knowIedge than that which is presented here, but they have produced exemplary empirical studies, ano these are always grist to the mili of theoretical perspectives like that of the sociology of knowledge.

I
1

XII

Scicllfifrc

Knowledg

Although this is a dernanding introduction to our subjecr, it is nonetheless an introduction. It concentrutes on the basic toundations from which an understanding of more elaborate ideas may grow. Hence it gives little prominence lo currently rnuch debated issues iikc relativism. reflexivity and self-reference, the role of new literary forms and post-rnodernist 'deconstrucrion - issues which are best approached after some extended considerarion of simpler matters, Nor is there any attempt to survey the whole range of views and perspectives currcntly
10 be found in our tield. When the sociology of scienrific knowledge first got under way over twenty years ago, there was general agreement

Acknowtedgements

about its basic approach. The concem at that time was mainly to oppose the arguments of rationalist philosophers who wished to treat science as a unique forrn of human activity, one which required no empirical understanding other than that implied by describing it as rational. Now, however, differences of view in sociology of science are every bit as extreme and far reaching as those in philosophy of science. It is no longer possible either to give an account of 'the approach ' established in the field, or a properly detailed accounr of the many and varied approaches actually to be found. It rnust be remembered therefore that the 'sociological analysis' of the books title is our own. If it is occasionally referred lo as 'the ' sociological approach, the reader is asked lo indulge US, or lo look lo our notes into altemative perspectives. and references for ways

We should like to express our gratitude to our students al Edinburgh Universiry, particularly those who took the MSc course in the Sociology of Scicntific also extremely hixtorians of of this book. very valuable Knowledge upon which this book is based. We are indebted to the various sociologists, philosophers and science whose work we have drawn upon in the writing Two (anonymous) publishers readers made a number of and helpful observations, and we have done our best to Carole Tunsley, for typing and retyping

meet their points. Finally we wish 10 thank numerous efticiency

drafts not only without cornplaint bur also with great and good humour, and Brian Southam of The Athlone when we missed several deadlines, Edinburgh 1995 Barry Barnes David Bloor John Henry

Press for his forbearance