© F. Enke Verlag Stuttgart Zeitschrift für Soziologie, Jg. 4, Heft 1, Januar 1975, S. 34-45

Controversies in science: Remarks on the different modes of production of knowledge and their use*

Helga Nowotny Europäisches Zentrum für Soziale Wohlfahrt, Wien

Wissenschaftliche Kontroversen: Über verschiedene Arten von Wissenserzeugung und Wissensgebrauch*

Inhalt: In der Wissenschaftssoziologie haben bisher Probleme der Konsensfindung vergleichsweise starke Beach­ tung erfahren, während selten danach gefragt wurde, warum wissenschaftliche Differenzen weiterhin bestehen. Eine Untersuchung wissenschaftlicher Kontroversen könnte zur Beantwortung dieser Frage beitragen. Kontroversen ent­ stehen nicht nur, wenn kognitive Differenzen bestehen, sondern vor allem, wenn diese Differenzen unterschiedliche Auswirkungen haben. Dies geschieht dann, so lautet die vertretene Hypothese, wenn immer die Art der kognitiven Beziehung zu anderen kognitiven Strukturen - und daher auch der Gebrauch des Wissens - ein anderer ist. Diese Hypothese wird durch Beispiele aus der kognitiven Anthropologie und der Wissenschaftsgeschichte illustriert. Schließ­ lich wird noch das Phänomen der „Nicht-Kontroversen“ erörtert, worunter jene Fälle verstanden werden, in denen kognitive Differenzen zwar bestehen, aber nicht — oder noch nicht - aktiviert wurden. Aus dieser Interpretation der verschiedenen Arten von Wissenserzeugung und Wissensgebrauch lassen sich eine Reihe von Schlußfolgerungen für die sogenannte Wissensakkumulation und für Engpaßsituationen in der Wissenschaftsentwicklung ableiten.

A bstract: In the sociology of science relatively much emphasis has been accorded to the problem of how agreement in science is produced, while it has seldom been asked why disagreement should persist. An analysis of controver­ sies in science should therefore throw light on these questions. Controversies arise not only when cognitive diffe­ rences exist, but when these differences come to matter. This happens, it is hypothesized, whenever the patterns of connectedness to other cognitive structures and therefore to the use to which knowledge is put, will differ. This can be illustrated by examples taken from cognitive anthropology and from the history of science. Finally, the cognitive status of ‘non-controversies’ is discussed, i.e. those occasions where cognitive differences exist but are not - or not yet - activized. Consequences can be derived from this interpretation of the different modes of pro­ duction of knowdledge and their use for the so-called accumulation of knowledge in science and certain bottleneck situations in the development of scientific thought.

I. Introduction through which intellectual debates could be sett­ led which proved otherwise undecidable through In the sociology of science as well as in philoso­ the conventional methods of scholarly, in parti­ phy much attention has been accorded to ques­ cular scholastic, debate. According to BEN-DAVID tions how agreement in science is produced. In the superiority of the new guild of scientists, the the sociological camp this interest has focused on incipient scientific community, over the contem­ the functioning of the by now legendary scien­ porary guilds of doctors, theologians and philo­ tific community, whose prove of existence has sophers, lay precisely in the fact that the former turned out to be so elusive (CRANE 1972, KLIMA were willing to abide by the rules of new scienti­ 1973), while among philosophers the belief fic method which assured them that agreement seems still widespread that procedures for reach­ could be produced. It is quite plausible that this ing clear-cut solutions to any kind of problem promise opened new visions of an accumulation worth its name ought to exist and enable agree­ of knowledge, as the newly acquired knowledge ment to be established. BEN-DAVID (1973) has would come to be built upon the firmly cemen­ recently put forth the intriguing interpretation ted foundations of previously agreed upon know­ that the institutionalization of science in 17th ledge. It is not without significance that the new century England was directly linked to the widely method was explicitly compared to a set of juri­ felt need for the possession of a secure method dical norms, thus reviving the old idea of a Court of Reason, and constituting a guarantee for both * This paper was presented at a working conference parties that a consensus producing procedure on “Determinants and Controls of Scientific De­ was at hand. Upon this skeleton of agreement velopment” held in Graz, Austria, in June 1974. Most producing norms and rules, flesh has been put of these contributions edited by K. KNORR, H. STRASSER, H.G. ZILIAN in Dordrecht, Holland, by the sociology of science through its emphasis Reidel, 1975, will appear under the title of the con­ on the functioning of the reward system, the so­ ference. cialization of scientists into their roles and the esta­ H. Nowotny: Controversies in science 35 blishment of effective communication structures ‘macroscopic and microscopic types of analyses’, — lines of inquiry that can be re-interpreted as are but a small sample of the types of conflict direct contributions to the question of how, in a scientific context which MERTON classifies through what social and cognitive factors, agree­ as social rather than intellectual. “ . . . When we ment is reached. In spite of this heavy bias, it is consider the current disagreements among socio­ nevertheless surprising that the question why logists, we find that many of them are not so there should be disagreement at all, has seldom, much cognitive oppositions as contrasting evalua­ if ever, been asked before. Viewed against the tions of the worth of one or another kind of so­ impressive results of decades of research in the ciological work. They are bids for support by the sociology of science (MERTON 1973) with all its social system of sociologists” (1973: 58). Con­ emphasis on the communal advances of know­ troversies in sociology, or polemics, as they are ledge, one is left wondering why disagreement called, are said to be “less a matter of contradic­ still persists. If scientists are socialized in long tions between sociological ideas than of compet­ years of intimate and intensive training, if they ing definitions of the role considered appropiate absorb and internalize the elaborate norms and for the sociologist. Intellectual conflict of cour­ standards of scientific thought and practice, if se occurs; an unremitting Marxist sociology and they have at their disposal an effective commu­ an unremitting Weberian or Parsonian sociology nication structure geared to their needs for the do make contradictory assumptions. But in con­ exchange of rewards and views, if authority is sidering the cleavages . . . we should note whether vested into institutions like Nobel prize winners the occasion for dispute is the claim that this or and eminence is easily recognized — why should that sociological problem, this or that set of there be disagreement and about what? ideas, is not receiving the attention it allegedly deserves. I suggest that very often these polemics have more to do with the allocation of intellec­ II. The ambiguous nature of controversies tual resources among different kinds of sociologi­ cal work than with a closely formulated opposi­ Answers to this deliberately provocative question tion of sociological ideas” . (1973: 55). Although will tend to be of either of two kinds: one set MERTONS observations pertain specifically to so­ of answers will emphasize the social nature of ciology, there is no inherent reason why the dis­ any scientific controversy. According to this view tinction of what controversies and polemics are controversies are what might be called ‘conflict about might not be extended with the appro­ by proxy’ — they are debates fought for reasons priate modifications to other disciplines as well. which ultimately are to be found outside the realm in which they purport to be fought and Another set of answers stressed the inevitability for reasons that are extraneous to the intellec­ of the ideological nature of any scientific contro­ tual content of the debate. Pushed to the extre­ versy, whereby ideology can either be taken in me, controversies and social conflict in science an all-inclusive sense or be limited to scientific appear to be inevitable, because there are only norms which bind together scientists who share limited resources in terms of positions, funds or them. The first view has been expressed most even immaterial rewards like professional recog­ clearly by ROSE and ROSE (1974) who have nition. Their scarcity leads to the emergence of emphasized the specific ideological issues around intellectual disagreements as the only legitimate which controversies evolve. Speaking about ideo­ mean of carrying out an otherwise inevitable com­ logical inputs into science, they note the broa­ petition for status and rewards. This kind of con­ der social pressure towards the comissioning of troversy is at times contrasted with controversies research, the engagement of scientists, and the of a supposedly purer version which are fought ideology of science and of its institutions. Their for their intellectual merits only. Such a view analysis of the types of conflict which are con­ has been expressed by MERTON (1973), limited ducted in contemporary neurobiology between though to sociology as a discipline. Charges like different groups, clearly reveal the strong ideolo­ the one that ‘others are busily engaged in the gical component contained in each paradigm. The study of trivia while the important questions are inborn view of schizophrenia, for instance, held neglected’, that ‘alleged cleavages between sub­ by the reductionists, refuses to admit criticism stantive and methodological issues’ exist, between of social structures while at the same time it en­ 36 Zeitschrift für Soziologie, Jg. 4, Heft 1, Januar 1975, S. 34-45 courages a manipulative view of treatment. Ac­ ween techniques, axioms, interpretative mles and cording to ROSE and ROSE controversies cannot explanations, WHITLEY arrives at a typology of be neatly separated on grounds of their purely competition in science. According to it, controver­ intellectual merits from the wider social signifi­ sies can be characterized by opposing views of cance which they must assume, willingly or un­ the object under study, of the appropriate mode willingly. Controversies of the type described by of understanding and explaning it, of the diversi­ these authors therefore tend to be of global sig­ ty of research techniques, of their validity and nificance, as not different views of the subject utility, of the different meaning of results obtain­ matter are opposed, but causes and treatment ed from applying these techniques and finally of alike. ROSE and ROSE regard such ideological the appropriate application and extension of a inputs in science as normal. Controversies appear technique or a model. The genesis of controver­ to be the outcome of divergencies which exist sies, as I infer it from WHITLEYS typology, ap­ at this inclusive ideological level which in turn pears to be the result of an uneven development are a result of the societal pressures to which within each research area or speciality with re­ science finds itself exposed. gard to the degree of the existent institutionali­ zation and the prevailing division of labour, lin­ But controversies cannot only be looked upon ked not only to the competition between ideas, as bringing to the fore the ideological comitt- but between scientists as well. ments of the opposing sides, but also as particu­ larly revealing with regard to the common norms The ambiguous nature of the interpretation of and standards of scientific behaviour shared by controversies in science as either a form of social the adherents of one side in a dispute. It is in conflict which manifests itself in intellectual dis­ this sense that BÖHME (1974) has recently re-ana­ guise or as a form of genuinely intellectual dis­ lyzed a historical controversy in psychology and agreement which entails social consequences can has reached the conclusion that controversies, also be found in the writings of some philoso­ like twin research, “should make explicit rules and phers of science. Notably KUHN (1962) has main­ norms which otherwise direct the secure march tained that the normal scientific tradition which of science only in an inconspicuous way, and at emerges from a scientific revolution is not only the same time the degree of obligation and the incompatible, but often incommensurable with different function of such rules to neighbour dis­ that which has gone before (1962: 102). Diffe­ ciplines become obvious” (1974: 7). For BÖHME rences between paradigms can mark off different controversies are a direct expression of the kind worlds in science, whose respective content is a and extent of agreement on norms that can be closed book to the opponent who cannot even found in a scientific community. This view has understand what goes on within a given paradigm been criticized by KLIMA (1974) who has argued that is not his own. Taking KUHN literally would that it is inadmissable to infer the social cohesion mean that controversies are utterly futile, if they of a scientific community from the logical cohe­ are conducted between paradigms, because they sion of the subject matter. are unsolvable, while they should not really occur within a paradigm. This extreme view of the non- This last argument points to an interconnection compatability of certain paradigms has been mi­ between social and cognitive factors in contro­ tigated by other philosophers of science in favour versies which is tacitly present in all the writings of a more dynamic process ar work which may on the subject, but which has not found a satis­ eventually help to overcome stated incompatibi­ factory analysis yet. An interesting theoretical lities. Notably TOULMIN (1972) sees such a proc­ attempt to bridge the gap has been made by ess in the selection and variation of scientific con­ WHITLEY (1973). Starting from the assumption cepts. Although he acknowledges that the con­ that competition between ideas and scientists is ceptual variation and intellectual selection in inevitable, controversies would be one form in science are embodied in professional organiza­ which the ongoing competition manifests itself, tions and their social ecology, the question of being linked to the prevailing form of division controversies is more or less solved by an appeal of labour and to differing degrees of cognitive to the evolutionary force inherent in scientific and social institutionalization of specialities and development which will inevitably weed out research areas. By considering competition bet­ those concepts which are somehow less adequate H. Nowotny: Controversies in science 37 to a given intellectual climate. In a similar way, links not only within an individual, but also although on very different grounds, LAKATOS’ across individuals engaged in the process of know­ and ZAHAR’S (1973) theory of progressive re­ ledge production. Some of the new patterns of search programms would lead one to believe that connectedness which thus emerge may relate to controversies are eventually solved by the more cognitive structures which are already highly progressive research programme superseding the ideologized, others may become ideologized in less progressive one. the course of developments and some may iso­ late them successfully against such links. As a This brief excursion through the literature, which collective process new knowledge is produced was not meant to be exhaustive, underlines the through the impact of new information, put into ambiguous nature of controversies and their in­ circulation by an individual, on the existing cog­ terpretation. There is a marked tendency to ac­ nitive structure which manifests itself in a chan­ cord priority either to social factors, external ge of the patterns of connectedness. Controver­ to scientific thought properly speaking, or to sies arise, we assume, not so much between oppos­ ideas and how they work their way almost irre­ ing content of thought, as between different spective of social circumstances. Yet it is quite modes of connecting cognitive structures. obvious that controversies have a dual character: they are cognitive as well as social processes at At the same time the acceptance or rejection of the same time. But in order to show this duality ideas brings individuals into certain relations to at work, it seems premature to accord priorities each other and thus can be said to create a soci­ to either side. My own interpretation starts from al structure describable in terms of cognitive si­ acknowleding the phenomenon of disagreement milarity or dissimilarity (MULLINS 1969, SCHMUT­ in science. Controversies are an integral part of ZER 1974). The dominant view of communica­ the collective production of knowledge; disagree­ tion in science as purely instmmental, geared to ments on concepts, methods, interpretations and the effective diffusion of knowledge or to trace applications, are the very lifeblood of science and the channels through which ideas have been proc­ one of the most productive factors in scientific essed in retrospect, is a result of its development development. If argument would cease, this would out of the early public opinion studies (STORER soon mean the end of the scientific enterprise 1973: XXVI-XXVI). It has totally obscured the which would rapidly turn into the administration fact that communication can be looked upon not of knowledge under the patronage of a scientific only as channels through which anything can be bureaucracy. In order to learn how disagreements communicated more or less effectively, but that occur, it is however not sufficient to regard them it is meaningful to look at social ties among scien­ as mere reflections of the world of outside con­ tists as resulting from the fact that they share or re­ flict. Rather, controversies are to be looked at pudiate certain ideas1. It is in this sense that we as an endemic and vital motor in the production have maintained elsewhere that the acceptance of scientific knowledge: they arise from different or rejection of ideas creates a social structure, modes of production of knowledge and their use. although one not freely operating from various constraints (NOWOTNY & SCHMUTZER 1974). In what follows controversies will be interpreted as the outcome of different modes of produc­ In the collective production of scientific know­ tion of knowledge taking place in different minds ledge based on a social structure which is partly and contexts, for different uses and linked in dif­ a result of those communication processes ferent ways to already existing cognitive struc­ tures. This differentiation takes place because 1 This is only one factor which may lead to the esta­ blishment of communication links, but seems to be every individual prossesses his unique cognitive the most important in science. Others are neighbor­ map, as the representation of the outer world hood or social distance, norms and their costs, and sta­ can summarily be described. It will thus attempt tus differentials between the communicating individuals. to link newly received information - provided The implications of these factors have been dealt it is not rejected out of hand - to previously with in NOWOTNY & SCHMUTZER (1974) in their relevance for the creation of a communication struc­ received information, it will tend to evaluate and ture. A more detailled and technical treatment in fit it into its own cognitive system. Viewed col­ form of a simulation model can be found in SCHMUT­ lectively, this process will generate new cognitive ZER (1974). 38 Zeitschrift für Soziologie, Jg. 4, Heft 1, Januar 1975, S. 34-45 through which ideas are diffused — to be accep­ tion, we can go through all the explanations ted, rejected or modified2 — the emergence of given by the Karam themselves - such as that controversies signalizes that different modes in it is not a bird, because it doesn’t have plumes, producing new knowledge are at work. This but ‘hair’, that it has no ‘brain’ but only bones means that the cognitive ties established with on its head — only to discover that there are already existing cognitive structures, i.e. the other birds that also have ‘hair’ or ‘no brain’ overall patterns of connectedness, will tend to and are still regarded as birds, until all possible differ. Social ties between individual scientists morphological explanations have been exhausted may help or hinder these processes, but they and have proven insufficient to explain the casso­ themselves are dependent on what it is that is wary’s special taxonomic rank. Nor does an in­ to be connected. As the most simple case of vestigation of classification by habitat prove any connectedness is to be found in classification more successful; the cassowary seems to defy systems we shall first turn to the examination this kind of analysis. To make matters even more of such a case. complicated, if one looks at the objective physi­ cal and biological differences between the casso­ wary and other birds and bats, as they may be III. Why the Cassowary is not a bird or what perceived at the level of common sense natural potential controversies are about history, a number of differences stand out. But this is not, at least not for the Karam, a sufficient For the Karam, a highland New Guinea tribe, the reason to place the cassowary into a special ca­ cassowary is not a bird (BULMER 1967). The tegory. We also learn that some other New Gui­ Karam, we are told, place in the taxon ‘yakt’ all nea highlanders who have just as much common the 180 or so kinds of flying birds and bats they sense and just as much knowledge of the casso­ recognize by names. Cassowaries are large, ostrich wary as the Karam, nevertheless place it into or emu-like birds and are not included in ‘yakt’, the same taxon as other birds. The answer but constitute the constrasting taxon, ‘kobity’. to which we are finally led, is, very briefly, that Why, we may ask with the anthropologists, is the cassowary enjoys a special place in Karam the cassowary not a bird? taxonomy, because it enjoys a special and unique relation in Karam thought, to man. This is out­ To find an answer to this seemingly simple ques­ wardly manifest by an elaborate set of rules which regulate when and by whom the cassowary 2 So far, citation index measures have failed complete­ may be hunted and how it is to be eaten. ly to take into account the controversial character of knowledge. These measures, like the ones based It would be utterly naive to put aside this story on the Science Citation Index, are usually thought as an esoteric illustration of primitive thought. to measure the impact of a publication on the cor­ The construction of the world, as we see it, is in pus of knowledge. As a side effect they are believed to enable the evaluation of previous papers, by the the most simple case achieved through assorting amount they are invoked in the course of new works. objects and grouping them together into classes, It is generally assumed that any reference to previ­ thus resulting in more or less elaborate classifi­ ous work means positive evaluation, while I would cation system constitutive of social reality. It maintain that a citation could also imply negative can be mainly animals which are assigned to evaluation - in the sense that the present author disagrees and takes issue with the cited work - or special categories and kept there by rules of be­ neutral evaluation - in the sense that the work re­ haviour ( TAMBIAH 1969, DOUGLAS 1966) Or ferred to is merely cited to invoke a routine summary occupations which form the bases for images of or a position which is well-known to the readers. society, of classes, power and its sources (COXON For the citation technicians it should not be too dif­ ficult to broaden the depth of their tool of analysis & JONES 1974). They are systematic in the sense by incoporating this simple distinction. I can also of following their own internal logic (TYLER imagine that, technically speaking, citation indices 1969) which can be extracted and analyzed by could be employed to locate controversies in scien­ means of such methodological devices as the ce and to be of help in establishing the empirical method of free grouping and multi-dimensional patterns of variation to be found in different disci­ plines and across time. I doubt, however, that such scaling. From many studies in cognitive psycho- an analysis will lead very far, if it is not embedded logy (BERRY & DASEN 1974), it seems extreme­ into a theoretical framework. ly plausible to argue that the world of experien- H. Nowotny: Controversies in science 39 ce can be ‘sliced’ in different ways. Certain ways which are estabilished between cognitive structu­ of slicing exclude others, but there is always res, and the patterns of connectedness which some overlap, as well as there are gaps — objects, emerge from linking cognitive operations to each persons, animals, concepts, which do not fit ei­ other. It is not sufficient to arrive at different ther way are therefore given the status of the kinds of knowledge, to assign birds or concepts anomalous, either to be explained away, or to to different categories. This is merely the basis be considered as dangerous and thus to mark by for potential controversies. In addition, this their conceptual ambiguity the interstices of so­ different kind of knowledge has to be linked to cial realities (LEACH 1969). other cognitive structures such that the difference will matter. Scientific theories are certainly much more com­ plex, elaborate, multi-leveled than the relatively Differences will tend to matter, however, when ever simple classification systems which serve every­ the clashing of different views on subject matter, day purposes. But they also tend to slice the methods, rules or interpretation, has deeply-felt world of scientific experience in different ways, and far-reaching consequences — in other words, although most of the time within the framework when the uses to which knowledge will be put of a legitimized scientific pluralism. There are by connecting cognitive structures, will differ. partial overlaps and inconsistencies, even within The chain of consequences from the original in­ mathematics as we know for certain since GÖ- nocent theorizing to its plugging into other cog­ DEL’S proof. The way in which it is sliced may nitive structures and therefore to its use, may appear arbitrary, but it is, like the cassowary, be visible from the beginning or may become visib­ ultimately tied to its user by some special rela­ le only much later; it may be ideological, when tion to other categories in his thought. “Knowl­ it is tied to socially relevant action, or remain edge is for the sake of action, and action is roo­ within the realm of thought to impinge on ac­ ted in evaluation” , writes KARL PRIBRAM and tion only in indirect ways. The unfolding dyna­ this holds in a very elementary psychological mics of controversies tend to render visible the sense as well as for the more elaborate uses of connections between different sub-fields of the knowledge. Knowledge, even in its most esoteric, common collective map, either by creating new seemingly useless and unpurposive forms, is still links or by making visible what has been present tied to other knowledge. This stems from the as hidden connectedness before. Controversies dynamic, constructivist nature of the production are therefore cognitive upheavals, through which of knowledge, as PIAGET has demonstrated. Knowl­ the common collective map is restructured by edge can be stored, but then it becomes subject providing new links or by bringing out into the to the processes of recall. As an active process, open those that were hidden. The often observed however, knowledge has to be used, by building appeal to a widening circle of arguments in the up cognitive structures which in turn continue course of a debate is only one sign for it. The to produce new knowledge. This holds for the emotional commitment of the opponents is an­ production of individual as well as collective other which is not merely a blind ideological fol­ knowledge, where individually produced knowl­ ly, conducted for reasons quite foreign to the edge is shared and standarized through the work­ debate, but it is an engagement for a specific use ing of the communication structure. of the produced knowledge.

The cassowary is given a special place in Karam The interpretation of controversies as generating taxonomy because it enjoys a special relation­ and depending upon the generation of new and ship to man in the Karam view of the world. different links to other cognitive structures, im­ This makes its classification at least potentially plies the rejection of a current view: the view controversial, as it is easy to imagine contradictory that the logical structure of controversies, espe­ ways of placing it into a taxonomic system. In cially their logical incompatibility, is sufficient order to become activized, controversies are not to explain their emergence. Logical incompatibi­ only dependent on social factors, such as a pro­ lities, partial or total, may exist and yet go unno­ cess of interaction, but they must be mobilized ticed or not be responded to for a variety of rea­ also within each cognitive system. This happens, sons, as is also the case with the so-called incon­ we have suggested, through the kinds of links sistencies within an individual’s mind. The indi- 40 Zeitschrift für Soziologie, Jg. 4, Heft 1, Januar 1975, S. 34-45 vidual can live quite well with a number of in­ and neither told anyone about it. The reasons consistencies, simply by shifting them around in for this neglect are analyzed by FRIDAY, whose different contexts, by forming ad-hoc or some­ exposition shall be closely followed and serves times more general rules of exception, inclusion well to illustrate the more general remarks ma­ or exclusion, by developing an astonishing re­ de above. Basically, the reasons for the neglect sourceful casuistry of definitions of a situation were that both, HUXLEY and TYNDALL were in ways that avoid cognitive hardships. It is fully engaged in the controversy over sponta­ only under special circumstances - and the la­ neous generation which was, in FRIDAY’S words, boratory conditions under which balance theory “only one battle in a much bigger war” . The experiments operate are among them — that the abiogenesis debate was more or less settled in usual escape-hatches of time and circumstance 1876, when TYNDALL was able to explain away have been closed and the individual is forced to much of the remaining evidence in favour of cope with his inconsistencies. The same holds abiogenesis and as a consequence, the chief op­ for scientific theories: we have learned to live ponent, H.CH. BASTIAN, conceded defeat. From with their inconsistencies which are partially the documentary evidence presented by FRI­ explained away or may go unnoticed or not DAY, it becomes very clear that TYNDALL regar­ cared for most of the time. It is only under ex­ ded the abiogenesis dispute only as symptomatic ceptional circumstances that they become rele­ for the larger struggle between the forces of ma­ vant and that the difference matters (it might be terialism and idealism in science and society, and added here that information has once been des­ himself as a sort of fied marshal for the materia­ cribed in BRUNNER’S phrase as “the difference list forces. TYNDALL was so engaged in “clearing which makes a difference”). away the Bastian fog” , as he called it, that he did not react to HUXLEY’S discovery. It simply Thus, to look for the different uses to which remained an unanswered question. It would have knowledge will be put and to ask how poten­ been troublesome for TYNDALL to turn away tially controversial cognitive structure are tied from his central aim of fighting BASTIAN, just to others and thus to possibly contradictory to clear up the question of mould toxins, which uses of knowledge, gives us some very general must have appeared to him a very peripheral con­ lines of inquiry to follow. What ought to be cern. For HUXLEY too, there were obviously done therefore, is not merely to analyze differen­ other matters to attend to. “And finally”, FRI- ^ ces in the sense of locating them and pinning DAY concludes (1974: 69) the rejection of the them down, but to analyze the kinds of links Pasteurian explanation of the phenomenon in which emerge in the cognitive system of those terms of oxygen deprivation might have left who are involved in a controversy. other findings of PASTEUR in doubt. If so, a tactical advantage would have been won by the supporters of spontaneous generation. TYNDALL IV. Minor incidents and major struggles: Social was unwilling to yield such advantages”. perception in the production of scientific knowledge What the ‘minor incident’ in a major struggle il­ lustrates is not only that the conception of In a recently published account of a little known science of these involved in the debate was a incident in the debate on spontaneous generation political one in the sense of advancing progressi­ which confronted biogenesis with abiogenesis in ve over orthodox thought, as TYNDALL expressed it the 1870’s, FRIDAY (1974) shows how HUXLEY, himself, but also that neither HUXLEY nor TYN­ a major proponent in the debate, incidentally DALL were interested in natural phenomena for discovered in 1875 that a secretion from the their own sakes. They took interest only in so mould penicillium glaucum was able to inhibit far as the phenomena “fitted with their own bacterial growth, unconnected with oxygen de­ materialist view of the world and provided amu- privation. He recorded this observation in his nition for their war with the idealists”. Although notebook and in a single letter to TYNDALL, it is a case history, the case is not exceptional. which has recently been found and served as Only a very naive conception of the production starting point for FRIDAY’S work. Neither HUX­ of knowledge in science or outside science could LEY nor TYNDALL looked for an explanation assume that natural phenomena are looked at H. Nowotny: Controversies in science 41 for their own sake. All knowledge is tied to other uses of the lower level knowledge structures are knowledge, be it in the making, or seeking con­ regulated by the higher ones. firmation. The boundaries of how far the uses of knowledge are to be extended, are flexible and The connective ties can, as in the TYNDALL and malleable. Although there is no way of ascertain­ HUXLEY case, but need not lie outside what are ing what went on in TYNDALL’S and HUXLEY’S in any case rather arbitrarily defined boundaries mind, it is reasonable to assume that they both of science or a discipline. In a penetrating ana­ saw their concerns as legitimate scientific concerns. lysis of a controversy opposing two evolutionary When TYNDALL wrote that he wants “to act the theories in paleontology, GRENE (1958) has part of a conservative rather than a destructive shown the different structures of theories by by gradually preparing the public mind for inevi­ pointing out their different uses as theories. These table changes which without this preparation are different conceptions of what evolutionary might take revolutionary form” , it is not politics theory in paleontology should look for and what that he talks about, but science couched in the kind of theory a theory explaining evolution language of political metaphor. should be. After having drawn attention to the parallelisms between the Neo-DARWINian theory There can be no cognitive structure so isolated advanced by SIMPSON and the theory advanced and turned upon itself that it would not tend to by the Swiss paleontologist SCHINDEWOLF — pa­ link up with others. Even in the case of very co­ rallels which yield “a nicely balanced pair of hesive and closed theoretical systems, like the re­ theories”, as one theory stresses continuity and search programmes described by LAKATOS and the adaptive character of all evolutionary change, ZAHAR, the tendency to link up with others and while the other stresses discontinuity (novelty to be used, is observable. Contemplating the ques­ of forms) and the non-adaptive character of ma­ tion why EINSTEIN’S research programme superse­ jor changes (1958: 113). GRENE proceeds to ded LORENTZ’, ZAHAR asks at one point (1973: show how different levels of widening disagree­ 242). ment appear. The ones on the merely verbal plane could easily be cleared up, only to bring “If the merit both of Kepler and Einstein only con­ forth those on the plane of visual imagery as sisted in ridding physics of unnecessary epicycles, each proponent employs a different visual mo­ then the importance of these two physicists in the hi­ del. On the plane of attention, it is very obvious story of science is very much overrated: Copernicus that what are central concerns in one theory are and Lorentz did all the creative work, and Kepler at best peripheral ones in the other. On the con­ and Einstein only applied Occam’s razor in order to demolish the expendable methaphysical scaffolding used ceptual plane, it finally becomes apparent that by their predecessors. Moreover, Copernicus knew that the opponents stand at different places in rela­ the path of the planets were not circular, hence that tion to their subject-matter, their outlook is his epicycles were part of the scaffolding Lorentz very different. But from these differences emer­ realized that he did not need the Galilian coordinates ge different conceptions not only of what is to in order to deduce the null results which he set out to explain. If so, Kepler and Einstein contributed to the be explained — continuity or discontinuity — economy of thought and not to the growth of knowl­ and what is to be accorded priority as a conse­ edge. This is an unacceptable conclusion. “ quence, morphology or phylogeny — but how, through what kind of theory these differences What is of interest here from our point of view in outlook are to be translated. The DARWINian are the reasons provided by ZAHAR for the re­ explanation denies structure in the sense of reduc­ jection of the conclusion at which he arrived. It ing it to its conditions. It provides an explanation is unacceptable, he argues, because getting rid of of evolution in terms of the mechanistic tradition the COPERNICAN epicycles was not an end in it­ of explanation, i.e. it must be both logically self, but was “subordinate to the needs of the simple and automatic (1958: 188). It is simple new heuristics”. Thus, it are the heuristic devi­ in that it unites the relevant phenomena under ces of the new research programmes which regula­ one hypothesis which is conceived in terms of te and control from a higher level of thought the existence and non-existence. It is automatic since theoretical developments which take place on the its only hypothesis interprets organisms mecha­ lower level. We are presented here with a hierar­ nistically, and their evolution as produced by chical image of scientific connectedness, as the random errors and by natural selection which are 42 Zeitschrift für Soziologie, Jg. 4, Heft 1, Januar 1975, S. 34-45 both mechanistic automatisms. In the SCHINDE- mentally a clash of philosophical world views on WOLFian kind of theory, however, a duality of the nature of God, matter and force (ILTIS 1973: logical levels can be found, as between continu­ 343). The debates were conducted between two ous and small-scale conditions versus discrete and different organizations of knowledge based on comprehensive patterns, which means that the metaphysical and mechanical principles. From explanation must forego its unitary character. ILTIS’ analysis it seems that the factors of social interaction played a major part in colouring the What is of interest from our point of view are perception of the adherents. Emotional commit­ the parallels which emerge between these two ments and loyalities to their masters formed the highly complex and elaborate theories on evolu­ specific social conditions under which the follo­ tion and certain very elementary processes of wers of NEWTON and LEIBNIZ worked and led cognition, such as categorization. What are in them to defend their respective world views by SIMPSON and SCHINDEWOLF’S case procedural reinterpreting the challenging experiments from rules to create structure or to avoid it, to assign the other side in a way that supported their own priorities to different steps which must lead to philosophy. They were unwilling and unable to different procedural routes to be taken and see the other side’s valid arguments. But if social which, as an end product, manifest themselves factors influence perception to such a degree, if in different types of theories, is an elaborate and theological and metaphysical commitments on refined version of certain elementary cognitive the part of the experimenters caused them to processes. In the case of categorization, for in­ interpret the mechanical experiments in a manner stance, elements are divided into groups, the same consistent with their natural philosophies, des­ elements are grouped into classes and people have pite the errors and the lack of conviction in their rules on how clusters are to be formed and on results, as ILTIS demonstrates, then these under­ the meaning of these identical groups (COXON lying cognitive processes should be investigated & JONES: 7ff.). Categorization of any wholistic in their own right, in order to determine the image and producing theories on evolution can procedures upon which they are based and to both be conceived as problem-solving tasks, which discover when they are no longer self-supportive, must be analyzed in their procedures and proce­ but begin to collapse internally or to converge dural rules, if one is to understand their construc­ externally. The recognition that both viewpoints tion. could be valid began to take place some 20 It would be utterly misleading, and the danger years later, when integration between the two exists when one analyzes controversies in science, theories began to occur. Apart from the social to satisfy one’s curiosity after having been able dynamics involved in the processes which contri­ to state individual differences, such as those bet­ buted to a partial solution of the controversy, ween ‘lumpers’ and ‘splitters’ in categorization, we would again expect that what changed were or between ‘Unitarians’ (or mechanistic theorizers) the uses to which the opposing systems were put. and ‘multi-level’ theorizers in science. If such dif­ How could shifts in the evaluation of theories ferences emerge they can merely serve as starting be explained otherwise, even, or maybe, precisely point for the further study of the underlying cog­ in those cases where the incompatability of the nitive processes which lead to different theorizing two viewpoints was uncontested, regardless of and therefore to different world-constructions. höw strong the influence of external and notab­ Our previously advanced hypothesis would again ly social factors might have been in leading to lead us to assume that these differences — in cog­ the perception of incompatibilities. nitive style one is tempted to say — are linked to the further uses of the knowledge thus pro­ Such was the case, for instance, for EINSTEIN’S duced and are part of the overall connectedness equivalence in gravitation theory (WOODWARD patterns of different regions of thought. & YOURGRAU 1973). It is on the historical record that EINSTEIN himself thought of them as being The important part played by perception, and incompatible and we are told by one of MACH’S thus by cognitive processes, in controversies, can biographers (BLACKMORE 1972: 259) that “no also be illustrated by the debate which went on rational person could or should hold both” between LEIBNIZ and NEWTON and their followers (views). Nevertheless, after a generation of phy­ for almost 20 years. The controversy was funda­ sicists had pursued the vision of a unified theory H. Nowotny: Controversies in science 43 without much success between the 1930’s and V. ‘Non-controversies’ and mutual indifference 50’s, a kind of MACH renaissance can be witnes­ sed today. Although it is not clear yet where the This is how I would like to call the numerous onslaught of papers on MACH will eventually occasions where cognitive incompatibilities, in lead, whether there will be a revaluation of varying degrees, exist and are visible either retro­ MACH’S principle and a partial convergence bet­ spectively or to an outsider. Yet they are not ween the formerly incompatible theories, the articulated or have not — or not yet — been ac- development as such points again to processes tivized. The other side, against which contrast­ of cognitive restructuring through changing con­ ing interpretations or results appear to be direc­ nectivity patterns. It is not merely the fact that ted, fails to respond to the challenge, or is simp­ the younger generation of physicists, those who ly unimpressed by the differences. Such occasions have not personally participated in the frustrating are ‘non-controversies’, as they take the place of search for a unified field theory, are courageous a controversy that has not materialized despite enough to tackle the problem, as WOODWARD the cognitive potential being there. The lines of & YOURGRAU suppose, but also the fact that fission that separate the two sides can be as great “since 1960 a shift from relativity theory as a as those that become visible during a controversy, mathematical formal science, divorced from the yet neither side is willing or bothers to commit main stream of physics, to a science drawing in­ itself and to engage fully into a controversy. The spiration from observation and, in turn, stimulat­ difference does not matter. Why not? To be sure, ing observation” has occurred. This general chan­ explanations can always be found, by pointing to ge in direction of the further uses of knowledge the absence of some factor, most probably social may take the form of connecting a theory to in nature, which might account for the failure developments that go on in what is seen as the to light up a controversy. But it might be more mainstream, or by putting the existing relations fruitful, from our perspective, to look at non­ into question. At the same time we are forceful­ controversies as being due to different modes in ly reminded that incompatibility — in itself a the production of knowledge. Although the re­ perceptual phenomenon — is always relative and sults might be similar enough to yield potential may undergo varying evaluation in the course of for disagreement over interpretation and the like, time, not because new knowledge has emerged the methods by which knowledge is produced that could serve as unifying platform for the for­ are apparently too divergent in non-controversies. merly incompatible theories (although this could also happen), but because the outside cognitive One side, for instance, may consider as data what links have been altered. the other side completely fails to consider (this may be one of the reasons for the mutual indif­ The negative test case is provided by all those ference prevailing for instance, between anthro­ controversies which continue without ever fini­ pologists and psychologists, as neither is much shing, truly unexhaustable and intractable. For impressed with the instruments and methods for some reasons, these kinds of controversies appear producing knowledge of the other side, as COLE to occur particularly frequent in the Field of et al. (1971) have observed, or methods and philosophy, where arguments can go on without techniques may be used which are completely successful termination “even, or perhaps specifi­ unacceptable for the other side. Yet, hostility cally, by people who have been carefully trained does not take the place of mutual indifference, in the technique of logical thought” (CRAWSHAY- as there is no interference in the way how each WILLIAMS 1957: 3). In his analysis of a series of side produces knowledge it considers valid. There such intractable controversies, CRAWSHAY-WIL- is a cognitive overlap on the phenomena to be ex­ LIAMS reaches the conclusion that their charac­ plained, but it is neatly separated at the same teristic nature is their indeterminancy with respect time through the whole set of procedures, stan­ to context or purpose. They cannot be terminat­ dards, methods, concepts and techniques, which ed, we can translate, because they are not tied to consistently mark the mode of production of any specific use, because they have been cut off knowledge of each side. Interference arises, we from being connected to other cognitive structu­ would suggest, whenever the mode of produc­ res. Thus, they are merely shifted around and tion of knowledge itself is called into question, around, aloof in their cognitive unconnectedness. which is likely to occur only then, when the 44 Zeitschrift für Soziologie, Jg. 4, Heft 1, Januar 1975, S. 34-45

knowledge produced is put to uses at contradic­ edness in information processing has been shown tory ends. to be present already in the most elementary cog­ nitive processes, and manifests itself there as From the existence of non-controversies we may ‘foci of attention’ or ‘critical configurations of infer that it might be quite useful to analyze con­ stimuli’ and the like. They are operative in scien­ troversies and non-controversies alike from the ce, where they manifest themselves in controver­ more general point of view of knowledge produc­ sies. It is for this reason that controversies are tion as information processing (NEISSER 1967, endemic and inevitable in any production of KLIX 1971). In line with the application of some knowledge, including the production of scienti­ general concepts of information processing to fic knowledge. this specific area of knowledge production, theo­ ries could be compared to the cognitive mecha­ Accumulation of knowledge is therefore an illu­ nisms or “schemes” as they are sometimes called sion if it is meant in the sense of amassing, pos­ (NEISSER 1967), through which data that have sessing and storing more and more knowledge. been received as input stimuli, are recalled and This could merely be the case for input stimuli, transformed and thus processed to serve the pro­ which are only potentially useful in view of their duction of additional knowledge. Disagreements further processing occuring through organized will emerge whenever the “data” are processed ‘schemes’, notably through theories with their in different ways, which implies that they will operative research practices. Which organization be used — still in the realm of cognition — for schemes are better, however, depends on what different ends (NEISSER 1967: 286ff.). knowledge is to be produced for, including the extremely important cognitive links to be esta­ This view has also consequences for the so-called blished to other forms of knowledge production. accumulation of knowledge. What can be accu­ The success of the Western type of science pro­ mulated, in the sense of amassing and storing, is duction seems to depend partly on unification merely input information. This input, however of purpose and the wide-spread acceptance of a its size or content may be measured, is clearly general sense of direction, at least until recently. of minor importance in the overall production of In addition, Western science is characterized by knowledge, compared to the processing mecha­ a relatively high internal cognitive cohesion, mean­ nisms. For science, these processing mechanisms ing that the cognitive links within different re­ are essentially theories with their research para­ gions in scientific though are relatively well tied phernalia. They constitute the scarce factor in to each other, while there is a low degree of cog­ the overall production process, as they impose nitive cohesion and only weak ties to other forms limits on what can be done with the received of knowledge production outside science. This and stored input information. They are the ones makes it possible to regard science as another that create the typical ‘bottleneck’ situations in form of eco-system, beset by its own contradic­ the development of knowledge. Conflicts are tions which manifest themselves above all between therefore likely to emerge over what is to be the amount of energy needed for its maintenan­ adopted as the appropiate way of processing in­ ce and the energy needed for innovative proces­ put information, i.e. how it is ‘best’ to be organi­ ses (BLACKBURN 1973). Controversies emerge at zed. This again however, depends on some kind those strategic locations where not only the choi­ of end that is to be achieved. Although it may ce between alternative processing schemes of knowl­ be assumed that some processing schemes are in­ edge is given, but where the choice is linked to herently superior to others, because they unite a other cognitive structures and forms of knowledge greater number of acknowledged advantages, there production. Non-controversies seem to signal that are many others for which clear-cut criteria of the form of knowledge production itself is — their respective advantages cannot be found, nor still or already? - differentiated. hopefully be developed. This situation is the re­ sult of the heavy dependence of any scheme of Finally, we do not know what an analysis of the knowledge production on its later uses: this holds empirical patterns of variation in controversies as for the processing of elementary visual stimuli, well as non-controversies, when compared over for the classification of birds and for any scienti­ time and across disciplines or sub-fields, would fic organisation of knowledge as well. The direct- yield and how these empirical variations are to H. Nowotny : Controversies in science 45 be interpreted. Are there more or less controver­ KLIMA, R., 1973: Bemerkungen zur Theorie der ,scien­ sies, more or less non-controversies in science tific community4. Sozialwissenschaftlicher Ausschuß today, when compared with 20, 70, or 150 years des Vereins für Sozialpolitik. Unveröffentlichtes Ma­ nuskript. ago? And if there are more, or less, what does KLIMA, R., 1974: Sind Experimentalregeln »soziale Ver­ it mean? bindlichkeiten4. Zeitschrift für Soziologie 3, Heft 1, 103-106. These are only some of the many questions KLIX, F., 1971: Information und Verhalten. Bern, Stutt­ which this essay on different modes in the pro­ gart, Wien: Verlag Hans Huber. KUHN, Th., 1962: The Structure of Scientific Revolu­ duction of knowledge in science hopes to have tions. Chicago 111.: University of Chicago Press. raised. LEACH, E., 1964: Anthropological Aspects of Language: Animal Categories and Verbal Abuse. In: E.J. Lenne- berg (ed.), New Directions in the Study of Language. Cambridge, Mass.: MIT Press. MERTON, R.K., 1966: Conflicts over Styles of Socio­ logical Work. In: R.K. Merton, The Sociology of Science. Chicago, 111.: The University of Chicago Press. MERTON, R.K., 1973: The Sociology of Science. Chica­ References go, 111.: The University of Chicago Press. MULLINS, N., 1968: The Distribution of Social and BEN-DAVID, J., 1973: Probleme einer soziologischen Cultural Properties in Informal Communication Net­ Theorie der Wissenschaften. Unveröffentlichtes Ma­ works among Biological Scientists. American Socio­ nuskript. logical Review, vol. 33, 786-797. BERRY, J.W., & P.R. DASEN (eds.), 1974: Culture and NEISSER, U., 1967: Cognitive Psychology. New York: Cognition. London: Methuen & Co. Ltd. Appleton-Century-Crofts. BLACKMORE, J.T., 1972: Review of Ernst Mach: His NOWOTNY, H., 1973: On the Feasibility of a Cognitive Work, Life and Influence. Berkeley, Calif.: Univer­ Approach to the Study of Science. Zeitschrift für So­ sity of California Press. ziologie 2, Heft 3, 282-296. BLACKBURN, Th., 1973: Information and the Ecology NOWOTNY, H. & M.E.A. SCHMUTZER, 1974: Gesell­ of Scholars. Science 181, 1141-1146. schaftliches Lernen. Frankfurt a. M.: Herder & Her­ BÖHME, G., 1974: Die Bedeutung von Experimentalre­ der. geln für die Wissenschaft. Zeitschrift für Soziologie ROSE, H. & St. ROSE, 1974: Do not Adjust your Mind: 3.1, 5-17. There is a Fault in Reality. In: R. Whitley (ed.), Social BULMER, R., 1967: Why the Cassowary is not a Bird. Processes of Scientific Development. London: Rout­ In: M. Douglas (ed.), Rules and Meanings, 1973. ledge and Kegan Paul. Harmondsworth: Penguin Books, Ltd. SCHMUTZER, M.E.A., 1974: Social Crystallization. Es­ COLE, M. et al., 1971: The Cultural Context of Learn­ sex University. Unpublished Ph. D. Thesis. ing and Thinking. New York: Basic Books. STÖRER, N., 1973: Introduction to R.K. Merton, The COXON, A.P.A. & C.L. JONES, 1974: Occupational Sociology of Science. Chicago, 111.: University of Chi­ Categorization and Images of Society, Working Paper cago Press. no. 4, Project on Occupational Cognition. Edinburgh: TAMBIAH, S.J., 1969: Animals are Good to Think and Edinburgh University Press. Good to Prohibit. Ethnology 8, no. 4, 424-459. 9. CRANE, D., 1972: Invisable Colleges. Chicago & London: TOULMIN, St., 1972: Human Understanding, vol. 1. The University of Chicago Press. Oxford: Clarendon Press. CRAWSHAY-WILLIAMS, R., 1957: Methods and Cri­ TYLER, St. (ed.), 1969: Cognitive Anthropology. teria of Reasoning. An Inquiry into the Structure of New York: Holt, Rinehart & Winston. Controversy. London: Routledge and Kegan Paul. WHITLEY, R., 1973: Competition, Division of Labour DOUGLAS, M., 1966: Purity and Danger. London: and Autonomy in Scientific Development. Manchester Routledge and Kegan Paul. University. Unpublished Paper. FRIDAY, J., 1974: A Microscopic Incident in a Monu­ WOODWARD, J. & N. YOURGRAU, 1973: The Incom­ mental Struggle: Huxley and Antibiosis in 1875. The patibility of Mach’s Principle and the Principle of British Journal for the History of Science 7, no. 25, Equivalence in Current Gravitation Theory. British 61-71. Journal of Philosophy of Science 24, 264-270. GRENE, M., 1958: Two Evolutionary Theories. British ZAHAR, E., 1973: Why Did Einstein’s Programme Super­ Journal of Philosophy of Science 9, 110-127, 185- sede Lorentz’s? British Journal of Philosophy of 193. Science 24, 223-262. ILTIS, C., 1973: The Leibnizian-Newtonian Debates: Natural Philosophy and Social Psychology. The British Journal for the History of Science, Part IV, Anschrift der Verfasserin: HELGA NOWOTNY, Ph. D. no. 24, 343-377. A 1237 Wien, Breitenfurterstraße 510