Professional Documents
Culture Documents
Editorial Board:
G. Bohme, Technische Hochschule, Dannstadt
N. Elias, Universities of Leicester and Bielefeld
Y. Elkana, The Van Leer Jerusalem Foundation, Jerusalem
L. Graham, Massachusetts Institute of Technology
R. Krohn, McGill University, Montreal
W. Lepenies, Free University of Berlin
H. Martins, University of Oxford
E. Mendelsohn, Harvard University
H. Nowotny, European Centre for Social Welfare Training
VOLUME VI - 1982
SCIENTIFIC
EST ABLISHMENTS
AND
HIERARCHIES
Edited by
NORBERT ELIAS
Universities of Leicester and Bielefeld
HERMINIO MARTINS
University of Oxford
and
RICHARD WHITLEY
Manchester Business School, University of Manchester, England
e-ISBN-13: 978-94-009-7729-7
TABLE OF CONTENTS
Introduction
vii
xiii
PART I
Scientific and Other Establishments
N. ELIAS - Scientific Establishments
P. WEINGART - The Scientific Power Elite - a Chimera; The De-
3
71
Historical Analysis
89
Academy of Sciences
111
PART II
Establishments and Hierarchies in the Development of
Scientific Knowledge
E. YOXEN - Giving Life a New Meaning: The Rise of the Molecular
Biology Establishment
HOHLFELD - Two Scientific Establishments which Shape
the Pattern of Cancer Research in Germany: Basic Science and
Medicine
J. FLECK - Development and Establishment in Artificial Intelligence
A. RIP - The Development of Restrictedness in the Sciences
T. SHINN - Scientific Disciplines and Organizational Specificity:
the Social and Cognitive Configuration of Laboratory Activities
123
R.
145
169
219
239
vi
Table of Contents
PART III
Establishing Boundaries and Hierarchies in the Sciences
R. G. A. DOLBY - On the Autonomy of Pure Science: The Con-
struction and Maintenance of Barriers between Scientific Establishments and Popular Culture
D. CHUBIN and T. CONNOLLY - Research Trails and Science
Policies: Local and Extra-Local Negotiation of Scientific Work
R. D. WHITLEY - The Establishment and Structure of the Sciences
as Reputational Organizations
313
Index
359
267
293
INTRODUCTION
In recent years sociologists of sciences have become more interested in scientific elites, in the way they direct and control the development of sciences
and, beyond that, in which the organization of research facilities and resources
generally affects research strategies and goals. In this volume we focus on
scientific establishments and hierarchies as a means of bringing aspects of
these concerns together in their historical and comparative contexts. These
terms draw attention to the fact that much scientific work has been pursued
within a highly specific organizational setting, that of universities and academic research institutes. The effects of this organizational setting as well as
its power relations, and its resources in relation to governmental and other
non-scientific establishments in society at large, deserve closer attention.
One significant aspect of scientific establishments and hierarchies and of
the power relations impinging upon scientific research, is the fact that the
bulk of leading scientists have the professional career, qualifications and
status of a professor. As heads or senior members of departments, institutes
and laboratories, professors form the ruling groups of scientific work.
They are the main defenders of scientific - or departmental - autonomy,
accept or resist innovations in their field, play a leading part in fighting
scientific controversies or establishing consensus. Even where research units
are not directly controlled by professors, authority structures usually remain
strongly hierarchical. These hierarchies too deserve attention in any exploration of the social characteristics of scientific knowledge and its production.
Not only do heads of institutes and departments often direct and control the
careers of their scientific staff, but they also wield considerable power in the
evaluation of scientific research and its practitioners. As most scientific work
is conducted in a hierarchical setting whose apex is formed by a more or less
authoritative scientific establishment, the significance of this social structure
for the development of scientific knowledge is considerable.
Scientific establishments and hierarchies within departments and research
institutes need to be considered in the wider context of hierarchies between
departments and scientific disciplines. Different departments vary in the
vii
Norbert Elias, Herminia Martins and Richard Whitley reds.), Scientific Establishments
and Hierarchies. Sociology a/the Sciences, Volume VI, 1982. vii-xi.
Copyright 1982 by D. Reidel Publishing Company.
viii
Introduction
ix
also heralded major shifts in the way research was organized, administered,
and funded .. State direction of research has, of course, now become commonplace but its organizational implications are not always realized. In particular,
the development of full-time research laboratories which are partly oriented
to non-intellectual goals has led to the formation of new fields which transcend traditional academic boundaries and values. Furthermore, funding
patterns now affect intellectual development to a greater extent and more
directly. They are often short term, leading to greater variety of intellectual
goals and a strong emphasis on "efficiency" in the administration of research.
This in tum increases the division of labour and specialization as scientists
strive to acquire reputations in a narrow, highly specific area.
A major part of the dominant belief system in science is, of course, the
need for greater precision and control over phenomena. This is discussed by
Rip in terms of the "restrictedness" of objects in Chemistry which enabled
that science to develop standardized samples and techniques. Such standardization encouraged the extensive division of labour in modem chemistry
which is described by Shinn in his comparison of laboratories in different
sciences. As he shows, research goals differed in this fields, as did the degree
of control over materials - or tasks uncertainty - and these variations were
clearly connected to the authority and communication structures of the work
organizations.
The connections between differences in laboratory organization and
differences in intellectual structures are not always easy to discern - and in
many way have become more complex since the development of multi-goals,
multiple-funded, full time research laboratories. A number of studies have
emphasized the importance of "local organizational" factors in the development of research strategies, especially in the biological sciences, but few have
directly addressed the issue of how such strategies coalesce - or do not
coalesce - into coherent social groups around particular intellectual goals and
procedures. Chubin and Connolly suggest the metaphor of research "trails"
bunching to form specialties in particular periods and the utility of different
evolutionary models in dealing with this problem.
The dangers of over reifying social boundaries have, of course, been well
rehearsed in sociology as a whole, and this had led some writers to an extreme
voluntarism. However, the development and maintenance of social and
intellectual boundaries are themselves social strategies which have intellectual
Introduction
xi
Bielefeld
R. D. WHITLEY
Manchester Business School
PART I
SCIENTIFIC ESTABLISHMENTS*
NORBERT ELIAS
Zentrum fiir Interdisziplinare Forschung, Bielefeld
1. Introduction
The social characteristics of scientific establishments are bound up with those
of the social institutions where most of them are located - with those of
universities. Occasionally, though, in the development of sciences, extra
mural establishments have arisen, especially during the earlier phases. Among
the best known examples are the Paracelsians and the Freudians - both
representatives of medical sciences which may be significant. Of course,
the interplay, and particularly the battles, between intra mural and extra
mural establishments, like their distinguishing characteristics, deserve more
attention: But the simple reference to extra mural establishments alone is
enough to put the problem of the relationship between universities and
sciences into better perspective. Whatever the contributions of extra mural
establishments to the development of sciences have been in the past and may
be at present, university-related groups have .gained the ascendancy - in the
development of almost all higher branches of learning they play the dominant
part.
During the Middle Ages universities in Europe formed centres of higher
learning at its pre-scientific, mainly theological and philosophical, stage where
it was linked to the authority of revealed knowledge. In the Renaissance
humanist groups, still tied to authority - that of Greek and Roman antiquity
-- established themselves in some European universities as powerful, model
setting groups of learned men, thus replacing the Church controlled by
court- and state- controlled establishments. Then, with the take off into
science during the 15th, 16th, and 17th centuries, natural scientists, together
with science-oriented philosophers and mathematicians, slowly made their
way into the universities and established themselves there. Finally, university3
Norbert Elias, Herminio Martins and Richard Whitley reds.), Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 3-69.
Copyright 1982 by Norbert Elias.
Norbert Elias
Scientific Establishments
Norbert Elias
Scientific Establishments
the type of abstraction which dominates the thinking of metaphysical philosophers about science and human knowledge. From the philosophers'
perspective these two appear as pieces of nature to be caught like all others
in a conceptual net of static and law-like generalizations beyond time. Yet
sciences are not part of physical nature. To pretend that they are is a futile
exercise. One cannot hope to discover universal laws, similar to those discovered by classical physicists, of something that is in no way universal, that
is in fact the result of a long diachronic social process. Philosophical sciencetheory with its law-like generalizations obscures the fact that sciences form
part, not of physical nature, but of the human level, of what one often calls
"culture", of the symbolic universe of human beings. Its firmly-structured
development is as inaccessible to static abstractions such as "eternal laws",
"validity", or "truth", as to unstructured historical descriptions. It lies
beyond the reach of both philosophical absolutism and historicist relativism.
When Galileo discovered regularities of falling bodies by studying the
movement of downward rolling balls with the help of a clepsydra, he could
rely on a vast fund of human knowledge, which gave him the certainty that
the regularities of falling bodies, which he observed there and then and which
he concisely represented as a general rule with the help of a few mathematical
symbols, would be the same the world over. When Descartes or, for that
matter, Kant tried to examine fundamentals of human knowledge by introspectively examining their own knowledge, they made the same assumption.
They assumed, as a matter of course, that concepts at a very high level of
generality, which they found in their own possession - concepts such as
"reason", "natural law", "mechanical causation" or "substance" - must be
universal properties of human beings everywhere. They assumed, in short,
that these concepts formed part of the nature of men. This assumption,
however, was fictitious. Philosophers had learned these concepts, and others
of the same type, with their language as part of the common conceptual
stock -in -trade of their society or, at least, of the stock-in-trade oflearned men
of their time. Had they lived a couple of hundred years earlier, they could not
have used, and not reflected upon, the same concepts because they were not
available in their society. Even if the words "reason" "nature", "natural law",
or "cause" were in use, their meaning in that age - and thus the concepts were different. The assumption that immutable laws discovered in a particular
case are the same universally, which was reasonably well founded in the case
Norbert Elias
Scientific Establishments
in their society as part of its conceptual and linguistic heritage. Kant, in fact,
had learned the concept of cause in this manner, together with many others of
the same kind. It was prior to his individual experience because it existed in
his society prior to his own existence. In his society he learned the concept of
mechanical and impersonal causation from his elders. In his reflections, however, he disregarded the fact that he had learned from others the word "cause"
and its meaning in the specific sense it had attained at the time he was born.
As learning forms part of a person's experience, he disregarded the fact that
concepts such as "cause" came to him through experience in that sense.
According to current knowledge, the metaphysical idea that the concept
of mechanical causation and other similar concepts of connection are built
into people's unchanging nature prior to all experience, can only mean that
this type of relationship-concept is genetically implanted in man and thus
transmitted from one generation to another without learning as part of man's
biological constitution. In actual fact, these concepts developed in a society
over time as part of its developing fund of knowledge. They came into being
through a continuing process of intergenerational conceptualization of
experiences. They were transmitted from one generation to another, not
genetically but through learning, and changed - advanced or declined in
scope, precision, and fitness - according to their double function as a means
of communication and a means of orientation; they changed as a result of
changes in knowledge and collective experiences of such societies which, in
turn, were dependent on antecedent conceptualization.
Systematic enquiries into long-term processes of conceptualization as
well as theoretical models of such processes, are still in their infancy. That
the Greeks used a word with a meaning akin to our "guilt" (aida) as term
with a meaning akin to our "cause", no doubt, is suggestive of a process 'of
coceptualization leading from a more personalized and involved to a more
impersonal and detached stage (2), from the quest for someone who is to
blame to the quest for something that acts without purpose, according to
recurrent rules. There are many 'other examples of long-term processes of
conceptualization all pointing in the same direction. To study and to explain
processes of this kind would certainly fill a significant gap in our knowledge.
Philosophy, however, tends to block rather than to encourage enquiries of
this kind. The philosopher's assumption that humans have, as it were, innately
specific rules of experience prior to any experience, makes it impossible for
10
Norbert Elias
them to escape from the notion of such a barrier interposing itself between
"object" and "subject". This conception of rules is the barrier. As long as one
believes in transcendental rules or forms, and regards their discovery as the
principal task of a philosophy of knowledge in general, and of sciences in
particular, the latter, human knowledge - non-scientific and scientific must appear as permanently patterned by people's own constitution. On a
transcendentalist view, human experience, for all its variety, seems to be set
for ever into a rigid and unvarying mould.
The strong solipsistic tendency underlying that kind of philosophy has its
intellectual roots in this assumption. Kant gave expression to this tendency
when he spoke of the significant doubt as to whether the object "which
we suppose to be outside of us may not perhaps be within us" (3). Similar
solipsistic doubts can be found in the writings of other transcendental philosophers, among them Husserl and Popper. However much they differ in other
respects, sooner or later they sound a note of despair.
One may well ask why this odd kind of philosophy, from Descartes to
Husseri and beyond, has gained so strong a position in the world of learning
which is, perhaps, even stronger in the 20th century than ever before. The
question deserves a more extensive enquiry. Off-hand one can refer to a few
of the interdependent conditions which contribute to the ascendancy of
transcendental philosophy in the wQrld of academic learning and to the firm
hold which metaphysical beliefs have gained over science theories in our time.
One of these conditions, as I have indicated elsewhere (4), is the strong
civilizing spurt which set in during the 16th and 17th centuries and which,
with many cross- and counter- currents and with many swings of the pendulum, continues to this day. It made for a stronger armour of individual selfrestraint interposing itself, all round, more evenly and stably between people's
spontaneous impulses to act and the execution of any action. The continued
growth and the growing effectiveness of state control and specific changes in
the mode of control, such as the change from almost exclusive reliance on
direct command and constraint to greater reliance on indirect methods of
control, contributed to this civilizing spurt. So did the increasing, lengthening,
differentiation and stability of interdependence-chains binding individuals
to each other. A complex of social changes in this direction offered social
rewards in terms of status, career, power chances or income to men capable
of containing their short-term impulses in the pursuit of long-term aims.
Scientific Establishments
11
The ascent of sciences and the scientification of society, too, were closely
connected with this civilizing spurt.
Take the transition from a geocentric to a heliocentric view of the universe,
which was one of the most momentous achievements of early science. The
geocentric view, even in its most sophisticated form, was characteristic of
people's unreflected self-centredness which dominated the long sequence of
pre scientific stages. The heliocentric view marked the ascent to a level of
consciousness at which humans were able, in their reflections, to look at themselves as it were, from a distance. The Copernican innovation implied, not
merely an advance in people's substantive knowledge, but also the ascent to a
higher level of restraint and self-distanciation; in short, a specific change in
people's personality structure. It required a lessening of their involvement in
the experience of natural events or, expressed differently, a greater capacity
for detachment. They had to forgo the pleasure of experiencing themselves,
humans on this earth, as the centre of the universe and to accept instead, an
emotionally far less appealing, although better orientating, i.e., more realistic,
world view.
A higher degree of individualization was another aspect of the same
transformation. Increasing differentiation of society and the stronger social
pressure for individual self-restraint, threw people back on their own resources. At the same time, the greater measure of self-restraint in more and
more situations, required of people by the larger and increasingly complex
figurations they formed with each other, was, in self-reflection, experienced
by many of them as an invisible wall separating them from other objects and
persons.
Here lay some of the strongest social and emotional roots of the homo
clausus feeling and the solipsistic tendencies that went with it. Transcendental
philosophy reflected this experience. It gave expression to it in the form of
the paradigmatic model underlying its knowledge theory - of the model
according to which an existential barrier interposed itself between the "subject" of knowledge and its "objects", between the ideas ''within'' and the socalled "external" world'.
The intellectual construction giving expression to this feeling of a barrier
and a distance between the "world within" and the "world without", the
assumption of an unlearned and immovable panel of rules and forms, built
into men prior to all experience, was all the more inflexible as it had a very
12
Norbert Elias
high value for philosophers. Their whole social and professional existence as
philosophers was bound up with it. With few exceptions, philosophy to this
day embodies a specific hierarchy of values which dominates the cast of its
problems as well as of solutions that appear philosophically relevant. It is a
value scheme which philosophers share with theologians and classical physicists. Transcendental philosophy, in particular, is "set" at discovering behind
the changing knowledge and experience of humans, something that is not
changing - something eternal and immutable. The value scheme underlying
this design of philosophical problems and of the forms of thinking that go with
it, may appear simple and perhaps self-evident. According to it, that which is
unchanging and eternal, as an object of enquiries has a much higher value, than
that which is changeable and impermanent. The implied assumption, however,
is far from self-evident. If people in quest of knowledge, allow themselves to
be guided by this value scheme, they implicitly assume that whatever they try
to explore, is itself arranged in a specific way. The value statement, on closer
inspection, reveals itself as an ontological statement. According to it, the
understanding and explanation of all that is changing can and must be found
in something behind it that is timeless, immobile, and immutable. The value
scheme is linked to a substantive view about the nature of things; it has, in
other words, ontological significance. Concealed in it is a statement - or a
belief - about the existing structure of the world. The classical philosophers'
and the classical physicists' concept of explanations in terms of eternal laws
or lawlike statements corresponds to this implied ontological belief.
There is no need to assume that this belief and the implied value scheme
are the result of a clear cognitive decision. It is more likely that this value
scheme reflects specific wishes and experiences of people themselves, not the
structure of their objects. Pride of place among these wishes probably takes
men's desire to find relief from the awareness of their own transience, the
wish to discover behind the impermanence of their lives, symbols of something that is unchanging and imperishable. It is an old tradition that among
the groups who specialize in the management and production of means of
orientation, those who cater for this desire are most honoured among humans
and enjoy the highest prestige. Transcendental philosophers belong to these
groups. They aim at providing people with an orientation which can give them
the assurance of something death-defying and eternal. As physicists reveal
eternal laws behind the observable flux of natural events, so meta-physicists
Scientific Establishmentr
13
14
Norbert Elias
Scientific Establishments
15
reduction", a reduction of that which is changeable to something unchangeable and eternal, reveals itself on closer inspection as a philosophical trap.
If such unchanging forms of people's intellectual activities really existed, they
would form part of their natural equipment. Their existence would cast
doubt on one's ability to advance the congruence of one's symbolic representations with the "objects", the relationships they represent. One thus gets a
clearer view of the basic paradox of. transcendental metaphysics. Idealistic in
appearance, it is naturalistic in fact. Kant still could, in all innocence, identify
nature and reason, for the concept of human nature, in his time, was not as
firmly linked as it is today to the research programme of natural sciences.
Biology in particular was not as advanced as it is today. In our time statements
about aspects of humans which are unchanging, common to all and part of
their nature, can only be understood as something rooted in one's biological
constitution. A philosophical quest for immutable and universal conditions
of human experience or knowledge, today, inevitably has a biologistic ring.
The philosophers' dilemma, as one may see, is that they cannot easily 'revise
their basic aim without stepping outside of what they themselves regard, and
what is regarded by others, as specifically philosophical.
There is an obvious way of escape from the impass where, for centuries,
transcendental philosophers have found themselves trapped. That way,
however, is closed to them. They cannot use it without losing their identity.
They are like people enclosed in a room from which they try to escape.
They try to unlock the windows, but the windows resist. They climb up
the chimney, but the chimney is blocked. Yet, the door is not locked; it is
open all the time. If they only knew it, they could easily leave the room. But
they cannot open the door, because to do so would disagree with the rules of
the game which they as philosophers have set themselves. They cannot open
the door, because that would not be philosophical.
The scientific mode of acquiring knowledge as a philosophers' objective
can serve as an illustration. Science is -a fairly recent stage in the development
of human knowledge. To attempt the construction of a science theory by
abstracting from so short a period in the development of human knowledge
what pretend to be immutable universals of the human intellect, is an
interesting intellectual game - but not more. Inevitably it leads into the
solipsistic impass. If one implies that rules of scientific rationality or of
experience have been implanted into human independently of all experience,
16
Norbert Elias
one cannot escape the conclusion that they form a barrier between "subject"
and "object" and limit once and for all human adaptive capacity - the
capacity for developing perceptions and observations of detail in the light of
new integrating concepts and theories, and integrating concepts and theories
in the light of new perceptions and observations of details.
One might be less sure of that possibility if past developments, including
the take-off into science itself - first of the physical, then of the biological,
and now of the human sciences - would not provide telling examples of
such double-harness advances in the development of conceptual symbols
and experiences, of theoretical syntheses and explorations of details. So
far, no limits of these developments are in sight. There are only manmade blockages, among them those produced by process-reducing theories
of knowledge and of science. No adequate orientation for those engaged in
scientific work can be expected from any science-theory which does not
account for the continued advance which sciences can make over long periods
and which is, after all, the centre-piece of all scientific work. In fact, the
difference between a sociological and a philosophical theory of science can
be summed up by saying that the latter is geared to untestable conceptual
symbols beyond time such as "eternal truth", while the former is focussed on
testable concepts such as "advance in knowledge", whose frame of reference
is a diachronic long-term process, a long chain of generations where specific
advances made by earlier generations are the condition of those of later
generations. The model of a diachronic sequential order thus replaces, as the
heart of a science-theory, the abstraction of timeless law-like universals which,
at the most, become auxiliaries. Nor is it possible to confine attention to
advances in knowledge made during the last three or four hundred years, i.e.,
to scientific advances. One can go further and say: No adequate orientation
for those engaged in scientific work can be expected from any science-theory
which does not explain how and why the production of human knowledge
from being non-, pre- and proto-scientific has become scientific over a widening range of problem-fields and why it has taken the form of a production by
an increasing number of specialized science establishments. Only by working
out theoretical models of this long knowledge process and, as part of it, of
the long take-off into science with the following multiplication of science
branches can one escape from the solipsistic trap. But in order to do that,
one has to give up the fictitious assumption that any scientist goes about
Scientific Establishments
17
his business starting from scratch. One has to consider as relevant to one's
theory the demonstrable fact that all scientists start their work with attempts
to solve problems resulting from previous advances in knowledge, from a
structured sequence of earlier problems and problem solutions.
If knowledge, non-scientific and scientific, is seen as an intergenerational
process, if every individual problem-solver is seen to stand on the shoulders
of others, instead of being considered as an individual in a vacuum, then the
door of the philosophers' closed room can be easily opened. Then it becomes
possible to approach the central problem, which has occupied metaphysical
philosophers at least since Descartes, in a different manner - the problem
of how human symbolic representations "within" can fit, or can be made to
fit better, events in the world "without". In that case this problem loses
the character of an infernal trap where human beings run around in circles from
"subjectivism" to "objectivism" and back; it ceases to be a matter of philosophical speculation and becomes a matter of theoretical and empirical
research in the field of the human i.e. social sciences.
3. Philosophy and the Problem of Time
Reference to one example may be enough, briefly to illustrate the change.
If one studies the problem of time in the philosophical manner, it remains
insoluble. Opinions have changed from an objectivist view of time (Newton)
to a subjectivist (Leibniz, Kant) and back. By and large, the philosophical
subjectivists appear to have gained the upper hand. Accordingly, time is
widely regarded as a form of apperception a priori, as a mode of experience
laid on in the human intellect prior to all experience and thus unlearned. One
cannot imagine, in other words, that oneself could ever experience events
without reference to a tightly knit framework of time measurement, such as
hours of the day or the sequence of calendar years. In actual fact, however,
this specific time experience is bound up with a stage of social development
at which societies could not function without a differentiated and firmly
institutionalized framework of time measurements - a society with long
interdependence-chains binding the social functions of many thousands of
people to each other and thus requiring very close coordination of their
activities in terms of time. It is well known that people ofless differentiated
societies neither possess nor need timing devices of our kind. Most of them,
as long as they are left to their own resources, lack time concepts at the high
18
Norbert Elias
level of generality and synthesis to which we are used, among them the high
level concept of time itself. Among the greatest difficulties one encounters, if
one ever attempts to fit members of less differentiated societies into the
temporal rhythm of an industrial society, is the former's incomprehension of
the social demands made on them in terms of punctuality and the impersonal
time regulation of their watches and clocks. Members of simpler societies are
often not able to adjust to the regulatory demands which time makes, as a
matter of course on members of more complex societies. The latter in tum
are hardly aware how profoundly their own personality is structured in
accordance with the regulatory demands of their society's institutionalized
timing devices. The self-regulation in terms of the regulatory demands of their
social time becomes an almost inescapable habit, a kind of second nature.
That is the reason why many people structured in this manner, transcendental
philosophers among them, are inclined to think that their own time experience
must be, in fact, part of human nature, an unlearned property of humanity
and independent of all experience.
A theory of civilizing processes is necessary to gain enough distance in
relation to one's own personality structure to recognize its connection with
the structure of the society where one grew up, and to perceive both, not
simply as something which exists, but as something which, in the course of a
lengthy development, has become what it is. A theory of the development of
the kind of concept represented today by the concept "time" (6) and of the
wider social development related to it, forms part of the theory of civilizing
processes. In Simpler societies, no differentiated and ubiquitous self-regulation
in terms of time is required. The early forms of what we call timing are
discontinuous and intermittent, mainly passive in character, and geared to
pointlike events, such as the appearance of the sun at a particular spot in the
sky, and to events which one can directly observe here and now, such as the
coming of a new moon. The ancestral forerunners of our concept "time" are
cncepts representing a low-level synthesis such as: "We start sowing when this
kind of bird appears." They are characteristic of pre-state societies such
as bands or self-ruling villages. Most early-state societies develop unifying
timing devices and corresponding time concepts at a higher level of synthesis.
As social units increase in size, differentiation, complexity and levels of
integration, timing devices, and concepts, change accordingly. In the more
developed industrial nation states, timing devices have the character of a
Scientific Establishments
19
20
Norbert Elias
Scientific Establishments
21
Broadly speaking, one can say that the function of timing is that of
coordinating and connecting the one-after-another aspects of two or more
continuous sequences of change, one of which usually serves as a reference
sequence and as a means of determining the relative positions or intervals
between them within the sequential changes of the other or others. Thus,
a recurrent sequence of social activities may be timed in terms of recurrent
positions of the roving sun on his way through the skies; the change continuum of a living person may be timed in terms of the sequence of years in
the reign of kings; and the one-after-another sequence of a person's work and
leisure may be timed in terms of the little mechanical change-continua which
we call watches.
However, at an earlier stage of social development, people have few means
of coordinating and connecting change-continua in this manner and, by and
large, fewer needs to do so. Hence, the actual social practices of timing and
the corresponding concepts change over time - they change in the specific
direction I have outlined before. A small-scale, point-like, relatively loose and
intermittent ad hoc synthesis transforms itself into a highly differentiated
continuous, permanent and very wide ranging synthesis, embracing what we
call "nature" as well as "society" and represented by our concept of time.
The difference between a philosophical and a sociological theory of time
illuminates the difference between a philosophical and a sociological theory
of knowledge. It also indicates why it is pOSsible, by means of a sociological
approach, to fmd a testable solution of problems which philosophers, for
centuries, have tried to solve in vain. In transcendental philosophical terms
time is made to appear, explicitly or not, as representative of a synthesis a
priori. That is a speculative hypothesis not grounded on evidence and easy
to refute. According to it, one is compelled by nature to perform a synthesis
of events, i.e., to connect them, in terms of time. As a rule, philosophers do
not explain why human nature should have evolved in that specific way. The
mystery which they set out to solve becomes only heightened by the solution
they offer. The sociological approach, sketched out above, provides a testable
answer to the apparent mystery. It requires a higher degree of self-distanciation - it only makes sense if one is able to emancipate oneself from the
assumption that one's own way of perceiving events in terms of time must
be a universal way of all humans. Once one is able to free oneself from this
illusion, time ceases to be a riddle. The great mass of material available from
22
Norbert Elias
earlier ages as well as from contemporary societies, shows a clear developmental pattern in the direction I have briefly outlined above. The naturalistic
metaphysical hypothesis of a synthesis a priori is replaced by the testable
sociological hypothesis of time as representative of a developing conceptual
synthesis. The former is beyond the reach of evidence unless one discovers
biological evidence in its favour. There is ample evidence for the latter.
Transcendental philosophers are apt to argue that a sociological theory
can have no relevance to their own, to a philosophical problem. That is a
good example of an academic establishment's defence of its own autonomy.
The insistence on the necessity to use arguments, which are "philosophically
relevant", deflects attention from the question of whether or not an argument
is intrinsically relevant to the solution of the problem under discussion. If
it is relevant to that solution, it is quite irrelevant whether or not it is philosophically relevant, especially if the relevance of philosophy itself for any
theory of knowledge is in doubt. The demand for philosophical relevance thus
illustrates the tendency of institutionalized academic establishments to claim
a monopoly in the production of knowledge in their field and to fashion the
knowledge produced accordingly. Establishments of this kind form power- and
status-hierarchies with each other. Within these hierarchies establishments of
different disciplines hold different positions - at any given time some rank
higher, some lower. The fact is known by most of the people concerned. But
one rarely refers to these status- and power-differential in these terms; they
are discussed at a lower level of synthesis with a high degree of involvement
and mostly in personal terms: as tensions between professors, departments or,
more rarely, disciplines. They are seldom conceptualized at a higher level of
synthesis. Sociologically the power- and status- differentials between different
scientific establishments and their reasons still are a largely unexplored field.
So are the tensions and struggles engendered by them. Compared with many
other tensions and struggles, power- and status- rivalries between academic
establishments are usually fought out in a more civilized manner. As far as is
known, no-one has ever physically mained or killed in their course. Nevertheless, these interdisciplinary establishment struggles, often long lasting and
unresolved, can be fierce; they can be highly injurious to lower-status establishments, to the outsiders, and to the defeated. Though they do not normally
lead to physical injury, mental injuries are frequent and often severe.
At stake in these rivalries are the relative autonomy and independence
Scientific Establishments
23
24
Norbert Elias
Scientific Establishments
25
26
Norbert Elias
Scientific Establishments
27
28
Norbert Elias
science, in short, have persistently omitted to take into account such aspects
of scientific work as the process of increasing scientific specialization and the
problems connected with it, among them the problem of interdisciplinary
relationships and their effect on the production and advance of knowledge
itself.
The exclusive fixation of transcendental science theories on the search
for timeless universals and the congruent value scheme has prevented their
representatives from perceiving the relevance for any science theory of
diachronic processes, not only of an individual science but also of the increasingly differentiated and expanding network of sciences. Today, problems
of interdisciplinary relationships in the form of cooperation as in that of
rivalries and conflicts can be regarded as a normal part of scientific development. That they, like some of the other aspects of sciences mentioned before,
are almost completely disregarded by these traditional philosophical theories
is symptomatic of the severely restricted horizon their tradition imposes upon
them. It indicates how remote from the developing sciences themselves these
science theories are, and how large is the discrepancy between their claim to
be basic to all forms of knowledge and their actual cognitive value, though
the idiosyncratic language of philosophers may disguise it.
Even more important, it may disguise that the problem of human knowledge, as framed by transcendental philosophers, admits of no solution. If the
contention were correct that special forms of intellectual activities are built
into every member of the human race prior to experience, independently of
learning, there would be no way in which humans could convince themselves
that the world, as represented by their social symbols, and thus, as experienced by them, can ever be an accurate representation of the world as it is
independently of their experience - except perhaps on Leibniz's assumption
of a pre-established harmony, instituted by God, between people's "internal"
images and the "external" world.
Nor does the opposite view carry greater conviction - the view that man's
symbols representing objects in connection with each other, are simply mirror
images of these objects. "Open the eye and you'll get the right idea", is not
a productive prescription. To break the deadlock one has to acknowledge the
fact that humans learn social symbols of the world from childhood on, which
pattern their experience and which they may, in their lifetime, develop further,
What has been lacking so far, are model of the unique relationship between
Scientific Establishments
29
30
Norbert Elias
Scientific Establishments
31
32
Norbert Elias
which has been put forward here. Take, for instance, the philosophical
distinction between "systematic" and "historical" enquiries. It is another way
of referring to the distinction between enquiries aimed at timeless universals
and others aimed at diachronic sequences of change. In these terms, philosophical epistemology is "systematic" and sociological epistemology, as
understood here, is "historical". On the face of it, this distinction may appear
to be purely fact-related; on closer inspection one can recognize that it
represents a specific scheme of values. It means "only" historical, namely
directed towards the layer of changing events, and thus "not systematic",
not directed towards timeless universals and, therefore, not philosophically
relevant, not participating in the high value of the discovery of eternals. In
other words, this too is an example of an academic establishment trying
to defend the independence of its tradition, its sovereignty, and its interests
by means of a conceptual polarization implicitly reflecting its own axiomatic
hierarchy of values, buttressing its own superior status.
The re-casting of the problem of knowledge suggested here, thus not only
affects the traditional concept of the "subject" of knowledge and the whole
scenario of the thinking statues, it also entails a change in what is regarded
as "basic" in man's quest for knowledge; it entails a change in the balance
between the cognitive value attributed to two distinct aims of this quest,
between the aim of discovering timeless, lawlike universals and that of discovering the structure and explanation of continuous processes of change in
the sequence of time, the equivalent, as one may see, of the transition from a
Newtonian to an Einsteinian approach.
In the age which put its stamp on the secular philosophy tradition of
Western Europe, in the 17th and 18th centuries, philosophers as well as
scientists would have hardly hesitated, had they ever been asked to decide
which of these two aims, in their view, endowed research with a higher
cognitive value. Most probably, they would have said that the discovery of
eternal laws, such as those of Newton, had a very much higher cognitive value
than the chronicling of historical events or discoveries about the development
of humanity. The philosophical distinction between "systematic" and "historical" enquiries carries this view into the 20th century.
However, the development of sciences in the late 19th and the 20th centuries itself casts some doubt upon the traditional assessment of these value
differentials between two aims of research. Thus, in biology a process model
the theory of evolution, has long gained acceptance as a theoretical reference
Scientific Establishments
33
34
Norbert Elias
explore it, at that stage, people had hardly the means to do so. Thus,he was
quite content with discovering recurrent law-like regularities in the movements
of the planets. It would need a closer study of the development of physical
sciences in order to show why it became possible, in the course of the 20th
century, not only to raise, but also to start solving the problem of the evolution of stars, of galaxies and, indeed, of the whole universe. But whatever the
reasons, in that field, too, the recognition has gained ground that law-like
theories alone cannot satisfy one's scientific curiosity. They have their
cognitive value as largely descriptive means of orientation or as rules of thumb
for practical purposes. But on their own they do not provide any explanation
as to how things came to be what they are. Newton's laws and, still more,
Einstein's law-like theories of relativity are indispensable as an aid in the
human search for an answer to the question of how stars, galaxies, or the whole
universe evolved. But the theoretical vehicles of genetic explanations are
evolutionary theories, models of processes in the course of which stars or
galaxies form themselves or decay. Process models embody a different type
of explanation from law-like theories. They seek to explain why one thing
happens after another in a long sequence of directional, though not necessarily irreversible events. It is the demand for this kind of explanation, for
genetic explanations which accounts for the greater weight, which, on balance,
process-theories have gained by comparison with law-like theories in cosmology.
In using this formula, as one may see, I have been careful to avoid giving
the impression of a polarization - the impression that law-like theories
and process theories exclude one another. In many cases, they complement
each other. Their cognitive weight, however, as instruments of research their
weight in relation to each other, has changed and can change quite drastically.
The cognitive value of the quest for animal universals has not entirely disappeared, but it greatly diminished when Darwin presented his genetic theory
of biological evolution, compared with the value attributed to it before.
A similar shift may be expected with regard to the cognitive value of lawlike and of process theories of knowledge, of science and of the human level
generally. At the human level, unchanging universals do not play the same part
as they play at the level of physical nature. Humans can learn not only from
their own individual experience, but by means of an intergenerational transfer
of symbols, also from the experience and knowledge acquired by previous
generations. By means of linguistic and other symbols, human beings can
Scientific Establishments
35
36
Norbert Elias
Scientific Establishments
37
such a close fit to the observable connections of events, that they now can, in
some areas, control and manipulate these connections according to their
social aims, with a fair measure of certainty. What is needed, therefore, is a
sociological type of enquiry, capable of working out process models of the
development of knowledge, fitting into, but not reducible to, models of the
long-term development of human societies. In this case, the paradigm change
does not concern one particular discipline. It concerns the whole tree of
knowledge with all its branches. Inevitably, it will get entangled in interdiSciplinary tensions and disputes. It cannot fail to be affected by power and
status differentials between different disciplines. Thus, one cannot hope to
bring about such a shift without including into one's theory of knowledge
and of science the role played by interdisciplinary relationships in scientific
developments and especially by the relationship between different scientific
establishments. Transcendental philosophers often claim that they can
prescribe for sciences generally. Their claim ought to be firmly rejected.
Theirs is an esoteric enterprise of no relevance to the work of social scientists,
and probably not to that of natural scientists either.
5. Scientific Establishments and Control of the Means of Orientation
The central social function of knowledge is that as a means of orientation. As
the individual orientation of every member of a society depends on the means
of orientation available there, groups of people who are able to monopolize
the guardianship, transmission, and development of a society's means of
orientation, hold in their hands very considerable power chances, especially
if the monopoly is centrally organized. The mediaeval Church was a large-scale
organization of this type. The very heart of its high power ratio was its
monopoly of the basic means of orientation, of revealed knowledge in large
parts of Europe. One of the principal conditions of the emergence of a scientific type of knowledge production was the breakdown of the Church's monopoly of the basic means of orientation. At least since the Crusade against
the Albigensians, this monopoly had been carefully guarded against dissenters
partly by persuasion, but increasingly by fire and sword. All attempts to explain the rise of a scientific type of knowledge which do not take into account
the armed monopolization of the European fund of knowledge by the mediaeval Church and do not link it to the partial break-up of this monopoly, are
38
Norbert Elias
bound to fail. If it is taken into account the problem ofthe emergence of a new
type of monopolization of the existing fund of knowledge and of the production of new knowledge by a new type of establishment, by scientific establishments, stands out in fuller relief.
The increasing professionalization of scientific work during the 19th and
20th centuries and its concentration in the hands of highly placed professional
groups, has led to the formation of a type of scientific establishment which in
certain respects differs from that of previous ages. Even though, through
integration into the universities it still stands in the line of succession of the
long tradition of higher learning in Europe, there are a number of differences
which require investigation. How, for instance, did universities transform
themselves from a stage where they were largely Church-controlled clerical
institutions to a stage of - mostly - State-controlled scientific institutions?
How did it come about that the rising sciences became attached to this
particular type of institution, to universities? Perhaps one should not take it
for granted that they did. How did this scientification affect the institutional
development of universities? And how did the development of universities
affect that of the sciences?
An example may help to see this interplay in better perspective. European
universities traditionally had a double function. As part of their mediaeval
heritage, they were, at the same time, agencies of teaching and of enquiry.
Their members were simultaneously engaged in advancing knowledge, in
widening the scope of human orientation and in handing on knowledge,
old and new, to the rising generations. This double function of universities
has not been entirely without influnece upon the development of sciences;
in fact the influence may be greater than the construction of a purely rational
model of science indicates. The role of textbooks is symptomatic of this
double function. On the one hand, textbooks are devices linked to the
teaching function of universities. Yet, on the other hand, they are also
vehicles of the struggle for, or the achievement of, consensus among leading
representatives of a particular academic field or, in other words, of scientific
establishments. Textbooks indicate what the establishment of a particular
academic field or sections of it, at a given time, regard as the essentials of
the received knowledge in that field. Competing textbooks may point to
dissensions and rivalries. Changes in textbooks often, though not always,
indicate that, and when, a scientific innovation at the theoretical or the
Scientific Establishments
39
40
Norbert Elias
with others - advance in their career, their income is likely to rise. However, in
the case of economic establishments, such as those of commercial enterprises
or factories, their standing within the status hierarchy of competitors is, to a
large extent, determined by the size of their income, i.e., by the establishment's
economic success or failure. In the case of academic establishments, inversely,
the economic resources accruing to them largely depend upon the status and
the power resources they can command within the academic network.
These resources are, in a general way, of the same kind as those at the
disposal of other professional establishments. Scientific establishments are
groups of people who collectively are able to exercise a monopolistic control
over resources needed by others. They control, and engage in, the production
of a particular type of knowledge. In many cases they combine this activity
with that of administering a fund of knowledge handed on to them, in their
particular field, from previous generations and with that of controlling the
transmission of that fund, and whatever advances they themselves have
contributed, to the following generations. By virtue of their monopolistic
control of an existing fund of knowledge and of the skills needed for developing it, for producing new knowledge, they can exclude others from access
to these resources or admit them to their use selectively. That is to say, they
are establishments only in so far as there are groups of not-established people,
. of outsiders, who need the resources monopolized by them and who depend
on them for access to, or use of, them. However, the need is never, or only in
marginal cases, entirely one-sided. Those who are outsiders, in relation to
a given establishment, as a rule, have on their part resources needed by the
establishments' members (which does not imply that the outsider groups
have a centralized or even a monopolistic control over these resources).
Established and outsiders, in other words; have specific functions for each
other. No
relationship is likely to maintain itself for long
without some reciprocity of dependence. However, the dependence of the
established on the outsiders is smaller and, as a rule, very much smaller, than
that of the outsiders on the established. Members of an establishment usually
are very careful to maintain and, if possible, to increase the high dependence
ratio of their outsider groups and thus the power differentials between these
and themselves. If this ratio diminishes, the power surplus of the established
group vanishes and the group itself sooner or later ceases to have the character
of an establishment, as indeed happens again and again. Thus established and
Scientific Establishments
41
42
Norbert Elias
Scientific Establishments
43
44
Norbert Elias
Scientific Establishments
4S
establishments, their status rivalry and the differentials of their power resources are not without influence on the construction of theories, the framing
of problems and the character of the techniques used for solving
6. Political and Scientific Establishments
But before getting on to some of the problems generated by the struggle
for dominance and autonomy among scientific establishments themselves,
I feel I must say something about the relationship between scientific and
establishments, in particular political establishments such as
governments, their high-ranking bureaucratic agencies and the spectrum of
parties. The problem, too little studied, does not lack urgency. An instructive
example are the similarities and differences between the relationship of
scientific establishments and state governments or party and military establishments in communist and non-communist countries, in one-party and
multi-party societies.
In these one-party states scientific establishments at universities form part
of a career pattern and of a power hierarchy whose highest point lies outside
the universities. The highest point in the career pattern of a professor is
membership in the central scientific institution of communist states, the
Academy of Sciences. In Russia a professor who can call himself Academician
has a higher status and greater power resources than one who cannot. This
centralization of control over scientific work corresponds to the - organizationally - monolithic character of a one-party state. The Academy of Sciences
has close links with, and is closely controlled by, the Central Committee of
the Communist party. Thus "Academician" is, in fact, the highest rung a
scientist can hope to reach in his professional career.
In East Germany, for example, one encounters a career pattern according
to which a university professor can be promoted first to the position of a
corresponding member of the Academy of Sciences. He may also be promoted
to a position at one of several research institutes, of the AoS (which is not
identical with membership of the Academy itself) or, alternatively, to a
position in one of several Research Institutes of the Central Committee of the
ruling party. All these are possible stepping stones on the way to membership
of the AoS itself. (That there may be a hidden overlapping of the functions
of some of these high-level research institutes, and thus a measure of rivalry,
46
Norbert Elias
need not concern us here.) Tight central control thus limits rather severely
the autonomy of scientific establishments below that of the Academy level.
The Academicians themselves are subject to a fairly strict party and state
control. Ideologically, the need for a measure ofautonomy of scientists at all
levels is still recognized. It is frequently alluded to in official publications
under names such as "self responsibility". In actual fact, the working out of
research projects is firmly controlled from above. Formally, they appear
to require approval of members of the Academy. The same goes for the
evaluation of research results.
In Western countries the evaluation still has a rather informal character.
One speaks rather vaguely of a "scientific community", as the agency which
accepts or rejects research results and decides on their integration into the
common fund of knowledge of a scientific discipline. But that is not a firmly
organized body. In fact, the formation of an opinion among the various scientific establishments about the work done by, or under the aegis of, one of
them is not a formalized social process - and, by and large, an open problem.
In communist countries it is firmly institutionalized. It lies in the hand of
special bodies, not necessarily of members of the AoS themselves.
In short, universities in these countries have become closely integrated into
the hierarchic structure of a one-party state with regard to research as well as
to teaching. It would be interesting to learn more about the consequences of
this type of organization for the development of the various sciences. Maybe
the Lysenko affair has served as a warning against the possibility that scientists
produce research results which they know will please the powers that be.
In multi-party states trends pointing in the same direction are by no means
entirely lacking. The dependence of scientific establishments on state fmance
and thus on state agencies of one kind or the other is on the increase. The
two great wars of the 20th century were among the levers accelerating the
interdependence between scientific and state establishments. Already during
the First World War some of the status and power rivalries between scientific
and non-scientific establishments flared up. They are not unusual in the case
of increasing interdependence between previously independent groups. In the
Second World War, again, the steadily increasing scientification of warfare
contributed to the frequent blend of collaboration and rivalry between
scientific and non-scientific, particularly military, establishments of which
many stories are told. In fact, it has been said, somewhat hyperbolically, that
Scientific Establishments
47
the 1914-1918 war was a chemists' war, the 1939-1945 war a physicists'
war.
In away, thus, the wars heightened the differentials in social power and
status of different scientific establishments. Evidently physicists and chemists
were much more capable of contributing to the war efforts than, say, sociologists, historians, or students of language and literature. During the greater
part of the 19th century, the social usefulness of physics was hardly recognized. In America, practical men, such as Edison who had little mathemlftics,
gained very high esteem as a result of their tangible successes both as inventors
and businessmen. Mathematizing physicists were still regarded as people in an
ivory tower whose social usefulness was doubtful. Late in the 19th century
and in the course of the 20th century all that changed. The advances made by
physical scientists in theory and practice were felt almost everywhere in
society at large. Thus, the wars, and especially the Second World War whose
outcome was in some situations decisively influenced by the inventiveness
of physicists, powerfully reinforced a trend in the hierarchic relationship
between different academic specialisms, which, perhaps less strongly, made
itself felt in peacetime too. Within the status hierarchy of scientific establishments, the physicists' claim to the dominant position as trend setters of
scientific work generally gained conviction as a result of the staggering
transformations in peace and war they had helped to bring about in society
at large. The demonstration of their social usefulness enhanced their status
and their power resources within, no less than outside, the academic world.
However, this was in no way the achievement of a clearly envisaged goal.
The descriptive historical literature about the relationship of scientific and
state establishments, especially in America, France and England is growing.
But sociological enquiries into the science-theoretical significance of these
developments lag behind; research into the impact which social developments,
such as the changing relationships between scientific establishments and
governmental or military establishments, have on the development of scientific knowledge itself, is still in its infancy. That the emancipation from an
extraneous authority was one of the principal conditions under which the
acquisition of knowledge gained its specifically scientific character, is fairly
well known - as far as it concerns the past. Galileo stands out as a pioneer of
the emancipation of science from the control of a powerful non-scientific
establishment, which claimed a monopoly of the means of orientation.
48
Norbert Elias
However, the question of why and how far scientists were able, even in
the world of the old state absolutism, to gain and to preserve a measure of
autonomy for their new mode of producing means of orientation, has hardly
been raised and explored. Nor has the complementary question found much
attention - the question whether and how far the relative independence of
scientific establishments can be maintained in the face of growing dependence
of their work on non-scientific establishments, bureaucratic, military, industrial or whatever. That scientific enquiries can be successfully carried out, at
least in some fields, under the very tight state control practised in the Eastern
one-party states, seems to point to the fact that the scope of autonomy
regarded as normal for scientific work in the Western world can, be whittled
down without seriously impairing the scientists: powers of discovery. But
whether that is really the case and, if so, how it is achieved, for instance
in the Soviet Union, these and other related questions can hardly be answered
without more systematic sociological enquiries in the developmental trends of
the relationship between scientific and non-scientific establishments and their
impact on the knowledge produced by the former. There may be serious
flaws in an authoritarian set-up that will come to light only in the long run.
7. The Development of Scientific Establishments
Scientific establishments derive their high power ratio from the monopolization of a particular type of knowledge. They are, in other words, groups of
specialists whose social function it is to administer and to augment a specific
fund of symbolic representations which has been handed on to them by
previous generations - a fund of symbolic representations which can serve
people as means of orientation. Scientists are not the first and not the only
groups who collectively monopolize the administration and production
of a society's central fund of knowledge, of its basic means of orientation.
Their principal predecessors in that capacity and, in more recent times
often their competitors, were priests of all kinds. However, the type of
knowledge monopolized by the two groups is different, even though there
are hybrids, and many transitional forms, of which philosophy is one.
The type of knowledge which has been administered and produced by
priests through the ages is, in the last resort, always revealed knowledge,
knowledge held to be received by people through a revelation of non-
Scientific Establishments
49
50
Norbert Elias
Scientific Establishments
51
52
Norbert Elias
Scientific Establishments
S3
plored is that of the sons for the sociologists' difficulties. Sociology being
itself a science can serve as an example. It enables one to probe into some
aspects of the wider social setting within which sociological establishments
are working today and into the pressures to which they are exposed, intramurally as well as extramurally.
8. Sociology in The Shadows of Two Stronger Power-Blocks
Intramurally Sociologists work within a setting where one of the most powerful and most successful groups of scientific establishments, whose members
rank high in the academic status hierarchy, that of physiCists, have formed
an idea of science in their own image; they have succeeded in propagating
the belief that their own method of setting and solving problems is the only
scientific method and that research not undertaken in accordance with their
prescription ought not to be accorded the status of scientific research. It
would perhaps be easier to recognize this as a hegemonic claim put forward
by one highly respected and successful set of scientific establishments in
its status and power struggles with others, were it not for the fact that many
members of another type of academic establishment, also traditionally
endowed with a high academic status, members of philosophical establishments, plough the same furrow. They reinforce the belief that physics is
the science par excellence, that other disciplines in order to be recognized
as scientific should use as nearly as possible the same method as physics
and, generally, should proceed along the same line. They have, indeed,
succeeded in giving many an academic a bad conscience or a feeling of
inferiority if they do not succeed in following the example of the physicists.
Yet, philosophers never bother to explain the diversity of sciences, their
science theories are generally based on the axiomatic belief that all sciences
must be cut according to the same pattern. Some of them go so far as to
present a purely phYSicalist ideal type model of science as a universal "theory
of science". By constructing highly formalized models of a science abstracted
from physics they claim to establish norms and to make prescriptions for
sciences of all kinds. The differences which exist between different types of
sciences lie beyond their field of vision. They neither account for these
differences; nor do they ask why many non-physical sciences, among them
most human sciences, cannot be fitted into the Procrustean model of a physicalistic science theory.
54
Norbert Elias
As far as sociology is concerned, one can say that measuring, quantification, mostly in statistical form, and the use of mathematics, has a legitimate
place in its proceedings. But it is a very limited place. The central areas of
the sociological problem field cannot be reached by means of quantifying
methods. They require other methods of enquiry and other types of theory
of which process and figurational theories are two, but by no means the only
examples. They can be constructed with as high an accuracy and reliability
as the law-like mathematical formulae of physics. Power balances and conflicts play a central part in these models and so do problems of conflictcontrol especially of war-control. Almost all social institutions are born from
a structural social conflict; they contain the birth-conflict, as it were, in a
frozen form which, in some cases, can thaw, thus bringing the conflict once
more into the open. In others it transforms itself into a stabilized, uneven
balance of power, a hierarchic order of superior and subordinate ranks. None
of these or of other long-term social processes, of figurational changes and
their dynamics can be adequately represented by means of mathematical
symbols and operations. If a quantifying reduction ala physics is attempted,
the results, inevitably, are barren and vast tracts of the problem field remain
unexplored.
As an example of the still unrecorded power- and status-struggles between
academic and other establishments, the case of sociology is rather instructive.
If we were to draw up a sketch of the situation in which sociological establishmen ts find themselves today, we would have to say that they are hemmed
in between at least two blocks - two types of establishment with much
higher power- and status-resources than they themselves have. In one way or
another, therefore, sociologists, uncertain of their own task as well as of their
standing, feel impelled, as weaker establishments often do, to seek greater
certainty and to gain kudos for themselves by a kind of mimicry; they take
over concepts, methods, values and forms of thinking from these more powerful establishments around them. But by doing so, they achieve the opposite
from what they hope to achieve. They increase their own uncertainty, impress only each other and perpetuate their malaise.
I have already referred to one of these blocks, the establishments of
physicists and physicalistic philosophers. The other type of establishment
from which prescriptive categories, concepts and values are taken over by
sociologists are political groupings, especially party establishments. This is
Scientific Establishments
55
made possible because the party establishments of our age legitimize themselves not simply as flag bearers of specific personal or group interests, but
also as harbingers of a general way of ordering human societies. This blend
of sectional self-interest and generalized social programme, represented by
the standard party ideologies of our time, is at present very much taken
for granted. One is hardly aware of the intellectual untidiness of a type of
ideology which tempers and half conceals a programme for the furtherance
of a group's self-interest by means of a programme for the general good of
society. Moreover, these untidy doctrines of an intrastate class conflict have
now become battlecrics of an interstate conflict threatening the destruction
of humanity.
However, although to-day social ideologies, social beliefs and ideals of this
type must be perceived as antagonistic to sociological theory, they were one
of its roots, a stepping stone towards the perception of human societies at a
higher level of synthesis. As this is one of the characteristics which party
belief and sociological theory have in common, it is not surprising that one
can observe, ever since the inception of sociology early in the 19th century, a
continuous interplay between them. Both have come into their own at about
the same time and sociological theories, so far, have always remained tied to
shades of party beliefs. The social establishments which form the primary
representatives of these beliefs, and their fol!owers in society at large, form
a much more powerful grouping of people than the groups of sociologists,
the makers of sociological theories, who, since Durkheim's days, came to
establish themselves slowly at the world's universities among the older scien-
tific establishments. It would be a rewarding task to follow in greater detail
the long-term development of the relationship between these two types of
establishment as well as that between social beliefs and ideals in society at
large and sociological theories in a univerSity setting. Perhaps the future
sociology of sciences will take care of such problems. Curiously enough,
sociology is among its own objects of study. One could probably show that
social belief and sociological theory at present form different poles of the
same parameter of knowledge, the one characteristic of greater involvement,
the other of greater detachment, different in their emphasis; they are represented by different professional groups.
This, then, is the second set of establishments whose concepts, values and
forms of thinking impinge upon those of sociolOgists. Again, the balance of
56
Norbert Elias
Scientific Establishments
57
physics - as misleading as the separation of theory and belief in the development of early sociology. A fuller understanding of the development of science
in its early stages is only possible if one enquires how and why a scientific
mode of approach emerges from non-scientific beliefs. In the early stages of
the emancipatory process, people still take it for granted that belief as a form
of knowledge takes precedence over theory; they do not perceive clearly
the difference, nor do they clearly recognize contradictions between belief
and theory. The emancipation of the scientific form of knowledge and of
acquiring knowledge from the matrix of beliefs, usually, is a slow intergenerational process. Beliefs, at that stage are usually backed by powerful
social establishments. By contrast, early scientific forms of enquiry and of
knowledge, as a matter of course, lack any institutional shell of their own,
they are often represented by circles of people without other than personal
bonds, or, at the most, with loose institutional bonds whose power ratio,
compared with that of the establishments representing the beliefs of the
moment, is very small; and the same is, of course, true of the rise of a new
type of science in its relation with old established sciences or with competing
social beliefs. At first, emerging theory or, for that matter, a new type of
science, is likely to be weak in terms of the power chances of its promoters.
Representatives of established beliefs are apt to use a variety of intellectual
devices in order to counter the challenge implicit in the emergence of a
novel type of knowledge. They may at first try to disregard its existence,
to kill it through silence. They may argue that the novel type of knowledge
does not contain much information beyond that which has been known
before. They may adopt some peripheral aspects of the new knowledge and
disregard those crucial innovations which run counter to their own tradition. In accordance with the power differentials between the representatives
of an emergent theory and the old established beliefs, representatives of
the former themselves may be accommodating vis-a-vis those of the latter;
they may be in all sincerity, unable to cur themselves loose from the older
beliefs.
There are, thus, a great variety of ways in which an emergent scientific
theory can be influenced by the ruling types of belief. The existence of a
centralized church enforCing the belief in a single god was probably quite
an important determining factor in the emergence of a type of theory embodying the concept of unified nature ordered according to eternal laws
(7).
58
Norbert Elias
Scientific Establishments
59
avoid the somewhat confusing idiom, of their object, in relation to the problem field, the object of other academic groups. The question, in short, with
which one is confronted here admits of a perfectly neat and clear formulation.
It is simply the question of whether and to what extent the nexus of events
to which we refer as human societies,
a structure of its own, which
is not entirely reducible to the structure and regularity of that nexus, which
we call ''nature'' and which is the subject matter of the research of natural
scientists, particularly of that field whose representatives claim today that they
are able to provide models for all other research enterprises and that their
method can and must be used for the exploration of all other fields of study.
The hegemOnic claim of the physical sciences is to some extent supported
by the widely held idea that the recognition of ''nature'' as a relatively autonomous nexus of events with structures and regularities of its own, is obvious.
One need only open one's eye, so it seems, in order to be aware of the fact
that the sun and the moon move in regular circles, that falling bodies or gases
obey immutable laws; in short, that the nexus of physical events possesses
recurrent regularities of its own has, a fairly high degree of autonomy in
relation to the wishes and hopes of human beings. Many people would
probably argue it is much more difficult to maintain that the societies formed
by human beings have structures and regularities of their own which are
relatively independent of human wishes and aims. That indeed is the core
of the problem. If human societies in their structure and their course are
wholly determined by wishes and aims of the people who form them; if they
are, for instance, determined by the wishes and aims embodied in political
party programmes, then indeed, the chance that sociologists and other social
scientists may gain a measure of autonomy in relation to political establishments is small. The aspiration of sociologists for a higher measure, though
of course not for complete, autonomy provided the power structure of a
state allows it, would be a vain hope. Again, if the structure of societies and
the regularities of social processes could be adequately expressed and explained as derivatives of the structure of atoms and molecules of which humans are undoubtedly composed; if in other words, the structure of societies
is reducible to the structure of their smallest component parts, then indeed
the claim of physicists to set models for social scientists would be justified.
In that case social sciences and, in a wider sense, human sciences, could have
no degree of autonomy in relation to physics, as their subject matter -
60
Norbert Elias
Scientific Establishments
61
62
Norbert Elias
the development of knowledge has no intrinsic structure of its own and can be
fully explained in terms of the structure and development of the groups of
knowers. However, neither of these two approaches does work, least of all in
the case of scientific knowledge. In that case it is particularly clear how
inextricably interwoven and interdependent are the social value and thus the
status and power ratio of groups of scientists and the cognitive value of the
scientific knowledge produced by them. The same can be said with regard to
the relative autonomy of both. The autonomy of the scientific knowledge
and particularly of the theories of a specialized branch of sciences and that
of the group of scientists who administer and develop a particular type of
theory, again, are inextricably interwoven and interdependent.
The fact determines to a very high degree the relationship between different academic establishments. Most of them are busily engaged in the task
of proving to themselves and to all the world, the autonomy of their own
field of studies in relation to other fields thus trying to ensure their own
autonomy as a professional group. Thus psychologists are eagerly building up
theories of their own which protect them from the threat of being subdued by
physiologists and, more generally, biologists, while some biologists, especially
etholOgists like Lorenz, on their part are on the warpath engaged in some
highly successful forays into the psychologists' and even the sociologists'
territories and have begun to settle down there. On the other hand, biologists
themselves are on the defensive against a vigorous expansionary move of the
physicists who, in the form of microbiology, have begun to colonize some
branches of biology such as genetics, to transform them into provinces of the
great physics empire and to subject them to the rule of what is called the
"scientific method" , which is in effect, the method of the physiCists.
All this may give some insight into the magnitude of the task before us.
Some of the basic theories, some of the forms of thinking, with which we
have grown up, may have to be revised. The theory of abstraction is only
one of many examples. If I were asked to say briefly and superficially what
I regard as the linchpin of the reformation to which I have alluded, I would
say it is the re-thinking of our theory of knowledge in terms of evolving
figurations of people, of developing groups of interdependent individuals
as the subject of knowledge rather than of an isolated individual of the
homo clausus type. It is the break with the habit of discussing knowledge
without saying what it is, without stating clearly that knowledge is the
Seicntijic Establishments
63
meaning of human-made social symbols which are intergenerationally transmissable. It is further the re-thinking of all our problems and theories of
knowledge in terms of long-term processes of change, replacing the present
tendency towards reflections, concepts and theories which aim at reducing
all processes of change symbolically to static conditions.
That such a research programme challenges many existing habits of thinking I have already said. These habits of thinking are backed by the authority
of two types of high ranking academic establishments which show strong
affinities to each other, the establishment of theoretical physicists and the
establishment of transcendental philosophers. The latter are firmly wedded
to the concept of a subject-object divide which casts doubt either upon the
existence of objects or the human capacity to grasp them as they really are.
Kant's brief statement that each of us may assume the existence of other
minds operating in a similar way as ours, but, of course, we can never be sure
of it, is revealing. Some contemporary transcendental philosophers with the
same solipsistic bent as Kant, speak of inter-subjectivity, apparently without
any awareness that the very concept contradicts the basic assumption of
transcendental metaphysics, which reduces cognition not only of objects,
but inevitably also of other subjects, to the transcendental operations of an
individual mind.
The blockages imposed on many of our reflections by the establishments
of theoretical phYSicists are of a different kind. An example is the assumption
that knowledge of the structures and regularities of the smallest constituent
parts of a composite unit provide the ultimate key to the knowledge of
the structure and regularity of that composite unit itself. The conclusion
to be drawn from this assumption for the relative cognitive value of the
knowledge provided by all the various non-physical groups of scientists, is
obvious. Physicists are the specialized providers of knowledge about the
smallest constituent parts of which everything else in the universe, humans
included, consists. On the assumption that the properties of composite units
can be fully deduced and explained from the exploration of their constituent
elements, physicists hold in fact, the key to all other sciences. There are
many variations and permutations of the physicists' claim that their science
provides the key to all others. In some cases, their hegemonic claim is based
on the idea that their method of research is the only scientific method and
ought to be imitated by all other scientists. In other cases, physicists put
64
Norbert Elias
forward the claim that the forms of thinking, the categories, which they
develop, in the exploration of the constituent parts of matter are applicable
to all realms including the realm of life.
Thus, Werner Heisenberg in a series of lectures, published in 1973 said
(8):
the forms of thinking evolved with the development of atomic physics are wide enough
to provide scope for the various aspects of the problem of life and the direction of
research connected with them.
Scientific Establishments
65
not investigate more closely the power and status hierarchy of the different
academic disciplines and raise the question: What determines, what
legitimizes the claim of different academic disciplines to a relative autonomy
in relation to each other? Or, alternatively, what justifies the claim of some
disciplines such as physics, that none of the others can have any relative
autonomy and that they all, sooner or later, can be explained and investigated
in terms of one of them, in this case, in terms of physics?
If one removes all disguises, the questions before us are these: Can we
assume that sooner or later all the structures and processes investigated by
sociologists will be explainable in terms of atomic structures or, for that
matter, of the structures of the ultimate particles of the universe whose
secrets contemporary physiCists try to unravel more eagerly than any of
their ancestors? Or, alternatively, shall we assume that social structures and
processes, because of their greater complexity, do not admit of any reliable
scientific investigation and must for ever remain a mystery?
It is not uninteresting in this context that Heisenberg in the same series of
lectures, stated even with regard to physics that (9)
In the great majority of cases, the complete mathematical calculation of a set problem
will technically not be possible for all too great complications can no longer be mastered
in mathematical terms.
66
Norbert Elias
component parts holds good only at those relatively simple levels of integration with which physicists are concerned. That is to say, with objects reaching
from the smallest subatomic particles of matter to the larger molecules. At
that level, in most cases, though not in all, processes of integration are
reversible. The properties of the component parts do not fundamentally
change when a given composite unit at the higher level disintegrates. When
a water molecule disintegrates into it atoms, the properties of the atoms
are not different from their properties as part, of a molecule. As soon as one
ascends to levels of integration which we call biological; represented, for
instance, by an organism consisting of a single cell, the rule established at
the physical levels no longer holds good. If a cell disintegrates the constituent
parts of the cell at the next lower level, e.g., its nucleus, its membrane, etc.,
disintegrate, too. They disintegrate, broadly speaking, into physical units,
into atoms and molecules. The difficulty for many people is to think in terms
of not only three or four but, perhaps, 20 or 30 levels of integration superimposed upon each other like the part-units of a Chinese box.
Such units require, in fact, different forms of thinking. Here there are, at
every higher level of integration, regularities which cannot be explained alone
in terms of the properties of the lower constituent parts. Thus, for example,
at a certain level of integration, one encounters a type of events for which
no parallel can be found at a lower level. Concepts such as life and death,
birth and heredity, for example, are characteristic of the biological but not of
the physical levels. We have separate words for these distinct properties
of different levels, one is often inclined to explain the events to which they
refer in terms of an addendum, a kind of substance added to or taken away
from the other component parts. Thus, some people try to explain life in
terms of an invisible force or substance added to the visible forces or substances. But that is only because a reifying physicalistic mode of thinking
blocks the awareness of the fact that forms of integration and the configuration of component units have explanatory functions, in addition to the
explanations to be derived from the properties of the component parts.
But perhaps I have said enough to indicate what that means with reference
to human societies themselves. In their case, too, the figuration of constituent
units, of human beings has an explanatory function of its own in conjunction
with the explanatory function to be derived from the properties of the constituent parts. In this case of the biological properties of individual human
beings.
Scientific Establishments
67
I had to cover my ground rather rapidly, but for the time being it may be
enough to put in fuller relief the basic problems with which we are concerned.
I am trying to state the fact that social structures have a relative autonomy in
relation to biological structures, that they represent a nexus of their own
different from the nexus which we call nature, though derived from it and
related to it. Present reflections of physicists and philosophers are still tied
to a scheme where the universe consists of physical eVents into which humans
are set inexplicably as explorers of objects. Neither physicists nor philosophers
so far recognize the distinct order of human beings, which we call societies,
as an order with structures and regularities of its own, as a semi-autonomous
level of the Universe. To establish this fact requires a struggle against many
established views and against the groups of people who are the holders of
these established views, of the older scientific establishments. Sociologists
should, I think, examine the basis of their own field of studies much more
thoroughly than they have done so far. They should know that a rising
science cannot assert itself unless it is able convincingly to free itself from
the inappropriate models of older establishments, to develop theoretical
models and forms of thinking commensurate with the relative autonomy
of its subject matter, and to fight for its own relative autonomy in relation
to the older sciences and the old philosophy which even when seemingly
hostile to them, is in terms of its categories and its de-humanizing mode of
abstraction their companion-in-arms.
68
2.
3.
4.
5.
Norbert Elias
hat?' Hrsg. von D Friedrich Theodor Ring, Konigsberg 1804 in: Imm. Kant's Werke,
Bd. viii, Cassirer Berlin 1,!22, S. 245:
"Die Form des Objekts, wie es allein in einer Anschauung vorgestellt werden kann,
griindet, sich also nicht auf der Beschaffenheit des Objekts an sich, sondern auf
der Naturbeschaffenheit des Subjekts ... "
See N. Elias, Problems ofInvolvement and Detachment.
Imm. Kant, ibid. p. 256 ..... Ob das Objekt, welches wir ausser uns ansetzen,
nicht vielleicht in uns sein konne und es wohl ganz unmoglich sei, etwas ausser
uns als ein solches mit Gewissheit anzusetzen." One is again and again surprised
at Kant's repressive innocence which allowed him not to notice that the very use
of the personal pronoun "us" referring as it does to a plurality of people "outside
of himself" vitiated his whole argument. The same goes for his use of a common
language directed at a plurality of human beings. A language, i.e., a specific pattern
of learnt sound signals symbolizing, and at the same time serving as a store of, the
common knowledge of a human group, presupposes a plurality of human beings. It
shows a remarkable lack of consistency in thinking - or, to use the fashionable word,
of "logic" - to try fitting the intrinsically social datum of language into the framework of a transcendental philosophy of whatever kind unless one postulates that the
specific language which "language communities" speak is inborn and biologically
inherited by each of the members of that community individually.
The essence of transcendentalism is the assumption of a highly specific
pattern either of thinking or of speaking which is present in every individual human
person prior to all learning. The concept of a human being, underlying transcendental
philosophy is that of an isolated individual, a homo clauses, endowed with specific
ways of connecting events which act as an impenetrable wall between his own consciousness and whatever is outside of it - natural objects or other persons. Every
specific language on the other hand presupposes learning from other persons, presupposes in fact the existence of other persons. Languages change over time. Every
individual person learns the language in the form which it has assumed at the time he
or she enters the "language community". Languages, like the societies where they
flourish, are in a condition of flux. Transcendental philosophy makes assumptions
about specific aspects of people's mental capacities which are unchanging or, in other
words, innate. These assumptions can hardly be reconciled with the observable changing languages spoken in observable changing societies.
In fact, expressions such as "language community" or "communication community" are a good example of one of the principal characteristics of transcendental
philosophy. In a large measure it creates its own objects. The fact, to which this
expression refers, is simple: as children human beings learn a communal language
from other human beings, and without learning it, they can neither argue nor communicate with others. By conceptualizing this observable datum in the form of a
noun at a very high level of abstraction, a philosopher gives it the appearance of a
datum whose discovery is of the very highest cognitive value. A new technical term
suggests in some way a reflection upon an "object", a condition of humans existing
prior to all experience. That is a good example of a philosophically invented object.
N. Elias, The Civilizing Process, Vol. 1 Preface, Blackwell, Oxford, 1979.
"Science does not rest upon solid bedrock. The bold structure of its theories rises,
as it were, above a swamp ... ", Karl Popper, The Logic of Scientific Discovery,
London,1968,p.59.
Scientific Establishments
69
6. Norbert Elias: 'Een essay over tijd' (An Essay on Time), De Gids, CXXXVII
CXXXVIII Amsterdam, 1974/75.
7. Joseph Needham has drawn attention to the possibility that the Chinese did not
develop the concept of a unified natural science partly because they did not have the
concept of a single god ruling the whole universe, see. "Why didn't China develop
modern science?" International Herald Tribune, Oct. 24, 1979, p. 98.
8. Werner Heisenberg, Wandlungen in den Grundlagen der Naturwissenschaft, Stuttgart,
1933, p. lxx.
9. " .. , dasz in den allermeisten Fiillen die vollstiindige mathematische Durchrechnung
eines gestellten problems technisch nicht mBglich sein wird, denn allzu grobe Komplikationen konnen wir eben mathematisch nicht mehr bewiiltigen" (Heisenberg, ibid.,
p.156).
PETER WEINGART
Universitiit Bielefeld
72
p",
,f
Weingart
If it is true then, that the "knowledge domain is growing", that scientifictechnical and professional knowledge has assumed considerable functional
importance and legitimating power, it will also have a growing impact on the
structure and contents of political problems. It may, therefore, be inferred
that knowledge conveys political power insofar, and only insofar, as it becomes a major ingredient in the definition of political problems. It does so
because a certain body of knowledge will then assume an orienting function
which entails the acceptance of implicit assumptions, the pre-determination
of the possible range of solutions, the exclusion of other knowledge and thus
alternative problem perceptions and solutions, and, as a result of all these,
the distribution of life chances in a certain pattern rather than another. On
the other hand, the instrumental function of knowledge alone, i.e., its role in
determining problem solutions, does not convey power because it leaves the
scientists in a state of dependency on those who do or do not call on them
for advice. This is the classical image of the scientific advisor, which, of
course, depends on the assumption, that politics and science, values and
knowledge, can be neatly separated.
Looking at the implicit and explicit assumptions of the writings on the
rise of the scientists to power and referring them to that notion of power
which alone can be meaningfully connected with knowledge as a resource,
I want to show that these assumptions are flawed. It can be argued that the
73
74
Peter Weingart
is limited, that the power to defme problems rests with what we have called
'hybrid communities', experts from different institutional backgrounds rather
than with the scientists alone. Secondly, I will show that the very condition
for the scientists' rise to political power, the growing instrumental importance
of scientific and technical knowledge leads to a politicization and consequent
de-professionalization of academic science.
2. Who Dermes Policy-Problems? - The Emergence of 'Hybrid Communities'
(a) The Institutional Basis of 'Experts'
In order to show that experts from a diversity of institutional backgrounds,
rather than scientists alone, are involved in the process of problem defmition
two types of indicators will be used: the formal representation of scientists
and other 'experts' in policy-defming bodies, and the differential impact of
types of knowledge held by these groups on the defmition of problems.
Regarding the first indicator the theoretically possible continuum of cases
ranges from the scientists exclusively initiating and structuring policy to their
being completely excluded from any influence. The latter case implies that
'experts' from industry, interest groups or public administration itself take
the leading role, if, as one can assume nowadays, there are always 'experts'
involved in policy-making. Examples for the first case can usually be found in
the area of science policy where a strong impact of scientists is least surprising
(8). I will deal with an 'intermediate' case, however, which seems to be more
representative: the formulation of the German government's programme
for environmental protection of 1970. The scientific-technical component
in this case is still very important but just one among others: aspects of
economic and fiscal policy as well as problems of regional planning and
political administration in general playa major role.
First of all, the initiative for the formulation of the programme emerged
within the government and was motivated both by objective pressures (international efforts coordinated by the U.N.) to reach agreements on environmental protection which would affect the competitiveness of industry and
internal political considerations (the administrative reorganization of old and
new competencies for environmental protection as part of a 'strategy of the
smaller coalition partner to gain more influence).
75
Outside advice was first obtained after the schedule for the programme
formulation had been fixed, when the Battelle-Institute, a commercial thinktank, received a grant from the Ministry of the Interior. After negotiations,
Battelle offered to work out an 'indicator matrix' for the evaluation of
objectives and scientific-technical considerations concerning environmental
protection. The research was supposed to support the formulation of objectives in the area of environmental protection. At this time a crucial decision
had already been made in the administration: formulation of the programme
was to be divided between 'project groups' organized "vertically", i.e.,
according to medial sectors of the environment such as air, water, open seas,
noise etc. Only three "horizontal" groups were planned, one on economic
problems, another one on problems of public information and a third on the
organization of scientific advice for environmental protection. Only the first
and the third were eventually established.
It is not difficult to see that this structure of the apparatus which was to
formulate the programme was a result of the jurisdictional structure of the
administration. Neither the cabinet committee nor the steering committee
set up in the Ministry of the Interior had the necessary competence for the
identification and processing of problems which would have allowed a more
encompassing approach to the problem of environmental protection.
The task of the project groups was supposed to be defined in detail by
the steering committee set up for this purpose. Their function, in general
terms, was to establish programmes for their respective subject matters which
allowed integration into the general programme. These sub-programmes
should contain an analysis of the status quo and a projection of likely developments for the next five years as well as a concept for protective and
counter-measures, the definition of threshold values, budgetary requirements
and economic implications. It was at the level of the project groups that
scientific, legal and economic expertise had to be relied upon, they were the
central mechanism with which to organize scientific advice to government in
establishing the programme on environmental protection.
The pattern of recruitment into the project groups was diverse, to say the
least, and no clear picture emerges. The group on 'environment oriented
technology', and its sub-groups, was composed almost exclusively of experts
from industry. Its work was virtually delegated to industry, in order to tap
the experience assembled in it, but obviously also for legitimation reasons,
76
Peter Weingart
77
that their work had no influence at all; the formulation of the programme
was completed before the report of their project groups. If, nevertheless, it
can be said that parts of the work had an impact, it is probably due to the
high proportion of administration experts in the project groups.
Looking at the composition of the project groups it is worth noticing
the different institutional backgrounds apart from the academic scientists.
One prominent group are the specialists from public administration. These
"experts" participate in policy formulation whenever the issue entails problems of regulation, jurisdiction, and implementation. Here, the Federal and
State administrations are directly concerned and their staff alone commands
the pertinent knowledge. The important function they serve is the integration
of that knowledge with that of the outside experts but their's is strongly
affected by both their departments' jurisdiction, their immediate area of
responsibility and the political strategy of their respective department heads.
A further group involved in policy formulation are scientists who do
not work in an academic setting, but either in government laboratories, big
science installations, or in industry. There is no systematic information to
my knowledge on how the different institutional settings in which they
work affect their perception of problems and the expertise they provide
in advisory capacities other than that they are known to internalize their
institutions' objectives (10). With the few exceptions of basic researchoriented laboratories, it is a common characteristic of the industrial and
government scientists, that their institutional base is non-academic. This
shows that even strictly scientific advice in policy-making is increasingly
provided by a part of the scientific community which has its institutional
base outside the universities and which works for objectives not set by itself.
Obviously, membership in advisory, planning and other ad-hoc groups
employed in policy formulation is only an indicator of the type of knowledge
that serves as an input and of the (institutional) interests that determine its
biases and selectivity. It is possible, however, to generalize on the
of the
material presented and evidence cited (11) regarding the role of scientists as
compared to other professional groups.
Not surprisingly, the degree to which scientists are coopted, and are being
delegated part of the function of problem definition and policy formulation,
depends on the nature of the 'political problem' or 'issue' at stake. In other
words, the more limited the problem, and the more focused on science policy
78
Peter Weingart
aspects, the more important is the role granted to the scientists. As the
scope of the problem extends to other areas of policy-making, and thus
other institutionalized interests are affected (in particular industrial), the
scientists have to share their position with "experts" from industry and
political administration, if they are called in at all. This means that judging
on the basis of institutional affiliation, the influence of scientists on policymaking beyond the boundaries of their competence and, more importantly,
areas of legitimate interests may occur as an exception but can hardly be
claimed to be a rule.
What we observe is rather a configuration of "expert groups" with different
institutional bases participating in the policy-making process which we have
called "hybrid community". (The term was chosen to differentiate it from
the scientific community although it is probably even less a "community"
than the academic scientists.) "Hybrid communities" represent the institutional expression of the increased communication pressures between the
differentiated systems of politics, science and the economy. Their function
is to help defme policy problems in terms of systematic knowledge, to translate (operationalize) them into technical goals, to tum them into research,
strategies, development programmes and correlate policy measures, all of
which feed back into the perception and defmition of the policy problems
themselves. The significance of the "hybrid communities", therefore, lies in
their cognitive function as brokers of expert knowledge and political values.
79
80
Peter Weingart
On the level of the project groups, similar examples can be cited. For
81
82
Peter Weingart
83
human subjects eliminates the very value-neutrality that is the basis of the
legitimation of free inquiry (14).
The same phenomenon occurs with the extension of research into areas
which are considered to be risky to human life. While this is, of course, a
familiar experience with respect to technologies (whether derived from
scientific research or not) and has led to prolonged attacks particularly on
technology in the case of nuclear energy, it is more recent in relation to
scientific research proper. So far, very little is known, how a general awareness of risks of certain types or research crystallized into a political issue. In
the case of the DNA-controversy it has been shown that it originated among
biologists in a technical discussion over safety, only then developing in to
a debate over values and was finally extended to a political discussion of
authority and trust (15).
There are numerous indications that the politicization of science is becoming a prevalent phenomenon. One is the recurring appearance of "political"
movements within the scientific community which can be summarily identified with a deepening division between the establishment and the dissenters.
The first incident of this kind, perhaps, was the conflict among the "atomic
scientists" over the use of the A-bomb (16). Since then other issues have led
to a politicization of science in the same manner: the involvement in the
Vietnam war, Fluoridation, environmental protection, nuclear power and
genetic engineering, to name the most well-known.
The pattern of the emergence and eventual disappearance of the political
movements within science, together with their respective political issues, is
fairly regular. As an issue begins to crystallize, two factions of scientific
experts emerge, aligning themselves with the 'non-scientific' groups taking
part in the debate. There are those who take sides with the 'establishment'
- usually government, industry, and military or all three of them - and
there are the dissenting, anti-establishment scientists who challenge the
establishment's position. The latter usually tum to the general public, seeking
to become advocates of its interests and mobilizing it for support. What
follows is an extended and often heated public debate which reveals an often
unbridgeable dissensus among the scientific experts over the interpretation
of the pertinent knowledge and of the interests and implicit political and
ethical convictions guiding these interpretations.
What, then, are the consequences of this politicization of science? Within
84
Peter Weingart
85
which is being overlooked by all those analysts who believe that the scientists
as a group, an establishment or elite or as a "new class" rise to political
power. They fail to recognize that by virtue of its very success, science is
being transformed. In providing knowledge that is instrumentally superior to
all other kinds of knowledge (e.g., experience, amateur knowledge, tradition
and belief) and in becoming involved in the defmition and solution of societal
and political problems, as an institution it loses its boundaries and, thus, its
identity. While, indeed, theoretical (or systematic) knowledge has become the
strategic resource of the post.industrial society, as Bell claims, this implies
neither an undue political power of the scientists, nor the disappearance of
value conflicts, political strife and conflicts of interest, let alone the achieve
ment of a higher rationality based on scientific knowledge.
Notes and References
1. Among others of the titles resounding this theme more than a decade ago were:
D. K. Price, The Scientific Estate, Cambridge, 1967; R. Gilpin, R. Wright (eds.),
Scientists and National Policy Making, New York, 1965; R. E. Lapp, The New
Priesthood, New York, 1965; S. A. Lakoff (ed.), Knowledge and Power, New York,
1966. That this tradition of thought continues to attract sociologists and social
critics is documented in A. Gouldner's The Future of Intellectuals and the Rise of
the New Class, London, 1979.
'
2. Dan Greenberg aptly observed: "The question ... isn't whether the scientists have
found a place in the upper councils of government ... Rather, the question is
whether the scientists' ... new role has swamped the traditional political process,
or whether we have been afflicted by some confusions between' presence and
power" (D. S. Greenberg, 'The Myth of the Scientific Elite', The Public Interest, I,
1965, p. 53).
Considerable confusion also results from the failure to differentiate between the
scientists' influence in 'policy for science' and 'science for politics'. Examples are
abundant, cf. among others, R. C. Wood, 'Scientists and Politics: The Rise of an
Apolitical Elite', in: R. Gilpin, R. Wright (eds.), op. cit., p. 55; A. M. Weinberg,
'Criteria for Scientific Choice', in: Knowledge and Power, S. A. Lakoff (ed.), op.
cit., p. 409; D. K. Price, op. cit., pp.12 and 97.
I also miss the needed clarity in N. Elias' notion of the power of scientific
establishments being derived from the monopolization of the means of orientation
- as he acknowledges himself - but, of course, the basic idea is the same as my
own. Cf. N. Elias, in. this volume (pp.41, 43, 45 ff.).
3. R. E. Lane, 'The Decline of Politics and Ideology in a Knowledgeable Society',
American Sociological Review 31, 658 (1966).
4. D. K. Price, op. cit., pp. 67, 68.
86
Peter Weingart
87
11. Cf.note8,above.
12. For the involved story about the crystallization of this issue, cf. Kiippers et al.,
op. cit.
13. Science, 181 (Sept. 23, 1973), p. 113.
14. Cf. also J. Katz, Experimentation with Human Beingr: The Authority of the
Investigator, Subject, Professions, and State in the Human Experimentation Process,
New York, 1972, esp. p. 312.
15. Cf. D. Nelkin, 'Threats and Promises: Negotiating the Control of Research',
Daedalus 107,2 (1978),200.
16. Cf. A. K. Smith, A Peril and a Hope, The Scientists' Movement in America 194547, Chicago, 1965.
17. For a very detailed analysis of the structural patterns of this process, cf. H.
Nowotny, Kemenergie: Gefahr oder Notwendigkeit?, Frankfurt, 1979. Also B. J.
Culliton, 'Science's Restive Public', Daedalus 107,2 (1978), 147-56.
HOMA KATOUZIAN
89
Norbert Elias, Herminio Martins and Richard Whitley (eds.) , Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 89-109.
Copyright 1982 by D. Reidel Publishing Company.
90
Homa Katouzian
91
Greece or Rome; religious beliefs and metaphysical systems have not been
peculiar to medieval Europe or European scholasticism (we can once again
think of classical Greece and Rome, among other times and places); and
'ridiculous' or 'irrelevant' topics of research and discourse are familiar from
all other ages, not least our own, even though their specific features, formulations and 'solutions' have inevitably varied from one to another.
The absence of a clear, explicit and precise line of demarcation between
scholasticism and science is, I believe, due to the scientific self-centricism
which I mentioned in the previous section. For, once we identify science
with the Renaissance and post-Renaissance experience which we think has
essentially remained the same until this day, it will be automatically assumed
that 'scholasticism' - i.e., that which was confronted by, fought against, and
eventually overthrown by the Renaissance and its aftermath - was also an
essentially homogenous phenomenon, a unique (even though long) historical
experience, and a thing of the past. In other words, the implicit line of
demarcation between science and scholasticism is neither analytical nor sociohistorical; it is, in fact, chronological: scholasticism is medieval, and science is
classical and modem! This, however, begs the original question.
I have already pointed out that our references to the specific methodological, technological, or even substantial characteristics of medieval scholasticism
do not, in themselves, make up a clear mark of distinction between scholasticism and science. As for the purely chronological line of demarcation, it will
be sufficient to pose the following questions to indicate its lack of sufficiency:
if we believe in a progressive theory of history, then it is surely odd that a
thousand years of social development (however gradual as well as discontinuous it may have been) should have been accompanied with a thousand years
of non-science, even anti-science; if, on the other hand, we hold a cyclical
theory of history (or, for that matter, no theory of history at all), then there
is no reason why we should think of scholasticism as a uniquely time-bound
phenomenon.
The key to the problem is I think to be found in our implicit, even unconscious, counterposition not of the experience of a thousand years of medieval
against several centuries of modem learning, but of the late medieval scholasticism against the early modem (i.e., the Renaissance and post-Renaissance)
knowledge and science, which - as a matter of historical fact - co-existed
over a long stretch of time: the dawn of modem science predated the
92
Homa Katouzian
93
had been once supplied by classical science, the 'heirs' to which were merely
repeating or mystifying, rather than adapting and developing it in order to
meet the challenge of their own time.
Early medieval knowledge was - regardless of its relative truth content or
'technological success' - by no means scholastic: 'it posed serious questions,
even including those about the relationship between man and God (as well as
man and man); it had brought God down to earth (or sent man up to heaven)
but - unlike its scholastic 'heir' - it was not keeping man, as well as each and
every man, 'to his station'; it did tackle many questions about the heavens
above, but it seldom asked 'how many angels could collectively stand on top
of a pin'. And, likewise, its founders and propagators were committed, even
dedicated men seeking and promoting such knowledge without fear or favour;
they, like many of those who were to lead the scientific revolution against
their decadent 'heirs' and degenerated 'legacy', forewent moral and material
comfort, and suffered moral and material persecution for the sake of their
theories, their methods and their principles. 8t Thomas Aquinas may have
been a very knowledgeable, certainly a very clever man; but 8t Augustine is
likely to have been a greater and more original thinker.
3. Science and Scholasticism in Space
I said earlier that our self-centric view of science is both temporal and spatial.
I will now make a few remaks on the spatial aspects of the problem, with a
brief reference to Islamic science and Islamic scholasticism. As it happens,
Islamic science began to rise and (in its own context) attain maturity, at
about the same time as (European) medieval scholarship began to decline and
decay; while, on the other hand, its decline and degeneration into Islamic
scholasticism took place in the successive periods of the gradual emergence
and rapid development of modern European science.
Islamic science and philosophy was, rather like the early (European)
medieval scholarship, a predominantly semetico-Hellenistic synthesis of its
own kind. It extensively drew on the neo-Platonics of Alexandria, GraecoRoman gnosticism and stoicism, Graeco-Persian medical and allied sciences,
as well as a both direct and indirect knowledge of Hebraic, Arabian, Christian
and Iranian metaphysics, politics, ethics and mysticism. It was 'Islamic' both
in so far as it was a direct result of the Islamic social revolution, and in as
94
Homa Katouzian
95
96
Homa Katouzian
became great, no country had enough space for me; as my value increased, I
lacked a purchaser.
I refer to Avicenna's voluminous, but relatively unknown, writings on
'scientific method', where he says:
When you are told that an 'Ari! [Islamic gnostic] refrains from having his daily food for
a long time, you should easily accept it and realise that such a thing is a well known
possibility in nature ...
And further
You may hear reports from the [practices] of the 'Ari! community which, being extraordinary, you may reject; for example, such claims as: an 'Ari! asked for it to rain ...
and it did rain ... Reflect, and do not rush into judgement! For such things are part of
the secrets of nature ....
He then proceeds to bring out the significance of these remarks by telling his
reader not to base his attitude on an automatic rejection of whatever he fmds
to be extraordinary, because, he says, he who rejects everything which lacks
an immediate basis in reason or observation is not very different from he who
accepts everything without argument or evidence:
It is therefore your duty to reflect upon all such [extraordinary 1 cases, and tend to
regard them as being within the realm of possibility, until you have better arguments
[than those offered in their favour] for rejecting them ... (6).
The great scientific spirit contained in these simple and clear precepts has,
I believe, never been surpassed; and if it is re-Iearned it will be to the great
advantage of our contemporary science and society. Avicenna was a scientist
and philosopher; he was neither a mystic nor (very probably) a committed
Muslim. And it is in that role and capacity that he is warning his fellow scientists against dogmatism - even including 'scientific' dogmatism. This is an
11 th-century rational philosopher who regards nature as being capable of
phenomena which are not immediately testable by human senses and reason,
emphasizes argument and evidence as the most important means of scientific
investigation, and tolerates metaphysical theories and claims as being 'within
the realm of possibility', not to be dismissed without patient and deliberate
reflection. Compare with the methodological criteria of logical positivism, the
most influential among 20th-century scientists, which regard 'unverifiable'
statements as 'meaningless noises', etc., etc. No wonder that when, a few
97
years ago, it was claimed that a performer could bend forks and spoons by
the touch of his fingers, many natural scientists were shocked into declaring
(for example, in the correspondence columns of The Times) that if this was
true, then the whole basis of their scientific beliefs would have to be false.
Indeed, this small and apparently insignificant event should give one a lot of
food for thought, for a science that is easily frightened by an alleged fork
bender must be in a very peculiar state.
4. Hallmarks of Science and Scholasticism
Medieval scholasticism emerged with the growing 'success', that is, the increasing social and intellectual respectability and fashionability of the new
synthesis. And it became firmly established as a social and intellectual institution, concurrently with the total establishment of the received doctrine
as a body of dogma. This was a social institution both in the concrete and
in the abstract senses of the term: in its concrete sense, it consisted of a
community of scholars in various seats and institutions of learning; in its
abstract sense, it was a rigid framework for the maintenance as well as defence
of the established dogma. The two faces of this institution together made
up the most distinctive hallmark of medieval scholasticism (in fact of any
scholasticism), that is, dogmatism. There can be a significant distinction
between dogmatism and mere belief in a given body of dogma, because (a)
dogmatism is an outlook which may be applied to any body of knowledge,
and (b) it dermes an active commitment to a body of dogma, and it involves
not only a rejection of all alternative systems of thought (both old and new),
but also resistance against any real addition to, or extension of the established
body of knowledge: it fights the 'infidel', hounds the 'heretic' and hunts the
'witch' all alike.
It was this (concrete and abstract) social and intellectual institution which
led to the more specific and more familiar characteristics of medieval scholasticism: a general preoccupation with dynamically useless and irrelevant, but
statically safe and 'solvable' puzzles; a rejection of every source, method and
topic of inquiry other than those contained within the established dogmatic
framework; a dominant (if implicit) view of knowledge and learning in terms
of their 'success' and 'usefulness' - both for the individual scholars and for
those whom they served; the excommunication, damnation and persecution
98
Homa Katouzian
of anyone who - in the very spirit of the accredited founders of the established dogma - dared to break the scholastic rules, and propose new methods
and ideas (not necessarily in conflict with the basic tenets of the established
dogma), or by pointing out the basic differences between the outlooks of the
acclaimed founders of the dogma and the established patriarchs of scholastic
dogmatism. The cases ofWycliffe, Jan Hus and Martin Luther merely provide
well-known 'religious' examples in an otherwise much wider movement both
in quantity and in quality. They generally fought from without the established
framework, and they eventually destroyed it.
The most basic hallmark of science in history is its anti-dogmatism, its opposition not merely to specific faiths, but to faith in general. And for this very
reason, there is no basic conflict between science and religion: anti-dogmatic
science does not have to reject all religious beliefs as nonsense because it has
no claim to total universal knowledge. On the contrary, it is the science which
itself is based on a faith - even if that is the faith in reason - which is essentially in conflict with religious beliefs, for the faith in any religious system
automatically excludes the faith in another. And this is an important reason
why some recent and contemporary scientists would tend to suffer from a
conscious intellectual schizophrenia if they happen to have any religious
beliefs; for they would like to hold on to both their scientific and their
religious beliefs as matters offaith, which are, by definition, contradictory (7).
It may be argued that I have merely produced a 'definition' of science
according to my own value system, and that my 'defmition' has no validity
in historical fact: that 'normal science' is bound to be a dogmatic system of
thought, that it is precisely this dogmatism which accounts for its 'success',
and that the science of which I am speaking is an abnormal, rare, extraordinary or revolutionary phenomenon (8). This argument which, in a different
form, has been very influential among modern academics can be put in a
different way, namely, that scholasticism is a much more deeply entrenched
and tenacious phenomenon than is science. I readily admit that dogmatism
(regardless of the dogma involved) has, up to now, been a more 'normal'
phenomenon than its opposite, just as well as tyranny has been historically
more 'normal' - i.e., much more frequently experienced - than democracy.
But this does not mean that tyranny is necessary for social progress, and that
does not prove that the growth of scientific knowledge is inevitably bound up
with dogmatism (9).
99
100
Homa Katouzian
surprising that those who put a lot of emphasis on the social and technological
usefUlness of science, also tend to be the strongest advocates of 'the neutrality
of science'.
However, the argument can be analytically split into two parts. First, that
a critical and undogmatic search for natural and social knowledge is likely to
be the surest and most efficient method of solving technological and social
problems. This is a statement of fact (not value) although it may be incorrect.
Second, that this is what we should do, which is a statement of value, analogous to those concerning the promotion of freedom, democracy, welfare etc.
It is in this sense that the hallmarks of science are unchangeable social norms
and values, even though their specific implications and applications do change
through time and space.
The basic problem with some recent historical generalizations about the
process of scientific research - and, in particular, those of Kuhn and Lakatos
- is that they generalize not so much about the (timeless) scientific vision
and approach which has always been critical, non-conformist and, if you
will, revolutionary, but about professional, institutionalized and established
activities which make up a well-known trade in history: some people used
to become Masters of Arts in order to gain admission into the scholastic
establishments (or receive permission from them for independent teaching),
just as well as others had to become master craftsmen before being accepted
by their relevant guilds. They now become university lecturers and professors,
while the others become skilled workers, managers, etc. Normal science tends
to be a description of the institutional behaviour and 'rules of the game' of
this particular trade - a trade which (as I will try to show in the following
section) made very little contribution to the development of modern European
knowledge and science until the 19th century. Indeed, it would be very
interesting to try and discover how much of the work carried out in those
institutions has stood the test of the time, especially in comparison with what
was produced outside and often in defiance of them.
It is sometimes believed that great scientific achievements must await the
arrival of extraordinarily intelligent and hardworking individuals. This at best,
is a half truth, however. There are always such men everywhere, and - in
particular - it is very unlikely that the European thinkers and scientists of
the past few centuries had greater native abilities than their contemporary
official scholastics. The reason why even the most learned, intelligent and
101
s.
102
Homa Katouzian
103
104
Homa Katouzian
the Marxes, etc. - is too well known to merit a detailed discussion: modern
philosophy and social science posed a more direct threat to the scholastic
dogma (and its social base) than many of the natural sciences; and their social
relevance as intellectual channels for social change was thought to be not
'useful', but positively dangerous. Adam Smith -- a Glasgow graduate and
professor, and a scholar at Balliol College, Oxford - might appear to be a
prominent exception to this list. Yet, his even-tempered approach to life
and learning notwithstanding, he left Oxford after six years of unhappiness
(without the slightest prospect for the future), and later resigned his chair at
Glasgow (which was an unusually progressive university for its time) when he
was forty (11).
In fact, it is from the 19th-century (albeit with some variations among
different European countries), that universities begin to become centres of
modern learning; from the end of the First World War when they (together
with similar research institutions) begin to become so, more or less uniquely;
and from the end of the Second World War when they (and the other institutions) became the established and 'mass-based' academic power centres as we
know them (12). And although there are no statistics available, there now
seems to be a growing tendency for the departure of some of their actual or
potential members who are ill at ease with their 'rules of the game', either
into other profeSSions, or into alternative (formal and informal) intellectual
institutions. Such a tendency would, in the short run, consolidate their
internal position; but, if there are any lessons to be learnt from history, its
long-run consequences could be of a very different kind.
Up to now, I have been discussing the institutionalist aspects of contemporary science and scientific frameworks. I will now try and explain, and
trace the developments of what I mean by their dogmatism. By dogmatism I
do not mean a belief in any given body of knowledge (or dogma), but an
uncritical commitment to any body of dogma, regardless of its relative truth
content, technological products, or - for that matter - ideological implications. That is to say, I regard dogmatism not as a mere belief in a set of ideas,
which would be different in different ages and among different societies as
well as individuals; but as a unique, timeless, attitude which may be shared
by all of them regardless of time, place or content. Medieval scholasticism
was dogmatic not because of what it believed in, but how and on what basis
it believed in it. To put the point in a different way, I am proposing an
105
analytical (as well as social) distinction between the beliefin something, even
though it may be 'false', and the faith in anything, even though it may be
'true': a belief in 'falsehood' is, by itself, not dogmatic, because it implies the
possibility of modification, correction and even rejection; the faith in 'truth'
is, however, dogmatic, because - by definition - it is final and unalterable. I
have purposefully put the case as strongly and categorically as possible in
order to make my understanding of the concept of dogmatism absolutely
clear. Otherwise, belief and faith are not entirely separable; and if we ever
knew 'the truth', the distinction between them would (and should) completely
disappear.
I now assert that - apart from their institutionalist aspects which I have
already discussed, and in spite of all their obvious and significant differences
which I have repeatedly emphasized - what makes the contemporary approach to knowledge and science (or the new scholasticism) fundamentally
comparable to the old scholasticism is that they are both based on a faith: the
one in God, the other in Reason. The historical process which led to the
decline of the old faith (though not necessarily its associated beliefs) and the
rise of the new, also describes the process of the rise of modern European
science and its subsequent decline (though not necessarily its associated body
of knowledge): the rise and decline of an open and critical search for greater
knowledge of reality which had been in direct conflict with the dogmatism of
its own time.
The first part of this historical process, that is, the war of attrition (although ina series of winning battles) against the old faith, the old dogmatism,
is well known and beyond general dispute; but its second part, that is, the
gradual emergence and establishment of the new faith, has received very little
attention: our scientific self-centricism has automatically excluded it from
explicit discussion; and the avalanche of detailed and extensive debates on
'the scientific method' has effectively buried it under its great heap.
The faith in Reason began to emerge at about the end of the 18th-century,
when the triumphant (social and intellectual) war against the old faith was
nearing its completion. The French philosophes and their sympathizers arid
fellow travellers everywhere, who had played a great role in the closing stages
of this victory, also sowed the seeds of the new faith which, however, can
have been no part of their intention. Voltaire's implicit hints that Christ may
have been a complete fiction, his explicit claim that Muhariunad, the prophet
106
Homa Katouzian
of Islam, must have been a cunning charlatan, and his pragmatic interpretation of religious beliefs - 'if God did not exist, we would have had to invent
him' - are probably the simplest and clearest examples of the approach to
knowledge which would automatically dismiss whatever is not immediately
available to reason and sense-perception as pure lies, inventions and fabrications (13).
To be fair, there was no science of psychology at the time, and theories of
history such as Vico's and Montesquieu's had not managed (and they have
not even yet managed) to temper universalist views of science and society.
Yet we know that, seven centuries earlier, Avicenna had succeeded in piercing
through the veil of dogmatic rationalism, of the faith in reason. There is,
however, a stronger historical explanation (even justification) for the attitude
of Voltaire and the others, namely, the fact that they were still in opposition,
still engaged in an ongoing struggle against the established dogmatism. But
this justification cannot be extended beyond its time-spatial context, for
example, to those contemporary scientists who would be seriously disturbed
by the performance of an alleged fork bender. For Voltaire's intention
been to defeat, not to defend, the established dogmatism.
However, the active and positive cult of worship of reason begins towards
the end of the 18th, and the beginning of the 19th-centuries. Voltaire's (and
the others') faith in reason had been implicit; and it had served as a weapon
in the fight against the old faith. Whereas the faith in reason (regardless
of otherwise significant differences within itself on 'the correct scientific
method' - on Objectivism, Subjectivism, and the like) nurtured and developed
by Condorcet, Laplace, de Tracy, St Simon, Comte, Bentham and others was
active, positive and assertive: it promised - in various, sometimes conflicting
ways - the coming of the Kingdom of Reason; and, both then and later,
it used similar methods to those of the old scholasticism in castigating its
critics. Yet the process was bound to be incomplete so long as the upholders
of this new faith (who, I emphasize, had and still have serious quarrels among
themselves, just as this was true of the adherents of the scholastic faith) were
still 'free floating'; that is, as long as the new faith had not been completely
internalized by scholastic institutionalism. I have already discussed the
process of its completion.
Science and scholasticism are timeless values and, in practice, established
academic frameworks may be based on either of them. The relative truth
107
content and/or the social and technological utility of any body of knowledge
does not, by itself, determine its scientific or scholastic status. The crucial
test is the relative prevalence (or absence) of dogmatism, including an effective machinery for the uncritical defence and preservation of established
doctrines and methods. Science promotes and scholasticism retards the
progress of knowledge.
Notes and References
*
1.
2.
3.
4.
5.
6.
7.
I am grateful to Alex Dolby, Herminio Martins, Helga Nowotny and Richard Whitley
for helpful comments on various drafts of this paper. The views and the mistakes
are entirely my responsibility, of course.
For example, even our observations on the Graeco-Roman scientific experience and
legacy, (which in any case we tend to regard as a preparatory exercise for modern
developments), are usually coloured by our own contemporary experience: ourown
participation in, and identification with contemporary science. What I have in mind
is something like Vico's (and, perhaps Montesquieu's) view that historical events
can be properly understood in their full contexts, and that a mere 'knowledge' of
them is often misleading. See Isaiah Berlin, Vico and Herder, London: Hogarth
Press, 1976: and Against the Cu"ent; London: Hogarth Press, 1979.
Adam Smith once observed that alth<.>ugh China was a richer country than England,
the economic and social characteristics of the English society in the 18th-century
provided it with much better prospects than China. Clarence Ayres, founder of
the American neo-institutionalist school of economics, has associated economic
progress in the whole of human history with a 'frontier' existence and mentality.
By analogy, scientific progress may be the result of a dynamic process of groping
for new knowledge, rather than a static act of grasping it.
See 'Aqilf, AtMr al-Wuzara, Teheran: Muhaddis, n.d., pp. 230-1. The form and
content of both these letters leave little doubt about their authenticity, although
(for a variety of detailed historical reasons) they may, in fact, have been exchanged
between Gazzali and Niziim al-Mulk's son, or his successor Taj al-Mulk.
See his Gulistan for this particular reference, and the whole of his CoOected Works
for the broader argument.
E. G. Browne, Arabian Medicine (The FitzPatrick Lectures delivered at the College
of Physicians, November 1919 and November 1920), Cambridge: Cambridge
University Press, 1962.
See al-Ishamt wa al-Tanbihat (Hints and Warnings], Cairo: Sulayman Duniya, n.d.,
pp. 850-902.
Bertrand Russell's contrast between religion and science implies much the same
thing (i.e., science as dynamic discovery in contrast to religion as static faith)
because he implicitly identifies all religion as dogmatic and all science as critical.
That is, his concept of religion is exclusively that of medieval scholasticism and his
concept of science is exclusively that of the classical as well as early modern science.
He therefore overlooks the fact that official science may become dogmatic, and
108
8.
9.
10.
11.
12.
13.
Homa Katouzian
unofficial religion could be critical. In other words, the wider and more basic
contrast is between dogmatism and criticism, almost regardless of their formal
frameworks. See his Religion and Science, London: Oxford University Press, 1960.
Berlin's reference to 'Averroist lines' - i.e., a complete separability of scientific and
religious beliefs held by the same individual - is of some relevance to my argument,
although I would advocate the possibility of a lack of basic conflict only as long as
the body of knowledge contained within a religious (institutional) framework is not
held dogmatically (there would, of course, have to be a few basic principles which
must be so held, or there would be no religion nor science; but this is not what I
have in mmd). 'Averroist' is probably an adjective of Averros (Ibn al-Rushd) the
great Islamic philosopher, but in any case, many of the great Islamic thinkers held
such a view, which was not usually or predominantly a matter of expediency; for,
in taking this view, they opened up Islamic doctrines themselves to wide interpretation and extensions in the service of intellectual development. See Isaiah Berlin,
Vicoand Herder, op. cit., p. 79.
I am obviously alluding to Thomas Kuhn's The Structure of Scientific Revolutions
(Chicago: The University of Chicago Press, second edition, 1970).
See my critique of Kuhn's thesis in H. Katouzian, Ideology and Method in Economics, London: Macmillan, and New York: New York University Press, 1980,
Chapters 4 and 5.
See further, ibid., Chapter 5.
Adam Smith gave up his eleven-year scholarship at Balliol College, Oxford, after
six years, and returned to his home town in Scotland. His letters of the Oxford
period are highly critical of that experience, and betray symptoms of alienation
and depression. Many years later, he resigned his chair in Glasgow in order to take
up the tutorship of a young nobleman with the immediate prospect of a long
journey to France, and the companionship of the leading French intellectuals, none
of whom were university scholars. He has aired some of his critical views of the
contemporary British universities in the Wealth ofNations. See John Rae, The Life
of Adam Smith, Jacob Viner (ed.), pub. A. M. Kelley, New York, 1965; and, Edwin
Cannan (ed.), The Wealth of Nations, London: University Paperbacks, 1961, Vol. 2.
Book V., Chapter 1, Part III.
It should be clear that the reference here is not to universities as opposed to
other (complementary or competitive) institutions such as the state, independent
and publicly-endowed research institutes, but to all of them taken together as
opposed to unprofessional learning and research by individuals and voluntary
associations through which most of the foundation stones of modern science
had been laid. And to forestall misunderstanding, it should perhaps be emphasized that no part of this paper's argument is intended to mean that knowledge
can be properly acquired via what is usually dubbed as 'pseudo-science' - i.e.,
an undisciplined 'search' in the 'world of reality', often leading to a grand vision.
On the contrary, the point being made is that precisely that which we recognize
as science and scientific progress has seldom flourished in the formal institutions
of learning, and that we appear to be living in an age when, once more in history,
these institutions
to be failing in their primary task of advancing knowledge.
These and other examples concerning 'religious matters' are intended purely for
short-hand illustrative purposes, and carry no other significance. For all we know,
109
Voltaire (and others) may have been right on these specific issues; but, although a
good deal of rational argument and evidence can be brought against them, my
purpose is simply to point out the dogmatism of their outlook and approach, rather
than the falsehood (or truth) of their views.
CYNTHIA HAY
The origins of academies of science have been located in the 17th and 18th
centuries, as scientific institutions linked with science as the pursuit of gentlemen and amateurs. The development of science as a profession, which
gathered pace throughout the latter half of the 19th century, was accompanied by the formation of an appropriate institutional basis, in the form of
research laboratories, and the acquisition and expansion of a foothold for
science in universities (4). In this context, the survival and adaptation of
academies of science in the 20th century, as more than vestigial scientific
institutions, attracts attention.
The history of the National Academy of Sciences is part of a broader
history of American science, which has been extensively explored by historians "acutely conscious of the importance of the problem of an elite in a
111
Norbert Elias, Herminio Martins and Richard Whitley (eds.), Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 111-119.
Copyright 1982 by D. Reidel Publishing Company.
112
Cynthia Hay
113
and technology (13). The expansion of the Academy was part of this development. It was fostered by a sequence of energetic officials who were officials
in the Academy, as well as being involved in other institutions in the government and science network.
Academies of science have adapted to the 20th century in a variety of
ways. West German academies of science have undergone a shift, from their
18th-century origins as honourary discussion societies, to undertaking longterm, collective efforts on the behalf of scientists (14): tasks which might be
described as those of a gentlemen's trade union. The Soviet Academy of
Sciences is closely linked with the government; it has survived at the pinnacle
of a hierarchy of institutions for implementing policies for science, subject to
occasional sudden changes in its organization and powers, at the government's
decision (15).
Some of the particular indigenous factors which have contributed to the
adaptation of the National Academy to science for policy can be identified
by comparison with efforts to establish an institution analogous to the
Academy in Japan. Occupation authorities, in setting up the Science Council
of Japan in 1949, "envisaged" a role for it "equivalent" to the Academy in
s'cience and government relations (16). The Science Council never achieved
such a position, for a number of reasons. These have to do with differing
political ambiances, and the related availability of scientists to take an active
part in science policy.
The Japanese scientific community traditionally has been distant from the
government; academics have been largely opposed to the ruling political party
in Japan. Consequently the Science Council took on the role of critic of the
government, rather than that of scientific adviser acting at the behest of the
government. Scientific statesmen have been thin on the ground in Japan, not
only because of the political attitudes of the scientific community, but also
because "collegial" decision-making procedures are customary in Japan,
which reduce the importance of particular individuals (17). The intransigent
Science Council was effectively kicked upstairs, with the establishment of a
more accommodating body, the Council for Science and Technology,in 1959,
with this body the government has maintained its traditional dominance in
science policy decisions.
Contrasts with the political environment of the Academy are marked. The
willingness of some highly-regarded American scientists to seek and maintain
114
Cynthia Hay
close links with government has long been a feature of American science, and
contributed to the founding of the Academy. The political views of American
scientists are various, and not, on the whole, as sharply opposed to the
government as in the Japanese situation. Individual scientists who have been
opposed to particular policies of a government have nonetheless been willing
on occasion to be linked with the government through science advisory
institutions (18).
The make-up of the Academy fosters its position of political cooperation
in government service. The backgrounds of elected members reflect the
characteristics of a scientific elite. The membership is heavily weighted
towards individuals from premier academic institutions, who have flourished
professionally, starting from their early recognition as especially able scientists (19).
The Academy is widely recognized as a conservative body of scientists,
both in its membership, and in the general tenor of its recommendations.
The lengthy and complicated procedures of election to membership tend to
favour non-abrasive scientists (20). The Academy, then, may be described
as an establishment in a classic sense: that is, a body of people integrated into
a network of major institutions, whose attitudes reflect this position. This
description stops short; the Academy as a scientific advisor, can more fruitfully be analyzed as a scientific establishment.
II
The Academy as a scientific advisor illustrates many of the features of scientific establishments which have been analyzed by Norbert Elias, Richard
Whitley, and Peter Weingart,in their papers for this volume (21). It commands
and controls resources for studies of the uses of science and technology,
through its privileged position as a quasi-official body, with a right to private
deliberations and with access to information through its relationship with
government, it commands resources in a different sense, by means of its
prestige, which enables the Academy to attract and to draw on the services
and expertise of scientists.
The position of the Academy as a "Supreme Court of science" whose
activities are largely financed by the government agencies who bring cases for
judgment reflects the way in which the visibility of a scientific elite may
115
exaggerate its political power. The Academy provides advice on issues whose
identification as policy problems stems from elsewhere.
The second set of processes which are reflected in the workings of the
Academy as a scientific adviser derive from the development of scientific
establishments. Weingart's analysis of scientific establishments links them
closely with the growth of science for policy, which is the direction in which
the Academy has deliberately sought to develop in recent years; a major aim
of the reforms of the Academy's structure in the early 1970s was to equip the
Academy better to contribute to science for policy discussions.
A number of writers have discussed the characteristic features and problems of science for policy: they describe the locus of scientific advice of this
kind, by terms such as Saloman's technonature and Weinberg's trans-science,
which identify its issues as being of a different order from those of science,
and not susceptible to resolution by scientific means (22). Weingart argues
that issues of this order have emerged in political arenas as a result of two
correlative processes, which he calls the de-institutionalization of science and
the politicization of science. The de-institutionalization of science refers to
the emergence of science as a free-floating approach, which has encroached
upon areas and issues traditionally the province of other institutions, such
as religion or the family; science is seen as a source of prescriptions and
judgments on topics such as child-rearing or healthy diets. The range of
scientific advice has been vastly extended, but the other side of the coin is the
politicization of science: with issues of this order, the grounds for decision
and the powers of decision are not matters of science. The location and
incorporation of scientific advice in policy processes reflects this ambiguous
position, which might be said to be one of authority without power.
The notion that scientists with particular competences have especial
qualities to provide advice on broad questions illustrates the emergence of
the scientific establishment as a locus of science above and beyond the
achievements of particular sciences: the hypostatization of science is its title
to authority. Science has acquired an authority surplus to the achievements
of particular sciences; the aura of the scientific establishment overshadows, in
more than a figurative sense, the lack of adequate criteria or institutions to
resolve cross-disciplinary questions on a level with the procedures adopted
within disciplines (23). The dearth of such procedures is critical in science
for policy. A recent controversy about an Academy report suggests how
116
Cynthia Hay
117
118
Cynthia Hay
119
23. Don Price, 'Money and Influence: the Links of Science to Public Policy', Daedalus
103 (1974) 97-114.
24. Food and Nutrition Board, Division of Biological Sciences, Assembly of Life Sciences, Towards HealthfUl Diets, Washington D. C.: National Academy of Sciences,
1980.
25. Jane E. Brody, 'When Scientists Disagree, Cholesterol is in Fat City', New York
Times 17 June 1980.
PART II
EDWARD YOXEN
University of Manchester
1. Introduction
1980 saw the appointment of the distinguished physiologist, Sir Andrew
Huxley, to the Presidency of the Royal Society in Britain and the joint award
of the Nobel prize for physiology or medicine to three molecular biologists,
Frederick Sanger, Paul Berg and Walter Gilbert for their various contributions
to the practice of genetic manipulation. Huxley's selection as the titular head
of the British scientific establishment is evidence of its deep conservatism (1).
The Nobel award illustrates where the newer fields of biology are now moving,
towards lucrative and contentious industrial involvement, and the ruthless
competition of corporate research. Even amongst this tiny group of rewarded
scientists there are some striking contrasts. Huxley comes from an older
discipline than molecular biology and is the scion of an established intellectual
family. As a pure scientist at the head of an- elite institution, he can draw
upon a rich cultural vocabulary to reaffirm a traditional, seemingly autonomous role for science in resolute opposition to the economic imperative for
change. Sanger, a Cambridge biochemist without a public identity except
as the winner of a second Nobel prize, epitomizes productive research in
a patrician academic context. On the other hand, Berg has shown more
Zivilcourage and innovative skill in the politics of science by organizing
an unprecedented moratorium on their new research field as an exercise
in the management of public concern. Finally, Gilbert, who crossed from
theoretical physics into molecular biology, has shown a desire for power and
fmancia! gain from the applications of his research. His involvement with
the newly formed research corporation, Biogen, has led to controversy at
Harvard (2) and reveals an attitude to science quite different in spirit from
123
Norbert Elias, Herminio Martins and Richard Whitley (eds.), Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 123-143.
Copyright 1982 by D. Reidel Publishing Company.
124
Edward Yoxen
But I also want to stress the importance of the activity which Norbert Elias
describes as "the administration of a specific fund of symbolic representations, which can serve man as a means of orientation" (4). It is precisely
the claim to be able to conceptualize and act upon the most fundamental,
constitutive processes of life, a claim that itself has a history, that has conferred upon molecular biology its power in particular social conditions. The
formation of its establishment has come about, I shall suggest, through a
certain type of interaction between specific conceptions of life and a process
of institutional change in biomedical research (5).
125
trace the roots of the former back to the second and third decades of the
20th century, to a period of expansion and institutional change in science
and medicine, as the dominant classes of the major industrialized nations
sought to generate the technical expertise required to run the new business
corporations, to administer new government institutions and to supply new,
more scientifically-oriented hospital-based forms of medical care. In America,
because of the opportunities and challenges of its industrialization and the
relative weakness of constraining political and cultural traditions, by contrast
with Europe, this process of educational and profeSSional development was
pursued with particular vigour (6). To this end, industrialists like John D.
Rockefeller, and Andrew Carnegie, created evolving, managed philanthropic
institutions, which were to direct resources into secondary and higher education, public health, scientific and medical research, medical education and
aesthetic culture, according to some kind of very general strategy (7). Their
aim as members of a newly-emergent corporate ruling class with enormous
private fortunes drawn from the oil and steel industries was to transcend mere
charity by creating a scientific philanthropy under the centralized control of
a group of Trustees and Foundation officials. The resulting disbursement
of funds was to produce institutions, skills and policies that would sustain
commitment to the changing social relations of urban and industrial life.
The managed, scientific programme of donation br such Foundations was
intended to stabilize and consolidate the corporate state. In Germany, similar
initiatives were more closely tied to the provision of government funds, as in
the founding of the Kaiser-Wilhelm-Gesellschaft in 1911 and the subsequent
appearance of its various research institutes. In Britain the relative lack of
enthusiasm of the long established ruling elites for science and technology
weakened the effect of any such reforms with the result that when the
Research Councils were set up after the First World War, they were dominated
by a very traditional, academic views of science and medicine, and had not
the money to effect major changes in the structure of research.
American foundations were therefore the expression of a particular view
of society and politics, namely the ideology of progressivism. They have
126
Edward Yoxen
enormous funds at their disposal and from their inception produced changes
in public health, medical education, and higher learning on an international
scale. As organizations they developed through time, passing through phases
of rationalization and reorientation and were organized around particular
management principles. Very soon the expertise on which they drew, and the
resources they developed were such that specific policies could be forced
upon recalcitrant disciplinary or professional groups as, for example, in the
reform of American medical education (8).
In the 1920s, the Rockefeller philanthropic trusts disbursed large sums of
money to build up American universities (9). By 1929 this policy came under
review, and the structure of the enterprise was itself re-cast by Raymond
Fosdick, J. D. Rockefeller's chief counsel, and a single Foundation created
with five separate divisions. Under Fosdick's influence, the strategy of general
support for scientific and professional education was shifted to one of more
concentrated support for specific areas of science, conceived as an increasingly
highly specialized, differentiated and professionalized activity (10).
The Rockefeller exercise in philanthropy had since its inception placed a
major emphasis on scientific medicine and public health. There can be no
doubt that the accumulated experience in the Foundation in the medical
domain, in dealing with eminent researchers as advisers, in setting up major
public health programs in America and elsewhere, in promoting educational
and clinical reforms, in building up medical schools and confronting dissenting
medical groups was an important influence on the outcome of the policy
deliberations of the late 1920s and early 1930s. In 1931, a Natural Sciences
Division was set up, to which a young mathematical physicist, Warren Weaver,
was brought in as Director, who in due course set up a programme of research
into 'vital processes', which subsequently became known as 'molecular biology'. When Weaver arrived, it was a period of tension between the Trustees
and Foundation Officials.
To consolidate his reforms, Robert Kohler has suggested, Fosdick established a Committee of Appraisal in 1934 to review the scientific rationale of
the Natural Sciences Division (11). Simon Flexner, the influential director of
the Rockefeller Institute for Medical Research in New York and chairman
of the committee, though not a Trustee, was critical of the directions and
policies. But Weaver's programme, albeit with some criticism of his support
for work in psychobiology, was endorsed by other committee members who
127
128
Edward Yoxen
The phrase 'vital process' ... was the label which I wanted to have tied to the programme
because I wanted to get it away from ancient limitations adhering to the word 'biology'
... But for some reason or other, some of my biological friends found this phrase 'vital
process' a little precious. The trustees never quite understood whether they were being
sold a bill of goods or not ... But it was biology, in its purpose; it was biology in its
ultimate orientation. It was not by any manner of means biology in terms of its classical
technique (14).
The aim was to defme a set of phenomena common to all organisms, which
could be analyzed experimentally in quantitative physico-chemical terms, the
understanding of which would throw light on more specific issues like human
behaviour, reproduction or the causes of disease. As Weaver's project gained
legitimacy and coherence, his attitudes and methods changed, and the contribution of his programme to an intellectual transformation in biology became
significant in its own right. As the 1930s progressed, he was able to rely
upon the general support of the Rockefeller Foundation Trustees and a
confidence in his skill and sensitivity granted to him by a network of leading
scientists, whose advice he sought, and to whom he channelled research
funds, often on a scale that no other agency, governmental or private, would
have been able to match. This'is not the place to consider in detail how
Weaver fulfilled this role in science, but it is important to emphasize that
he should be seen as a kind of 'product champion', transmitting ideas and
rhetorics and negotiating between two influential groups, the Trustees and
university scientists (15). Furthermore, he was forced to be selective in his
patronage and became able to enforce and consolidate his intervention in
biology.
So far I have considered the economic, political and institutional preconditions for this kind of development. In the next section I want to consider the
state of technical and theoretical developments in the life sciences by the late
1920s and early 1930s, and why molecular biology specifically should have
resulted from this kind of initiative. We then go on to consider what kinds of
institutions molecular biologists built up in the post-war period, and the kinds
of actions they pursued, having 'established' their science, given its rhetorical
and technical potential as research organized around a new conception of life.
3. Conceptions of Life and a New Strategy for Biology
One of the ideas I want to emphasize is the shifting view of what counts as
129
130
Edward Yoxen
131
132
Edward Yoxen
133
his desire to stimulate new fields and new methods (29). Microbiology also
had a close connection with medicine, but in the 1930s could scarcely have
seemed a likely source of general biological understanding. But, at least, in
the very special context of the Institut Pasteur in Paris, an important school
of molecular biologists emerged from this disciplinary base (30). Finally,
evolutionary theory and systematics seemed too classical in approach to
Weaver, too loosely related to experiment and analysis and too distant from
physics. Embryology had flowered in the 1890s, but the 1930s was in difficulties. In any case, Weaver's task as a patron was to evade or transcend the
divisions of existing disciplines, not least because it would shift the balance
of power in his favour, and to strengthen the analytic and methodological
rigour of biological research, according to criteria operating in physics and
chemistry. What he chose to do was to cluster together a set of projects of a
transdisciplinary character amongst them Pauling's structural chemistry of
biological macromolecules, Astbury's crystallography of DNA and protein,
Schoenheimer's use of radioisotopes in biochemistry, Beadle's work in
biochemical genetics, Svedberg's development of the ultracentrifuge, Perutz's
X-ray studies of haemoglobin. The intention was to intensify research, to
allow people to travel to the more progressive labs, to promote the use of
physical theories and methods in biology and to interest physical scientists
in the life sciences (31). The effect was to organize interest and resources
around specific fundamental problems, particularly those concerned with
molecular structure, and to alter the standards of what counted as an explanation in biology away from questions of adaptation, behaviour, macroscopic
form and the process of evolution, and towards functional questions that
were considered in terms of molecular structure. It co-ordinated a set of
approaches to the general question of how to derive biological insights from
knowledge of the structure of biological macromolecules. As the 1930s wore
on, increasing attention was paid to the structure of deoxyribonucleic acid
(DNA), although the bulk of the evidence suggested that the main constituent
of the hereditary material was protein and not DNA. By the late 1930s one of
the physicists sponsored by Weaver, Max Delbrtick, had begun to use bacterial
viruses as the simplest possible system in which to study the mechanism of
gene replication (32). By the late 1940s, Delbrtick had refmed his ideas to a
point where he and his collaborator Salvador Luria, who worked in Muiler's
department in Indiana, were able to establish a research network and training
134
Edward Yoxen
135
This situation began to change as the more ambitious lobbyists for science
began to exploit the opportunity afforded by the impact of vastly increased
wartime expenditures on science, and the levels of funding increased. In 1944
the statutes regulating the U.S. Public Health Service (PHS) were revised,
which allowed the service to carry out the role played by the CMR. In 1945
the plans for a National Science Foundation in America were drawn up by
Bush and his advisers. In the event, it took a further five years of political
debates for this institution to be established. Stric1dand has suggested therefore that in such a situation expansionist officials in the PHS saw the need to
conduct a campaign of offensive appropriation of research monies, by taking
over in early 1946 the medical contracts placed by the OSRD, using money
made available by the fall in the world price of penicillin (38). Very quickly
the money allocated to the National Institute of Health (NIH) as the research
arm of the Public Health Service began to increase. In fiscal 1947 the figure
was almost $8 million, ten times the level in 1939; in fiscal 1948 it was $26
million. But despite this kind of Congressional endorsement and despite the
involvement of leading academic figures in medical research in the allocation
of State funds, disagreement continued over how research should be organized
and conducted strategically. This was linked to a political struggle over the
plan for a nationalized health care system in the United States. In particular,
the American Medical Association (AMA) used its power to obstruct governmental reforms that would subsidize hospitals, health insurance, and expand
medical education. Given this situation, Strickland has suggested that the
rapid growth of biomedical research support in the 1940s and 1950s can be
explained in terms of the interaction of a number of political lobbies and
factions (39). They included the AMA, which was prepared to accept the
State support of biomedical research whilst vigorously resisting State subsidy
to the hospital system; a private, very wealthy, well-organized research lobby,
with close links with leading NIH officials and influence on Capitol Hill; a
136
Edward Yoxen
bureaucratic elite within the NIH, which was able to expand its own power by
working in alliance for a long while with the research lobby; and a group of
Senators and Congressmen, whose political influence and status derived from
a particular kind of interest in health and its promotion via scientific research.
Thus one can view the expansion of research support in the United States
at the expense of any government intervention into the market for health
care as a resolution of forces, as a policy founded on compromise. One con"
sequence was the formation of a network of laboratory directors controlling
the flow of funds into universities and medical schools (40); they became in
effect a research establishment. Another consequence was the recurrent call
for greater centralized control over expenditure, and the introduction of a
management structure within the granting agencies like NIH and NSF to
achieve more and more precise regulation of specific programmes and policies
(41). Indeed the academic advisers - what Brown calls the 'superacademic
general staff' - have increasingly found themselves constrained by centrally
imposed research strategies (42). This situation has also appeared in Britain
with the initiatives and policies symbolized by the Trend and Rothschild
reports and the continuing attempts to restructure and redirect State-supported scientific research. Throughout the world, the control by research
elites over money, status, and access to research opportunities within an
increasingly specialized set of biological disCiplines has intensified the degree
of competition and assisted the formation of a very esoteric, introspective
research culture. Heirich has described the institutional consequences of these
economic relations upon university research and the fmancial necessity to
form big labs in which hierarchically-ordered researchers produce results
that are continually exchanged for new grants (43). Pickvance has portrayed
the psychological and emotional effects of undergoing and reacting against
socialization into these sorts of roles as a biologist (44). The effect of this
increasingly intense drive for results was to enforce an exclusive concentration
on just these problems, phenomena, or aspects of an organism that ftlled the
needs of specific research programmes. Molecular biologists learnt to take an
increasingly instrumental attitude to the living material with which they
worked and were forced through the pressures of international competition
to intensIfy the degree of specialization in the problem-solving skills required
to stay in the field. As a result, their relationship with, and conception of,
nature changed. Ufe came to be viewed in informational terms.
137
138
Edward Yoxen
and the functional (48). But its power within a theoretical rationale is also
a means to power on a social or professional level. Molecular biology offers
a general language for thinking about disparate biological problems in a
historical context where increasingly there has been a need for research
bureaucracies to order research strategically, to compare different highly
specialized lines of analysis, and manage the production of results. In this
sense a unifying, informational, systems biology serves a managerial research
system. This is not to say that the concept of a genetic programme is only a
social construction, but that its organizing role within a system of metaphors
has been developed through the interaction of the State, the biomedical
research establishment and an interdisciplinary alliance in biology in the
post-war era.
s.
Finally, I want to return to the ideas of spectable and display with which I
began, since I believe that the rise of molecular biology was, in one sense, a
peculiarly public process. In other words, one part of the negotiation for
status by molecular biologiSts was a more general cultural intervention,
mobilizing public sentiments in support of their conception of life. Indeed it
is interesting to reflect that the post-war period in which television rapidly
emerged as a major medium of mass communication, was also the period in
which molecular biology emerged as a new speciality in the life sciences.
Television offered a new cultural resource to be used in the process of institution building, by playing upon the theme of an attack on the 'secrets of life'
and molecular biologists were quick to seize upon this opportunity. Two
examples are presented here. There are just two instances of a whole genre of
popularization and media reporting about 'The Biological Revolution', that
appeared in the late 19508 and is still around today, in forms that increasingly
stress the technological possibilities (49). Its influence on how the general
public thinks about biology, evolution, genetics and the nature of human
existence has been profound. It amounts to an attempt to inculcate a new
sensibility about the power and role of science.
In the 1950s the BBC began to experiment with science programmes
on British television. The success of some of the early series on science,
such as Your Life in Their Hands which was devoted to surgery, encouraged
139
enterprising producers within the BBC to make more expensive and spectacular programmes at a time when its patrician styles were being challenged
by the new commercial television stations. The fIrst such Science Spectacular,
What is Life?, was shown on the 1st December 1959, presented by Raymond
Baxter and Professor Michael Swann. One of the main themes of the programme that appears in the drafts and planning discussions is the idea that
biologists are facing a redefmition of 'life' through their work in the cell
biology and genetics (SO). This was portrayed by a vast hemispherical model
of the cell, through which Baxter and Swann walk at one point, pointing to
ribosomes and mitochondria and so on. It was a carefully guided tour around
the new world of the cell, and various sections of the journey are overseen by
other scientists, mmed in their labs or with their equipment into the studio.
One has the strong impression, watching the programme twenty years later,
of a kind of cultural alliance behind the programme, between ambitious
young producers in the BBC exploiting the visual possibilities of a new
medium and a new topic for broadcasting and a group of scientists, of junior
professorial rank, recently admitted to the Establishment, doing biology in
a new way at the molecular level and seeing the need to mobilize cultural
support for their work to assist the consolidation of their research programme.
Some of these people went on with these exercises in popularization. John
Kendrew (now Sir John and until recently secretary-general of the European
Molecular Biology Organization) presented a series on molecular biology
in the mid-6Os called The Thread of Life. Lord Swann was until recently
Chairman of the BBC.
In the United States, one can fmd similar instances of this use of a new
resource to develop a new public sensitivity to the power of contemporary
science to analyze and control the fundamental features of living matter,
indeed to the extent that 'life' could be created in a test tube. In 1967 a
group of Californian scientists succeeded in creating an artificial copy of a
virus on a natural template. Their institutional sponsors, Stanford University
and the National Institutes of Health, carefully orchestrated the publicity of
this scientific result so as to maximize its coverage as 'the creation of life in
a test-tube'. President Johnson, also seeking to redeem his political image,
added to the publicity with the result that this item reached the front page
of virtually every U.S. daily paper, in the rather ambiguous terms of 'the
synthesis of life'. In this instance, the deliberate playing on a new conception
140
Edward Yoxen
of life in the media almost jeopardized the results it was supposed to achieve,
namely the increased legitimacy of biomedical research institutions. One can
see the same kind of attempts to exploit the possibilities of media attention
with the 1974 moratorium on recombinant DNA research and the same kind
of ambivalent response. I have elsewhere discussed this incident in greater
detail (51). It provides an interesting example of the highly complex and
culturally variable responses to the deployment of what Elias has called 'a
specific fund of symbolic representations' as part of a strategy for professional
or institutional legitimation.
6. Conclusion
Molecular biology came into existence through the interaction, on the one
hand, of a specific promotional initiative in research, by a private philanthropic foundation seeking to sponsor new developments in biomedical
research and to extend a system of control over that research by transcending
the constraints of existing discipline boundaries, and on the other, a reductionist strategy for biology drawn from genetics, that re-ordered the
conceptual field of the life sciences, so as to place the problem of gene
structure in a central position. This interaction produced a cluster of projects
concerned with the structural analysis of biological macromolecules, which
formed a programme labelled 'molecular biology'. As the programme developed, the genetic elements in it acquired an organizing role at the research
front in biology. In the post-war conditions of rapid institutional growth
the field of molecular biology produced highly significant technical results.
The formation of a research establishment controlling governmental funds
followed from the political and economic conditions that made such unprecedented sums of money available and the ability of molecular biologiSts
to offer a rationale for thinking about and ordering the life sciences and to
mobilize various kinds of support for their conception of life.
Acknowledgements
This paper has been through a number of drafts and in its revision I have been
helped by discussions with a large number of people. I am particularly grateful to Graham Cox, to the organizers of the conference in June 1980 on 'The
141
Recasting of the Sciences between the two World Wars' in Florence and
Rome, to Robert Seidel, Pnina Abir-Am, Richard Whitley, Jon Harwood,
John Pickstone and Barbara Wilkinson.
Notes and References
1. Anon, 'Democracy and the "Royal" (editorial)" New Scientist 88 (4 December
1980) 618; A. Huxley, 'Evidence, Clues and Motives in Science', Times Higher
Education Supplement (2 September 1977) 4-6; R. M. Young, 'Can We Really
Distinguish Fact from Value in Science?' ibid. (23 September 1977) 6;A. Huxley,
'Fact and Value Must Not be Confused', ibid. (7 October 1977) 27.
2. Anon, 'Harvard backs off recombinant DNA' (editorial), Nature 288 (4 December
1980) 423-4.
3. R. D. Whitley, this volume.
4. N. Elias, this volume.
5. These ideas have also been developed in E. J. Yoxen, 'The Social Impact of Molecular Biology' (unpublished Ph.D Thesis, University of Cambridge, 1978); .and in
E. J. Yoxen, 'Life as a Productive Force: Capitalising the Science and Technology
of Molecular Biology', in R. M. Young and L. Levidow (eds.), Studies in the Labour
Process, Vol. 1 (London: CSE Books, 1981), pp. 66-122.
6. N. Reingold (ed.), The Sciences in the American Context: New Perspectives (Washington: Smithsonian Institution Press, 1979); D. J. Kevles, The Physicists: the
History of a Scientific Community in Modern America (New York: Random House,
1979); D. Noble,America by Design: Science, Technology and the Rise of Corporate
Capitalism (London: Knopf, 1977).
7. J. Weinstein, The Corporate Ideal in the Liberal State, 1900-1918 (Boston: Beacon
Press, 1968).
8. E. R. Brown, Rockefeller Medicine Men: Medicine and Capitalism in America
(London: University of California Press, 1979).
9. R. B. Fosdick, Adventure in Giving: the Story of the General Education Board
(New York: Harper, 1962).
10. R. E. Kohler, 'The Management of Science: the Experience of Warren Weaver and
the Rockefeller Foundation Programme in Molecular Biology', Minerva 14 (1976)
279-306; R. S. Seidel, 'The Evolution of Science Policy in the Foundation: The
Rockefeller and Carnegie Philanthropies' Support of the Physical Sciences' (paper
given to the conference on 'Recasting Science between the Wars', Rome, 1980).
11. Kohler,op. cit., pp. 291-6.
12. /bid.
13. L. Hodgkin, 'The Politics of the Physical Sciences', Radical Science Journal 4
(1976) 29-60; see also Yoxen, 1981 (note 5 above).
14. Warren Weaver, Transcript of Oral History Memoir, Oral History Office, Butler
Library, Columbia University (Record No. 343, 3 vols., 1961), pp. 333-4.
15. The term 'product champion' was suggested by Robert Seidel; see note 10 above.
16. T. S. Hall, Ideas of Life and Matter: Studies in the History of General Physiology,
600 BC. -AD 1900 (London: University Of Chicago Press, 1969).
142
Edward Yoxen
17. F. Jacob, The Logic of Living Systems: a History of Heredity (London: Allen Lane,
1974).
18. R. Dubos, The Professor, the Institute and DNA (New York: Rockefeller University
Press, 1976).
19. D. J. Haraway, o-ystals, Fabrics and Fields: Metaphors of Organicism in TwentiethCentury Developmental Biology (London: Yale University Press, 1976).
20. G. Allen, Life Science in the Twentieth Century (Cambridge: Cambridge University
Press, 1978).
21. G. Allen, Thomas Hunt Morgan: the Man and his Science (Princeton: Princeton
University Press, 1978).
22. H. J. Muller, 'The Gene as the Basis of life', Proceedings of the International
Congress of Plant Science 1 (1926) 897-921; reprinted in H. J. Muller, Studies
in Genetics (Bloomington: University ofIndiana Press, 1964) pp. 188-204.
23. H. J. Muller, 'The Need of Physics in the Attack on the Fundamental Problems of
Genetics', Scientific Monthly 44 (1936) 210-14.
24. E. A. Carlson, 'An Unacknowledged Founding of Molecular Biology: H. J. Muller's
Contribution to Gene Theory, 1910-1936', Journal of the Hisotry of Biology 4
(1971) 149-70; N. Roll-Hansen, 'Drosophila Genetics: a Reductionist Research
Program', Journal of the History ofBiology 11 (1978) 159-210.
25. E. J. Yoxen, 'Where Does SchrOdinger's What is Life? Belong in the History of
Molecular Biology', History of Science 17 (1979) 17-52.
26. R. C. Olby, The Path to the Double Helix (London: Macmillan, 1974).
27. J. D. Bernal, 'W. T. Astbury', Biographical Memoirs of FeUows of the Royal Society
9 (1963) 1-36.
28. R. E. Kohler, 'Medical Reform and Biomedical Science: Biochemistry - a Case
Study', in M. J. Vogel, C. E. Rosenberg (eds.), The Therapeutic Revolution
(Philadelphia: University of Pennsylvania Press, 1979), pp. 27-66.
29. Interestingly, in the British research council system physiologists acted as sponsors
of molecular biological projects, which seemed likely to push physiological understanding down to the molecular level.
30. A. Lwoff, A. Ullmann, Origins of Molecular Biology: a Tribute to Jacques Monod
(London: Academic Press, 1979).
31. C. H. Waddington, 'Some European Contributions to the Prehistory of Molecular
Biology', Nature 221 (25 January 1969) 318-21.
32. J. Cairns, G. S. Stent, and J. D. Watson, (eds.),Phage and the OriginsofMoiecular
Biology (New York: Cold Spring Harbor Laboratory 1966).
33. N. C. Mullins, 'The Development of a Scientific Speciality: The Phage Group and
the Origins of Molecular Biology',Minerva 10 (1972) 51-82.
34. J. P. Baxter, Scientists Against Time (1946: Cambridge: MIT Press, 1968).
35. C. Pursell, 'Science Agencies in World War II: The OSRD and its Challengers', in
Reingold,op. cit. (note 6 above).
36. Olby,op. cit. (note 26 above), p. 328.
37. S. P. Strickland, Politics, Science and Dread Disease: a Short History of u.s.
Medical Research Policy (Cambridge: Harvard University Press, 1972), p. 35.
38. Op. cit., pp. 25-6.
39. Op. cit., Chapters 3-7.
40. M. Heirich, 'Why We Avoid the Key Questions: How Shifts in Funding of Scientific
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
143
RAINER HOHLFELD
Institut {iir Gesellschaft und Wissenschaft an der Universitiit ErlangenNiimberg, Erlangen
I. Introduction
During the last five years, the war against cancer has become a serious issue in
political debates among health politicians, science policy-makers, the medical
profession, scientists and the general public in the Federal Republic of
Germany. The German debate started with a time lag of at least ten years
compared to the U.s. cancer crusade and cancer debates which culminated
in the biggest research programme ever known in the biosciences - the U.S.
National Cancer Program Plan of 1974 (1). The questions in the German
discussion were: are we lagging behind the American effort? What remains to
be done in view of this American crash programme? Does it make any sense
to deal with the cancer problem as if it were a 'biological' moonshot? How
can German cancer efforts best be coordinated? (2) The German government
and science administration's response to these questions was to set the priorities and goals within the frame of the "Programme of the Federal Government for the Promotion of Research and Development in the Service of
Health Care" (3).
By derming the health policy demand for research, this programme gave
rise to intervention in the self-regulation operation within health research and
related areas. Cancer as a chronic disease won high priority within the frame
of the programme. Both basic scientists in biological research and medical
scientists issued strong warnings against planning cancer research at alL
But at the same time, the scientific communities concerned did make some
proposals and recommendations. In view of the requirements to be met by
research oriented towards so complex a goal as reducing the incidence of
145
Norbert Elias, Herminio Martins and Richard Whitley
Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982.145-168.
Copyright 1982 by D. Reidel Publishing Company.
146
Rainer Hohlfeld
147
of organisms in the smallest units which could be identified by light microscopy: cells and cell organelles, e.g., chromosomes. For hereditary phenomena,
the Mendel-Morgan chromosomal theory of heredity is still a firm base for
animal and plant breeding today (8).
The theoretical revolution of biological research in our century consisted
in the replacement of the terms of classical biological theories by the terms of
physics and chemistry: atoms and molecules. The goal of this new way of
conceiving living matter is to reduce biological units and entities such as genes
or membranes or cells to chemical terms. Thereby, biological phenomena are
inserted within the range of chemical subjects and biological systems are
conceived as chemical systems. Biology as 'molecular biology' (9) becomes a
particular type of chemistry. In relation to traditional biology, the classical
macrotheories which conceive living matter in units of cells and cell elements
are 'reduced' to or 'replaced' by molecular microtheories (10). The reductionist concept constitutes the paradigm of molecular biology and defines a
researchprogramllle.
The dynamics of the reductionist programme can be characterized by the
intention of molecular biologists to theorize the particular subject areas of
biology, which can be phenomenologically distingUished from one another, in
molecular terms. The various ranges of phenomena can be ordered according
to a
of biological structures: bacterial cell, higher cell
(eucaryotic cell), cell system, organism, population. According to the biological functions, this hierarchy can be defmed as follows: metabolism, heredity,
cellular differentiation and development (11) (see Figure 1).
The reductionist programme started with an explanation of cell composition and the molecular mechanisms underlying metabolism in biochemistry
(12). With the enzyme theory, biochemistry reached the level of theoretical
maturity (13). Historically, this programme was continued by the molecular
theory of the gene - molecular genetics - which started in the thirties and
was accomplished for the molecular genetics of bacteria about thirty years
later (14).
At present, the strategy of molecular biologists is to extend the validity
of the molecular genetics paradigm to biological phenomena of higher complexity than bacteria, i.e., to the molecular genetics of the higher cell and
the processes of cell differentiation and cell development, which take place
in cellular systems (15). I have tried to localize the research frontiers of
I.
cell (e.g.,
sporulation)
differentiation
Integration of
different genetic
systems
Ontogenesis
Morphogenesis,
embryology
Biochemical
integration
(e.g. humoral
regulation of
glucose-metabolism)
Organism
Developmental
genetics
BioChemi!uy of
cellular systems
(e.g. blood cells)
Cellular systems
Phylogenesis
Population
genetics
Biochemical
ecology
Population
Fig. 1. Diagram of Research Fronts in Molecular Biology in Relation to Biological Subject Areas. The Scheme is ordered by levels of
growing complexity of biological structure (from left to right) and of biological function (from top to bottom).
development
Cytogenetics
Genetics of
bacteria
heredity
Cellular
/
biochemistry
Higher cell
(eucaryotes)
Bacterial
biochemistry
Bacterial cell
(procaryotes)
Function
metabolism
Structure
Degree of
Complexity
5:
g:
00
149
150
Rainer Hohlfeld
dynamics and research strategies to the extent that these components are
involved in the cancer issue (20).
3.1. The Ideology of Basic Research: Science Cannot be Planned Towards
Political Goals
Scientists in molecular biology still hold to the idea of science for its own
sake. In their eyes, the laws of scientific progress are determined by the
structure of nature itself, the results cannot be anticipated. Science is an
enterprise of experts and research can be regulated by no one but the scientists themselves. For all these arguments science policy planning is viewed
with considerable scepticism:
Directing research, whatever kind of method used, results in bad research performed by
people who will let themselves be influenced by science policy because they lack any
true scientific motivation (21).
Nonetheless, the belief system of autonomous science offers an interpretation of the guide-line for the utilization of knowledge: First, since the success
of research ca1inot be programmed, society must wait for theories to reach
maturity, whereby a new base is built for technological applications, and,
second, a clear distinction must be made between basic and applied science.
The 'supply-model' of science and with it the dichotomy of basic and applied
science result from this interpretation.
Combining both these elements - the claim that science must be directed
by the experts themselves who are to maintain their authority in the planning
of research goals and that the dichotomy of basic and applied science be
taken account of - the German Research Association, as the spokesman of
the republic of science in Germany, concluded:
Targeted research can only begin when the requisite knowledge has been established on a
flIm basis. Accordingly, the relationship between basic research and targeted research
must not evolve to the detriment of the former. Basic research must be promoted on a
wider scale. This requires additional funding but certainly not research programming
(22).
Cell biology is the field which is thought to produce the knowledge essential to understanding the cancer problem. In the opinion of the biologists, a
precondition for the solution of the cancer problem is the clarification of the
laws of cell biology.
151
Basic research, in particular cell biology, must generate the necessary knowledge before
any real breakthrough can be expected to occur. Impatience, no matter how justified
and understandable on the part of millions of cancer patients, should not make either
scientific organizations or the politicians adopt measures which ultimately swallow up
vast amounts of money without bringing any real success (23).
This way of thinking reduces the cancer problem to key events in biological
processes, which must be explained by molecular theories. Only the theoretical breakthrough is seen to provide the rational base for any real improvement
in fighting cancer as well as for a transition from a half-way technology to a
medical high technology. In the eyes of basic scientists the theoretical strategy
in cancer is supported by the fact that 'steel, beam and drug' as the conventional tools in fighting the disease have left research marking time instead of
moving ahead.
3.2. The First Step Towards Application: Experimental Cancer Research
Experimental cancer research is a type of research located behind the frontiers of 'true' science, structured by the still unsolved fundamental theoretical
questions. Scientific activity in experimental cancer research is not determined by the internal dynamics of scientific advance but by goal orientation.
Compared to basic research, experimental systems and research problems are
selected from the perspective of solving problems of practical relevance rather
than that of accumulating pure knowledge. Scientists working in this field,
however do share with basic scientists the idea that health problems of this
type must be solved by scientific instruments on the basis of the clarification
of the underlying biological mechanisms and that this in tum requires 'high
technology' .
Concerning the biology of cancer, we know too little in the largest sense of this word to
be able to work out a rational chemotherapy for cancer, i.e., the pharmaceutic products
we are working with today mean that we are using crude sledge-hammer methods which
kill healthy cells as much as diseased cells. We are not yet able to strike at cancer while
protecting the healthy parts of the organism which we need to protect to maintain life
(24).
The other idea held in common by basic scientists and workers in experimental cancer research, is that work is the enterprise of experts who must
have free choice in selecting the problem areas they intend to attack:
152
Rainer Hohlfeld
There are but a few experts in this field. Who should be entitled to tell us what we
should be doing - policy-makers in the public health system or perhaps journalists who
have taken up this or that trend in this business or computer experts, or anyone else for
that matter? (25)
Once one goes beyond the level of the common features of experimental
cancer research, it is possible to distinguish three different types of research
enterprises, different both with respect to the maturity of theory development and to the 'intrinsic' or 'instrumental' role of the scientists concerned
(27).
These researchers, then, are working with the high standards and methodology of the academic disciplines. For this reason this work is not regarded as
being 'bread-and-butter research' .
A structure of the research enterprise which is best described as a specialty
has evolved (29), with such features as its own reward system, communication
exchange, workshops and meetings. The forming of specialties of this type is
exemplified by the. case of research on the metabolism of nitrosamines, a
group of carcinogenic agents synthesized in the acid environment of the
153
stomach, which cause cancers of the liver, throat and oesophagus. In the
words of one of the workers in this field:
There are perhaps some 120 scientists working on the nitrosamines. We all know each
other and what each of us is doing. Once a year, or perhaps every two years, we arrange
meetings to debate the relevance of our results and to coordinate our work in a very
informal way. So we do not need any external research programming (30).
Applied research behind the fronts of theory building. When intrinsic scientific motivation combined with the 'external attitude' of goal orientation
occurs in research fields in which the theoretical key questions are still
unsolved - as is the case of molecular cell biology, developmental biology
and immunology - a type of research evolves which most closely corresponds
to 'applied' research in the usual sense of the term. Scientists doing this type
of research try to transfer the progress made on pure systems at the research
fronts to clinically relevant experimental systems, by trial and error methods;
they attempt to elaborate systems which are more complex but immediately
relevant for cancer treatment. Such systems are often deSignated as 'dirty'
systems. This designation refers to systems which are too complex and too
difficult to reproduce to use the highly sophisticated methods of the research
fronts and which therefore offer no scientific breakthrough. This type of
'transfer research' is research behind the fronts of scientific progress (31).
The scientists concerned with this kind of cancer research, and who
are from their training biologists, have split loyalties. The external goal is
convincing and it is easy to identify with it. But cancer research is not frontline research, it has no great status. No clear-cut identification with disciplines
and research fronts is possible here.
'Cancer researcher' is nearly an invective. The cancer researcher is placed between two
worlds. These are the clinic on the one side and prestigeful basic research, for instance
within the Max Planck Society, on the other. The people engaged in basic research will
claim that cancer research is not a solid basis, that such preoccupation with cancer as the
main theme is a restriction of large-scale basic research (32).
154
Rainer Hohlfeld
Nonetheless, the scientists concerned will defend their work, or rather its
ranking, by stressing its importance for the patient:
As I see the matter - and this may be a subjective view - decades of biochemical and
virological research involving great expense has in no way helped the patient in the
clinic. As far as I am concerned, therefore, chemotherapeutic research has greater value
than does biochemistry or virology (34).
The daily business: strategic planning of biomedical research in the pharmaceutical industry. For researchers in the pharmaceutical industry it is
not regarded as improper to be involved in practical matters. The strategic
organization of experimenting and coordination of activities of departments
are such everyday practical matters. All kinds of scientific models, methods
and systems, such as 'finalized' type of research in biochemistry, 'transfer
research' to develop new inhibitors of tumor growth, and simple trial-anderror empiricism in drug screening are undertaken and funded. The one
exception is basic research concerned with the fundamental theoretical
problems. Here one does not know in advance where and how to set the
priorities and this is considered to be too risky for the industry.
Instrumental reason defines the identity of the industrial research worker.
He Or she will not hesitate to cooperate with public institutions, as for instance
with the U.S. National Cancer Institutes, in programmes for cancer drug
development. Proceeding in line with the economic objectives set by the
155
156
Rainer Hohlfeld
with major problems arising from the career pattern of the medical scientists.
So far the physicians' and clinicians' career-pattern do not encompass research. The clinical researcher has been trained as a physician and must then
specialize so that he is never able to free himself from the constraints of his
'original' training and career. This actually means that the clinical researcher
cannot simply pursue his own research interests and develop competences
comparable to those of workers in basic research. In medical training, research
is regarded as a dead-end street. In addition, the physician who is not engaged
in research but who 'treats' patients, i.e., manipulates them directly, in the
eyes of people who are doing research enjoys high prestige and is sure of a
high income. Research in hospitals for all these reasons is not considered
attractive for a German physician.
3.4. The Belief System of the Medical Profession: The Medical Model
In their self-perception, clinicians feel obliged to adopt a bed-side orientation
in their work. Not the originality of the scientific background but the pressure
of immediate and responsible action gains priority in their everyday work.
So far, in their eyes there has been no real breakthrough nor even a real
contribution to medical progress in fighting cancer from molecular biology
(40).
The strategy of treatment is still steel, beam and drug. But it is acknowledged by leading scientists that so far there has been no real standardization
of methods, tumor classification, documentation, and consequently no
possibility to control the success of therapy. Randomized clinical trials are.
recommended as examples for necessary and desired planning:
. . . this includes national and international studies ... on therapy control and therapy
comparisons. Comparisons of the effects of one specific therapy as against the successful
outcome of another while at the same time taking account of the various side-effects, the
remission rates, duration of remission and time of survival require very careful planning
and the performance of cooperative studies (41).
Yet this assessment which focus on the treatment of cancer only is countered by the scientists who have the entire course of the disease in mind, that
is the epidemiologists. The latter draw attention to the restricted vision of the
physicians:
157
All the physicians want to do is to treat diseases once they have made their appearance.
Mostly the physician is not motivated to prevent them. All he sees is the event as such
and all his efforts will be put into treating it; all his thinking and action is directed to
curing (42).
The focus on only one or two aspects of the natural history of cancer
diagnosis and treatment reflects an established tradition of medical thinking
which still governs the medical profession: the 'medical model' (43).
Theoretically, the medical model takes its orientation from the objectifying
sciences, i.e., from physics, chemistry and biology and has absorbed their
empirical-hypothetic procedures. Phenomena are worked on, processed to
be stripped out of their practical context and idealized into laboratory
phenomena, to make them reproducible. The scientist becomes the observer
who fmds himself confronted by an object which he himself has stylized.
Thus, the patient is isolated from his life context and is objectively examined
and observed by the physiCian, guided by the measurable and localizable
phenomena of the disease. The medical perception is somatic, organ-centered,
and aimed strictly at the physiological and pathophysiological functions
(44).
This theoretical model accords with a specific practice of treatment of
disease which takes a purely reactive line: the sick organ, the sick cell must be
cured. Accordingly curing can only set in for the localizable diseased entity,
the cell, the organ, the individual who is ill, but never can be applied to the
environment and to the living conditions of humans.
This scientific concept of disease implies a fragmentation of therapy, as is
exemplified from the canon of the specialities established on the basis of
organs and systems of organs in the medical disciplines and in their special
domains. At the highest level of technical power of control available to the
specialist, treatment of disease is split up into a multitude of disconnected,
singular practices (45). This disciplinary separatism can clearly be identified
in the attitudes and practice of physicians working in the specialities involved
in cancer diagnosis and treatment.
In Germany one-track specialism takes the form of the patient winding up in the hands
of the internist although he needs radiological treatment or needs to be transferred to the
surgeon. But the internist hesitates to send the patient on to radiology because of the
pecuniary losses that this entails (46).
158
Rainer Hohlfeld
for the same motives: they are afraid of losing their patients to the new
oncologists.
The ignorance prevalent in one clinical domain about the work and orientation of the others and the tendency of the heads of clinics to set up 'fiefdoms'
were further demonstrated by the controversies about which discipline is to
have primacy in clinical oncology teams. For these reasons so far it has not
proved possible to establish combined tumor treatment in Germany of the
type current in the Swiss oncological services.
The separation of medical chairs from each other even affects the chances
of cooperation evolving between clinical and empirical research.
Nowhere have I seen so much animosity and at the same time so much cronyism as
is now current in medical practice and in the medical clinics. This is a state of things
which leaves its mark even on the attitudes and work of the assistant physicians, which
affects research, which gives rise to schools of enemies and schools of friends, and which
ultimately has an extremely inhibiting effect (47).
The fight for dominance in cancer research and treatment not only characterizes the disastrous state of the communication between the clinical
domains, but gives rise to sharp controversies between science and medicine
regarding their fields of competence.
Quarrels and squabbles prevail in all the areas in which science impinges on the clinical
domains, except for those in which medicine plays a primary role ... The physician's
social position is a unique one and thereby a priori creates difficulties for cooperation
with the sciences (48).
On the other hand, when scientists argue for tumor centers with integrated
research units and particular 'research patients' physicians reply "We cannot
allow the patients to be left in the hands of the scientists ... We will not be
reduced to the role of agents of the scientists" (49).
The object of cancer epidemiology, is man, not rat nor mouse. Nor is
the object of cancer epidemiology the individual cancer case but groups of
individuals who, compared to other groups within the population, share a
159
high risk of getting cancer. The thinking of the cancer epidemiologist, therefore, focuses not only on the categories of illness but also on the categories of
non-illness. In other words, the epidemiologist's question is what makes
the healthy ill. But his way of thinking creates difficulties for him. "The
clinicians are not able to think in this dichotomy, their thought is always
centered around those who are already ill" (51).
Epidemiology as a science dealing with the human population must start
with risk factors and go on through to the social causes of disease. Its results
can encompass the actual causes of illness and lead to cancer etiology. Etiological research on the basis of population studies may have consequences
which politically are 'uncomfortable'. Stopping production of polyvinylchloride-based plastics in Norway, because it was found that this process
caused occupational cancer, can be seen as a case in point.
The status of cancer epidemiology has been improved by the cases of
occupational cancers and cumulating evidence that many substances which
are annually put on to the market or are by-products of industrial society
may act as carcinogens.
After all, this is the population we are really interested in, namely the human population
... Because this is the case, cancer epidemiology has in the past few years been receiving
increasing attention and this in turn has led to increased activities (52).
Despite this increasing attention, the state of data collection as a base for
epidemiological work in Germany is viewed as being disastrous. There are
only two cancer registers, which is wholly inadequate to supply representative
data for a population of some 60 million. Various reasons are put forward to
explain this wholly inadequate situation for starting any proper work. On the
one hand, there is the question of data protection:
Epidemiological analysis must relate to persons, must encompass the identity of the
patients. This is particularly difficult in our country, one can only say: A burnt child
dreads the fire. The experiences of the Third Reich in this respect after all are not a
pleasant memory. This has to some extent brought epidemiologists into disrepute.
And so we keep on running, trying to catch up with the international standard. Still,
these matters are extremely difficult in all the countries of the West. Here we enter the
domains of privacy and confidentiality, the legal requirement concerning confidential
medical communication, the right of the person to anonymity regarding the private
personal data (53).
On the other hand, this state of affairs is attributed to the strong resistance
160
Rainer Hohlfeld
161
162
Rainer Hohlfeld
On the other hand, for researchers doing this advanced work, research
in mature areas like cell biochemistry and molecular genetics of bacteria is
regarded as 'homework', done to solve the respective problems down to the
last chemical detail. Once the main problems of a subject area are solved, the
leading scientists start migrating with the perspective that a new research
front can now be tackled (56). Less attractive, both from the point of view
of rewards and career opportunities, is work with more complex and 'dirty'
systems.
Even more complex is bed-side oriented clinical research, which is most
relevant for the patient. This type of research is recommended as highly
important, but nobody wants to do it. The methodological standards are
regarded as low and theoretical breakthroughs cannot be expected. So far,
this type of research has not been undertaken on any large scale.
The subject area focussed on the environmental-body relationship and the
etiology of the disease (and which therefore represents the prevention phase
of the disease process) is degraded to a system of medical statistics by the
curatively-oriented medical and basic scientists and enjoys little scientific
recognition.
The attractiveness of the different approaches within cancer research for
'good' scientists and the resultant options for recruitment of new manpower
for the non-academic type of approaches depends largely on this hierarchical
pattern, which is determined by the scientific recognition gradient from 'hard'
to 'soft' science (57). The most rewarded scientific activity is the one which
reduces complex phenomena to immobile stable micro-units - the molecules.
The least rewarded activity is that which deals with the non.reproducible,
changing and mobile units - the psychological and social conditions in the
etiology of cancer (58).
The field of force of basic science and the overall reduction scheme in
conceiving illness is supplemented by an independent center of enormous
institutional and political power: the medical profession and its institutions.
Medical practice and health care, in West Germany monopolized by the
medical profession (59), are not directly linked to the research programme
of the biosciences. The intellectual identity of the medical profession is still
determined by the traditional medical model. The high status of the profession and of the physician is due to his expertise in the exercise of a public
function which has high priority. The reward system within the profession
163
164
Rainer Hohlfeld
Also beyond doubt is the fact that in spite of all the efforts of cancer
organizations, research funding associations, and of the Federal Government,
the present state of German cancer research and cancer policy can still be
characterized by one basic feature: pluralism without organization (64).
Notes and References
*
1.
2.
3.
4.
5.
6.
165
7. I have interviewed 29 scientists both in leading and non-leading position in the areas
of molecular biology (basic research); experimental cancer research, clinical cancer
research, cancer medicine, and epidemiology in the years 1975 and 1976. For a
more systematic foundation of the classification see Section 3. (The author has the
tape recordings and transcribed records of the interviews.) I refer to these records
as, e.g., 'Expert Interview Experimental Cancer Research'.
8. For the development of classic genetics theory, see A. H. Sturtevant,A History of
Genetics, New York and Tokyo: Harper & Sons, 1966.
9. The term molecular biology was introduced by Astbury, 'Adventures in Molecular
Biology', Harvey Lectures 46 (1950) 3-44, to denote the very complex molecules
playing the key role in cellular processes like proteins, and has since then often
been used synonymously with the term 'molecular genetics', e.g. the molecular
conceptions of the phenomena of heredity. The biochemists have used the term
synonymously with biochemistry, that is the approach that traditionally investigates the metabolism of cellular and organismic compounds. From a systematic
point of view, it is by far more conclusive to use the term molecular biology for
the entire reductionist programme and to differentiate between the particular
theoretical approaches such as 'biochemistry', 'molecular genetics', 'molecular
embryology' to exemplify.
10. For my use of the term 'reduced' and 'reductionism', I am referring to K. Schaffner,
'The Peripherality of Reductionism in the Development of Molecular Biology',
Journal of the History of Biology 7 (1974) 11-119. For historical reconstruction I
would prefer to emphasize, the incommensurability of classical and molecular
genetics and in the light of the fate of the old theory to use the term 'replacement'
(cf. D. Hull, cited by Schaffner, ibid., p. 119). The replacement of a theory by
another is equivalent to a paradigm revolution in the sense of T. S. Kuhn; The
Structure of Scientific Revolutions, Chicago: The University of Chicago Press,
1970.
11. This listing makes no claim to completeness but only attempts to grasp the hierarchies inasfar as they are relevant for cancer biology. The principles of 'assembly
of molecules' which characterizes the higher degrees of order of biological processes
and structures defined as 'boundary conditions' (L. L. Gatlin, Information Theory
and the Living System, New York: Columbia University Press, 1972, 14-17) or as
levels of emergence (M. Polanyi, 'Life's Irreducible Structure', Science 160 (1968)
1308-12.) were used by Polanyi to argue against reductionism by pointing out
that precisely these structural principles cannot be reduced to chemical theories.
This position was rejected by K. Schaffner, ibid., 113, who stated that even the
structural relations in biological systems can be conceived in chemical terms.
12. For the developments of theories in biochemistry, see R. E. Kohler, 'The History
of Biochemistry: A Survey', Journal of the History of Biology 8 (1975) 275-318;
G. Allen, 'The Chemical Foundation of Life', in G. Allen, Life Science in the
Twentieth Century, Cambridge: Cambridge University Press, 1978, pp. 147-85.
13. Theoretical maturity of a theory shall be preliminary deImed as the state of a
theory, where the basic principles of explanation and constitution of a definite
range of subjects are coherent with the empirical data., cf. G. Bohme, W. van den
Daele, and W. Krohn, 'Finalization of Science', in Social Science Information 15
(1976) 306-30.
166
Rainer Hohlfeld
14. For the establishment of the molecular genetics approach and its science policy
preconditions, see E. Yoxen, 'Giving Life a New Meaning: the Rise of the Molecular
Biology Establishment', this volume. For the history and periodization of molecular
genetics, see R. Olby, The Path to Double Helix, London: Macmillan, 1974; G.
Stent, The Coming of the Golden Age, Garden City, N.Y.: The Natural History
Press, 1969. For a more systematic interpretation in the sense of a phase model of
scientific development (cf. Bohme, van den Daele, Krohn, ibid.) see R. Hohlfeld,
'Theory Development in Molecular Biology', in: W. Callebaut, M. de Mey, R.
Pinxten, F. Vandamme (eds.), Theory of Knowledge & Science Policy, Ghent:
Communication and Cognition, 1979, pp. 346-57.
15. See E. E. Luria, 36 Lectures in Biology, Cambridge, Mass.: The MIT-Press, 1975,
pp.213-73.
16. The term was proposed for a new level of technology in medicine, achieved by
applying theories of molecular biology to medical subjects (cf. note 19) by L.
Thomas, Aspects of Biomedical Science Policy. Washington: National Academy of
Sciences, 1972. Systematically development of high technologies means that on the
basis of a mature fundamental theory of the subject area a specialized 'daughtertheory' is constructed to deal with the particular technical constraints. This case of
goal oriented theory construction has been termed 'f"malization' in Science; cf.
Bohme, van den Daele, Krohn, op. cit., 1976, Note 13; Hohlfeld, op. cit., 1979,
Note 14.
17. cr. B. Hartley, 'The Bandwagon Begins to Roll',Nature 283 (1980), 122.
18. Luria,op. cit., 1975, (Note 15), p. 248.
19. In this sense, the term is used by one of the leading protagonists of the philosophy
of combining molecular biology and medicine to achieve a high technology in
fighting disease, C. L. Thomas, op. cit., 1972, Note 16.
20. By this classification I refer to P. Weingart, 'On a Sociological Theory of Scientific
Change', in: R. Whitley (ed.), Social Processes of Scientific Development, London
and Boston: Routledge & Kegan Paul, 1974, pp. 45-68.
21. Breuer, H., zur Hausen, H. Oettgen, H. F. Schmidt, C. G. (Hrsg.), Bericht uber ein
Expertentreffen zur Frage der lftiologie und der therapeutischen Beeinjlussung
maligner Tumoren. Schriftenreihe der Deutschen Stiftung fliI Krebsforschung, Bd.
1, Bonn, 1977, p. 41.
22. Deutsche Forschungsgemeinschaft, Bestandsau/nahme Krebsforschung in der
Bundesrepublik Deutschland 1979, Boppard: Harald Boldt Verlag, 1980, p. 200.
23. Ibid., p. 199.
24. Expert Interview Experimental Cancer Research I.
25. Ibid.
26. Ibid.
27. To discriminate between the social roles of scientists with respect to their motives
for doing research, I would hold to the classification of S. Box and S. Cotgrove,
'Scientific Identity, Occupational Selection and Role Strain', The British Journal
of Sociology 27 (1966) 20-8, which draws the line of distinction between the
'intrinsic' role of scientists mainly operating within specific scientific communities
and the 'instrumental' role of scientists working mainly in industrial research and
development.
28. Expert Interview Experimental Cancer Research II.
167
168
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
Rainer Hohlfeld
Ibid.
Expert Interview Epidemiology II.
Op. cit., Note 52.
The basic idea of this reduction scheme is very clearly outlined by Engel, op. cit.,
1977,Note44.
Historically, as reported by an eye-witness, the consensus reached among scientists
at the Cold Spring Harbor Symposium in 1963, at which the clarification of the
genetic code was demonstrated, made molecular geneticists attack the eucaryotic
cell as the subject following area (Expert Interview Molecular Biology II).
In the reduction scheme outlined above I have focussed on the reduction of the
complexity of phenomena. With the connotation of 'dirty' and 'pure' systems, the
focus is on the experimental reproducibility of research objects. The hierarchy
defined from this more operational point of view is discussed by A. Rip under the
terms of 'restrictedness' and 'unrestrictedness' ('The Development of Restrictedness
in the Sciences', this volume). The clarification of the relation of both hierarchies
needs further investigation.
This scheme might serve as a model for the more general value scheme outlined by
N. Elias, 'Scientific Establishments', this volume.
For an analysis and criticism of the German medical establishment and the health
care system see P. LUth, Kritische Medizin, Reinbeck: Rowohlt 1972; H. U. Deppe
(ed.), Vemachliissigste Gesundheit, Koln: Kiepenheuer & Witsch, 1980.
Cf. M. Janicke, Wie das Industriesystem von seinen Miflstiinden projinert, Opladen:
Westdeutscher Verlag, 1979, pp. 82-7.
Cf. A. Weinberg, 'The Coming Age of Biomedical Science', Minerva 4 (1965) 3-14.
Expert Interview Molecular Biology III.
Deutsche Forschungsgemeinschaft, op. cit., 1980 (Note 23), p. 203.
Cf. also the present German parliamentary debate on cancer policy, 'Programm zur
Krebsbekiimpfung', Das Parlament, 31.1.1981, pp. 1-4.
JAMES FLECK
University ofAston in Birmingham
1. Introduction
In this paper, I discuss the role played by scientific establishments in the
development of a particular scientific specialty (1), Artificial Intelligence
(AI), a computer-related area which takes as its broad aim, the construction
of computer programs that model aspects of intelligent behaviour. As with
any discussion of a scientific specialty, the identification of what is involved
is not unproblematic, and the above serves as an indication rather than a
defmition. While the term 'Artificial Intelligence' is used in a variety of
ways (2), there is a discernable group (perhaps approaching the degree of
commonality to be called a community) of researchers who recognize the
term as descriptive of a certain sort of work, and who, if they themselves are
not willing to be directly labelled by the term, can locate themselves with
respect to it.
Unfortunately, there is little or no cOminonly available literature that
systematically charts the scope of this area. It is worthwhile, therefore,
to consider the distinctive socio-cognitive characteristics of research in AI
as a prelude to a fairly specific discussion of the social and institutional
processes involved in the development of the area (3), thus providing a basis
for exploring the usefulness and applicability of the concept of establishment.
2. Socio-Cognitive Characteristics of Artificial Intelligence
The patterns of research in AI exhibit distinctive characteristics, forming a
paradigmatic structure which includes such elements in the scientific activity
as research tools, practices, techniques, methods, models, and theories, as well
169
Norbert Elias, Herminio Martins and Richard Whitley
Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982.169-217.
Copyright 1982 by D. Reidel Publishing Company.
170
James Fleck
as the normative and evaluative aspects for selecting among them (4). They
serve as guidelines and a basis for future research, but are complexly interrelated, often encompassing contradictory facets in tension.
The elements in the paradigmatic structure of AI are as follows:
(1)
(2)
(3)
(4)
(5)
(6)
(7)
171
The craft knowledge of AI is deployed in the construction of computer program models - computational models - of some aspect
of intelligent activity. These models are generally pitched at the
symbolic level of meanings rather than at the physiological level of
the underlying mechanisms. This distinguishes AI from many other
cybernetic approaches, and from much computer simulation work.
The focus on intelligent behaviour provides a disciplinary context psychology - but, due to the great variety of social interpretations
and applications of the term 'intelligence', specific goals for research
are not thereby dictated. This lends AI a similarity with what can be
termed instrument or technique-based specialties, such as Xray
crystallography (6),
are free to be applied to various goals.
Associated with the wide variety of specific examples of intelligent
activity that have been modelled, a clear research area differentiation
has emerged since the early 1960s in which subareas have developed
their own particular specialist guidelines and techniques, focussed on
their own more circumscribed concerns. The research areas that
could be identified in the early 1960s were game playing, theorem
proving, cognitive modelling (an emphasis on models with psychological verisimilitude). natural language, machine vision, and a range
of specific applications (7) some of which have themselves subsequently differentiated out into well defined research areas. These
research areas (or strands of research (8 constitute a primary
setting for scientific activity, and consequently have been one of the
basic arenas for competition among practitioners, as will become
evident.
These cognitive characteristics, or elements of the AI paradigmatic structure are, of course, at a very general level. They open up a huge cognitive
space which offers wide opportunities for exploration, and were elaborated
at a fairly early stage in essentially their complete form, while subsequent
work has largely exploited the possibilities opened up. This overview of the
development of AI invites comparison with Edge and Mulkays' account of the
development of radio astronomy: the initial discovery of radio waves from
space opened up the possibility of a new source of astronomical information
- a new cognitive space - which was subsequently exploited by ever more
172
James Fleck
173
174
James Fleck
175
176
James Fleck
The Second World War acted as a melting pot for various quite different lines
of research and disciplines. In the intense concentration on the common goal
of winning the war, traditional disciplinary boundaries were broached and
new areas of research emerged, such as information theory, operations
research, cybernetics and, of course, the development of the digital computer
itself. These areas of research can be broadly characterized as the software
sciences, in that they focussed on pattern and organization rather than on
substance or matter - the concern of the natural sciences such as physics and
chemistry.
Cybernetics, a rather general field given a name and identity by Norbert
Wiener's classic book: Cybernetics - Control and Communication in the
Animal and the Machine, was concerned with the essential similarities between machines and biological processes (13). Work in the area developed
during the 1940s, and involved such approaches as the comparison of biological and neurophysiological processes with electrical circuits and networks of
artificial neurons, or the investigation of the general principles of adaptation
in self-organizing systems - systems which were rich in feedback connections,
and would settle into stable configurations after being disturbed (14).
The advent of the digital computer in the early 1950s heralded a new
approach which sought to build models of intelligent processes at the symbolic level (15). Concepts were represented and operated on directly in the
computer using high level programming languages. These 'symbolic' models
represented intelligent activity at the level of thought itself, rather than at the
level of the physiological mechanisms underlying thought, thus contrasting
sharply with other cybernetic approaches.
In 1952, a conference was held under the rubric 'Automata Studies' (16).
This conference, organized largely by John McCarthy, was intended by him
to attract proponents of the symbolic modelling approach. It failed in this
aim, and attracted contributions more clearly in the other cybernetic traditions. This determined McCarthy to "nail the flag to the mast the next
time", which he did by explicitly using the term 'artificial intelligence' in a
subsequent summer school held at Dartmouth College in the United States
in 1956, to discuss 'the possibility of constructing genuinely intelligent
machines' (17). The official title was 'The Dartmouth Summer Research
177
178
James Fleck
179
180
James Fleck
m
IIWiener -Rosenbleuth
Princeton
von Neumann
I
I
Morgetstern
7 L
\Sha'fmon -McCu1.1och
Newell
SRI
cCart
CMU
.)L
{FeigenbaUlll.}-- ---<
Hart
Duda
(Manna
Nilsson
(Feldman
(Evans
Lederberg
Buchanan
<Ems
f--(Fikes)
Walker
<- !---<Waldinger)
Sacerdoti
<- !---<ColeS)
Fischler
(Quinl
(Colby
Green)
Pople
Binford
<-
Shortliffe
<---
(Quillian
Slagle)
Banerji
Hewitt
Kaplan
Eastman
Robinson
Abelson
(Greenblatt}---
lTenenbaum)
Luckham
Wang
Raphael)
Earnest
Floyd
Elsewhere
Bobrow)
(Guzman
(Falk
(Montanaro
<Roberts
Samuel
<-
Berliner
Barrow
Minsky
Papert
Fredkin
Winograd)
Pohl
Schank)
Amarel
Charniak
McDermott
Bledsoe)
Sussman
Boyer
HWlt
Goguen_
Goldstein
Rieger
Marr
Sinunons
(Waltz
Winston
Rosenfeld
Nevins
Fig. 1. The Establishment in the United States: 1960-mid-1970s. The members of the
establishment were derived from a consideration of the editorial board of the Arti/icial
Intelligence Journal, conference organizing committees, invited conference s,Peakers and
panel members, supplemented by well known researchers as judged on the basis of a
reading of the literature. They include 73 out of a total of upwards of 500 contributors
to the area. Available data was limited, but indicated that only some 11 out of the 73
had not worked at some time or done a PhD at one of the big four Artificial Intelligence
centres: Massachussetts Institute of Technology (MIT); Carnegie Mellon University
(CMU); Stanford University (SU); and Stanford Research Institute (SRI). At least 24
of the 73 received their doctorates from one of MIT, CMU, or SUo There is no particular
significance in the ordering, nor is the record of movements complete. Intergenerational
and inter centre linkages are probably underestimated due to lack of data.
movement of personnel.
student of! worked for.
colleague relation.
Boundary between the wider establishment, and the Artificial Intelligence establishment.
181
prevalence of such links - links which have led to charges of nepotism being
levelled at the AI establishment (31). This structure of very strong intergenerationallinkages turns out to be characteristic of the development in Britain as
well, and in the section on the establishment in the United Kingdom, some
underlying reasons for the strong linkages are discussed.
The emergence of this group as the establishment in AI was undoubtedly
consolidated by their success in getting the backing of the United States
Department of Defense, mainly through the Advanced Research Projects
Agency (ARPA), which provided some 75% of United States AI funding for
the ten years from 1964, and through the Air Force (32). Furthermore, the
preference on the part of ARPA for concentrating resources in a few selected
centres guaranteed the position of the establishment, especially in view of the
great expense of adequate computing facilities, which effectively barred other
groups from competing.
Another aspect of development in AI that has been characteristic and of
importance for the field, and that has served to further reinforce the position
of the establishment should be noted. This is that the general aim of research
in the area to product intelligent machines haS excited extreme reactions and
has tended to lead to very strong for and against (33) alignments. Such a
reaction is not at all surprising given the sensitivity of such a goal to peoples'
images of themselves. Here Elias' comments about the competition for the
monopoly over the means of orientation are relevant (34). The AI approach
is seeking to establish and legitimate a view of intelligence and the nature
of mind which challenges the received commonsense view of mind and
intelligence as something rather special and certainly well beyond the reach of
scientific analysis. Moreover, this received view is very much under the sway
of the religious establishments, or where religious authority does not hold,
under the sway of a liberal humanist tradition. Strong reactions are commonplace in AI and, on the sociological level, have probably had the effect of
heightening the difference between those on the inside and those on the
outside, and consequently have reinforced and concentrated the position of
the establishment.
Thus it can be observed that the emergence of the American establishment
was very much bound up with the development of AI as a distinctive area
of research, and their position was consolidated by their success in gaining
backing from the Department of Defense. The American establishment was
182
James Fleck
not only involved in providing an organizational basis for research in the area,
but was also very closely concerned with the elaboration of a distinctive
cognitive basis for research in the area, the AI paradigmatic structure. In the
following discussion of the development of AI in the United Kingdom, some
of the themes already introduced will be reiterated, while other issues will
become evident.
6. Development in the United Kingdom
In Britain during the 1940s, there was a similar flourishing of interest in
general cybernetic concerns as occurred in the United States, and discussions
of the possibility of machine thought were common (35). A. M. Turing was
an enthusiast for the possibility of intelligent machines, and his 1947 and
1950 papers still stand in many respects as definitive surveys of the arguments
for and against AI (36). R. J. W. Craik, whose 1943 book The Nature of
Explanation is recognized as one of the texts marking the emergen.ce of
cybernetics, wrote passages that bore a remarkable foreshadowing of the
actual AI paradigmatic structure, as for example in the following passage:
... thought models, or parallels, reality - that its essential feature is not 'the mind', 'the
self', 'sense data', nor propositions but symbolism and that this symbolism is largely of
the same kind as that which is familiar to us in mechanical devices which aid thought and
calculation (37).
183
184
James Fleck
185
186
James Fleck
187
expansion of the early and mid 1960s. This expansion, in fact, started being
curtailed just after the establishment of the new department, and the resulting
squeeze contributed to the problems that beset the department, as the level
of University Grants Committee support could not keep pace with the large
Research Council funding that the new, rapidly growing area attracted. Had
the curtailment of expansion come a few years earlier, it is highly unlikely
that a separate department would have been approved. The second factor
was the support afforded by Swann: as Dean of Science he had backed
Michie's previous initiatives, and newly-elected in January 1966, as Principal
of Edinburgh University, he was very receptive towards new departures and
constantly promoted the status of Edinburgh as second only to Cambridge
in research (65). Without such sponsorship, it is again doubtful whether a
new department would have been instituted, or whether Michie would have
succeeded in attracting Gregory and Longuet-Higgins. However, the department was established in October 1966, and while in the event the institutional
attractions were evident, it is interesting to consider the scientific motivations
for these people with a non-AI background to change their area of research.
Gregory had engineering interests which led him to seek a new methodology involving a closer study of the physical basis in the brain for perception
and cognition than was usual in psychology at that time, and the AI approach
seemed to promise developments along these lines (66). He also brought with
him other members of his group, notably -S. H. Salter who had extensive
engineering competence and built the robot hardware (and was later to
become known for his wave power system - the Salter duck) and J. A. M.
Howe, a psychologist who was to explore the applications of AI in educational
research, and who became head of the AI department at Edinburgh in the late
1970s.
Longuet-Higgins was very much a scientific high flyer, achieving international distinction in his work in theoretical chemistry with C. A. Coulson
at Oxford, and gaining a professorship at the early age of thirty. For his
eminence in Chemistry, in 1958 he was elected a Fellow of the Royal Society,
and in 1968 a Foreign Associate of the United States National Academy of
Science, the highest American honour available to someone not a United
States citizen. Despite his great success in chemistry, or perhaps because of it
in that he was motivated to seek similar success in a new and potentially
exciting but unexplored field, he had joined with Gregory in planning a Brain
188
James Fleck
Research Institute. Negotiations were in hand for funding from the Nuffield
Foundation, and for accommodation at Sussex University, when Michie, at a
meeting with Gregory early in 1966, suggested that Edinburgh would be an
ideal centre in view of its already established AI work. With the institution of
the new department at Edinburgh, Gregory's group and Longuet-Higgins
moved to Scotland, and great hopes were entertained for the future of
coordinated research in the area.
The anticipated cooperation failed to materialize. Problems over accommodation, personal, political, administrative, and scientific factors were involved
in what became a very complex and confused situation during the late 1960s
and early 70s. Gregory never really settled in at Edinburgh nor became
involved with the computational approach though he remained favourably
inclined towards it, and in 1970 he finally left to go to Bristol University. The
engineering workshop in the Bionics Laboratory had proceeded, however,
with the building of the robot hardware, and a prototype was connected to
the computer for the first time in May 1969. Longuet-Higgins did absorb the
computational approach, but irreconcilable differences between him and
Michie over the installation of the new computer and the robot project, as
well as over their approaches to work in the area, soon emerged and resulted
in Longuet-Higgins moving into separate accommodation and thenceforth
running his unit (then called the Theoretical Section) quite independently
apart from access to the common facilities.
Michie favoured a rather swashbuckling style of directing large team
projects oriented to goals which could be linked with industrial applications
and, in fact, was involved in launching a university based company to market
compiler systems and other software for the POP-2 language, which was
developed in the department (67). In addition, he was extremely energetic
and persuasive and very successful in obtaining funding from many different
sources (68). Longuet-Higgins in contrast, favoured a more restrained, academic style, preferred an individual basis of working with a few colleagues on
research chosen purely for its intrinsic scientific interest, and was dubious
about the advisability of mixing commerce and industry with research. These
differences in style, aggravated by contrasts in personality, were associated
with conflicting views on AI: LonguetHiggins thought that 'artificial intelligence' was not a science or technology in its own right, but was a new way
of tackling problems in those existing sciences which were relevant to the
189
190
James Fleck
But despite the tension between the senior people at Edinburgh, there
was a thriving research environment in the late 1960s and early 1970s, with
frequent visits by people from elsewhere in the world, including the United
States and the up and coming Japanese AI-oriented groups (73). Young
researchers were very successful, sometimes gaining an international reputation before obtaining their doctorate: P. J. Hayes for example, became well
known after publishing a joint paper with J. McCarthy in 1968 (74), only
receiving his PhD. in 1972.G. D. Plotkin's work on inductive inference was
considered outstanding (75); R. Kowalski made a name for himself with his
vigorous promotion of the predicate calculus as a programming language in its
own right, an approach which became an independent strand of research
termed 'Logic Programming' (76); while R. M. Burstall, one of the original
members of the Experimental Programming Unit, and Michie's second in
command, became established as an outstanding computer scientist with an
international reputation in his specialist area - the Theory of Computation,
in which he built up his own group in the 1970s. In 1978, he was appointed
to a chair with the title 'professor of AI' despite the fact that computation
theory was by that time a general computer science research area, rather than
a specialist AI one. The robot project attracted considerable pUblicity with
some five television and ftlm crews visiting (77): indeed demonstrations
became so frequent as to interfere with the everyday research work and had
to be restricted (78).
However, the concentration of talent, the surfeit of publicity, and perhaps
more than anything else, the predominance of research over teaching in AI,
attracted hostility from other departments weighed down with heavy teaching
responsibilities (79), and strong pressures grew for the area to normalize its
activities. In addition the lack of a career structure for researchers on shortterm contracts, coupled with the increasing uncertainty over the future of the
centre due to the leadership tensions led people to start moving elsewhere:
Hayes went to Essex; Kowalski to Imperial College; several other researchers
to the United States; and Longuet-Higgins himself, along with members of
his group, moved to Sussex University. In the course of a couple of years,
therefore, many of the most highly respected researchers left Edinburgh, and
in some quarters Edinburgh was viewed as being in decline (80).
The problems facing AI were not restricted to Edinburgh alone, nor was
the division of the department the outcome of purely local politics. Rather,
191
these events were tied up with national attitudes especially on the part of the
SRC. The SRC had never been happy with the breakdown of cooperation
over the major robotics grant and had become dissatisfied with the progress
made in work on the project. This dissatisfaction stemmed to some extent
from a basic lack of sympathy with the goals of those involved in the project.
Michie's very ambitious plans for a seven-year industrially oriented programme of research in robotics failed to win favour with the SRC and was
never formally submitted (81). More modest proposals were put forward to
maintain the level of effort on robotics, but, despite a years very intensive
work on the project, in which programmable assembly using visual recognition
of parts was attained (82) (at that time one of the foremost achievements in
robotics in the world, comparable to leading work in America and Japan),
these proposals were turned down. At that stage, the SRC had become
very impatient with Michie, as his entrepreneurial talents did not fit in with
their expectations, and the previously-mentioned survey of AI by Sir James
Ughthill, FRS, Lucasian Professor of Applied Mathematics at Cambridge, and
an eminent hydrodynamicist, undoubtedly influenced their decision.
Lighthill's report created a major controversy and was published' in April
1973 along with other assenting and dissenting views. Lighthill was highly
critical of AI in general and suggested there were three basic categories of
research in the area: work aimed at advanced automation on the one hand,
and at computer based Central Nervous System research on the other, with
in addition a bridge category with the basic component of building robots,
which he saw as the essential underpinning for AI to have any claims to unity
and coherence. Progress in this cateogry Lighthill suggested was virtually nonexistent and the building of robots a mistaken enterprise possibly motivated
by a desire on the part of those concerned to 'minister to the public's general
penchant for robots by building the best they can', and possibly also by
'psuedomaternal' drives to compensate for male researchers' inability to
give birth (83). (It is not hard to detect a reference to Michie's polemical
enthusiasm in these comments.) Furthermore, what success there had been,
he suggested, was evident only in particular applications and derived from
knowledge contributed from the substantive fields modelled, rather than
from any AI component. In time, he saw the bridge category as withering
away, while work directed towards the two extremes would become integrated
with other research in their general areas.
192
James Fleck
Not surprisingly, this caused a major stir in the AI community across the
world (84), and the resulting controversy received much public airing in
the press and even on television (85). Without a doubt, despite Ughthill's
protestation that his report:
would simply describe how AI appears to a lay person after two months spent looking
through the literature of the subject and discussing it orally and by letter with a variety
of workers in the field and in closely related areas of research (86),
it delivered a blow to the prestige of research in the area from which it has
never fully recovered. While UghthiIl's comments on robots were directed
at the specifically AI category, it appeared they also had some effect on
inhibiting robot research and use in Britain in general (87), whereas in other
countries robotics has been a steadily expanding area throughout the 1970s.
In practical terms, the report did affect financial support for research in AI
in Britain, particularly in the case of Michie's proposals for robotics research
and also had some influence on funding in the United States where AI robotic
projects were cut back, and the Advanced Research Projects Agency , ARPA,
the main sponsor of American work in the area, started insisting on missionoriented direct research, rather than basic undirected research (88). These
cutbacks took place in the context of the general reduction in public spending
in the early to mid-1970s, which affected scientific research in all areas,
especially those not seen to be of 'social relevance'. However, the effects
were to some extent mitigated in the case of AI: partly by the variety of
funding sources supporting the area; partly by the SRC's identification of
machine intelligence as an important area of long range research in its 1972
Computing Science Review, which underlined the fact that there was, in
any case, no one predominating view on the value of AI; and partly by the
expanding nature of computing science in general. Consequently, particular
projects were able to get support, especially if their relevance was emphasized
and explicit reference to robotics avoided (89).
Furthermore, the debate over the Ughthill report also led to cognitive
science being recognized by the SRC, and a panel was set up to review applications in this area. Thus there was to some extent a shift in resources to
this area, rather than a straightforward cut back of AI as a whole.
That the reorganization of AI at Edinburgh, with the effective removal of
Michie from a central position, was not a purely local affair, but was rather
193
194
James Fleck
195
in AI the first place (101). This internal dynamic, coupled with the Theorem
Provers' assured confidence in their formal mathematics-based status, had
led to their being always rather autonomous; and under easier funding conditions, they would probably have become a completely independent specialty
of computer science, much as Pattern Recognition, itself based on a strong
internal dynamic, had done in the early 1960s (l02). However, the Theorem
Provers made somewhat of a comeback in the late 1970s with the logic programming approach embodied in the language PROLOG, thus re-establishing
themselves to some extent as a source of techniques of utility to AI in general
(103).
A third contributing factor to the perceived decline of Edinburgh as an AI
centre was that computation theory, another of the major research themes in
the department there, under the leadership of R. M. Burstall, professor of AI,
had become more central to computer science in general during the 1970s
and less of a specifically AI approach (104). This situation was rationalized in
the late 1970s with the transfer of the Computation Theory group from the
department of AI into the department of Computing Science. A fourth and
fmal contributory factor to the perceived decline of Edinburgh, was that
other major concentrations of AI interest had developed in Britain: at Sussex,
Essex, and later in the Open University. At the same time, numerous oneperson AI projects were pursued elsewhere, often based on 'colonization' or
'infection' from the established centres.
Professor R. A. Brooker, always an enthusiast for AI (he was one of the
panel members in favour of AI in the 1972 SRC Computing ReView), had
wanted to invigorate the research atmosphere in the computer science department at Essex, and to this end had recruited J. M. Foster from the Aberdeen
AI research group to a chair in the department (105). However, this did
not work out and Foster left after about a year, having done little on the
AI research side, but having developed the elements of a course in the area.
There was considerable interest in the AI approach on the part of young
researchers in the department such as J. M. Brady and R. Bornat, and when
P. J. Hayes arrived in the late 1972 from Edinburgh, bringing with him his
extensive familiarity with the AI literature and research front, he catalyzed
the development of research projects in the field at Essex. One such project,
supported by the SRC, was the development of a system to read handwritten FORTRAN coding sheets using high level knowledge to guide the
196
James Fleck
197
198
J ames Fleck
199
nervous system research on the one hand, and advanced automation on the
other, it is difficult to align his prediction with the continued coherence of AI
- that is, the continued identifiable existence of the AI paradigmatic structure. Rather than the established areas of linguistics, psychology, and philosophyabsorbing AI, it would seem that cognitive science is an emerging synthesis
based on the unifying computational modelling approach of AI. Indeed, one
could equally well argue, against UghthiU, the AI practitioner's extreme view
that what is happening is merely the process of colonization of other areas by
the AI approach:
I see the future of AI as a very long haul towards computational theories of physics,
chemistry, linguistics, sociology, visual perception, locomotion and every other aspect
of what it means to be human (122).
Clearly, therefore, views still differ over the assessment of the place and
future of AI.
7. The Establishment in the United Kingdom
One of the main features of the development of AI in Britain was the initial
and continuing strong American influence. Nearly every one of the leaders of
AI research in Britain had visited the United States and been impressed by
developments there, before moving into the area, or promoting it themselves
in Britain: Michie, Meltzer, Sutherland, and Clowes had all visited AI projects
in the United States in the early 1960s. These links with the American AI
community were maintained and strengthened in the ensuing years, and
continue today with a high international exchange of personnel between the
various centres.
These links also underlie the substantive similarity of research pursued in
Britain and America, at the general level of the paradigmatic structure, as
reflected for instance in the research profIle at Edinburgh which matched the
patterns evident in the United States. This has remained true since the mid
1960s when AI took off in Britain and, by and large, it has been the case that
the initiative in the development of AI in terms of the broad content of
research has remained in the hands of the establishment in the United States.
Consequently, the emergence of the establishment in Britain has been bound
up with organizational aspects of development to a greater extent than in the
200
James Fleck
201
202
James Fleck
203
ABERDEEN
EDINBURGH
AMERICA
SUSSEX
ELSEWHERE
ESSEX
64
!!L..
!!L..
Z!L
=
l-
llIi-
ZL.
I-
Tit
=
:55
1L.
BO
204
James Fleck
A major feature of development in Britain was that the initial AI establishment did not emerge from the strong pre-existing cybernetic or computer
science network, despite the clear prefigurement of AI research in the work
of Turing and Craik. Rather, it came from people external to such work:
Elcock at Aberdeen; Michie, Meltzer, and Longuet-Higgins at Edinburgh; and
later, Boden and Sloman at Sussex. Even Sutherland and Clowes at Sussex,
although they had links with the cybernetic tradition, certainly were not
centrally involved with it. Such an entry into the AI establishment from
outside the cybernetics tradition had also been evident in the United States,
with Simon and Newell. But, just as was the case in the United States, the
first generation members of the British AI establishment, although they were
marginal to cybernetics and computer science, certainly did not arrive from
nowhere. Michie, for instance, had some sixty publications to his credit in
his previous specialist areas of genetics, immunology, and reproduction, and
held the post of reader in the Department of Surgical Science at Edinburgh.
He was recommended by W. H. Waddington and M. Swann on his appointment to Edinburgh in 1958, and Swann continued to support him during
his AI activities. Meltzer, too, as a reader in the Department of Electrical
Engineering, had a solid reputation for his work on electron beam dynamics
(used by NASA for the design of ion propulsion for space vehicles) and
solid-state electronics; and Longuet-Higgins, with his international standing
in theoretical chemistry was clearly aheady a member of the wider British
scientific establishment.
The migration of outsiders with some standing, and hence the freedom to
move fields (127), seems to have had, therefore, as its major consequence
the construction of an organizational structure within which the pursuit
of AI research subsequently developed. In some cases (especially that of
Sutherland) the organizational leadership role merged with another, probably
necessary, role in the emergence of a new interdisciplinary area - namely a
sponsorship role. The institutional developments at Edinburgh would not
have been possible without some strong support from sponsors placed in
fairly influential positions with the university government, especially in view
of the tendency for the publicity attracting, research intensive, AI activities
to arouse suspicion and resentment: both Sir Edward Appleton and Sir
Michael Swann, who succeeded Appleton as vice chancellor of Edinburgh
University, were active in encouraging and supporting those developments.
205
And at Sussex there was also fairly widespread support for AI oriented
developments among many of those in positions of influence.
As well as this positive sponsorship within the wider establishment, which
was accompanied by a positive evaluation of the status of AI, there was a
negative sponsorship as well, as demonstrated by Sir James Ughthill's report.
Ughthill's opposition could clearly be seen to support the status quo: He
affirmed the value of the currently existing areas included in Advanced
Automation (e.g., control engineering) and in Central Nervous System research (e.g., neurophysiology), and moreover attributed any success in AI to
contributions arising from these areas. The validity of the emerging interdisciplinary area of AI was thus challenged and denied, explicitly in terms of the
aheady-established disciplines surrounding AI. In particular, Ughthill's focus
on the established Central Nervous Systems areas of research - neurophysiology and neurochemistry - and his use of the term 'central nervous system'
happened to align with the dominance in Europe of the neurosciences which
study the 'hardware' of the brain, over the cognitive sciences - linguistics
and psychology - which might be said to study the 'software' of the brain
- a dominance which is not as clear-cut in the United States (I28). In this
dominance relation we see a prestige hierarchy, with those sciences closest to
the physical sciences accorded most prestige. In this context, the study of the
brain at the reductive level of biochemical or_neurophysiological mechanisms
is considered more prestigious than the AI approach at the information
processing levels.
Moreover, it is difficult to account for the impact of the Ughthill report
(by his own admission a two months layman's view of the area) (129) except
in terms of the authority carried by Ughthill's eminence. It is interesting to
note that reference is still offered to observations that Ughthill made on the
difficulties for search arising out of the combinatorial explosion (I 30), as if
he were their originator, whereas, in fact, the combinatorial problems had
been regarded as the raison d'etre for AI - the huge size of the space of
possible moves in chess, for example, estimated by Shannon in 1950 as in
excess of 10 120 , dictated the need for heuristic strategies to restrict the
search space to manageable proportions (131).
But what is also of interest about the Ughthill report, apart from its
importance as an authoritative pronouncement on the status of AI research,
is that it is one among a multitude of attacks on the field (132). Such attacks
206
James Fleck
on AI have been commonplace, and while they purport to deal with the
particularities of the subject matter of research in the area, it is quite clear on
closer inspection, that they are more concerned with the general goal of
constructing an intelligent machine. It would be too lengthy to argue this
fully here, but it is perhaps pointed by the continued relevance of Turing's
comments on the arguments for and against the possibilities of constructing
intelligent machines (133), despite the fact that the distinctive AI approach
had not emerged when Turing was still alive. It is also pointed by the fact that
these attacks on AI are not the prerogative of any particular group: criticism
has come from all shades of political opinion, and from all areas of research;
scholarly as well as technical.
Ironically, this variety in attacks can only be matched by the diversity
in the sources of support for AI, or the range of, often conflicting, views
within the field itself. Something of this has already become evident in the
differences between LonguetHiggins and Michie, one favouring the cognitive
science defmition of AI, and the other the Machine Intelligence approach.
There are other divisions: those supporting a theoretical formal approach,
such as McCarthy, for instance, and those supporting the exploitation of
practical applications, such as Feigenbaum; those who see no problems with
accepting military funding (McCarthy and Feigenbaum) and those implacably
opposed (Meltzer, while Michie was opposed to classified work).
These divisions, which are legion in the area, are coupled with a rather
amazing state of substantive partisanship, or scientific ethnocentricity, in
which proponents of the various different research areas each tend to see
their own approach as the real AI approach - to the theorem provers, theorem
proving is central, for the natural language proponents, language is the basis
for reason, and so on (134).
Many of these differences can be related to the background competences
of the practitioners, and can be interpreted as competition between groups on
the research area level for resources and authority within AI as a specialty,
constituting perhaps the primary locus for competition over cognitive com
mitments, quite distinct from the individualistic level of competition for
recognition, long identified in the sociology of science as a basic motor of
scientific development. An important point to note is that such differences
are not precluded by the paradigmatic structure of AI, outlined earlier
as providing guidelines for the common computational approach and its
207
programming basis, but which does not dictate a dogmatic monolithic atti
tude, nor inculcate a unifying solidarity. Moreover, this variety of views in
and around AI can be related to its position as an interdisciplinary area, with
particular research areas associated with particular neighbouring disciplines for example, the natural language research area is associated with linguistics,
while theorem proving has links with metamathematics. And as an interdisciplinary area, the status of AI research is still very much in process of
negotiation. The cognitive science developments appear to have led to an
acceptance, on the part of those involved, of the validity of AI: indeed the
impression in that context is that AI is the 'hard' formal core, and therefore
of high status. But in a computer science or general scientific context, AI
is still seen very much as a 'freaky', rather dubious fringe activity, and con
sequently of rather inferior status (135). Moreover, there are two broad
categories of attacks which can be related to these contexts. On the one hand
there are attacks, often by philosophers and others, on AI for being reductionistic and impossible (136) - in a sense it is 'harder' than is appropriate
for the study of intelligent activity . On the other hand there are criticisms,
often by computer scientists, of AI for being morally wrong (137), bad
science (138), or undisciplined and sloppy (139).
Finally, perhaps the variety and depth of feeling of the many attacks on
AI derive, not so much from what is, in fact, done in AI research, but rather
from the fact that the very broad aim of research in the area, namely the
construction of intelligent machines, bears uncomfortably on our conception
of ourselves. In Elias' terms, AI research is seen to be involved with a very
sensitive area of the means of orientation: the area which is concerned
with the nature of mind. Furthermore, during several hundred years of
development and struggle with other competing establishments, a scientific
establishment has yet to succeed in gaining a monopoly over the means of
orientation in this area. In making its challenge in this area, therefore, AI is
inviting violent attacks, and its practitioners should hardly be surprised when
they suffer them.
208
James Fleck
appears to have been as follows. Around the period of the Second World War,
catalyzed by the war-time weakening of traditional disciplinary boundaries,
and brought into being by exigencies deriving from the unprecedented problems of organization and communication posed by the increasingly complex
social structures and conditions, there emerged the software sciences. These
had their focus on pattern rather than substance, and included operations
research, computer science, and cybernetics. In particular, the cybernetics
area, with its focus on the processes common to animals and machines,
promised a realization of the age-old desire to make an artificial man, a
machine that could think. Within the general area of cybernetics various
approaches were made to the construction of intelligent machines, some
based on electronic analogue of neuron networks, others on the simulation
of processes by means of the newly developed digital computer. In this
context, the paradigmatic structure of AI was articulated in the United States
during the late 1950s.
The American establishment in AI consisted essentially of those who had
contributed to this articulation, and who had provided the institutional
structure within which subsequent research based on the distinctive AI
paradigmatic structure, could be undertaken. The source of authority and
reputation of the establishment was derived from the effective demand which
developed for this distinctive approach, on the part of those who wanted to
follow the approach themselves and those who thought the approach was
worthwhile and promising. In addition, the emerging establishment secured
the backing of the funding agencies, aided by their good connections with the
wider establishment, in competition with other approaches within computer
science. The preference of the funding agencies to concentrate resources in
a few centres, coupled with the expense of the instrumental base required
(the digital computer), ensured that an effective monopoly over material
resources as well as the cognitive ones was maintained.
As the paradigmatic structure was already elaborated before it was exported
to the United Kingdom in the early to mid-1960s, it left less opportunity
for substantive contributions on this general level while there was ample
scope for extending the institutional facilities for carrying on AI research
in Britain: consequently the first-generation British establishment in AI
consisted of those who were able to set up organizational forms to exploit
paradigmatic structure, thus giving them a monopoly
the already
209
over the cognitive resources in the area. Because of the constructive nature
of AI, which was manifested in it being only one among several competing
cybernetic approaches, and since people already committed to a particular
approach tend to stay with that approach, members of the first-generation
establishment included people from outside the cybernetic and even computer science traditions. This was the case in the United States, but was
more marked in Britain. Moreover, the members of the first-generation
establishment, in fact, were drawn from those who already had some standing
or prestigious backing in another field, and being thus well connected, they
were able to secure the backing of the Science Research Council in competition with other approaches within computer science, ensuring their effective
monopoly over material as well as cognitive resources in the area.
This monopoly over resources was reinforced by the constructive and craft
nature of the AI paradigmatic structure, and the patterns of development
in both the United States and the United Kingdom followed the lines of
personnel mobility and contact, thus leading to a tight intergenerational and
intercentre structure of linkages. In particular, students and 'descendents'
of the members of the first-generation American establishment have tended
to dominate the field. The paradigmatic structure of AI, however, while
providing general guidelines for the methodological approach employed in
AI research, does not dictate the direction of research. Due to the very wide
ranging nature of the focus on intelligent activity, research in AI has become
involved with the subject matter of many other disciplines, and has therefore
developed as an interdisciplinary area. This interdisciplinary character of
AI has induced many mutually competing divisions within the area, as well
as leading to many external views on the status of the field, which has consequently been very much a matter for negotiation: from the point of view of
the 'soft' sciences, such as linguistics and psychology, AI has appeared 'hard'
and therefore of superior status; from the point of view of the 'hard' sciences,
such as computer science and physics, AI has appeared somewhat 'freaky'
and therefore of inferior status. Furthermore, as a newly emerging specialty,
AI has been in competition with already established disciplines: the Ughthill
report, critical of AI, can be interpreted in this context as an affirmation of
the status quo. Finally, because of the general aim of constructing intelligent
mechanisms, AI has been seen as challenging the monopoly on the means
of orientation with respect to the nature of mind. As this bears directly on
210
James Fleck
Acknowledgements
I should like to acknowledge the help given by the interviewees to whom a
draft of this paper was circulated. The comments of J. A. M. Howe, H. C.
Longuet-Higgins, D. Michie and N. S. Sutherland were particularly valuable.
I should also like to thank R. D. Whitley for his suggestions and the SRC for
their support.
211
212
James Fleck
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
so.
51.
52.
53.
54.
55.
213
op. cit. (note 7), pp. 389-405. This was also commented upon in interviews: R. A.
Brooker, Essex University, 28/3/79; and A. M. Uttley, Sussex University, 21/3/79.
A. M. Turing, 'Intelligent Machinery', (1947), in B. Meltzer and D, Michie (eds.),
Machine Intelligence S, Edinburgh: Edinburgh University Press, 1969, pp. 3-23
and A. M. Turing, 'Computing Machinery and Intelligence', Mind 59 (1950)
433-60.
R. J. W. Craik, The Nature of Explanation, Cambridge: Cambridge University
Press, 1952, p. 57.
W. R. Ashby, Design for a Brain, New York: Wiley, 1952; and An Introduction
to Cybernetics, London: Methuen, 1965.
W. G. Walter, 'An Imitation of Life', Scientific American 182, No.5 (1950)
42-5; 'A Machine that Learns', Scientific American 185, No.2 (1951) 60-3.
F. H. George, The Brain as a Computer, Oxford: Pergamon Press, 1961.
The Ratio club was discussed in an interview: A. M. Uttley, Sussex University,
21/3/79; and is also discussed by McCorduck, op. cit. (Note 18), p. 59.
The proceedings of this conference are published in: Mechanisation of Thought
Processes, London: HMSO, 1960.
J. McCarthy, 'Programs with Common Sense', ibid., 75-84.
D. Michie, On Machine Intelligence, Edinburgh: Edinburgh University Press,
1974, pp. 37,51, and 66-7; where he states: "It was from my personal association with Turing during the war and early post-war years that I acquired my
interest in the possibilities of using digital computers to simulate some of the
higher mental functions that we call 'thinking'."
This, and much of the following information pertaining to Michie's activities is
derived from interviews with Michie (Edinburgh University, 29 and 31/8/78),
backed up by other available sources.
For example, D. Michie, 'The Effect of Computers on the Character of Science',
University of Edinburgh Gazette 34 (Oct. 1962) 23-8; and 'The Computer
Revolution: Where Britain Lags Behind', The Scotsman, 12/7/63.
Computing Science in 1964, A Pilot Study of the State of University Based
Research in the U.K., Prepared for the Research Grants Committee by Dr Donald
Michie, London: The Science Research Council, 1965.
Interview: Lord Halsbury, London, 18/12/79.
Council for Scientific Policy/University Grants Committee, A Report of a Joint
Working Group on Computers for Research, London: HMSO, Cmnd. 2883, 1966.
Interview: Lord Halsbury, London, 18/12/79.
Reconstructed from information in the archives of the Department of Artificial
Intelligence, and the Machine Intelligence Research Unit, Edinburgh University,
and from interviews.
J. McCarthy, 'Review of the Lighthill Report', Artificial Intelligence S (1974)
317-22.
R. M. Burstall and R. J. Popplestone, 'POP-2 Reference Manual', in E. Dale and
D. Michie (eds.) , Machine Intelligence, Vol. 2, Edinburgh: Oliver and Boyd, 1968,
207-46.
Interviews: B. Meltzer, Edinburgh University, 1/8/77 and 23/1/79.
N. L. Collins and D. Michie (eds.), Machine Intelligence, Vol. 1, Edinburgh: Oliver
and Boyd, 1967. There have been nine workshops in the series to date.
214
James Fleck
56. E. W. Elcock, 'Report of the SRC Computer Research Group', Aberdeen University, May 1970; and interviews: A. M. Murray, Aberdeen University, 9/11/78;
P. M. D. Gray, Aberdeen University, 23/7/79; and J. M. Foster, R. R. E. Malvern,
19/6/79.
57. Information on the activities of Clowes derives from a letter: M. B. Clowes,
23/11/78; and interview: M. B. Clowes, Sussex University, 22/3/79.
58. Interview: B. Meltzer, Edinburgh University, 23/1/79.
59. Interview: P. J. Hayes, Essex University, 27/3/79.
60. These conferences, The International Joint Conferences on Artificial Intelligence,
have been one of the main organs of communication in the area, and a major
outlet for publications in the area.
61. R. L. Gregory, Eye and Brain, London: Weidenfeld and Nicholson, 1966.
62. Information on the robot is derived from numerous documentary sources in the
archives of the Department of Artificial Intelligence, and the Machine Intelligence
Research Unit, and from various interviews.
63. Interview: J. A. M. Howe, Edinburgh University, 24/1/79.
64. As was evident from various interviews, including one with S. Michaelson, the
director of the Computer Science Department, at Edinburgh University, on
18/7/79.
65. Lord Swann commented on the new departures in an interview: London, 10/10/
79. At a press conference in connection with an open day for industrialists, he
is reported as commenting that Edinburgh was among the top two centres of
research, despite having no costly 'big science' projects. University of Edinburgh
Bulletin 8, No.2 (Oct. 1971).
66. Interview: R. L. Gregory, Bristol University, 25/7 /79.
67. Conversational Software Ltd., launched in 1970.
68. Interview: D. Michie, Edinburgh University,.31/8/78. He commented that at one
stage some twenty sources were involved.
69. Computing Science Review, London: Science Research Council, 1972, p. 17.
70. Ibid., p. 19. Compare also D. Michie, 'Schools of Thought about AI', University
of Edinburgh, School of Artificial Intelligence, Experimental Programming
Report No. 32,1973.
71. Artificial Intelligence: A Paper Symposium, London: Science Research Council,
1973,p.i.
72. This review was carried out by a Special Committee, chaired by Prof. N. Feather,
and set up by the University Court. It reported in 1973, and extracts were made
available to me by C. H. Stewart, at that time secretary to the University, in a
letter,18/7/78.
B. As was evident from interviews and also from the Departmental Newsletter.
74. J. McCarthy and P. J. Hayes, 'Some Philosophical Problems from the Standpoint
of Artificial Intelligence', in B. Meltzer and D. Michie (eds.), Machine Intelligence,
Vol. 4, Edinburgh: Edinburgh University Pre8s,1969.
75. G. D. Plotkin, Automatic Methods of Inductive Inference, unpub. PhD diss.,
University of Edinburgh, 1971.
76. Interview: R. A. Kowalski, Imperial College, London, 15/6/79.
77 . From the Departmental Newsletter.
78. From the Minutes of the Round Table, a sort of departmental board.
215
216
James Fleck
97. Michie organized the 'AISB Summer School on Expert Systems', held at Edinburgh University, July 1979.
98. A. Bundy et al., Artificial Intelligence: An Introductory Course, Edinburgh:
Edinburgh University Press, 1978.
99. T. Winograd, Understanding Natural Language, Edinburgh: Edinburgh University
Press, 1972.
100. Various interviews, and Note 27.
101. J. A. Robinson, 'A Machine Oriented Logic Based on the Resolution Principle',
Journal ACM 12 (196S) 23-41.
102. J. Fleck, 1978, op. cit. (Note 3), pp. 44-8.
103. D. Warren, 'PROLOG on the DEC system-l 0', paper presented at the AISB
Summer School on Expert Systems, Edinburgh University, July 1979, notes
some applications. The position of theorem proving was also discussed in several
interviews.
104. Interview: R. M. Burstall, Edinburgh University, 10/8/77.
lOS. Interview: R. A. Brooker, Essex University, 27/3/79.
106. Interviews: J. M. Brady, ES'sex University, 26 & 28/3/79; P. J. Hayes, Essex
University, 27/3/79.
107. J. E. Doran, 'Knowledge Representation for Archaeological Inference', in Elcock
and Michie, op. cit. (Note 9S), pp. 433-S4.
108. Interviews: B. Anderson, Essex University, 30/7/79; Y. Wilks, Essex University,
8/8/79.
109. Information on movements through Essex was derived from interviews with
Brady,op. cit., Note 106, backed up by other sources.
110. Information on developments at Sussex University was derived from various
interviews and documentary sources, in particular the interview: N. S. Sutherland,
Sussex University, 21/9/79.
111. For example, A. M. Uttley, 'Simulation Studies of Learning in an Informon
Network', Brain Research 102 (1976) 37-S3.
112. M. B. Clowes, 'On Seeing Things', Artificial Intelligence 2 (1971) 79-116.
113. This organization is frequently commented on in the Annual Reports, Sussex
University, and is discussed by the first Vice Chancellor, Asa Briggs, in his 'Drawing
a New Map of Learning', in D. Daiches (ed.), The Idea of a New University: An
Experiment in Sussex, London: Deutsch, 1964, pp. 60-80.
114. Letter: Lord Briggs, 28/11/1979.
l1S. 'Working Party on School of Cognitive Studies', University of Sussex, June 1970.
116. Interview: M. A. Boden, Sussex University, 19/9/79.
117. M. A. Boden, Artificial Intelligence and Natural Man, Hassocks: Harvester Press,
1977.
118. Interview: A. Sloman, Sussex University, 20/3/79.
119. A. Sloman, The Computer Revolution in Philosophy, Hassocks: Harvester Press,
1978.
120. Interview: P. N. Johnson-Laird, Sussex University, 20/9/79.
121. For instance, approximately one quarter of the MSc computer science courses
outlined in Graduate Studies 1974-75, Cambridge: CRAC. 1974, mentioned
Artificial Intelligence, and it is often mentioned as an acceptable interest in job
adverts.
217
122. J. M. Brady, 'A Glimpse of the Future of AI', text ofa lecture delivered at 'Computing 79', Syndey, Australia, August 1979, p. 2.
123. B. C. Griffith and N. C. Mullins, 'Coherent Social Groups in Scientific Change',
Science 177 (1972) 959-64.
124. I do not want to argue a 'great man' view of history here. The situation can be
interpreted in terms of the emergence of socially defined possibilities which in the
event were exploited by Michie. If Michie had not been, development in Artificial
Intelligence in Britain would still have taken place with other people stepping in
to a greater extent. Perhaps what is required is a 'great opportunities' view of
history, with the emphasis on the socially given possibilities rather than on the
people who exploit them.
125. T. Shinn, 'Scientific Disciplines and Organizational Specificity: The Social and
Cognitive Configuration of Laboratory Activities', in this volume.
126. See for instance, P. Kraft, Programmers and Managers: The Routinization of
Computer Programming in the United States, New York: Springer Verlag, 1977.
127. M. J. Mulkay, in The Social Process of Innovation, London: MacMillan, 1972,
concludes: " ... intellectual migration, whether into established networks or
into virgin areas, will normally be led by mature researchers of known repute.
These men use their eminence to attract funds and graduate students; and in
various ways they try to use their existing knowledge and techniques as a point of
departure for the construction of the new intellectual framework" (p. 54.). The
pattern of development of Artificial Intelligence in Britain certainly conforms
with the first part of what Mulkay writes, but it is difficult to square it with the
second part. It would appear that the extent of articulation of the paradigmatic
structure in the United States preempted attempts by migrants in Britain to use
their existing knowledge in the construction of new intellectual frameworks, and
left them scope only for negotiation over the organizational aspects.
128. See N. S. Sutherland, 'Neuroscience Versus Cognitive Science', Trends in Neurosciences 2, No. 8 (1979) i-ii, which discusses some of the particulars of this
dominance.
129. Lighthill,op. cit. (Note 83), p. 1.
130. For example, Sir Geoffrey Allen (Chairman of the Science Research Council), in a
talk in the Department of Liberal Studies in Science, the University of Manchester,
17/5/79.
131. C. E. Shannon, 'Automatic Chess Player', Scientific American 182, No.2 (1950)
48-51.
132. See Note 33.
133. Turing,op. cit., Note 36.
134. Fleck,op. cit. (Note 3), pp. 138-9.
135. Such an impression was given by those in Artificial Intelligence (for example,
interview: A. Bundy, Edinburgh University, 17/7/79) and those outside, in the
course of interviews.
136. Dreyfus,op. cit., Note 33, is the classic example here.
137. Weizenbaum,op. cit., Note 33.
138. Taube, op. cit., Note 33.
139. E. W. Dijkstra, 'Programming: From Craft to Scientific Discipline', Proc. International Computing Symposium, 1977, Liege, Belgium, April 1977 , pp. 23-30.
ARIE RIP
University of Leiden
1. Introduction
The technical side of science and its contribution to the dynamics of scientific
developments are often neglected. As Derek de Solla Price complains:
It is unfortunate that so many historians of science and virtually all of the philosophers
of science are born again theoreticians instead of bench scientists (1).
In a more balanced view, theory as well as techniques should get their due,
as in Bernal's aphorism: "Science, in one aspect, is ordered technique; in
another, it is rationalized mythology" (3). The prevailing neglect of the
technical side of science then becomes something to be explained, one of the
reasons obviously being the higher status of mental labour compared with
manual labour in our culture.
Philosophers of science, apart from their professional interest in theory
and speculation, may be excused for their neglect of technical aspects because
the discourse of science, especially in its publications, is much more accessible
to outsiders than its empirical practices (4). The same excuse applies to
sociologists of science, although with less force. Those philosophers of science
who do give a lot of weight to the technical side, the actual control of nature,
219
Norbert Elias, Herminio Martins and Richard Whitley (ed,.), Scientific Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982.219-238.
Copyright 1982 by D. Reidel Publishing Company.
220
Arie Rip
are often more interested in the consequences for the foundation of scientific
knowledge (e.g., Janich, Mittelstrass (5)) or for ontological issues (e.g.,
Bhaskar (6)) than in elaborating a balanced view. B6hme and van den Daele,
lately of the Max-Planck-Institute in Starnberg, are the exceptions, but their
work has been discontinued for the moment (7).
For the sociologist of science, the important issues are how the technical
side of science is involved in the dynamics of scientific developments, and
what the impact of technical innovation and its diffusion is on scientific
practices. These are large issues, and in this article I shall limit myself to a
discussion of the concept of 'restrictedness' and how it may contribute to
these issues.
The characterization of sciences as 'restricted' or 'unrestricted' has been
introduced by Pantin (8) and elaborated by Whitley (9). Especially from
Whitley's work, it is clear that 'restrictedness' is a bridging concept, combining features of the objects of the sciences and of their social organization.
Neither Pantin, nor Whitley emphasize technical aspects in their discussion of
'restrictedness'. I shall argue that the restrictedness of a science depends on its
control over the relevant part of reality, its ability to transform circumstances
into parameters and variables: Such an interpretation of the concept of
'restrictedness' makes it possible to discuss technical dynamics in the sciences
while including their social aspects. The approach of this article is thus the
same as when Shinn argues that "the logic contained in experimental operations and the manipulation of instruments impinges significantly on the cast
of social relations" (10). The difference is that Shinn looks at the relation in
contemporary laboratories, while I focus on its development over time.
2. The Concept of 'Restrictedness'
Whitley has shown how Pantin's rather unarticulated distinction between
restricted and unrestricted sciences (for example, between physics and
geology or parts of biology) can be made more precise and sociologically
interesting by looking at science as productive labour (11). Scientific work is,
as is work in general, concerned with transforming objects with tools for
some goal. Differences between sciences will therefore at least to some extent
be related to the specific features of what is being transformed and how, and
such features will give rise to differences in work organization.
221
222
Arie Rip
223
control over the knowledge object. In restricted sciences, the behaviour of the
knowledge object can be restricted, often very narrowly, to those aspects the
researcher wants to study. In unrestricted sciences, this is not, or not very
well possible, so that the behaviour of the object is unrestricted.
Since the definition proposed here is still directed toward the reduction
of ambiguity, the implications for the nature of scientific work and its
organization derived by Whitley on the basis of his definitions will still be
applicable. In addition, increases in the degree of restrictedness of a science
over time can now be linked directly to an increase in control over reality.
This point will be developed in the next section.
In principle, all cases intermediate between a completely unrestricted
science (if that may still be called a science) and a completely restricted
science (if that still shows dynamic development) are possible. In contemporary SCience, however, unrestricted sciences are not free to choose whether
to develop or not towards a higher degree of restrictedness. The high status
of physics as the paradigm of a restricted science forces the practitioners
of unrestricted sciences (e.g., the life sciences, but to some extent also the
social sciences) to aim for more restrictedness, even if they have to import or
imitate it. The alternative is to react against the status of physics as the ideal
science and emphasize other approaches: holism in biology, configurations in
sociology (17).
3. The Production of Restrictedness
Control over the knowledge object can be seen to increase through the
successful production of pure samples (in chemistry and medicine), "pure"
effects (in the physical sciences) and through tightly formulated procedures
(18). The historical reconstruction of the dynamic leading to increased control
is made difficult by the way modem science takes its power to control for
granted. For instance, in chemistry pure compounds are considered to be
non-problematic technically (they can be made) and conceptually (their
defmition is straightforward), while in fact, many chemists are not able to
give an acceptable definition, and depend on commercial suppliers for their
pure compounds (19). In this section, I shall use a few secondary sources
to sketch an account of the development of pure samples and pure ef(ects,
without attempting a full reconstruction.
224
Arie Rip
225
The demand for drugs made it possible to gain a livelihood from the
preparation of substances in the 16th and especially in the 17th century,
while at the same time, a demand for chemical recipes and instruction grew
(27). In the growth of analytical techniques (mineral waters and assaying
being the most important), there is a continuity extending from the 16th into
the 19th century (28). Natural philosophy did play an independent role,
especially in the introduction of corpuscular notions and the Cartesian
explanation of the behaviour of acids and alkalis, but it was one component
of a matrix in which medicine, alchemy and technology played their parts
(29).
During the 18th century, especially in France, chemists who had been
trained as pharmacists or physicians found a new and steadily expanding use
for their skills in the solution of industrial problems (30). In Germany, Stahl
and his followers formulated the phlogiston theory, partly in attempt to
dissociate themselves from lowly pharmacists and disreputable goldmakers
(31). The rationalization of chemistry by Lavoisier and his followers could
build upon the results of these earlier developments. The notion of a pure
sample was then related to the new theoretical structure, and this guided
attempts to improve techniques and procedures.
The technical dynamic operated within a framework of theory, but after a
time it became institutionally distinct from the ongoing research efforts in
chemistry. When teaching laboratories were established in the course of the
19th century, and later, when chemical analysis for monitoring purposes (e.g.,
water quality, industrial tests, food additives) became important around the
tum of the century, the regular production of pure compounds became a
necessary condition for obtaining reproducible results, and a commercially
interesting proposition. At present, a number of big companies offer a large
catalogue of chemical compounds and reagents ("chemicals"), graded according to purity.
(b) Changing Know/edge Objects
This historical sketch of the production of pure samples can be interpreted
as an account of a changing knowledge object. Slowly gathering momentum,
the changes become more rapid from the 16th century onward. The end
result is that the outcome of experiments depends less on the skill of the
226
Arie Rip
227
228
Arie Rip
also procedures) like air pumps and Leiden jars that produce or demonstrate
certain effects are synthetic instruments, and should be distinguished from
analytic instruments that extend and calibrate observation (38). They emphasize the role of both kinds of instruments in the development of science
in the 17th and 18th century. Research into the way the instruments worked
and attempts to refine and extend them provided the basis for theories of the
phenomena produced or measured by them. Only after successful theory
development (and the consequent rationalization of the instruments) is it
possible to view instruments as embodied theory.
The first use of analytic instruments was to explore new domains (e.g.,
telescopes, microscopes); scales were used to indicate the occurrence of a
loss or increase of weight, and only later in measuring the extent of the
change. New instruments like barometers and thermometers were themselves
objects of research, and gave rise to important scientific developments.
Bohme and Van den Dae1e trace the design and performance norms regUlating
the technical-scientific dynamic, and show that universality, that is, independence of surroundings and of the materials used in the instrument, is more
important than accuracy as such. Synthetic instruments also, and obviously,
provide opportunities to study the effects they produce. This, as well as their
possible entertaining and technological usages, determines the dynamic of
their development.
From the 17th century until now, effects continue to be important: new
domains are opened for exploration and their discoverers are honoured by
eponymy. After a time, the production of the effect is brought under control
and embodied in standardized apparatus (which sometimes takes off to a
separate career in other sectors, for instance Rontgen apparatus in medical
diagnosis and technological monitoring). The dynamic of the development at
the research fronts, where new effects have to be domesticated, is still the
same as in the 18th century. But the instrument makers of that time, with
their close relations to the evolving scientific community, have been replaced
by a flourishing branch of industry. With the production of monitoring
apparatus for applied science, industry and government as its origin and
backbone, the scientific instrument industry has profited from the expanding
market provided by modern scientific research, and now produces sophisticated instruments like NMR-spectrometers and special gaschromatographs.
Industrial decisions and innovations influence scientific developments, as well
229
as vice versa; the present trend towards automation of analysis and measurement is a prime example (39).
4. Some Implications of Increasing Restrictedness
230
Arie Rip
There are obvious advantages in importing restrictedness. For an unrestricted science to really profit, however, it has to reorganize its practices.
Organic chemistry during the decades before 1940 is an example of a semirestricted science with accepted, "chemical", criteria for proofs of molecular
structures (43). After the Second World War, physical instruments were
increasingly used to provide information about molecular structure and
interaction, at first hesitantly and in the face of resistance from the old
elite. Later on, it became commonplace, and the interpretation of the data
provided was considered unproblematic, although chemical confirmation
was still sought. At present, the use of sophisticated apparatus has become
an indicator of respectable chemical work, and sometimes a criterion for
the acceptance of articles (which cannot be met any more by the poorer
laboratories).
The routinization of work and division of labour that is characteristic for
restricted sciences according to Whitley, becomes possible in unrestricted
sciences through importation of restrictedness, as well as indigenous growth.
The impact of routinization is felt at the lower strata of the scientific work
force, including research students. The career perspective of most scientists,
however, is to leave the drudgery of actually performing research tasks after
a time. This makes them accept, and later forget their earlier experiences,
and insufficiently aware of the influence of instrumentation and automation
on selection and outcome of research problems (44). Quality control thus
becomes difficult, but as an issue it normally remains part of scientific folklore. When restrictedness is imported, even the recognition of quality control
problems recedes. Only when controversies erupt, as happens every now and
then in food testing and environmental monitoring, will such issues become
manifest (45).
Restrictedness in the sciences also has effects outside science. Human
capacities to produce restricted situations outside the laboratory, transforming a local reality to be similar to the original experimental situation,
are the essential precondition for applying science. A lot of effort, technical
creativity and often social power is required to effect such a transformation.
The necessity of transforming reality to a restricted situation, before
science can be applied, helps to explain the "external", often unexpectedly
negative effects of the application of science. Agricultural science, for instance, although only partially restricted, needs monocultures to be applicable
231
- as it were, the experimental plots enlarged n times. The required intervention in the ecosystem has its repercussions, which can and should be
distinguished from the side-effects of the applied science (for instance, with
pesticides). Another example is that the scale and general infrastructure of
power plants is often just as important as the specific technology (nuclear,
oll- or coal-fired) being used - as is now slowly being realized in technology
assessments (46).
S. Concluding Remarks
Viewing scientific developments as being partially determined by technical
dynamics, and describing such developments in terms of restrictedness may
also shed light on processes of theory formation. Increasing restrictedness
leads to empirical generalizations and conceptual distinctions, i.e., "bottomup" theory formation (47). Other dynamics playa role, for instance, the
speculative dynamic of world views like the mechanical philosophy of the
17th and 18th century, or atomism untll the 20th century. Such "top-down"
theory formation may induce an explanatory dynamic or process of paradigm
articulation, when the problems in the original explanation of the knowledge
object of a research area are taken as the starting point for theoretical and
experimental studies.
The existence of different dynamics in science emphasizes its heterogeneity, in spite of unified science ideals. In this respect, Kuhn's seminal
article on the mathematical and experimental traditions in science is particularly interesting (48). According to Kuhn, the mathematical tradition, which
goes back to antiquity and covers the "rational ordering sciences" (49), was
transformed during the Scientific Revolution through the innovation of
puzzle-solving groups, while at the same time a new tradition was born, the
Baconian approach of "twisting #le lion's tall", that is, subjecting nature
to artificial experiments. The latter tradition carries the technical dynamic
supporting increasing restrictedness, but ideals of rational ordering provide
stimuli to the experimental tradition.
Kuhn emphasizes the separate evolution of the two traditions, sometimes
far into the 19th century, because he is arguing against a unitary view of
science (50). He also draws attention, however, to a reshuffling of the traditions from circa 1800 onwards, when (pure) mathematics split off, many
232
Arie Rip
233
a corresponding rise in the status of biology); and the social status of the
practitioners of the different sciences.
Although data on social backgrounds and social status of practitioners of
different disciplines are practically non-existent, some information about
physicists and chemists provides a tantalizing glimpse of historical processes
at work to keep a distance, both socially and intellectually, between the two
disciplines. Kuhn notes that during the 17th and 18th centuries, practitioners
in the mathematical tradition were more or less established, and held posts in
universities and the newly-founded academies. Amateurs were the mainstay
of the Baconian sciences, and even when practitioners of these sciences were
members of an academy, they were most often second-class citizens (58). The
role of different social groups and intellectual traditions in the origin and first
phase of the Scientific Revolution is difficult to disentangle. In any case, the
establishment of the royal academies in Britain and France may be seen as the
first step towards pacification between the elite of the new science and social
and political powers (59). Other such moves occurred at later stages; for
instance, Stahl's efforts in Germany to put theoretical chemistry on the map
(60).
When the mathematical and the experimental traditions become associated
with different social statuses, this difference _will be maintained by social
processes like co-optation, differential recruitment and social and intellectual
distancing. Medically trained men for a time provided a middle ground, socially
belonging to the higher classes, but intellectually often inclined toward the
experimental tradition (or to holistic approaches as in the Naturphilosophie).
Mter the complete mathematization of physics, the distinctions may become
quite sharp, as shown for example by Pyenson and Skopp for physics and
chemistry around 1900 (61).
Such thoughts are speculations, however. They have been set down
because they derive from looking at science from the point of view of
restrictedness, and combine social and cognitive aspects in a bridging
concept. The fruitfulness of the concept of restrictedness itself lies in its
bridging the gap between social and cognitive aspects of scientific developments, which is a necessary condition of achieving a real understanding of
science.
234
Arie Rip
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
235
236
Arie Rip
24. If Janich, op. cit., 1978 (Note 5), pp. 19-20, is followed, theory is considered as a
means for making knowledge communicable, including the knowledge obtained by
an experimenter in terms of what is and what is not successful. On such a view of
theory, it does play an important role.
25. Multhauf, op. cit., 1966 (Note 20), pp. 216, 222, also 281; also Allen G. Debus,
Man and Nature in the Renaissance, London: Cambridge University Press, 1978,
pp. 29,45-7,53. There is a curious parallel in Debus's terminology for humanist
methods, when he speaks about the humanist adherence to ancient philosophy
"provided that they were assured that their texts were pure and unadulterated"
(ibid., 7).
26. Debus,op. cit., 1978 (Note 25), p. 46.
27. Multhauf,op. cit., 1966 (Note 20), p. 353; also Marie Boas, The Scientific Renaissance 1450-1630, London: Fontana Books, 1970, p. 154. Other developments are
the wide circulation of recipe books and the start of industrialized drug production
(Multhauf,op. cit., 1966 (Note 20), pp. 258, 262).
28. Szabadv3.ry,op. cit., 1966 (Note 20), pp. 41-6 and 58-9.
29. Thus Multhauf, op. cit., 1966 (Note 20), p. 349.
30. Henry Guerlac, 'Some French Antecedents of the Chemical Revolution', Ambix 5
(1959) 73-112.
31. Karl Hufbauer, 'Social Support for Chemistry in Germany during the 18th Century:
How and Why Did It Change?' Historical Studies Phys. Science 3 (1971) 205-31.
The strategy of intellectual distancing noted here is treated also by R. G. A. Dolby,
'On the Autonomy of Pure Science: The Construction and Maintenance of Barriers
between Scientific Establishments and Popular Culture', in this volume.
32. The discussion in this paragraph is based on Alfred Romer (ed.), Radiochemistry
and the Discovery of Isotopes. New York: Dover, 1970. See also T. J. Trenn, The
Self-Splitting Atom. A' History of the RutherfordSoddy Collaboration, London:
Taylor and Francis, 1977.
33. Nicholas Wade, 'Guillemin and Schally: The Years in the Wilderness', Science 200
(1978) 279-82; also Bruno Latour and Steve Woolgar, Laboratory Life. The Social
Construction of Scientific Facts, Beverly Hills and London: Sage, 1979, p. 60.
34. Lois Wingerson, 'Mistaken Identity of Proteins Threatens Biochemists' Results',
New Scientist 86 (24 April 1980) 192-3.
35. Latour and Woolgar, op. cit., 1979, Note 33.
36. Bohme and Van den Daele, op. cit., 1977 (Note 7), p. 223 (my translation).
37. The experimenter in fact attempts to control chaos (compare Ziman, op. cit., 1978
(Note 14), p. 59) and in spite of the successful exclusion of ever more "disturbing
factors", ambignity remains: is the effect real, or a result of a "dirty system"?
38. Bohme and Van den Daele,op. cit., 1977, Note 7. The same distinction is made by
Janich,op. cit., 1978, Note 5.
39. Studies of the development of the instrument industry have mostly concentrated
on the economic aspects. The inter-relationships with scientific developments and
the problem of quality control are well worth further study. In chemistry, for
instance, the classical elemental mass analysis is now fully automated, and often
performed by special laboratories. Folklore has it that one in ten analyses goes
wrong, and that mistakes are not always spotted. Clinical analyses are troubled by
the same problem, aggravated by the greater variability of the samples and the risk .
40.
41.
42.
43.
44.
45.
_46.
47.
48.
49.
50.
51.
52.
237
238
53.
54.
55.
56.
57.
58.
59.
60.
61.
Arie Rip
TERRY SHINN
C N R S,Paris
240
Terry Shinn
----1
Fr:equcncy
of contact
--
241
infrequont
mod"rat"y frequent
Ir,qucnf
242
Terry Shinn
243
hand, to their materialization in the form of bench work, on the other. This
countenances two intellectual operations. First, individual scientists transform
specific aspects of the original theoretical options and hypotheses into cognitive terms suitable for experimental treatment. They thus translate sometimes
vague or inoperable theoretical phrasing into experimental language. Secondly,
the scientists corroborate and validate observation. To some extent they
supervise experimental procedures, and in those cases where findings do not
coincide with expectations, it is they who attempt to identify the cause of
discrepancy. (For a representation of cognitive patterning in the chemistry
research units, see Figure 2). The cognitive patterning of these research units
conceptual
theory
data I information
collation
thus features a marked degree of specialization as particular operations theorization, experimentation, validation of observations, etc. - accrue to a
predetermined segment of the personnel, and as activity in more than one
domain of research is uncommon.
In many respects the social and cognitive characteristics of laboratories
engaged in computerized vector analysis differ dramatically from those
involved in mineral chemistry. The source and tenor of authority is diffuse
244
Terry Shinn
and ill defined. The hierarchic apparatus is unsubstantial and attenuates group
distinctions within the laboratory. Exchanges between the personnel are
largely oral, multidirectional and profuse. In like fashion, cognitive specialization is minimal. Individuals take on one task or another in conformity
with the needs of a given research project. Work is neither allocated nor
permanently fixed.
While, in principle, the organization and operation of computer laboratories falls under the aegis of a research director, in reality, decisions are taken
collectively by the ensemble of the personnel and are arrived at informally.
It cannot be said that research directors "wield authority". In de facto termS
their prerogatives are largely nominal, as in most situations their voice counts
for no more than that of their colleagues. Here, authority is not an affair
of orders and coercion, but rather a matter of persuasion and communal
responsibility. Hence, as a matter of routine, scientists and technicians alike
are caught up in the process of establishing research strategies and selecting
appropriate methods. In almost all instances, these portentous decisions, as
well as lesser ones, are reached through casual and lengthy discussion which is
suspended only after some sort of consensus is reached. Decision-making is
therefore a relatively cooperative and flexible venture in which peers come
together to share their individual reflections, and agree to a common course
of action.
The hierarchic configuration of computer laboratories is both unimposing
and uncomplicated. They feature only three categories of personnel: research
directors, a single stratum of scientists and one contingent of technicians
(see Figure 3). Scientists constitute the largest group, as they account for
roughly 60 percent of the staff, outnumbering technicians by 2 : I. The
geometric representation of this hierarchic disposition could be thought of as
an elongated parallelogram, where the negligeable distance separating the
peak and the base evokes the tightly-knit nature of the organizational setting.
The setting is fluid, semi-amorphous and free of encumbering barriers and
divisive intermediary strata. This organizational malleability even extends
to professional mobility, as almost one scientist in two becomes a research
director before retirement, and over 80 percent of technicians are promoted
to the position of scientist. It is not incidental that the technical personnel of
these laboratories enjoy a much more advanced education than the analogous
categories in mineral chemistry labs. While the two contingents of chemistry
245
Inlrequent
technicians have, at best, one or two years' training, and far more often do
not even possess a, secondary school diploma, those in computer centers
usually have one if not two university degrees.
The integrating dynamic of these research units is also manifest in their
propensity for unprogrammed and unfettered communication. All three
categories of personnel stand equal as initiators and recipients of administrative and scientific information. The place occupied in the communication web
by scientists and technicians is no less salient than that of research directors.
Data and instructions move with the same regularity along the vertical and
horizontal axis. The quantity of information exchanged among the research
staff is prolific and contact tends to occur in the most unsystematized manner. As a rule, exchanges take the form of spontaneous conversation involving
from two to four individuals. Scheduled meetings are rare, and written
documents are practically regarded as anathema. This suppleness of contact
applies not merely to intra-laboratory exchanges, but also to inter-research
unit communications. Albeit research directors seem somewhat privileged
in their dealings with other research centres, scientists, and for that matter,
technicians as well, also maintain working relations with outside scientists and
occasionally with government bodies and industry.
In computer laboratories, intellectual activity is not subjected to the
severely constricting effects of the division of labour. The full range of tasks
246
Terry Shinn
engendered by research is undertaken at different times by different individuals in conformity with their momentary tastes, temperament and current
state of intellectual preparedness. People are not assigned work; they gravitate
toward particular operations for the duration of a specific project and then
shift to different types of work as new projects emerge. Within the confmes
of the computer research centre there exists, therefore, a bastardized variety
of the "butterfly society" where intellectual migration comprises an overriding feature of cognitive inquiry.
Research themes are sometimes proposed by a scientist and at others by a
research director or member of the technical staff. In the three laboratories
studied for this project, about half of the themes were suggested by scientists,
a fifth by research directors, and a third by technicians. The task of translating
the research topic into a coherent and mathematically manipulable set of
preliminary hypotheses is similarly shared by the entirety of the research
team. This extremely difficult, abstract, and often abstruse work, does not
All
belong to the research director or to a select group of senior
of the personnel are actively included in the process. The same is true for the
assessment and interpretation of the numerical values provided by the computer. It is virtually certain that over the span of two research programmes,
each person within the research unit will participate in the fmal step of
research, attempting to make sense of the quantitative data, to fit it into the
outline provided by the working hypothesis and to relate it to questions
of theory either in the domain of physics or mathematics. Hence, in these
laboratories the domains of conceptualization and analysis are not restricted
to the few, since it is seen as fundamental that each individual contributes
fully in the creative dimension of mathematics-based investigation (see
Figure 4).
In the same vein, all levels of personnel are involved in the more tedious
and less imaginative work of the laboratory. Technicians are not alone in
carrying out such tasks as secondary programming, transferring arithmetic
functions on to tape, and later collecting and collating computer data sheets
and claSSifying the information they contain. Scientists and research directors
are also customarily occupied with such activity. Granted, the amount of
time they invest in routine operations is less than that of technicians, yet,
they nevertheless shoulder over half of the more banal work; and this,regardless of their seniority or formal position in the research unit. Inside computer
247
conclptual
theory
data I information
coliaiion
laboratories, it seems then, that cognitive operations are not, or are but slightly, bound to a set of rigid conventions. It is the established norm for each
individual to take up the challenge of the complete research programme, which
implicitly necessitates contact with the more arduous aspects of investigation
as well as the opportunity to take part in the creative conceptual tasks.
While the social and cognitive patterning of physics research units evince
some of the traits common to chemistry and computer laboratories, they also
manifest a host of unique aspects (5). Leadership is semi-participatory, as
various factions continually contest the power of decision. The hierarchy is
segmented, resulting in internal cleavage, and while communications are
definitely channelled, they nevertheless frequently follow a circuitous path.
Cognitive operations are also torn between the effects of centripetal and
centrifugal influences. Subordinate personnel are formally relegated the task
of carrying out and overseeing some of the less creative work, yet they also
initiate opportunities to take part in conceptual operations. Conversely,
senior staff seek to maintain a grasp over matters touching on the formulation
of hypotheses and the assessment of fmal results, but they are often compelled
to take on less purely cerebral tasks.
248
Terry Shinn
I
I
I
249
,
I
,
,,
I
I
I
I
I
,,
I
I
,
\
painfully formulated. For their part, junior scientists and technicians similarly
coalesce as they strive to affect or penetrate the
of decision-making. Indeed,
attacks on the bastion of power rarely prove felicitous.
The lower echelons seldom succeed in reorienting laboratory planning or
operations. This apparent impasse is partially mitigated, however, as a fairly
sizeable group of junior scientists and technicians climb into the upper echelons and thereby become full-fledged participants in policy determination.
Almost one junior scientist in three rises to the post of senior scientist or
research director as does one technician in six.
The communication network mirrors the compartmentalized dimension of
this hierarchic arrangement. The network consists of two semi-hermetic
conjoint communication circuits; the one composed of senior scientists and
a research director, the other of junior scientists and technicians. Within each
of the circuits, contact is irlcessant and personal, and interactions take both a
250
Terry Shinn
horizontal and vertical path. But as already reported, while in the subordinate
echelons, the spirit of solidarity prevails and colours discourse between
individuals, the bias of exchanges in the upper echelons is far more erratic and
problematic. Despite this dual organizational configuration, a substantial
body of information nevertheless does move between the two circuits. This
generally transpires during the process of daily work, and the agents responsible for the flow are primarily senior and junior scientists whose research tasks
inevitably bring them together on a regular basis. Of no less importance in
these physics laboratories, a balance is struck between written and oral
exchange. Interactions inside each circuit are largely oral, as are most of the
contact between the two categories of scientists. But perhaps in an effort to
neutralize informal exchange, research directors make certain that written
memoranda represent an alternative communications channel, and one of no
little importance at that.
In these physics research units, intellectual tasks tend to be both narrowly
defined and predesigned. Each category of personnel is responsible for a
specific mental operation. But in addition to this, there exists a parallel
covert structure where certain areas of research activity escape the strictures
of official rules. Here, different members of the staff address given elements
of the research project in terms of the programme's inherent logic, rather
than as a function of their formal position in the laboratory hierarchy.
Theory-related questions are without exception coopted by research
directors and senior scientists, as are the bulk of matters associated with
scientific strategy and methodology (see Figure 6). Together, these two
groups constitute a clique which excludes other laboratory agents from
abstracting activities. This interdiction is as unwaveringly upheld at the close
of research projects as at their genesis. Thus, a restricted group takes full
charge of the entirety of conceptual synthesis and analysis. At the opposite
end of the spectrum, junior scientists and technicians are called on to design
and set up experiments, to conduct them, and to present the resulting observations in an orderly and comprehensible manner. Put in another way,
they are responsible for the experimental apparatus, its effective operation,
and for the quality and codification of experimental results.
These intractable cognitive matrices do not, however, seem to provide a
wholly adequate intellectual framework for this kind of phYSics research. The
distance separating conceptualization on the one hand, and experimentation
251
conceprual
th.ory
on the other, is seemingly too vast. The mechanism required to link these
two elements frequently proves inadequate. In fact, this linking mechanism
introduces two associated processes: (a) the implementation of extraordinarily precise experimentation. This necessitates scrupulous and unflagging
procedural care as well as more than a modicum of insight into the phenomena
under investigation; (b) the conversion of frequently esoteric and disjointed
observational commentary into significant data. These twin sensitive intellectual operations are dependent on advanced familiarity with both technique
and theory, and the required mix of the two components is shaped by the
exact parameters contained in the research programme. This aspect of mental
activity thus cannot be preprogrammed and in these research units informal
and makeshift arrangements are consequently elaborated so as to neutralize
the shortcomings of the officially advocated cognitive organization.
To summarize the main argument up to this juncture: scientific research
activity gives rise to a range of social and cognitive organizational configurations. Moreover, cognitive patterning and social dispositions are symmetric.
In each one of the three cases described above, the internal structure of the
252
Terry Shinn
253
discussion of some factors that infuse different domains of scientific enterprise with their specificity of practice and structure.
2. Thoughts on Causality
Inevitably the sum of factors contributing to the social and cognitive morphology of scientific activity is incalculable. Certainly, however, elements like
public policy, the availability of funding, and the formulation of priorities
by highly sophisticated research institutes have, in recent years, played an
important role in shaping scientific processes; and the dreams, ambitions and
eccentricities of individual investigators have, always, been central to the
mould of research operations. Yet while determinants such as these undeniably merit much additional attention, this analysis of organizational causality
focuses on two alternative agents. First, social disposition will be assessed in
the light of the imperatives and inhibitions engendered by specific experimental procedures and by the use of certain kinds of experimental apparatus.
In short, this hypothesis posits that the logic contained in experimental
operations, and the manipulation of instruments, impinges significantly on
the cast of social relations. Secondly, cognitive patterning will be examined
in terms of the various components which constitute the epistemic structures
of different sub-disciplines. The key idea is that distinct areas of research are
built around a specific mix of thought categories which delimit the contours
of cognitive organization.
To take the case of the chemistry laboratories described above: the correspondence between mineral chemistry research and the mechanistic organizational model stems from both distinctive experimental practices and mental
processes which, in their turn, are contingent on the category of matter and
the particularity of phenomena under investigation.
Conventionally, mineral chemistry deals exclusively with a narrowly
defined class of objects and their. most immediately ascertainable manifestations; or, as expressed by the historian and philosopher of chemistry, F. A.
Paneth, it directs attention to "elements" and "properties" (7). The class of
matter examined in this instance is mineral, or mineral-associated, and the
goal of the researcher is to identify, claSSify and qualify such substances
and to understand their metamorphosis when in mixture or solution with
analogous or dissimilar chemical agents. Mineral chemists are often taken up
254
Terry Shinn
with the study of mineral-based compounds and alloys. In addition, this subdiscipline is preoccupied with the physical character of mineral substances,
for example, their colour, transluscence, texture, density, elasticity and so
on. This is achieved either through direct unmediated visual observation or
through the use of a battery of rather traditional laboratory equipment which
ranges from interminable numbers of beakers, test-tubes, balances, etc. to
somewhat more elaborate devices for gauging light defraction or mechanical
distortion. In a word, the area of inquiry referred to as mineral chemistry
tends to concentrate on the most palpable and directly measurable features of
a given group of elemental substances.
The unavoidable reliance of this form of research on extensive versus
intensive laboratory bench work and on an abundance of germane chemical
substances, tends in itself to dictate the bounds of what might be regarded as
the only genuinely feasible social framework for its effective operation. As
inquiry in this domain necessitates the simultaneous deployment of multiple
experimental parameters, and the innumerable repetition of each type of
experiment, a relative abundance of subordinate personnel is required to
prepare and conduct experiments, and to record observations (8). Furthermore, since experimental procedures are normally straight-forward, and
the nature of observation is generally unproblematic, the educational and
conceptual qualification of the technicians and laboratory assistants, who
carry out the major part of this bench work, is elementary. This experimentation and observation necessitates a large-scale fIltering and arrangement of
fmdings which frequently prompts supplementary bench work in order to
control and validate recalcitrant items. Such verification is always carried out
directly by the better trained and more learned intermediate-level scientific
personnel. The work of the scientist is thus not as mechanistic as that of the
technical staff, although it is often painfully time-consuming and arduous (9).
It is also essential to point out that in this type of laboratory, logistics-related
questions also have their importance. The need to assure an uninterrupted
and massive flow of appropriate chemical agents to the laboratory, and to
guarantee an adequate stock of experimental instruments, becomes a monumental administrative task when demand is urgent and requirements diverse.
In the same manner, because of the simultaneity of a host of experimental
paradigms, and the presence of several strata of personnel within these
research centres, the task of liaison sometimes takes on looming proportions.
255
From this, it should be clear that the sheer volume of experimental practices,
the tedium of translating observation material into sound data and the need
for an assertive brand of authority, go far to establish the type of centralized,
restrictive and formal social relations found in chemistry research units (10).
like the social configuration, the epistemic. structure, and hence the
cognitive organization of mineral chemistry laboratories, is also powerfully
influenced by the importance and specificity of observation data. The fundamental cognitive aim of this research is to identify elementary substances and
to establish categories of permanent relationships. Several mental operations
enter into this process: (a) the perception of bits of information which are
held to represent defmite relationships between clear-cut parameters; (b)
the instantaneous selection of pertinent elements of information from the
total observational universe; (c) the transcription of observations into an
accurate and communicable set of signs and the assembly of these signs into a
manageable format, (d) the analysis of observation data and the extraction
of significance. Of course, the initial suppositions underlying the research
programme are founded on indirect knowledge of previous work, or on
immediately preceding experimentation. The process of research is thus
circular in character; and this is particularly true in mineral chemistry which
is highly dependent on the veracity of anterior conceptual canons.
It seems from this that while theory is just as crucial for achievement in
mineral chemistry as it is in other scientific fields (11), the place which it
occupies is nevertheless specific, and different from that of many other
sub-disciplines (12). In mineral chemistry the prinCiple of "reductionist
development" constitutes an important methodological tenet. This principle
stipulates that a maximum of observations must be encapsulated in a minimum
of theoretical axioms (13). This orientation is further strengthened by the
principle of "theoretical clusters" which submits that a given zone of inquiry
retains validity if the bulk of the field's axioms are revealed as adequa}e even
though a number of its fundamental propositions have been shown to be
invalid (14). In effect, the application of the reductionist and cluster tenets
radically attenuates and sometimes altogether obviates, the impulsion for
fresh theory, for the active development of new conceptual canon is frequently seen as inconsistent with the epistemic thrust of the sub-diScipline.
The dominant mental operation in these research units is instead bound
to analogy construction. This exercise gives focus to the myriad bits of
256
Terry Shinn
257
258
Terry Shinn
259
260
Terry Shinn
the exact nature of the experimental problems and with the details of observational expectations. Indeed, although problems like the notion of "non-separability" and the Heisenberg indeterminacy principle are rarely voiced, they
nevertheless do represent very real methodological and practical obstacles
to sound research. For this sort of reason, the social organization of these
research centres must, by all means, institutionalize mechanisms which
oblige senior scientists to collaborate closely both with junior scientists and
technicians, and furthermore, which force them to enter into the experimental
operations themselves. This trait constitutes the essential dimension of the
organic organizational model which ensures a high degree of integration, but
within the confines of functional specificity.
As the terrain of investigation of solid-state physics spans a variety of
physical phenomena and forces, research activity, ipso facto, incorporates
a broad scope of mental processes, and hence episternic structures. Theory is
assuredly central to the growth of knowledge in this sub-discipline, as is the
amassing of refmed and detailed observational particularities; and between
these two operations, there are situated equally crucial thought categories like
retroductive hypothesis formulation and metaphor-based model construction
(21).
Within this field, narrowly declarative descriptive statements, expressing
mechanistic elements of observables, are granted some legitimacy. Acceptance
of this type of statement is, however, sometimes illusory, as its long-term
viability is strictly contingent on integration into a broader context whereby
it either adds to, or provides a basis for, comprehensive explanatory formulations. Thus, while declarative descriptive statements do not in themselves
constitute advancement, they are regarded as satisfactory if they eventually
contribute to a global representation of phenomena (22). Theory development
is similarly precarious, and theory lacks the sacrosanct status which it enjoys
in many other sub-disciplines. On the one hand theory is patently transient; it
is routine to see sets of ideas brutally forced aside by competing notions. On
the other, it is remarkably stable since seemingly faltering theories are often
retained through fine adjustments and reconstrual (23). In solid-state physics,
anomaly tends to figure importantly in the process of theory construction
and in the credibility accorded to a given theory. Depending on specific
circumstances, some anomalies are viewed as sufficiently damning to lead to
the outright suppression of a theory. Under other conditions, the presence of
261
262
Terry Shinn
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
263
264
Terry Shinn
15. See Max Black, Models and Metaphors: Studies in Languages and Philosophy,
Cornell University Press, Ithaca, New York, 1962, chapters 3 and 13, particularly
pages 220-6; also see Mary Hesse, Models and Analogies in Science, Sheed and
Ward, London and New York, 1963, pp. 30-54; 65-75; 84-98; see also Hesse,
The Structure of Scientific Inference, Macmillall, London, 1974; see also Harre,
The Principles of Scientific Thinking, Macmillan, London, 1970, chapter 2.
16. See A. Sloman, The Computer Revolution in Philosophy, Harvester Press, Hassocks,
1978.
17. See Benson Mates, Elementary Logic, Oxford University Press, New York, 1965;
see also Ian Mueller, 'Euclid's Elements and Axiomatic Method', The British Journal
for the Philosophy of Science 20 (1969) 229-308.
18. It is unclear as to whether Sloman agrees with this analysis; at certain points he
seems to suggest the highly integrated nature of mathematical operations, while at
others he apparently posits a more disaggregated approach. See Sloman, op. cit.
19. Here the term "hybrid" is used as a metaphor and does not allude to Ben-David's
concept of "hybridization". See J. Ben-David and R. Collins 'Social Factors in the
Origins of a New Science: the Case of Psychology', American Sociological Review
31 (1966) 451-65.
20. For a discussion of "tool experts" see Ravetz, op. cit., pp. 90-1.
21. See Black, op. cit.; see also N. R. Hanson, 'The Logic of Discovery', Journal of
Philosophy S5 (1958) 1073-89; 'More on the Logic of Discovery', Journal of
Philosophy S7 (1960) 182-8; Patterns of Discovery, Cambridge University Press,
Cambridge, 1958.
22. For an example of this process, see Nicholas Georgescu-Roegen, The Entropy Law
and the Economic Process, Harvard University Press, Cambridge, Mass. 1971,
chapters Sand 6.
23. For an analysis of several theory types'see A. Rappaport, 'Various Meanings of
"Theory" ',American Political Science Review S2 (1958) 927-88.
24. Hanson, 'Logic and Discovery', 'More on the Logic of Discovery', Patterns . ..
op. cit.
25. For an evocative, albeit fumbling reductionist approach in the analysis of science,
see B. Latour and S. Woolgar, Laboratory Life. the Social Construction of Scientific
Facts, Sage Publications, London, 1979.
26. In another article when treats the organizational specificity of research laboratorjes,
Shinn utilizes the social variables of education as an explanatory device. See T.
Shinn, 'Division du Savoir et Specificite organisationelle des laboratoires de recherche industrielle en France', Revue Franfaise de Sociologie 21 (1980) 3-35.
27. A new study, just under way, employs these five analytical variables. See T. Shinn,
'Transparence et opacire scientifique', project outline for a research project for the
DGRST-CNRS programme science technologie societe.
PART III
R.G.A.DOLBY
University of Kent at Canterbury
1. Introduction
Modem science has grown into a large-scale and complex activity. In an age
of increasing bureaucracy and specialization, the knowledge construction
industry is also bureaucratized and segmented. The intellectual divisions
within science and between science and comparable activities, now hallowed
by tradition, have become locked into the institutions of science, to be
exploited by the relevant interest groups. Cognitive boundaries have thus
been turned into social barriers. When these barriers were being built up, they
could be justified in terms of a convergence of the cognitive requirements of
efficient knowledge production and the social advantage they conferred upon
insiders. The arguments can conveniently be encapsulated in terms of the
notion of expertise. In pure science, the more expert an individual is on
a topic, the more he can be trusted as an authority. Expertise is acquired
through specialist training, association with other experts and by making
recognized contributions to knowledge relevant to the topic in question.
The institutionalization of knowledge advancement particularly involves
organizing the activity of individuals high in hierarchies of expertise. Any
contribution to knowledge is ideally directed to those who can best appreciate
it, that is, to those who are most expert. And contributions from those who
are most expert are the most readily appreciated. Informal communication
networks naturally emerge among the elite of science; more formal institutions
try to limit themselves to the higher part of hierarchies of expertise by setting
267
Norbert Elias, Herminio Martins and Richard Whitley (ed,.j,Scientijic Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 267-292.
Copyright 1982 by D. Reidel Publishing Company.
268
R. G. A. Dolby
269
As custodians of knowledge, they control a means of orientation for individuals in our society and, in practice, put pressure on sociologists to conform
to their methods and standards of production of knowledge. Elias's paper
illustrates the perceptions of the articulate outsider, and offers a revealing
repertoire of the strate gems of the heterodox figure in attacking the social
institutions of the orthodoxy. However, it is not obvious that we should
simple replace the ideology of the insider by that of the outsider. And we
should certainly not take Elias's historical claims at face value (2).
Any exploration of the barriers around science must cross territory scarred
by a grand issue: that of the autonomy of pure science. Although now
increasingly under attack, the doctrine of the autonomy of pure science
is still well entrenched in science and in the disciplines studying science.
Apologetics for science of the mid-century put the traditional view in a sharp
and now familiar form: pure science has its own methods and standards of
knowledge advancement, and any external social interference in science will
either be resisted or will lead to a degenerate form. Nazi rejection of Jewish
science and Stalinist Russian rejection of bourgeois capitalist science are
therefore to be seen as unacceptable state interference; the success of Lysenko
in promoting his eccentric agricultural ideas by relating them to official state
ideology is to be seen as producing mere pseudoscience. That is how the
autonomy of pure science was defended in the time of totalitarian regimes
in Germany and Russia; echoes of the same rhetoric can still be heard (3).
The issue of the autonomy of science and in particular of the production
of scientific knowledge is complex and multifaceted. Empirical or theoretical
study of the processes of boundary construction around science will inevitably
be coloured by prior commitments to its disciplinary manifestations. Where
some see the rational institutionalization of objective features of science,
others see the ideological workings of professional self-interest.
My purpose in reviewing these difficulties of coming to a satisfactory
understanding of scientific establishments and the barriers that surround
them is to suggest that the issues are so blurred that those who claim to
provide clear and practical answers are likely merely to be attempting to
legitimate their own interests. This paper is limited to seeking to enrich our
background knowledge of the construction of intellectual barriers and to
finding ways in which methods of learning about such issues can become
more decisive. The intellectual barrier I will examine in most detail is that
270
R. G. A. Dolby
271
avoiding the escalation of controversy and for giving priority to limited and
more manageable forms of disagreement. T. H. Huxley, for example, could
counter Bishop Wilberforce's appeals to general sentiments in the audience
by proclaiming the status of Darwin theory to be something which only
the expert biologist was competent to judge. Huxley was able to use this
strategem persuasively for that audience, because Wilberforce's descent to
ad hominem rhetoric against Huxley (he asked whether it was on his grandfather's or his grandmother's side that he was descended from a monkey)
licensed Huxley to make an ad hominem reply. Huxley could expose the
errors in the Bishop's speech, and thus suggest that here was one non-biologist
who was not competent to judge such a technical biological matter. Thus
Huxley was able to sharpen up a relevant distinction between expert and
non-expert and give persuasive reasons why the more confined setting should
be given epistemological priority.
The strategy of constructing intellectual distance between the forms of
discourse prevailing on two sides in a dispute can, then, be very effective
when linked to acceptable hierarchical notions of relevant expertise. It can be
used (not always successfully) in conilicts between rival disciplines within
science, and as we shall see, it has had great success in confrontations between
science of the establishment and popular science.
I now wish to introduce one of the battlegrounds in which the rhetoric of
true and false science has been employed in the construction of intellectual
barriers. If we accept that the scientific establishment is contained within
the institutions of the elite of hierarchies of scientific expertise, we must
accept that a proportion of the interest in science lies outside such exclusive
institutions, in the numerous less expert individuals whose primary access to
science is through popular culture. Much of the science of the popular culture
popularization of orthodox science, but in addition, intellectual enterprises
are mounted which fail to penetrate the scientific establishment. These are
one of the targets to which the rhetoric of the philosophical images of science
are regularly applied. Barriers are often invoked or constructed to separate
such enterprises from true science, and the notion of intellectual distance is
employed in showing just how different in kind and quality they are from
true science.
In the modern age of professional science, the barriers between the qualified practitioner of a scientific discipline and the unqualified amateur are
272
R. G. A. Dolby
so clearly present that problems of exclusion are usually minimal. But the
case we will look at in detail, that of evolution just before Darwin, took place
when there was no relevant professional science. There were certainly exclusive
institutions for the elites of the sciences, but in amateur science, many members of each new generation must earn their own admission to the elite. The
battle was thus over who deserved to be admitted. The different institutional
settings for orthodox and popular science encourage some differences in the
prevailing styles of reasoning and argument. Orthodox science typically
strives for consensus through rigorous argument from agreed observational
data with modest theoretical postulates. It is specialized, and each region of
inquiry is deliberately limited to that which is recognized as achievable by
the accepted methods used. In contrast, popular science aspires only to
plausibility, for there is insufficient expertise for consensual judgements
to rigorous standards to be made. Rather than limit itself to what can be
achieved by established methods, popular science seeks systematic generality
and maximum significance on questions that trouble its supporters, and it
flourishes by maintaining popular interest. New ideas are often transient in
popular science, appearing only while they are newsworthy, but those ideas
which are successfully related to the lasting interests and commitments of a
sector of society may persist in that sector even though they have dropped
from wider attention.
Very often the forms of popular science against which the scientific
establishment finds it desirable to construct barriers has a style more suited
to the popular forum than to that of scientific orthodoxy. It is ,wide ranging
in coverage, offering insight into and explanations of many popular issues.
It is visionary and programmatic rather than rigorous and testable. It gains
its support without the mediation of the scientific expert. This deSCription
applies especially well to the case we will go on to consider, of evolution just
before Darwin, but we are encouraged to take the case seriously because,
after Darwin, very sirnilar ideas became the scientific orthodoxy.
273
complex and highly differentiated. One of the fuller and better known of
such presentations was that of J.-B. Lamarck. But by the 1830s, considered
scientific opinion, especially as represented by such influential men of science
as Georges Cuvier and Charles Lyell had rejected Lamarck's idea of the mutability of species. Lamarck was thought to have produced mere speculation
which lacked a satisfactory causal mechanism for species change, and which
was against the general run of available evidence. In spite of this scientific
rejection, evolutionary speculations continued throughout Europe. In Britain,
discussion focussed for a time on an anonymous work, Vestiges of the Natural
History of Creation (1844). The essential claim of this book was that the
world is regulated by law and that God works in the world by means oflaw.
The author sought to show that there is a universal law of progressive development which applies to the formation of the solar system from a primordial
nebula, to geological change Ipld to the creation and transformation of life on
earth. life is spontaneously generated under favourable physical conditions,
and under suitable conditions each species is gradually modified, generation
by generation, from simpler to more complex forms. Man is the highest
product so far of this development process. The author did not attempt to
provide a mechanism for the evolutionary process, but had assembled a much
richer collection of supporting arguments than Lamarck had been able to
do some decades earlier. Vestiges generated considerable popular interest.
In Britain it had sold over 20,000 copies in eleven editions by 1860 (5).
Reviewers noted that it was particularly influential among those who had no
specialized knowledge of its subject matter. Mllny of its enthusiasts were
women or artisans seeking to educate themselves, but it was also much
discussed in the general culture of the influential classes of society. The
ideas it expressed were rich in their human implications and stimulating in
their controversial status. And discussion was frequently rekindled by new
rumours as to its authorship. It was not until some years after his death in
1871 that it was fmally revealed that the author had been Robert Chambers,
an Edinburgh publisher.
In contrast to its popular interest, Vestiges was vehemently condemned by
men of science. We shall examine the reasons the reviewers offered for their
rejection and the social and cultural interests they defended. We will then
look more closely at the distinction between popular and scientific forums
of discussion, as it was made by the critics of Vestiges and accepted by its
274
R. G. A. Dolby
author in his reply (6). After looking at the ways in which Vestiges differed
from orthodox science by being in the popular domain, we will examine its
influence - its lasting positive influence, and its negative influence on the
strategy forced upon Darwin and his supporters in the subsequent presentation of Origin of Species (1859).
275
Thus Sedgwick was not merely prepared to describe the conclusions of the
book as mischievous nonsense, but as antisocial nonsense (I 8).
Sedgwick made it clear that in attacking Vestiges, he was defending social
and cultural interests - those of the orthodox man of science, committed to
the existing methods and results of science, and to their harmonious relationship with true religion. He interpreted Vestiges as insidiously undermining
those interests, as being the more dangerous because it simulated real science.
He did not hide his social outrage, because he was appealing to others who
276
R. G. A. Dolby
shared his interests, though they might not have had sufficient scientific
knowledge to be aware of the corruption the bland packaging of Vestiges
concealed.
Vestiges as Popular Science
The scientific critics of Vestiges insisted that its author could not be a man of
science and that he was writing for a popular audience. David Brewster, for
example, contrasted the single-minded search for truth of the astronomer or
naturalist with
those revellers in speculation who practice their orgies in the temple of science, ransacking its storehouse for the materials of hypothesis, and not infrequently adulterating
them for popular taste, or fashioning them for vulgar apprehensions ... (19).
Adam Sedgwick confessed to having thought that the book might have been
written by a woman, explaining in a mixed metaphor,
We were led to this delusion by certain charms of writing - by the popularity of the
work - by its ready boundings over the fences of the tree of knowledge, and its utter
neglect of the narrow and thorny entrance by which we may lawfully approach it above all by the sincerity of faith and love with which the author devotes himself to any
system he has taken into his bosom (20).
277
"laudable industry and zeal, there was also an intellectual timidity rendering
all the results philosophically barren?" (22). And the conclusion was finally
drawn,
Do men of science have their minds suitably prepared to receive with candour, or treat
with justice, a plan of nature like that presented in the Vestiges of Creation? No, it must
be before another tribunal, that this new philosophy is to be truly and righteously
judged.
Chambers' biographer suggests that this 'other tribunal' was presumably the
educated layman who was buying out every edition of Vestiges (23).
We see, therefore, that Chambers and his critics agreed that Vestiges was
a work of popular science rather than a contribution to the community of
specialized scientific expertise. Even though this occurred at a time before
science was professionalized, the distinction between the institutionalized
single-minded search for truth and the unspecialized contribution to popular
culture, was clear to the participants. Once we recognize clearly the popular
nature of Vestiges, we can understand better some of its characteristics and
some features of the reaction against it.
Vestiges sought, not to establish modest conclusions with disciplined
rigour and inductive certainty, but to stimulate interest in topics which were
regarded as important in the wider society. Its technique, therefore, was to
be suggestive and thought provoking rather than to prove its points. Knowledge claims in orthodox science build only on work which is widely accepted
or which the author hopes to persuade his audience to accept readily. For in
that way they can become authoritative. Scientists avoid drawing points into
their argument which have generated inconclusive controversy. An individual
mid-19th-<:entury scientist might have been inclined to accept the nebular
hypothesis, or spontaneous generation, or phrenology. But to invoke any of
these areas of controversial interest in a separate different scientific argument
would only help bring the latter, too, into disrepute. In contrast ,Chambers
appealed to all these areas in his argument, and maintained his commitments
to each of them as long as he thought them defensible in successive editions.
1his was a very bad strategy in orthodox science, but it was exactly what a
popular audience wanted. The development hypothesis gained its interest by
being shown to be relevant to a wide range of existing issues, and by drawing
upon examples which caught the popular imagination. Furthermore, as
278
R. G. A. Dolby
popular science, Vestiges did not limit itself to the scientific ideas it covered.
Unlike Darwin's presentation of evolution fifteen years later, Chambers
spelled out clearly his conception of the implications of the theory for man
and for man's relation to God. A popular audience was less interested in
science as such, than in the implications of a scientific idea for matters of
greatest human importance.
Some modem commentators on Vestiges have asked themselves why it
was that scientists reacted so strongly to the work when they could have
condemned it simply by ignoring it (24). And part of the answer seems to be
because it was a form of popular science that threatened the main thread of
social justification of science which had built up in popular science literature.
For several decades all the respectable leaders of scientific opinion, of whatever school,
had been actively engaged upon the project of popularizing natural philosophy among
mechanics and disseminating science to untutored multitudes. The result, it was thought,
could only elevate the moral sentiments of the working class and stabilise its situation by
demonstrating the providential and material necessity and the comprehensive beneficence
of industriallU"rangements. But now, it suddenly appeared, it made a great deal of
difference what sort of scienCe was popularized. Hitherto scientists themselves, misunderstood to be sure, had been unfairly charged with encouraging infidelity and moral
anarchy. And just as they were successfully solidifying the religious cement of industrial
society, a real heretic, cleverly and falsely got up in the guise of science, came to undo
their labours and to. demolish the framework (25).
Vestiges was not a threat as a contender for scientific knowledge - most men
of science rejected it out of hand. But its general readers, who had not the
advantage of expert knowledge could easily be seduced and corrupted by it.
The criticism came because it was so plausible and so lucid and was generating
popular interest, and although attack gave it still further publicity, perhaps
it would eventually discredit it in the public domain.
Vestiges and Later Evolutionary Theory
One of the problems arising in the vast body of historical discussion of the
Darwinian revolution has been to clarify the relation between Darwin and
his evolutionary predecessors. On the one hand, it is possible to read such
early evolutionists as Lamarck as providing a well thought out theory which
differed from Darwin's only in that it did not employ the particular mechanism of natural selection (26). On the other hand, it is possible to regard
279
Darwin's as the first truly scientific theory of evolution and his predecessors
as merely vague speculators. Modern historians of science have developed a
dislike of looking for precursors of the recognized proponents of scientific
ideas. Such chains of precursors can be forced back indefinitely into the past,
provided the documents are reinterpreted in the perspective of a modern
framework, rather than in terms of the interests of the time in which they
were written. The search for precursors has been ridiculed as the disease of
'precursoritis'. But in the case of Darwin, to refuse to recognize the significance of precursors is to take sides in a late-19th-century-debate - a debate
in which Darwin's supporters stressed his originality and his critics (such as
Samuel Butler) accused him of plagiarism. I think that the present study can
show why it was. that Darwin and his supporters tended to exaggerate the
difference between The Origin of Species and its precursors; in brief, it was
because those precursors, and especially Vestiges of the Natural History of
Creation, had come to be regarded as popu1ar pseudo-science. The strategy of
gaining acceptance for Darwin's theory required him to stress the intellectual
distance between his own work of scientific natural history, which was
'proved' by the standards of the science in question, and the imperfect
popular speculations which preceded it. The Origin of Species kept to a
minimum the references to man and to God, on which controversy was most
heated. It was not open to Darwin to appeal to the authority of an earlier
tradition of evolutionary ideas, for that would have encouraged his audience
to regard Darwin as having merely produced yet another unproven speculation. T. H. Huxley, for example, in an earlier commentary, described the
author of Vestiges as a dreamer, "by whose well-intentioned efforts the
Lamarckian theory received its final condenmation in the minds of all sound
thinkers" (27), and in contrast [mis] described Darwin as one who "abhors
mere speculation as nature abhors a vacuum ... The path he bids us follow
professes to be, not a mere airy track, fabricated of ideal cobwebs, but a solid
and broad bridge of facts" (28).
One of the interests of more recent history of biological science has been
to explore the extent to which biological theories were influenced by contemporary social ideas. Has the history of biology followed its own internal logic
of problem generation and problem solution, or has it merely reflected the
state of the external culture? This has been a special interest because of the
frequency with which biological theory has been applied back to arguments
280
R. G. A. Dolby
281
from it and so provoking further wrath from his scientific critics. When
spiritualism became the rage later in his life, he showed enthusiastic interest
in that. Spencer wrote early articles on phrenological topics. He had a critical
interest in positivism, and wrote major treatises on psychology and sociology.
The causes championed by A. R. Wallace, whose paper on evolution by
natural selection stimulated Darwin into publication, are comparable. In
his youth, Wallace had dabbled in and become committed to phrenology,
mesmerism, and phrenomesmerism. He had become convinced of the fact of
evolution by reading Vestiges. This interest was a spur to taking up a career
as a naturalist. Later in life, he took up other unorthodox interests, including
spiritualism and psychic research. Wallace's evolutionary work illustrates one
of the routes for popular ideas from unorthodoxy to scientific respectability.
Although Darwin did not belong to this social circle of intellectual adventurers, he was undoubtedly influenced by it in the presentation of his ideas,
particularly in his later writings. But not all the influence was positive. When,
in 1859, Darwin fmally published the theory he had been gestating for more
than twenty years, he went to some trouble to separate it from the popular
discussions. He made only minimal reference to man and to religion, although
he had always been personally interested in the implications for man. Instead
he concentrated on an inductively arranged argument and a response to the
scientific objections he had had ample time to anticipate. Having won over a
small group of supporters in his own area of natural history, they helped him
convert more of the scientific orthodoxy. Darwin was not actively involved in
the confrontations in the controversy which was generated, always pleading
ill health. But T. H. Huxley, in particular, was an effective advocate. Huxley,
who had earlier written a savage review of Vestiges (32), maintained his
intellectual consistency by stressing the differences between Darwin's theory
and its popular precursors and by insisting on the methodological superiority
of the present theory. Although Darwin's supporters did not avoid the wider
implications of an evolutionary biology, they insisted that the status of
the theory should be settled as a purely scientific matter. I have already
mentioned Huxley's opportunistic exploitation of Bishop Wtlberforce's
mistakes at the Oxford meeting of the British Association for the Advancement of Science in 1860. It appears that Huxley, like Wtlberforce, had not
desired or planned for this meeting (33). The day before the exchange with
the Bishop, Huxley had sought to avoid discussion of the theory because,
282
R. G. A. Dolby
" ... a general audience, in which sentiment would unduly interfere with
intellect, was not the public before which such a discussion should be carried
on" (34).
Interest in evolutionary theory rapidly became widespread in British
society. Among the many currents of evolutionary discussion, writers such
as Spencer were prominent. Spencer's evolutionary ideas now had a new
scientific status. Spencer became a philosopher of the age amplifying newly
fashionable ideas into a world view. Many late-19th century intellectuals
passed through a phase of interest in (and often later reaction against) his
systematic expositions. When the leading natural historians eventually turned
to the detailed examination of the evolution of man, they found it impossible
to limit their discussion to 'pure biology'. Inevitably, the themes, preoccupations and prejudices of the popular literature found some expression in their
work. The inheritance of acquired characteristics, particularly habits returning
as instincts, a Lamarckiat). idea given full support by Spencer, was increasingly
accepted by Darwin in his writings on man. Ideas of inequality between the
sexes, social classes and races, which had been much discussed in the popular
literature were often given unthinking recognition in the work of the most
serious scientists. The idea that there is a single line of evolutionary development (part of the chain of being), which had been a common feature of
popular discussions since Lamarck, and which made it natural to suppose
that one could study other cultures and races as corresponding to earlier
evolutionary stages of European man, inspired mid- and late-19th-century
anthropology. The idea that the developing child, like the unborn embryo,
recapitulates the evolution of the species, was carried via the popular literature from its pre-Darwinian origins into late-19th-century psychology (35).
According to this doctrine, children instinctively climb trees and hide in caves
as our ancestors habitually did. The idea that nature is (like modern capitalist
society) a place of open competition, of struggle for survival in which only
the fittest perpetuate their kind was applied by social Darwinists to justify
laissez-faire capitalist society. All these features of evolution came, not
from the pure biology of Darwin and his friends, but from the popular and
semi-popular evolutionary philosophy which preceded and accompanied
their work, and from which they could not completely insulate their own
reasoning.
The common context argument, then, is effective for those aspects of
283
evolutionary theory which were primarily developed as popular and semipopular literature. But scientists of the late-19th century became increasingly
concerned to separate their own activity from popular science and to ensure
that all of the new science followed a rigorous methodology. Let us return
for illustration to the writings of T. H. Huxley.
In Huxley's generation, British science was for the first time dominated
by those gaining a living from science. There was a new self-consciousness
among this group, and Huxley's campaigns and enterprises on occasion
defended the interest of the new occupation (he used this word rather than
'profession'). Huxley presupposed a sharp distinction between the established
science to which he contributed and science in the popular culture. He
frequently addressed both audiences and was occasionally obliged to defend
his willingness to devote effort to popularization of science. But he also
played his part in strengthening and maintaining the boundary between the
two. For example, he spoke against the difficulties produced by those "who
think that what they have picked up from popular exposition qualifies them
for discussing the great problems of science" (36). In his writings on pseudoscience, he deplored
the waste of time and energy bestowed on the endeavour to deal with the most difficult
problems of science, by those who have neither undergone the discipline, nor possess the
information, which are indispensable to the successful issue of such an enterprise (37).
The author of Vestiges was among those he was alluding to. What he wished
to attack was
a process of mystification, based upon the use of scientific language by writers who
exhibit no signs of scientific training, of accurate scientific knowledge, or of clear ideas
respecting the philosophy of science, which is doing very serious harm to the public (38).
284
R. G. A. Dolby
Huxley, though he was a friend of Spencer's, was among those who objected.
The need to separate scientific knowledge from popular speculation could be
accomplished in this case by pointing to the fallacy of arguing from 'is' to
'ought'. Spencer had only succeeded in constructing his laissez-Iaire values
from evolution by assuming that society should take a form in which competition rather than cooperation is crucial for survival (39).
3. Sociobiology
No attempt is made in the discussion which follows to provide a comprehensive introduction to or analysis of the recent controversy over sociobiology.
However,it is of interest to apply the understanding developed in the previous
section to a contemporary issue, one in which the heat of battle has yet to
die away, and with which many readers will already be familiar. The present
controversy has focussed around the 'new synthesis' of E. O. Wilson, who
sought to redefme the disciplinary boundaries between studies of animal
behaviour and population genetics by drawing the material into a single
quantitative theory. However, in his large textbook (40), Wilson did not limit
himself to a synthesis of existing scientific learning; he added a fmal chapter
in which sociobiology was applied to man. It was mainly this extrapolation
which triggered the controversy. The central claim was that Wilson was
resurrecting biological determinism, a favourite ideological device for rightwing viewpoints. Wilson's presentation, it was claimed, failed to separate
out the effects of "the personal and class prejudices of the researcher" (41).
Although Wilson claimed to feel intimidated by the hostility (42), which he
insisted was based on misconstrual or misrepresentation (43), he responded to
the challenge. Indeed, his readiness to maintain his side of the controversy led
him to produce a whole book on the sociobiology of man (44).
One problem Wilson had to face was that, although he claimed to offer a
new synthesis, he was by no means operating in an intellectual vacuum. The
efforts to biologize human social behaviour were increasingly discussed within
popular culture. There, publications could attract appreciative attention by
being internally plausible, comprehensive and giving a new sense to sociallysignificant issues. Many of them appealed to similar sectional social interests
to their predecessors earlier in the century. But some new ones were added.
There was, for example, Elaine Morgan's The Descent 01 Woman (45) which
285
286
R. G. A. Dolby
conflict into a general forum. This was not a matter to be settled by some
narrow group of experts. The devices Wilson had used to distance himself
from the immediately preceding popular writers were turned against him. He
had, for example, accused them of advocating unfalsifiable theories (52). His
most outspoken critics, the Sociobiology Study Group of Science for the
People argued in one criticism:
When we examine clIIefully the manner in which sociobiology pretends to explain all
behaviours as adaptive, it becomes obvious that the theory is so constructed that no
tests are possible: There exists no imaginable situation which carmot be explained; it is
necessarily confirmed by every observation (53).
287
Human sociobiology should be given a chance to prove its worth. If it cannot deliver on
its promises, it will collapse soon enough .. ; but if it does prove viable, then its success
could pay scientific dividends of the highest order (55).
fu terms of my account, this amounts to proposing that the human sociobiologists should be allowed to set about creating a new scientific community,
the leaders of whom we might come to accept as the experts in such matters.
That is what 'proving viable' amounts to in social terms. And the way in
which this would normally be done is by sociobiologists making and negotiating specific knowledge claims in such a way that critics can no longer
find a SOCially acceptable position from which rational objections can be
made. The cumulation of relevant uncontested knowledge by a community
is normally sufficient for us to concede to it an area of expertise. I think,
however, that it will be extremely difficult to establish human sociobiology in
this way. The conflict has already been widened to a popular forum in which
new claims of expertise are regarded with suspicion. In the absence of some
widely appreciated achievements upon which subsequent research could
build, the strategems available for de-escalating the conflict will take time to
be effective.
4. Concluding Remarks
Professionalism is undoubtedly a factor in the enforcement of the barriers
between and around the established sciences in the present age of professional science. But we have seen that such barriers were first created by the
elite of amateur science. Furthermore, professional self-interest tends to
conserve existing boundaries, but as the range of science has extended and its
internal linkages have been restructed, the boundaries have often had to be
renegotiated. Even within professional science, the boundaries that otherwise
function so readily as barriers do not force conformity upon elite innovators.
Wilson, for example, was quite prepared to cross existing disciplinary boundaries in creating sociobiology. Our studies of the boundary between popular
and established science show that although it is exploited as a barrier on
many occasions, it must be crossed by those who wish to create new outposts
of scientific orthodoxy. The pioneers of social science had to face this problem
in an acute form. Although they could invoke the methods of professionalized
288
R. G. A. Dolby
289
290
4.
5.
6.
7.
R. G. A. Dolby
5.
8. For example, one reviewer said that Vestiges claimed its hypothesis to be true
because it is "drawn from the response of nature, and the response of nature is
only true when it gives back the watchwords of the hypothesis" (A. Sedgwick,
'Vestiges of the Natural History of Creation', Edinburgh Review 82 (July 1845)
41).
9. See, for example, (F. A. Bowen,] 'Vestiges of the Natural History of Creation',
North American Review 60 (1845) 465.
10. Anonymous reviewer of Vestiges-in The Athenaeum, No. 897, Jan.4.1845, p. 11.
11. (A. Gray,] 'Explanations of the Vestiges', North American Review 62 (1846)
503. Detailed criticisms of the use of resemblances in Vestiges are also made by
Sedgwick,op. cit., 1845 (Note 8), pp. 5-9, and by Bowen, op. cit., 1845 (Note 9),
p.451.
12. Anonymous reviewer, op. cit., 1845 (Note 10), p. 11.
13. Idem.
14. (D. Brewster,] 'Vestiges of the Natural History of Creation', North British Review
3 (1845) 470.
15. A. Sedgwick, A Discourse on the Studies of the University afCambridge, 5th edn.,
London and Cambridge: J. W. Parker, 1850, p. xix.
16. A. Sedgwick,op. cit., 1845 (Note 8), p. 63.
17. Letter from Sedgwick to Lyell. J. W. aark and T. M. Hughes, The Life and Letters
of The Reverend Adam Sedgwick, Cambridge: Cambridge University Press, 1890,
Vol. 2, p. 84.
18. A. Sedgwick, op. cit., 1845 (Note 8), p. 3.
19. D. Brewster, op. cit., 1845 (Note 14), p.4 73.
20. A. Sedgwick, op. cit., 1845 (Note 8), pp. 3 -4.
21. R. Chambers, op. cit., 1845 (Note 6), p. 175.
22. Ibid., p. 176.
23. Millhauser,op. cit., 1959 (Note 5), p. 144.
24. See, for example, C.C. Gillispie, Genesis and Geology (1951) New York: Harper,
1959, p. 162.
25. Ibid., pp. 150-1.
291
26. As an example of the search for precursors, see B. Glass et al. (eds.), Forerunners of
Darwin: 1745-1859, Baltimore: Johns Hopkins Press, 1959. The classic of the
genre of precursor'ileeking histories of science is P. Duhem, Le Systeme du Monde;
Histoire des Doctrines Cosmologiques de Platon du Copemic, 10 Vols, Paris:
Hermann, 1913-59.
27. T. H. Huxley, 'The Darwinian Hypothesis' (1859), reprinted in his Collected Essays.
Vol. II, London: Macmillan, 1894, p. 13.
28. Ibid., pp. 20-1.
29. See, for example, R. M. Young, 'Malthus and the Evolutionists: The Common
Context of Biological and Social Theory', Past and Present, No. 43 (1969) 10945; R. M. Young, 'The Historiographic and Ideological Contexts of the NineteenthCentury Debate on Man's Place in Nature', in M. Teich and R. M. Young (eds.),
Changing Perspectives in the History of Science: Essays in Honour of Joseph
Needham, London: Heinemann, 1973, pp. 344-438.
30. R. M. Young, 'Natural Theology, Victorian Periodicals and the Fragmentation
of the Common Context', in C. Chant and J. Fauvel (eds.), Darwin to Einstein:
Historical Studies on Science and Belief, Harlow, Essex: Longman, 1980, pp. 69107.
31. See, for example, the works of Young just cited in Note 29 and 30.
32. T. H. Huxley, 'Vestiges of the Natural History of Creation', British and Foreign
Medico-Chirurgical Review, 13 April 1854.
33. Altholz,op. cit., 1980, Note 4.
34. Quoted by F. Darwin from The Athenaeum, July 14,1860. Darwin, op. cit., 1888
(Note 4), p. 320.
35. See, for example, the writings of J. M. Baldwin and G. S. Hall.
36. T. H. Huxley, Discourses: Biological and Geological. Volume 8 of his collected
essays. London: Macmillan, 1894, p. viii.
37. T. H. Huxley, 'Science and Pseudo-Science' (1887), in Science and Christian Tradition. Volume 5 of his Collected Essays. London: Macmillan, 1894, p. 116.
38. Ibid., p. 177.
39. T. H. Huxley, 'Evolution and Ethics' (1893), in Evolution and Ethics and other
Essays. Volume 9 of his Collected Essays, London: Macmillan, 1894, pp. 46-116.
40. E. O. Wilson, Sociobiology: The New Synthesis, Cambridge, Mass: Harvard University Press, 1975.
41. E. Allen, et al., 'Against "Sociobiology" " New York Review of Books, November
13, 1975, pp. 182, 184-6. Reprinted in A. L. Caplan (ed.), The Sociobiology
Debate: Readings on Ethical and Scientific Issues, New York: Harper and Row,
1978,p.264.
42. N. Wade, 'Sociobiology: Troubled Birth for New Discipline', Science 191 (1976)
1151-5. Reprinted in Caplan, op. cit., 1978 (Note 41), p. 332.
43. E. o. Wilson, 'Academic Vigilantism and the Political Significance of Biology',
BioScience 26 (1976) 183, 187-90. Reprinted in Caplan,op. cit., 1978 (Note 41),
p.291.
44. E. O. Wilson, On Human Nature, Cambridge, Mass: Harvard University Press, 1978.
45. E. Morgan, The Descent of Woman, London: Souvenir Press, 1972.
46. See, for example, R. Ardrey, African Genesis: A Personal Investigation into the
Animal Origins and Nature of Man, London: Readers Union, 1963; The Territorial
292
47.
48.
49.
50.
51.
52.
53.
54.
55.
R. G. A. Dolby
Imperative: A Personal Inquiry into the Animal Origins of Property and Nations,
London: Collins, 1967.
See, for example, D. Morris, The Naked Ape: A Zoologist's Study of the Human
Animal, London: Cape, 1967; The Human Zoo, London: Corgi, 1971.
WIlson,op. cit., 1975 (Note 40), pp. 27-31.
L. Tiger and R. Fox, The Imperial Animal, London: Secker and Warburg, 1971,
p. x.
Ibid., p. xi.
Wilson,op. cit., 1978 (Note 44), p. x.
Wilson,op. cit., 1975 (Note 40), p. 28.
E. Allen, et al., 'Sociobiology: A New Biological Determinism', in Sociobiology
Study Group of Boston (eds.), Biology as a Social Weapon, Minneapolis: Burgess,
1977. From the quotation in M. Ruse, Sociobiology: Sense or Nonsense?, Dordrecht: Reidel, 1979, p. 111.
I have in mind in particular the very helpful book by the philosopher of science M.
Ruse,op. cit., 1979, Note 53.
Ibid., p. 214.
1. Introduction
At any given point in the development of a scientific specialty, there exists
some finite set of research topics which are seen as legitimate, interesting,
and feasible by the members of the specialty. How do the members of the
specialty achieve an allocation of their efforts across these topics? What does
the aggregation of these individual decisions imply about the development of
the specialty? To what extent are these consequences shaped by the activities
of scientific 'hierarchies' (Le., formal and informal organizational structures,
policies, and leaders)? And might these consequences be shaped differently
(and perhaps more effectively)?
These questions have been posed before, in various guises (1). Typically,
however, each is addressed separately with little regard for its companions.
For example, the prevailing view on the question of individual topic choice
appears to favor a 'competitive grazing' model. The specialty is pictured as
a bounded meadow into which individual scientists, competitors for recognition and priority, allocate themselves. Areas of heavy over-grazing become
less attractive, virgin pasture is sought out, and the overall territory represented by the available research topics is Uniformly exploited. Thus,according
to Hagstrom (2), those who discover important problems upon which few
others are engaged are less likely to be anticipated and more likely to be
rewarded with recognition. As a result, scientists tend to disperse themselves
over a range of possible problems.
While Hagstrom suggests purposive choice in the matter, Price (3) in
examining the phenomenon of multiple discovery, suggests a simple random
model, in which the ripe apples of discovery are seized by the reaching hands
293
Norbert Elias, Herminio Martins and Richard Whitley
Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 293-311.
Copyright 1982 by D. Reidel Publishing Company.
294
of the blind scientific harvesters. (It should be noted that Price's demonstration of reasonable fit of his model to Merton's (4) multiple-discovery data
should not be over-interpreted; a similar fit would be found if the harvesters
picked from only a single tree, neglecting the rest of a huge and fruitful
orchard.)
2. Research Trails and Scientific
The simple agricultural images cited above accord poorly with our impression
of how scientific problems are actually chosen - with the passion and concern
shown by individual scientists. These scientists select their projects under the
influence of a complex of organizational, cultural, political, and intellectual
conditions. Our intent in this essay is to sketch the outlines of a richer
and more complex model, in which individual choice of 'research trails' is
embedded in such issues as the distribution of research effort across problems
in a specialty, the development of a specialty over time, and the role of
hierarchies and elites in shaping these processes.
Our central concern is with the implications for an aggregate of scientific
work, or specialty, of the choices of individual researchers - what Schelling
(5) refers to as the relationship between 'micromotives' and 'macrobehavior'.
Our micro motivational assumption will be that individual researchers attempt
to choose their research topics 'sensibly' - neither randomly nor omnisciently
- in pursuit of a bounded self-interest. This assumption, we will suggest,
leads directly to an important and unfortunate set of implications for the
development of specialties: that there exist important pressures which lead to
undue persistence of individuals in some research trails rather than others;
that the social processes associated with the development of these trails tend
toward conservative pressures for intellectual continuity on new entrants; and
that the aggregate result of these processes as is that, far from a wide dispersion
of research effort around the boundary problems of a specialty, there will be
unproductive over-concentration on some few problems, while high-potential
areas go underdeveloped (6).
The present relevance of specialties is their thoroughly relativized definition (7). Once relativized, they are easily reified. To posit their existence and
classify researchers relative to them does not make them real. Yet as Kuhn (8)
intimated, and several others (9) have demonstrated, innovations in a science
295
often come from outside its perceived boundaries. Edge and Mulkay (10) call
this 'marginal innovation'. But what is crucial is the perception that a novel
idea and its bearer are marginal to a particular specialty. If we remember that
novel ideas are routinely transmitted via the published literature, preprints,
and the spoken word, then the analytically decisive process may be the
transi tion of a researcher from one specialty to another.
Despite the fluidity of specialty defmition (11), the migration of researchers, we would argue, is a discernible process which results in the exploitation
of problems. The attraction of migrants may indeed identify a specialty as
'hot', though policy -makers with discretionary resources to allocate are likely
to foster this attraction, and once fostered, to reinforce it (12). Relativism
notwithstanding, then, the construct 'scientific specialty' implies extra-local
intellectual linkages amongst an aggregate of researchers at some point in
time.
Alternatively, a specialty can be conceptualized as the confluence of
several research trails, each representing a sequence of work by an individual
or a small team of researchers. These research trails may be distinguished by
some continuity of focus - be it methodological-theoretical and/or problemoriented - in published research (13). A research trail thus directs attention
to the coherence and development over time of a series or program of research
projects undertaken by an individual or group (the latter often consisting of
local colleagues), while the notion of a specialty directs attention to the
coherence amongst several such trails at a particular point in time. As in the
cable bundles found in complex electrical devices, a particular research trail
may be part of one bundle or specialty for a period of its history, then branch
off to become part of another bundle, and so on. A slice through the bundle
at some point reveals some semblance of the membership and interrelationships between the current members; following each wire from its source to
its destination shows the linkages over time within one single trail. The
'res.earch trail' notion, then, emphasizes the development over time of the
activities of an individual or team of researchers.
If a research trail corresponds to a body of work at the level of a 'problem'
or 'problem area', then a specialty is an aggregation of related problem areas
(14) defined retrospectively. It is the reciprocal and relative nature of trails
and specialties which render them empirically 'slippery' but analytically
fruitful (see below). That is, we observe a distribution of research effort
296
297
(2) The costs to the researcher will generally tend to peak early in the
research trail, declining to a much lower, and roughly constant, level thereafter. A number of familiar items are included in this non-recurring initial
set-up cost: time and effort involved in reviewing the relevant literature;
learning the techniques of research in the area chosen; acquiring suitable
equipment, instruments, colleagues and facilities; securing research support
as an outsider with no prior track-record of performance in the area; and so
on. Once incurred, these costs generally do not continue, but drop quickly
to a much lower level.
(3) Payoffs to the researcher are not limited to intellectual yield alone. We
assume that publication is positively valued by most researchers, even if
intellectual yield is modest. Further, an additional publication within an
established research trail has the effect of maintaining interest and visibility
- a veneer of success - for the body of work as a whole, so that connected
publications may be valued more highly than are isolated ones. Finally,
we suggest that research trails rarely meet an unambiguous dead end; they
typically peter out in studies of small impact or marginal variation.
Any formal consideration of .the combined effect of these factors would
suggest that investments in research topics, i.e., problem choice, involve
elaborate calculations. We postulate no such elaborate choice process. Rather,
as Knorr observes, "opportunities for success emerge routinely and naturally
from the flow of research" (18). This keeps the risks, or at least the team's
calculation and perception thereof, low. In Knorr's words, "once made,
investments tend to stabilize the effort" or to generate experiments that are
"on the safe side" (19).
298
akin to those of 'a tinkerer who uses everything at his disposal to produce
some kind of workable object' (21).
What Jacob's imagery connotes may well appear more descriptive of
the basement inventor than of the gleaming equipment of the professional
researcher. But if we consider the imagery as describing the intellectual,
rather than the physical, equipment of the researcher, the picture is more
vivid. At any point in time, a researcher has available a stock of intellectual
'odds and ends' - techniques, conceptual notions, insights, and intuitions generated by the training and research trails in which she has participated to
date. Further, these 'odds and ends' are not merely free-floating entities,
open to recombination into whatever intellectual objects the researcher may
choose. They possess a strong measure of internal coherence. Just as complex
eyes evolved through a series of simpler, but still functional, primitive photoreceptors (22), so scientific novelty is 'tinkered' onto the end of previous
functional constructs. The history of the research trail, in short, places
powerful constraints on where it may next develop.
While admitting their power, we cannot allow these constraints to be
absolute. Novel tinkering is always possible; and the form of the new intellectual objects is shaped by a second set of constraints, arising outside the
intellect of the individual researcher. It is these external environmental forces
that shape the social production of scientific work, that select from the many
potential novelties those few that will become 'real science'. Of course, the
results of this interaction at a given time will affect the materials available for
'tinkering' at some later time. In this sense, we embrace both the 'internalist'
and the 'externalist' views of scientific development. The internalist view,
here represented by the scientist's perception of coherence within a research
trail, sets limits on the range within which tinkering is possible; the externalist
view stresses the selection, by forces external to the individual scientist, of
the novelties that will be retained and those that will be discarded. Thus,
with evolutionary theory we embrace a contextual view of scientific work
(23).
We hasten to note, however, that the internalist-externalist issue is closely
mirrored in the recent re-emergence within evolutionary theory itself of the
debate between the 'adaptationist' and 'structural integration' traditions.
In Gould's (24) sketch of the extreme positions, one sees the direction of
evolution as channeled "by adaptive requirements of local environments,"
299
while the other stresses the role of "the nature of variation and the morphology of the system in which it arises" (25).
Gould's comments on the differences in emphasis between the two programs are sharply relevant to our concerns here. Natural selection (externalist
pressures) are, of course, the fundamental engine of change. However,
the possible routes of selection are channeled by inherited morphology, building material,
and the amount and nature of variation itself. Though selection moves organisms down
the channels, the channels themselves - rather than the paucity of weU-designed outcomes - impose primary constraints on the direction of change .... Selection on one
part of a structure may impose a set of correlated and nonadaptive modifications on
other parts of an integrated body. Many features, even fundamental ones, may be
nonadaptive (though not, to be sure, strongly inadaptive) either as developmental
correlates of primary adaptation or as 'unanticipated' structural consequences of primary
adaptations themselves (26).
300
petitive struggle for resources and acclaim with various hierarchies - local and
extra-local alike. What we now consider is how these hierarchies, especially
through policy intervention, impinge on problem choice, thereby locking
researchers into and out of research trails.
5. Trails and Hierarchies: Research Persistence and Policy Pressures
The investment model sketched above proposes that two contexts - local
and extra-local (e.g., national policies, international specialties) - structure
opportunities for the pursuit and abandonment of research trails. A retrospective assessment of opportunities and successful local tinkering with them
may be equated with the aggregate growth of specialties and the knowledge
claims therein. However, the amenability of problems to policy-manipulation,
goal-direction and eventual translation of research into utility can be questioned (29). Is certain knowledge resistant to manipulation? Must theories be
'finalized' before they can become policy tools? Does political action - an
attempt to use scientific knowledge - vindicate its claim to truth? Such
questions can be construed as investment rationales. They are part of the
rhetoric that policy-makers and researchers use to justify the funding support
they, respectively, give and receive. Such 'vocabularies of justification' (30)
vary with the context, the audience, and the research problem in question.
Presently, the two contexts of analytical concern are the local research site
and the wider social and intellectual environment in which knowledge claims
are negotiated with political, professional, and lay audiences. The viability
of research trails can be seen as an outcome. of such negotiation. Prior to
it, however, we may ask what minimal conditions must be satisfied for a
researcher or team to pursue a trail.
5.1. Conditions Favoring Pursuit ofa Trail
Four conditions, each with local and extra-local repercussions, can be distinguished as predisposing a researcher to pursue a research trail: legitimacy,
funding, access to local resources, and training capacity. To be faithful to our
relativistic argument, we qualify these conditions as being peculiar to postWorld War II U.S. science.
301
302
(iv) Training capacity. One resource that warrants separate discussion is the
training capacity of the organizational site. Academic labs typically serve as
training and certification sites for neophyte scientists. As several authors have
observed, the presence of such pre- and post-doctoral personnel enhances
research continuities (41). Put another way, the practices of training and
entry to research careers will tend to exacerbate the tendencies to undue
persistence in worked-out or low-yielding research trails. Graduate students,
for example, are likely to seek dissertation advisors who can provide well-
303
304
and Hagstrom sees it as naturally dispersing, we see it as excessively concentrated. We also see it - problem choice that begets trails which, in turn,
converge to form specialties - as extraordinarily sensitive to local exigencies
and extra-local pressures that must be socially negotiated to ensure intellectual viability.
5.2. Political Pressures and the Evolution of Specialties
In retrospect and at an aggregate level, research trails sustain scientific specialties. Specialties envelop a set of research problems and a cadre of researchers
whose local tinkerings attract national investments. Science policies and
funding programs are formulated to rationalize those investments. Investments are made on a competitive basis and mediated through mechanisms
such as peer review. Scientists, as appointed government advisors, erstwhile
consultants, and proposal reviewers, playa collusive role in defining 'scientific
merit', encouraging certain research trails and creating 'hot' specialties. This,
too, is 'interpretive' science practice. For policies make a difference - both
in scientific careers and in the evolution of those knowledge-producing
communities we call specialties. And scientists - not all, but the advisingconsulting-reviewing elite - mediate those policies and the funding decisions
they warrant (44).
The elite (which is demographically older, more productive, and located at
more prestigious sites than the community at large (45)), is in the advantageous position of helping both to dispense and to receive extra-local resources.
The elite legitimates and funds. like all gatekeepers these elite scientists wield
power by allocating and withholding both the approval and the information
that is construed as 'specialty growth and decline'. The point is this: with
such a concentration of power among a relative few, the scientific elite plays
a functional role in th::: amelioration of political pressures and the evolution
of research. This elite represents, indeed is employed by, an interlocking set
of institutions which, at least in the U.S., garner the lion's share of available
resources, students, and faculty talent (46). Through reputation and iocation
alone, this elite can direct or redirect resources so as to affect the activity
perceived to be within a specialty and the subsequent viability (through
local project funding and pUblication) accorded it. The elite can, in short,
determine what problems are chosen and whose trails are sustained.
305
Such (re)direction imbues research goals and science policies with a fluid,
i.e., 'negotiated', character. Scientists may neither originate nor ostensibly
satisfy these goals and policies, but they intervene in significant ways. Just
as local teams generate idiosyncratic interpretations of literature and ongoing
research, extra.J.ocal hierarchies and elites operate so as to shape policyformulation, -implementation and -evaluation. Their role, while advisory, is
manipulative and selective; it is not purposefully or systematically conspiratorial. But in the evolution of a specialty, elites wield a disproportionately
large influence relative to their number or the richness of their research
tradition(s) (47).
Consider this in the perspective of evolutionary theory, as discussed earlier.
In this theory there exists a clear tension between processes that lead to
variation and those that lead to retention. A reproductive system that allows
no variation is doomed, because there is no chance of increasing adaptation
to the species' environment. Conversely, inadequate retention processes imply
that the selective filter is insufficiently sharp, and that unsuccessful as well as
successful variations are retained, with no overall gain in adaptation. A species
that produces only mutations is as doomed as a species that produces none
(48).
In the 'evolutionary-investment' model we have outlined, a scientific
specialty that retains only its currently-active research trails is doomed to
extinction; as the trails become exhausted, the practitioners move on, and
new entrants grow scarce. Conversely, a specialty which is excessively open
to novelty fails because of a lack of consensus as to what is known, what is
worth knowing, and what knowledge-generating practices are acceptable;
coherence is lost, faddism reigns, and the intellectual development of the
specialty is foreclosed. Therefore, elites must exercise some selectivity. But
as our case for undue persistence attests, the balance between variation and
retention favors the latter. Each study is likely to be all too similar to its
predecessors in the research trail; promising alternative trails are li1,{ely to be
ignored or inadequately explored; and the specialty drifts into pre-revolutionary rigidity, from which crises - the resistance of elites defending particular
orthodoxies and their status prerequisites (advisory and gatekeeping) notwithstanding -lead to radical restructuring of the lines of inquiry.
306
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
307
308
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Studer and Chubin, op. cit., 1980 (Note 9), chapter 3; M. Useem, 'Government
Influence on the Social Paradigm', The Sociological Quarterly 17 (1976) 146-61;
and Edward Yoxen, this volume.
For a similar usage of a pUblication series as a unit of analysis, see Diana Crane,
Invisible Colleges, Chicago: University of Chicago Press, 1972.
M. J. Mulkay, G. N. Gilbert and S. Woolgar, 'Problem Areas and Research Networks
in Science', Sociology 9 (1975) 187-203; also see R. Whitley, 'Cognitive and Social
Institutionalization of Scientific Specialties and Research Areas', in Richard D.
Whitley (ed.), Social Processes of Scientific Development, London: Routledge and
Kegan Paul, 1974, pp. 69-95.
Karin D. Knorr, 'Producing and Reproducing Knowledge: Descriptive or Constructive?' Social Science Information 16 (1977) 677. Barry Barnes' (Scientific Knowledge and Sociological Theory, London: Routledge and Kegan Paul, 1974, p. 58)
discussion of the scientist as 'bricoleur' is relevant to the notion of tinkering as
creating opportunism.
Bruno Latour and Steve Woolgar, Laboratory Life: The Social Construction of
Scientific Facts, Beverly Hills: Sage, 1979; K. D. Knorr and D. W. Knorr, 'From
Scenes to Scripts: On the Relationship between Laboratory Research and the
Published Paper in Science', Social Studies of Science 11 (1981), forthcoming; also
Shinn, this volume.
P. Bourdieu, 'The Specificity of the Scientific Field and the Social Conditions of
the Progress of Reason', Social Science Information 14 (1975) 19-47.
Knorr, op. cit., 1977 (Note 15), p. 680.
Ibid., pp. 680-81.
Francois Jacob, 'Evolution and Tinkering', Science 196 (1977) 1161-6.
Ibid.
T. Dobzhansky, Evolution, Genetics and Man, New York: Wiley, 1963.
The hazards of applying evolutionary theory to scientific work and intellectual
change are outlined in L. J. Cohen's essay review of Toulmin's Human Understand
ing, 'Is the Progress of Science Evolutionary', British Journal for the Philosophy of
Science 24 (1973) 41-61. In what follows, we have tried to heed Cohen's critique
of Toulmin's 'evolutionist philosophy of intellectual history' without succumbing
either to the pitfalls of the Darwinian metaphor as it describes a population of ideas
and the scientists who advance them or to Cohen's niggardliness in demanding more
than a parallel between the development of a scientific discipline and the evolution
of a biological species. Also see Ron Johnston, 'Contextual Knowledge: A Model
for the Overthrow of the Internal/External Dichotomy in Science', Australian and
New Zealand Journal of Sociology 12 (1976) 192-203.
Stephen Jay Gould, 'The Evolutionary Biology of Constraints', Daedalus 109
(1980) 39-52.
In an intriguing echo of our current concerns, Gould (Ibid., p. 40) notes in passing
that while "no one would, of course, embrace either extreme position in its entirety" the choice of one or the other as a starting point for inquiry critically shapes
the course of study. The exfoliation of the two research trails has been independent
to the point at which structural integrationist notions are almost entirely European,
adaptationist models almost entirely American.
Additional details from the evolutionary debate might clarify the parallelism we
26.
27.
28.
29.
30.
31.
309
310
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
311
44. See M. J. Mulkay, 'The Mediating Role of the Scientific Elite', Social Studies of
Science 6 (1976) 445-70; J. J. Salomon, 'Science Policy and Its Myths: The
Allocation of Resources', Public Policy 20 (1972) 1-33.
45. L. Groeneveld, N. Koller and N. Mullins, The Advisors of the U.S. National Science
Foundation', Social Studies of Science 5 (1975) 343-54.
46. H. Zuckerman, 'Stratification in American Science', in E. O. Lauman (ed.), Social
Stratification: Theory and Research for the 19708, Indianapolis: Bobbs-Merrill,
1970, pp. 235-57.
47. According to A. W. Gouldner, 'Prologue to a Theory of Revolutionary Intellectuals',
Telos 26 (1976) 3 -36; also see Elias, this volume.
48. Yet, in W. Schafer's ('Finalization in Perspective: Toward a Revolution in the Social
Paradigm of Science', Social Science Information, 1979, 915-43) words, .. 'Finalist'
alternatives in science invert the sequence postulated by classical evolutionary
theory. Variation no longer precedes the selection of new mutants but results from
prior selection" (p. 918).
RICHARD WHITLEY
Manchester Business School, University of Manchester
313
Norbert Elias, Herminio Martins and Richard Whitley
Establishments
and Hierarchies. Sociology of the Sciences, Volume VI, 1982. 313-357.
Copyright 1982 by D. Reidel Publishing Company.
314
Richard Whitley
315
316
Richard Whitley
317
318
Richard Whitley
319
320
Richard Whitley
321
322
Richard Whitley
differentiation within research laboratories which were integrated and controlled by an administrative hierarchy organized on the joint basis of technical expertise and knowledge and academic rank. Problems which required
collaboration between highly skilled equals or which were relatively diffuse
and general so that they could not easily be divided up into separate parts and
tasks were less likely to be taken up and worked on in these systems of work
orgimization and control than in more "amateurish" systems (28).
The systematic organization of scientific research in employing institutions
with clear hierarchies of authority based on differences in expertise and
knowledge demonstrated how the arcane and esoteric activity of discovering
the laws of the universe, undertaken by a few exceptionally gifted individual
"geniuses", could be rationally administered and controlled. It also showed
how discoverers could be produced on a relatively large scale so that when
science became seen as useful knowledge, and as a method for producing
useful knowledge rather than as a system of fixed knowledge (29),
did not need to wait for "geniuses" to appear but could simply expand the
existing system of scientist production. In a sense, the academic domination
of science through the production of knowledge producers, and their systematic use as research workers solving differentiated tasks which were coordinated by the laboratory head, provided the model of research organization
for employers in other spheres. By showing how scientists could be trained
and how neophytes could be organized to produce knowledge, the development of research schools in universities demonstrated not only that science
could be useful as systematic, fixed knowledge but also that the actual
process of knowledge production could be planned and organized. Rather
than waiting for the fruits of a few individual geniuses to fall off the tree of
knowledge and then be applied to extra scientific problems, the organization
of research schools demonstrated how scientific work could be administered
as a process oriented to a variety of goals. It was thus capable of being rationally organized and, by extension, planned for goals other than purely
academic ones. Science as a system of knowledge about the world which was
stable, true and coherent became transformed into a process or method of
knowledge production which could be organized and planned. The academic
professionalization of science thus paved the way for the large scale use
of science in non-academic environments and for non-academic purposes
(30).
323
324
Richard Whitley
because they control more resources. While not all research programmes are
equally "successful", and not all sciences are equally programmable, academic
establishments are more able to define what is "successful" because they
control the writings of the "official history" of the sciences in textbooks
and their use in training new recruits. Also, because they control the bulk of
the scientific labour force, both staff and students, they are more able to
dominate interpretations of the major goals of a science than are individual
"amateurs". As well as controlling the labour market they dominate the
definition of tasks and coordination of task outcomes through their control
of the major part of the knowledge production system.
Control over a substantial part of the scientific labour force is not the
only means of control over knowledge production exerted by academic
establishments. As part of the process of increasing precision, purity and
standardization of tasks and technical procedures, which partly developed
from the use of students in research, the technical apparatus required for
undertaking scientific research which resulted in competent contributions
became more complex and expensive. The raw materials similarly became
more refined and restricted in availability so that the individual practitioner
found it more difficult to conduct scientific research without collective
resources. The technical means of producing scientific knowledge increasingly
outgrew the capacity of individual provision and control so that they had to
be collectively organized and controlled. The more concentrated they were in
universities and under academic control, the more academic establishments
dominated reputational communities. The reduction of task uncertainty
through increased precision and standardization of measurements and observations rendered science more esoteric and professional, and more under the
control of academics.
The standardization of materials and procedures in many sciences in the
19th century was part of the general process of professionalization of scientific work. It aided the academic domination of the sciences through enabling
trainees to make contributions to knowledge while distancing the objects
and procedures of science from everyday, lay concepts and substances. The
simultaneous transformation of scientific research into an activity involving
systematic use of esoteric standard techniques conducted by teams andgroups
in a relatively extensive division of labour and into an activity necessitating
collective, and collectively organized, resources which were controlled by a
325
single major institution, the university, was a crucial part ofthe development
of professionalized academic science. The standardization and formalization
of much scientific research was also important in the development of professional reputational communities. As Kuhn points out (32), the mathematization of a science enables, ceteris paribus, controversies to be speedily resolved
and "normal" science to progress smoothly. By standardizing technical
procedures and symbol structures, communication across social and spatial
boundaries is facilitated and the reputational community more able to control
work practices and outcomes. The reduction of what might be termed technical task uncertainty enabled practitioners with a common set of relatively
standardized techniques to communicate task outcomes quickly and easily
and so coordinate them. Competence in such fields then becomes defined
as competence in the use of these techniques for dealing with appropriate
problems. Relatively clear social and cognitive boundaries can be demarcated
and reified and strong collective social control over the details of work
processes exerted. While sciences existed as reputational organizations beforehand, it was their identification with training and employment units built
around research programmes using standard procedures which strengthened
their identities and power. Equally, the domination of academics over employment and training units which inculcated and applied these procedures
facilitated their control over reputational organizations to the extent that
scientific fields have become synonymous with academic "disciplines" or
"specialties" .
4. Variations in Academic Domination of the Sciences
326
Richard Whitley
327
328
Richard Whitley
329
that the sciences only vary in their similarity to, say, sociology or philosophy
on the one hand and physics and chemistry on the other hand. In other words
they are either "polyparadigmatic", and apparently highly disputatious, or
highly restricted, centralized and bureaucratic. The problem with relying
entirely on this one dimension is that many other differences between fields
are glossed over or ignored and that some areas simply do not easily fit; e.g.,
economics, molecular biology, mathematics. Furthermore, there is a strong
implication in adopting this dimension on its own that the sciences inevitably
tend to develop in the direction of physics which is exactly the sort of
unidirectional assumption under attack in much recent sociology of science.
The plurality of the sciences, and of their paths of development, are better
understood by viewing them as reputational organizations which differ on a
number of dimensions.
Treating the sciences as one form of work organization in which control
of work practices, and sometimes goals, is dominated by the professional
reputational community, suggests two major dimensions for comparing and
analyzing changes in the structure of sciences: task interdependence and
task uncertainty (38). The extent to which tasks need to be coordinated
for significant contributions to be made to knowledge goals, and the predictability, visibility and repeatability of tasks and task outcomes are key factors
in organizing and controlling work, in science as elsewhere. In this section
I will sketch the use of these dimensions' for analyzing the sciences and
subsequently outline how they are interrelated and change.
(a) The Degree ofMutual Interdependence
Task interdependence among scientists as members of reputational organizations is derivative from their need to rely upon each other to produce
knowledge claims which are significant for organizational goals, and receive
validation from their professional colleagues. The more dependent practitioners are upon each other, as opposed to other groups, for reputations, the
more they have to coordinate their work and interrelate task outcomes. High
dependence on colleagues for validating contributions as significant and
competent implies dependence on the work of others and their goals and so
a high degree of ''functional integration" (39). It also leads to the objectification of knowledge objects and their detachment from the immediate interests
330
Richard Whitley
331
formalized procedures and languages, such as mathematics, and so specialization encourages the development and extended application of standard
techniques and highly formal media of communication. Interdependence and
coordination needs thus increase as the number of practitioners rises and
specialization intensifies, to the extent in physics of creating an entire class of
theoreticians which is itself subdivided (40). These theorists are "needed" to
make sense of the mass of task outcomes produced by highly specialized
experimentalists working on very specific problems and sub-problems, in
terms of the dominant goals of the field which are interpreted and "guarded"
by the establishment of theoretical physicists.
This greater dependence on professional colleagues arising from increased
specialization is further intensified by the concentration and centralization
of research facilities, journal space, funding agencies and other necessities
for conducting research and obtaining reputations. The more limited these
facilities are, and/or the more they are concentrated in a few centres or under
the control of a few people, the more crucial it becomes for scientists to
demonstrate that their contributions are important for the accomplishment
of major organizational goals. Their dependence on established colleagues for
validation of their work is therefore much greater under these circumstances
than when facilities are cheap and widely available as in, perhaps, mathematics. The hierarchies of specialties and sub-fields in physics - as contrasted
with chemistry - are partly a result of the high degree of concentration of
resources and funding sources. By successfully institutionalizing particle
physics as the primary field of physics and thus ensuring that major reputations in physics depended on access to very complex and expensive apparatus,
contemporary phYSicists so concentrated and centralized resources that a
relatively small number of scientists controlled the careers of large numbers
of physicists. Only by convincing established leaders and administrators can
practitioners obtain access to facilities that are essential to make major
contributions to the dominant goals and so acquire major reputations. Furthermore, the sheer size of the facilities necessitates a high degree of division
of labour and administrative coordination in the workplace so that the
independent scholar has to depend on a host of technicians and colleagues
simply to generate some output and organize it into a form where some
sense can be made of it. Not only, then, are scientists here highly dependent
on a small group of colleagues for the opportunity to do research but they
332
Richard Whitley
333
not always associated with strong centralization of control and the domination
of a single theory. While centralization of control over key resources in the
reputation system encourages increased dependence upon one's colleagues
in the sciences, the reverse is not necessarily the case. Strong reputational
organizations with control over jobs and careers may exist without being
organized around a central goal and approach to problems. A common educational background which develops similar skills and ensures that practitioners
share technical procedures here leads to a common professional identity and
enables the community to exclude outsiders from jobs without centralizing
control over facilities. Work in these fields, such as mathematics (42), is
highly specialized and involves formal communication systems for the award
of reputations which affect career prospects and yet no single view of the
field is dominant to the extent that particular problems or topics are regarded
as central. Resources for conducting research are cheap and widely available,
as are jobs and journals, so different goals may be pursued with little need to
order them into an overall hierarchy. Increased numbers of practitioners in
these fields simply results in greater differentiation and specialization of skills
and tasks which are coordinated through the formal language and common
educational background. As long as the various topic areas can obtain access
to jobs, pressure to organize them into some sort of hierarchy of importance
will remain slight. If jobs become scarce,
and relevance to disciplinary
on the other hand, some means of ordering contributions and topic areas in
terms of their importance will probably develop and more efforts will be
devoted to general syntheses of the field as a whole.
The development of strong reputational organizations in the sciences
with a high degree of interdependence among practitioners is associated with
increasing autonomy and distance from lay and non-scientific concerns and
goals. By definition, the relevance of exoteric audiences decreases as the
importance of esoteric goals and criteria for the award of reputations rises.
Equally, the development of a strong reputational community controlling
access to jobs and research facilities implies increasing autonomy from other
sciences and their goals and technical procedures. For a distinct reputational
community to form around separate goals and practices some autonomy must
be granted; once organized, such strong professions (43) seek to increase
and extend their autonomy by controlling access to jobs through insisting
on particular technical skills which can only be acquired through training
334
Richard Whitley
programmes controlled by that community. This in itself increases practitioners' dependence on each other as appeals to external audiences and
funding sources become channelled through the professional establishment.
Direct assess to the lay public and/or powerful non-scientific establishments
is seen as non-professional behaviour and sanctioned accordingly, as in the
Velikovsky case. Autonomy is thus both a condition and a result of increasing
mutual dependence among scientific workers.
Strategies for social and intellectual closure are not always successful
though. The social sciences show how difficult it is for some fields to establish
autonomy from lay concerns and approaches, as Elias and Katouzian discuss
in their papers for this volume. Attempts to professionalize around a common
body of techniques and methods allegedly derived from, or consonant with,
more established fields have been only partly successful. The relative success
of economics and psychology in distancing themselves from lay control and
problem formulations has been accompanied by attacks from philosophers
and other establishments as well as internal dissension during the expansion
of university posts. The extent to which the goals and dominant approaches
of these fields is indeed separate from lay goals and purposes is, in any case,
open to dispute (44). Furthermore, even when autonomy from non-scientific
groups has been achieved, to the extent that particular technical competencies
become necessary for access to jobs and reputational communities control
careers, autonomy from other scientific groups may still be limited. As
Fleck's study of the development of Artificial Intelligence in this volume
shows, existing scientific establishments were able to exert considerable
influence on funding agencies and the direction of much work in this area
in Britain.
The degree of autonomy over goals, techniques and jobs possessed by any
reputational community, and thus the extent of mutual dependence among
its members, is obviOusly dependent upon its ability to monopolize access
to posts and research facilities, as well as its control over prestigious journal
space. As I have already discussed, the academic model of reputational
organizations exhibited strong tendencies towards such monopoly, at least for
some fields, but the dominance of this model has declined with the growth
of non-academic research institutes. Where the goals of work organizations
do not easily fit in with those established in academic training programmes,
and other established groups have direct interests in the goals pursued by
335
employment organizations, then the degree of dependence on a single, particular reputational community will decline. Plurality of work organizations
and goals represented in them is likely to lead to a plurality of reputational
communities available to practitioners such that their dependence on anyone
group of colleagues is reduced and the boundaries of such organizations
become more fluid. Intellectual and social closure here become difficult to
enforce and scientific identities are often multiple and overlapping. While a
common body of technical approaches may produce a degree of "normative
integration" (45) among, say, biochemists, the goals to which these techniques may be put are so various in many of the bio-medical sciences that
task interdependence and functional integration among this group of practitioners is considerably reduced. The increasingly attested fluidity of research
boundaries, goals and identities among many biological scientists is an exemplication of this phenomenon, as is the "tinkering" described by Knorr and
others (46). The general development of "directed" research in industrial
nations is likely to accelerate these sorts of changes in the organization of the
sciences and promote the formation of new reputational communities and the
dissolution of old boundaries and knowledge structures.
336
Richard Whitley
337
338
Richard Whitley
339
the less need is there for intensive collaboration between scientists during the
data production phase and the more routinized can this become - as shown
by Shinn's paper here. The more skills can be related to a variety of goals, on
the other hand, and the more the particular way they are combined in likely
to produce novel and uncertain outcomes, the more important it is to monitor
the generation of information very carefully and
less can it be left to
neophytes and technicians. Different patterns of collaboration and control
thus arise at different stages of the research process in different fields.
6. The Organization of Scientific Fields and their Establishments
Differences between scientific fields, and changes in their organization, can
be summarized and analyzed in terms of these two dimensions of structure
- degree of mutual dependence among practitioners and task uncertainty.
Although they are conceptually independent, changes in one dimension
are quite likely to lead to changes in the other in modern professionalized
sciences and some examples of these changes will be discussed later. First,
though, I will briefly summarize some of the major characteristics of fields
with particular combinations of mutual dependence and task uncertainty as
sketched in Figure 1. In all these cases, technical task uncertainty is assumed
to be sufficiently reduced to enable some comparisons of research results to
be drawn and some sort of communication system established so that the
reputational system does function to a minimal extent.
1. In the first case, scientists do not depend greatly on each other for their
reputations which in turn, are not critical for obtaining jobs and facilities.
Standardization of technical procedures is not high, and the meanings of
research outcomes relatively unclear and only tenuously linked to theoretical
goals. General and diffuse interpretative schemes are produced which rely
extensively on tacit skills for their understanding. These skills are likely
to be produced in a limited number of research sites by highly personal
means and will be associated with particular research goals and conceptual
approaches. The formation and reproduction of research schools which
effectively ''talk past" each other is a notable feature of these scientific
fields. Intellectual conflicts are wide ranging and frequently involve personal
disputes. Because the need to rely on all practitioners "in" the field is low,
scientists can focus on local reputations and compete for jobs and facilities
340
Richard Whitley
LEVEL OF
MUTUAL
DEPENDENCE
Low
High
Low
4. Smoothly functioning
formal bureaucracy,
e.g., contemporary
Chemistry
1. I nformal crafts
organization with
unspecialized
intellectuals producing
diffuse interpretative
schemes, e.g., traditional
humanities, European
sociology
High
2. Collegial-professional
complex pureaucracy
with demarcation
between central areas
and peripheral ones,
e.g., post 1945
physics
Fig. 1. Levels of task uncertainty and mutual dependence in the sciences and their
organizational correlates
within one school of thought and practice rather than developing more
general and widespread skills and procedures which would enable more object
centered and formal assessments of contributions to be made, and reputations
allocated across schools and employment organizations. Here, facilities are
relatively widely available and simple to use so that extensive technical
staffs are not required. Resources can largely be supplied on a local basis
without substantial recourse to external funding agencies. Also, the degree of
autonomy from lay audiences and their goals and concepts is likely to be
relatively low in these fields and many practitioners will seek direct access to
these audiences. Some examples of such areas are the traditional humanities
and many of the social sciences where the professional community as a whole
does not dominate the allocation of jobs.
Establishments in these sciences will be relatively decentralized and control
knowledge production and validation through largely local and personal
means. Because the reputational organization as a whole has only limited
control over careers, its establishment's ability to set goals and systematically
341
342
Richard Whitley
which share a common core of technical expertise and "craft" skills but
manifest a plurality of goals. Depending on the extent and diversity of
funding sources, and the cost of facilities, these goals will be more or less
closely related and mutually ordered. In mathematics, for example, where
most work is carried out in university departments which also train new
generations, a greater coherence of goals would be expected because employment is concentrated in a single institution and hence dependence on
colleagues is quite high with a strong degree of "normative integration"
provided by a common curriculum. As Fleck shows, the AI community is
more diverse in its employment sites and funding sources. Consequently,
dependence on a single relatively coherent, community is less - as is its
autonomy from lay concerns and other scientific establishments. Fragmentation - both social and cognitive - is therefore higher.
Establishments in these fields validate contributions in terms of the
standardized technical procedures which constitute the core identity of the
field. While competing groups conflict over the importance of particular
theoretical goals, and the overall significance of task outcomes for the field,
they all concur on the common body of skills and expertise which distinguish
practitioners and provide a means for the allocation of reputations. Insofar as
research facilities and/or funding sources are concentrated, one establishment
may dominate and impose its goals on other groups through its control of
resources. Where, however, facilities are widely available and journal space
not controlled by a single elite which also control jobs, groups adhering to
different goals will compete for domination of the field without any single
one becoming established as central. Different subgroups in different employment organizations are able to set their own goals and objectives within the
broad constraints of the disciplinary identity. Also, although the standardization of techniques and development of an esoteric and formal communication
system enable control of work to be exercised at a distance by rules, personal
contact and patronage remain important means of inculcating skills and
controlling task performance as Fleck's discussion of AI shows. Control is
thus less immediate and personal than in the first case, but personal authority
is still a crucial element in setting priorities and realizing objectives.
3. In the third case, theoretical task uncertainty is lower but the degree of
mutual dependence is still high, often because of the limited availability of
343
344
Richard Whitley
and procedures rather than personal authority and direct example. Assessment of task performance and the significance of task outcomes is a relatively
impersonal and formalized activity carried out by scientists dispersed over a
wide geographical area with few, if any, direct personal ties to the originators
of knowledge claims. Although tacit skills and judgements are still important
here, as they are in all the sciences, and personal patronage and contacts are
equally crucial to scientific careers, as they are in business, control processes
in these highly centralized and standardized sciences are less directly personal
than in those previously considered. Whether an experiment has been currently carried out, and the theoretical meaning and significance of the results,
are more easily and readily decided by relatively impersonal and formal
procedures in these sciences than in others (52).
4. Finally, in fields where task uncertainty is low, and the degree of mutual
dependence among practitioners is also fairly limited, planning of work with
a high degree of division of labour is possible and overt conflict over goals
unlikely. If funding is fairly easy to obtain and is available from a number of
different sources - e.g., industry and government as well as from national
research agencies - individuals' dependence on their professional colleagues
will be less than when it is impossible to carry ouf research without obtaining
funds from a single source. Central reputations in the science as a whole,
will therefore be less important in these fields than where resources are
monopolized by a single agency. Coordination of results into a coherent
theoretical scheme, which orders sub-fields into a dominant hierarchy, is less
important here, as in chemistry perhaps, than in physics. Goals, concepts and
techniques are fairly well integrated in these sorts of fields but their ordering
into a strong hierarchy is not of great concern. Theoretical work is, therefore,
not as crucial in these sciences and the involvement of theoretical issues and
goals in the day to day conduct of research not as high as it seems to be
in physics (53). Patterns of collaboration and communication within and
between research organizations tend to resemble other formal structures of
work allocation and control with strong distinctions drawn between levels of
expertise and competence and fairly sharp differences noticeable between
skills. Shinn's
of work organization in solid state physics in contrast
with that in mineral chemistry highlights these points.
Establishments in these fields control both work goals and procedures to
345
a high extent through largely impersonal and formal means, but they are not
as centralized and hierarchically structured as in the previous case. Instead,
segmentation of sub-fields with their own establishments and separate goals
seems to occur with little overt ordering of these into a single scale of prestige
and power dominated by the establishment. Within an elaborate and formalized body of knowledge and skills which defmes and orders tasks, and
evaluates the significance of results, in a relatively uncontested manner,
different groups develop research programmes in a variety of directions which
are integrated through the commonly accepted rules.
7. Changes in the Organization of Scientific Fields and the Emergence of the
Scientific Establishment
Changes in the degree of mutual dependence among practitioners in a scientific field have consequences for the degree of task uncertainty in that area
and for its intellectual and social organization. Equally, the development
of standard techniques and/or theoretical consensus and coherence change
relations between scientists and the organization of the reputational system
and its relations to employment organizations. Both of these sorts of changes
are related to the intellectual and social autonomy of the sciences and their
funding structures.
The degree of mutual dependence among scientists is related to: (a) their
dependence upon external audiences and sponsors as opposed to esoteric
and professional audiences, (b) the number of
whose work is
oriented to the goals of a particular reputational community, (c) the scarcity
of resources for accomplishing relevant tasks and, (d) the scope of the problems which can be adequately dealt with by single individuals. These factors
are interrelated as we have seen and an increase in one is likely to lead to
changes in others and in the overall structure of the field. If, for instance,
the number of practitioners seeking reputations in a particular field, which
already has a fairly formalized communication system, increases, the scope
of problems tackled is likely to decrease as specialization of tasks and skill
becomes higher. Also, technical precision and standardization are likely to be
emphasized, thus reducing technical task uncertainty, as means of ensuring
comparability of results and coordination of task outcomes. Relative scarcity
of resources - which is likely to be reinforced by the move to greater precision
346
Richard Whitley
and complex techniques - will force scientists to orient their work more
towards each other and claim reputations on the basis of contributing to
community goals. Common and standard technical procedures facilitate such
claims and the award of mutual recognition. Provided the environment
is fairly benign, then, overall autonomy is likely to increase and the field
become more professionalized. To some extent, this seems to have happened
in some of the humanities since the 1950s; work has become more technical
and specialized, remote from lay concerns and language and specific to
scholarly goals. Diffuse interpretative schemes seem less manifest, at least in
Anglo-Saxon countries.
The impact of increased numbers of practitioners seeking reputations from
a single community varies in effect according to the extent of technical and
theoretical autonomy of the field, the relative scarcity of resources, especially
jobs, and the existence of a technical, standardized communication system.
This variability can be seen by a brief comparison of Anglo-Saxon sociology
and economics. The expansion 0 f university posts in both of these disciplines
in the 1960s and early 1970s has resulted in a fragmentation of the former
field and a stronger emphasis on mathematization and technical competence
in the latter.
The lower technical autonomy of sociology, its greater reliance on tacit
skills and personal relationships and the lack of a strongly institutionalized
symbol system for communicating and comparing results meant that the
general availability of jobs and other research resources decreased mutual
dependence and encouraged goal conflict and differentiation. Because posts
were much more widely obtainable, the need to acquire a reputation across
the whole field through conducting research according to the central, established tradition decreased, and sub-groups based on divergent goals,
methological beliefs and distinct technical procedures were able to control
access to jobs and so reputational organizations were established around
sub-goals. Personal contacts and allegiances among the establishment were
insufficient to control the expanded organization and the development of a
more formal and standardized style of research which could be inculcated
throughout the training system as the basis of "normative integration" was
inhibited by the expansion of jobs. Central authority declined, specialization
and differentiation occurred through general methodological orientations and
conceptual frameworks - often linked to ideological views and political
347
348
Richard Whitley
349
350
Richard Whitley
longer as crucial as they were. Through their control of school curricula and
textbooks, academics still play a major part in the public image of science,
and still decide how resources within universities - especially jobs - are
allocated to different fields, but science as a major social institution producing
and validating the "means of orientation" in Elias' phrase, seems to have
transcended the university. Scientific establishments controlling reputations,
setting knowledge goals and controlling task allocation and performance are
no longer identical to academic establishments and academic science not
identical to the whole of the "best" science. Universities still produce the
knowledge producers but employing organizations also train them in specialist
skills, often for novel purposes, and coordinate their work for specific tasks
which often have no counterparts in academic science. Insofar as work goals
in these organizations are partly controlled by reputational organizations,
new scientific fields become established around non-academic goals, which
are regarded as being equally "scientific" as to those which correspond with
academic boundaries. Indeed, some of these fields, such as cancer research,
later become established in universities.
The development of new reputational organizations controlling resources
outside the universities, and on fields which cut across established academic
boundaries and priorities, is part of the general extension of science to new
areas and concerns and its redefinition as a generalized set of procedures for
producing knowledge rather than a fIXed body of knowledge about a given
reality. As science attempts to monopolize knowledge production and validation and impose its criteria of rationality and understanding upon societies,
scientific ideas increaSingly come to dominate the production of all forms of
knowledge, especially once major social establishments such as the military
and business have become convinced of the utility of the dominant form of
scientific knowledge. Forms of understanding which are antithetical to these
ideals are not scientific and so do not lead to knowledge. Consequently
nascent groups of knowledge producers attempt to prove their scientific
status, and thus gain access to jobs and research facilities, by imitating the
current image of science and organizing their work in a "scientific" manner
with the apparatus of profeSSionalized science. As science grows in influence
and prestige, so too do the ideals and methods of science become important
to all knowledge producers and the reputation of being scientific a key resource in the struggle for legitimacy. The mutual interdependence of scientific
351
352
Richaxd Whitley
8. Concluding Remarks
In suggesting that the sciences should be considered as one form of work
organization - reputational organizations - which are a subset of professional
organizations, I have tried to show how intellectual change and differences
between scientific fields can be understood in terms of variations in their
organizational structure. The professionalization of the sciences in the 19th
century tied jobs and statuses to intellectual reputations for the first time in
a systematic way. This meant that the organization of work, ofreputational
groups and of employing institutions became inextricably entwined so that
changes in the one involved changes in the others. Subsequent developments
in the employment of scientists and generalization of science to novel areas
both decreased the institutional specificity of science in universities and
revised the status of science as a fixed body of truths. As the "scientific
method", science became an abstract set of norms for producing knowledge
oriented towards a variety of goals and a variety of employing organizations
arose to control and direct such knowledge production. This variety of relations between employers, scientists and universities means that the academic
model of scientific work and "col'nmunities" is inadequate for understanding
the professionalized sciences. Viewing them as reputational organizations
which vary in their structure and relations to other organizations seems to
me to be a more useful approach than assuming either their autonomy as
self-governing groups producing certified knowledge or their inevitable
reduction to market interests.
Notes and References
1. See, for example, G. Lemaine et al. (eds.), Perspectives on the Emergence ofscientific disciplines, Paris and the Hague: Mouton, 1976; D. Edge and M. Muikay,
'Fallstudien zu wissenschaftlichen Spezialgebieten'in N. Stehr and R. Konig (eds.),
Wissenschaftssoziologie, Kohn: Westdeutscher, 1975.
2. G. Bohme, W. van den Daele and W. Krohn, 'Finalization in Science', Social Science
Information IS (1976) 301-30; R. Hohlfeld, 'Theory Development in Moleculax
Biology', in W. Callebaut et al. (eds.), Theory of Knowledge and Science Policy,
Ghen t: Communication and Cognition, 1979.
3. Especially his Structure of Scientific Revolutions, University of Chicago Press, 2nd
ed., 1970; 'The Function of Measurement in Modem Physical Science' in the
collected essays The Essential Tradition, University of Chicago Press, 1977 and
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
353
354
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
Richard Whitley
19th Century', W. Krohn et al. (eds.), The Dynamics of Science and Technology,
Sociology of the Sciences Yearbook Vol. 2, Dordrecht: Reidel, 1978.
J. B. Morrell, 'The Chemist Breeders: The research schools of Liebig and Thomas
Thomson', Ambix 19 (1972) 1-46.
Partly because of the need to monopolize the production of producers and to
justify that monopoly through institutionalizing knowledge and skill barriers, and
also to claim separate, esoteric expertise which can only be obtained through the
professional certification agency.
This was especially the case in 19th-century Germany of course; see the books by
McClelland and Ringer already cited in Notes 7 and 13 respectively and the two
papers by R. S. Turner, 'The Growth of Professorial Research in Prussia, 18181846, Causes and Context', Historical Studies in the Physical Sciences 3 (1971);
'University Reformers and Professorial Scholarship in Germany', in L. Stone (ed.),
The University in Society, Princeton University Press, 1974.
McClelland clearly links the particular form the institutionalization of scholarship
took in Germany, and its subsequent high social prestige, to 'an attempt to stabilize
and legitimate the rule of the more flexible part of the aristocracy with the aid of a
small elite recruited from the middle class', op. cit., Note 7, p. 98.
McClelland,op. cit., Note 7; Morrell, op. cit., Note 15.
As suggested by Terry Shinn, op. cit., Note 6.
See, among others, M. Berman, , "Hegemony" and the Amateur Tradition in British
Science', Journal of Social History 8 (1975) 30-50; R. Porter, 'Gentlemen and
Geology: the Emergence of a Scientific Career, 1660-1920', The Historical Journal
21 (1978) 809-36.
Morrell,op. cit., Note 15.
The academic professionalization of the field sciences developed later than in
physics and chemistry and, in general, occurred later in England; see, for example,
J. G. O'Connor and A. J. Meadows, 'Specialization and Professionalization in
British Geology', Social Studies of Science 6 (1976) 77-89; Porter, op. cit., Note
21.
Morrell,op. cit., Note 15.
C. Jungnickel, 'Teaching and Research in the Physical Sciences and Mathematics in
Saxony, 1820-1850', Historical Studies in the Physical Sciences 10 (1979).
Ibid., pp. 41-2.
See the papers cited in Note 21 and D. Edge, 'Some Sociological Reflections' in
M. Hoskin (ed.), General History ofAstronomy Vol. 3, Cambridge University Press
(forthcoming). S. F. Cannon has a detailed discussion of preprofessional science in
England in her Science in Culture, New York: Neale Watson Publications, 1978,
.
Chs. 5, 8 and 9.
Cf. Jungnickel, op. cit., Note 25; Berman, op. cit., Note 21 and his Social Change
and Scientific Organization, London: Heinemann, 1978; Porter, op. cit., Note 21.
As discussed by Berman in his historical account of the Royal Institution in Social
Change and Scientific Organization, op. cit., Note 28; see also R. H. Wiebe, The
Search for Order 1877-1920, New York: Hill and Wang, 1968, especially Chapter
6.
30. The standardization and simplification of experimental techniques by Liebig and
others was a key condition of the "finalization" of science. As W. van den Daele
31.
32.
33.
34.
35.
36.
37.
38.
355
356
Richard Whitley
53.
54.
55.
56.
57.
58.
357
Pickering and Harvey in the Sociology of the Sciences Yearbook. Vol. 4. Dordrecht:
Reidel. 1980.
As reported by Shinn in this volume. Others comment on the growing separation of
theoreticians and experimentalists in physics. which may lead to theoretical physics
becoming organized like mathematics as a distinct reputational organization: see
Grabner and Reitner. op. cit . and Levy-Leblond. op. cit . Note 41.
Ethnomethodology is the most obvious example of this. cf. N. Mullins. 'The Development of Specialties in Social Science: the Case of Ethnomethodology. Science
Studies 3 (1973) 245-73. See also K. Knorr. 'The Nature of Scientific Consensus
and the case of the Social Sciences'. in Knorr et al. (eds.). Determinants and
Controls of Scientific Development. Dordrecht and Boston: Reidel. 1975. In his
Theories and Theory Groups in Contemporary American Sociology. New York:
Harper and Row. 1973. Nicholas Mullins suggests that causal modelling will become
the new integrating force in North American Sociology which fits in with my view
that increased numbers can increase mutual dependency and lead to technical
standardization. He also claims that (p. 305) the new theories are method centred
rather than being differentiated by topics. If so. this suggests that largely common
sense foci are being replaced by more esoteric. "professional". ways of differentiating tasks. While I think this is partly what has happened. the relative lack of
autonomy from common sense concerns and issues in sociology mitigates such
professionalization tendencies as Mullins shows in his discussion of the "radicalcritical". "social forecaster" and other new theory groups.
See. C. J. Lammers. 'Mono- and Poly-paradigmatic Developments in Natural and
Social Sciences'. in R. Whitley (ed.). Social Processes of Scientific Development,
London: Routledge and Kegan Paul, 1974; Martins, op. cit., Note 4.
Katouzian,op. cit . Note 44.
As in Geology and Physics, see Porter, op. cit., Note 21. O'Connor and Meadows,
op. cit . Note 23; Moseley, op. cit., Note 31.
This is scarcely a recent phenomenon, of course, but the accommodations made by
natural philosophy in the 17th century .were not those of professionalized science.
The development of organized, directed knowledge production linked to reputational organizations, and the monopolization of knowledge production and validation by professional scientists, have resulted in the entrenchment of particular
scientific ideals - which are supported by other establishments - throughout
the research system. The more dependent scientific fields are upon each other for
legitimacy and reputations. the more the image of science which is congenial to
dominant groups will dominate all knowledge production as access to jobs and
facilities is denied to "amateurs", "pseudo-scientists" and "ideologists".
INDEX
Abir-Am, P. 141
academic establishments 62
institutions, dogmatism 104
repressive nature 103
success, disillusion 102
Academy of Sciences of U.S.A. ix
advice from 111-119
advisory role 114
dietary advice, controversy 116
expansion of 112
origins of 111
supreme court of science 11"4
Academy of Sciences of U.S.S.R.
development of 113
links with state 45
Academy of Sciences, Prussian 112
West German, development 113
Agassiz, L. 118
alchemy 224
Aleksander, I. 211
Allen, E. 291,292
Allen, G. 142,165,217
al-Mulk, Nizam 94
Alt, F. L. 211
Altholz, J. L. 290,291
Ambler, P. A. 215
Anderson, B. 196,216
Appleton, E. 204
'AqnY 107
Aquinas, St. Thomas 93
Ardrey, R. 285,291
Armer, P. 212
Armistead, N. 356
artificial intelligence ix, 169-217
attacks on 205
"big four" in 178
conflict in 172
department of, breakdown of 189
first 186
development
conferences and 186
359
360
Beadle, J. 133
Bearden, C. 196
Becquerel, H. de 226
Bell, D. 85, 86
Ben-David, J. 118,264
Bentham, J. 106
Berg, P. 123
Berlin, I. 107,108
Berman, M. 289, 354
Bernal, J. D. 137, 142, 143, 219, 234
Bhaskar, R. 220, 234
biochemistry, service-like 132
biology, differing notions 129
programme management 134
research strategy 131
biomedical research 226
growth, from political in-fighting
135
strategic planning of 154
biotechnology as trend in molecular
biology (q.v.) 124
Bitz, A. 263
Black, M. 264
Blume, S. S. 86, 234, 307
Boas, M. 236
Boden, M. A. 197,198,204,216
Boffey,P. 117,118
Bohme, G. 165, 220, 227, 228, 234,
235,236,307,309,352
Bold, A. M. 237
Bornat, R. 195
Bourdieu, P. 308
Bowen, F. A. 290
Box, S. 166
Brady, J. M. 195,196,212,217
Breuer, H. 166
Brewster, D. 276, 290
Brickman, R. 118
Briggs, A. 197,216
Brody, J. E. 119
Broesterhuizen, E. 237
Brooker, R. A. 193,195,200,213,215,
216
Brooks, H. 73, 117
Brown, E. R. 136,141,143
Brown, S. C. 118
Browne, E. G. 95, 107
Bulthaup, P. 235
Index
Bundy,A.216,217
bureaucratization 4
Burns, T. 263
Bush, V. 134
Burstall, R. M. 190,195,197,213,216
Cairns,1. 142
Caldin, E. F. 263
Callebaut, W. 166,238,352
cancer, occupational 159
research ix
in Germany 145-168
basic vs. targeted 150
clinical 155
epidemiology and 156, 158
experimental 151
medical opinion and
156
molecular biology and 146
patient-oriented 154
political debate 145
political planning 149
power structure 160
Cannan, E. 108
Cannon, S. F. 354
Cannon, W. B. 127
Carlson, E. A. 131,142
Carnegie, A. 125
cause, Kant on 9
cell biology, cancer research 151
Chambers, R. 273, 277, 278, 280, 290
. Chargaff, E. 143
chemists, social status 233
Chubin, D. E. x, 164, 211, 293-311
civilization, acceleration of 11
time perception and 18
Clark, J. W. 290
Clowes, M. B. 185,197,199,204,214,
216
cognitive organization, limitations 253,
257
cognitive psychology 198
cognitive studies 197
Cohen, L. J. 308
Cole, J. 310
Cole, S. 310
Collins, H. M. 222,235,237
Collins, N. L. 213
Collins, R. 264,355
361
Index
Dreyfus, H. 1. 212,217
Drosophila 130
Dubos, R. 142
Dupree, A. H. 118
Dijkstra, E. W. 217
Edge, D. O. 171,211,295,306,352,
354
Edison, T. A. 47
Edsall, D. 127
education, computers in 194
effects, restricted 227
Einstein, A. 14, 32, 34
Elcock, E. W. 185, 193, 204, 214
Elias, N. D. viii, xi, 3-69, 85, 86, 111,
114,117,118,124,141,168,
174,175,181,207,211,212,
235,261,263,268,269,289,
307,311,353
elites, admission to 272
Emerson 103
Engel, G. 1. 167
environmental protection 74
epidemiology 156
of cancer 158
Ernst, G. W. 212
establishments, political and scientific
45-48
scientific 3-69
development of 48-52
historical 3-4
orientation control by 37-44
theory of 25-37
evolutionary theory 33, 297
debate, 19th century 282
Vestiges (q.v.) and 278
evolution (pre-Darwin) 272
experimental operations, social relations
and 253,257,259
expertise, institutional basis 74
policy making and 78-80
Faraday, M. 103,336
Feather, N. 214
Feigenbaum, E. A. 206,211
Feldman, J. 211
Ferreira, P. 237
Fisher, C. S. 356
Index
362
Fleck, J. ix, 169-217, 239, 261, 310,
334,342
Flexner, S. 126,127
Flowers, B. 189
fork bending, scientific belief challenged
by 97
Fosdick, R. B. 126,141
Foster, J. M. 185,193,195,214
Fox, R. 285, 292
funding
influence on intellectual development
x
Gregory, R. L. 186,187,188,197,214,
215
Griffith, B. C. 217
Groeneveld, L. 311
Groenewegen, P. 237
Guerlac, H. 236
Gummett, P. 143
Hagstrom, W. O. 293, 307, 353, 356
Hall, G. S. 291
Hall, T. S. 141
Halsbury, Lord 182,213
Hanson, N. R. 264
Haraway, D. J. 142
Hardy, S. 196
Hargens, L. L. 356
Harre 264
Hartley, B. 166
Harwood, J. 141
Hausen, H. zur 166
Hay,C. ix,I11-U9
Hayes,P.J.190,195,214,215
Heirich, M. 142
Heisenberg, W. 64,65,69
heliocentric world-view 11
Henry, J. 112,118
heredity, genetics and 130
Herschel, W. 274
Hesse, M. 264
hierarchies
in scientific establishments vii
research trails and 300
Hill, A. V. 134
Hill, K. 118
Hodgkin, L. 141
Hohlfeld, R. ix, 145-168, 309, 352
Holden, C. 310
Hollis, M. 356
Homburg, E. 238
Hoskin, M. 354
Howe, J. A. M. 187,194,210,214
Hubble, C. 14
Hufbauer, K. 236
Hughes, T. M. 290
Hull, D. 165
Hume Hall, R. 237
Hutchins, E. 86
Index
Hussed, E. 10
Huxley, A. 123,141
Huxley, J. 289
Huxley, L. 290
Huxley, T. H. 270, 271, 279, 281, 283
hybrid community 78,290,291
ibn Sina, Abu Ali (Avicenna, q.v.) 95
ideas, innateness of, Kant's opinion 8ff
inquiry, morphology of 240
institutionalism, hallmark of scholasticism 102
instruments, analytic 228
synthetic 228
intellectual barriers, construction of
270
public/scientist 267-292
intellectual elitism 269
Jackson, D. 143
Jacob, F. 129,137,142,143,297,308
Jannich, P. 220,234,236
Janicke, M. 168
Johnson-Laird, P. M. 198,216
Johnston, R. 143,308
Jolliffe, C. 183
Judson, H. F. 143
Jungnickel, C. 320, 354
Kaiser-Wilhelm-Gesellschaft 125
Kant,I. 7,8,13,17,67,68,289
Kargon, R. 355
Katouzian, H. ix, 89-109, 356
Katz,J.87
Kendrew, J. 139
Kerr, S. 307
Kevles, D. J. 118,141,353
Kilminster, R. 67
Kitschelt, H. 86
Klerman, G. L. 167
Knight, D. 238
Knorr, K. D. 211,234,296,297,307,
308,310
knowledge, scientific
acquisition of 15
institutionalization of viii
theories of 14, 62-65
363
Kohler, R. E. 126,127,132,141,142,
165
Koller, N. 311
Konig, R. 352
Kowalski, R. A. 190,214
Kraft, P. 217
Krohn, R. C. 86
Krohn, W. 86,164,165,234,309,352
Kuhn, T. S. 36,50,100,101,102,108,
165,224,231,232,233,235,
237,238,294,307,310,313,
316
Kuklick, H. 289
Kunze, P. 67
Kiippers, G. 86,87
laboratory
cognitive organization in 254, 258,
260
cognitive pattern 242, 243, 246, 250
data and materials flows in 254
hierarchy 240, 244, 248
communication ladder in 241,
245,249
lower echelons, importance 242,
246,250
liaison in 254
morphology 239,244,247
organization 240,244,247
Lakatos, I. 100
Lakoff, S. A. 85
Lamarck, J. B. 273,278
Lane, R. E. 72,85,86
Laplace 13,106
Lapp, R. E. 85
Larson, M. S. 353
Latour, B. 227, 236, 264, 308, 353
Laudan, L. 310
Lauman, E. O. 311
Lavoisier, A. 225
Law, J. 211
Layton, E. T., Jr. 234
Leibniz 8, 17, 28
Lemaine, G. 307,309,352,355
Lepenies, W. 238
Levidow, L. 141
Leydesdorff, L. 238
364
Libich, S. 164
Liebig, J. W. 318,320,321
life
as information process 137
differing notions of 129
genetic reductionist concept of 132
synthesis of - front-page news 139
Lighthill, J. 190, 193, 200, 205, 215,
217
Lillie, F. R. 127
Linnaeus 235
Locke, J. 103
Loeb, J. 129,130
Long, T. D. 118
Longuet-Higgins, H. C. 186, 187, 188,
197,201,204,206,210
Lorenz, C. 62
Lundgreen, P. 86
Luria, E. E. 166
Luria, S. 133
Li.ith, P. 168
Lwoff, A. 142
Lyell, C. 273,275,290
Lysenko 269,289
MacDonald, D. K. C. 263
MacKay, D. 182
MacLeod, R. 309
Macpherson Stalker, G. 263
Madow, W. G. 309
Malvern, R. R. E. 214
Manegold, K. H. 353
Markle, G. E. 309
Martin, B. R. 211
Martins, H. 107,353
Marx, K. 104
mass communication, molecular biology
and 138
Mates, B. 264
mathematization 65
McAlpine, A. 263
McCarthy, J. 176,177,178,179,182,
190, 206, 211, 212, 214, 215
McClelland, C. E. 353
McCorduck, P. 212
McCulloch, W. S. 179, 182,211
Meadows, A. J. 354
Index
medical opinion of cancer research 155
medicine and cancer research in Germany
145-168
Medvedev, Z. A. 118
Meltzer, B. 184, 185, 189, 194, 199,
200,204,206,213,214
Mendel, G. 130
Mendelsohn, E. 118,164,234,262,289
Merton, R. K. 212,289,294,307,310
Meselson, M. S. 118
metaphysics
ascendancy of 10
beliefs of 13
philosophers and 6-17
Mey, M. de 166,238
Michaelson, S. 214
Michie, D. 182, 183, 185, 187, 188,
193,199,200,204,206,210,
213,214,215,216,217
microbiology 133
Mikulich, L. 1. 215
Mill, J. S. 103,274
Millhauser, M. 290
Minsky, M. L. 177,178, 179, 182, 194,
212
Mitroff, I. I. 310
Mittelstrass, J. 220,234
Mittermeir, R. 310
molecular biology ix
as goal-oriented research 136
cancer research and, in Germany
146
establishment of 123-143
as discipline 124
fundamental nature of 137
historical development of 125, 147
mass communication and 138
new trends in 123
research frontiers of 148
T.V. shows on 139
molecular structure of biological molecules 133
Montesquieu 106
Moore, W. E. 86
Morgan, E. 132,284,285,291
Morgan, T. 130,131
Morrell, J. B. 320, 354
365
Index
Morris, D. 285, 292
Moseley, R. 355
Muller, H. 1. 131,142
Mueller, I. 264
Mulkay, M. 1. 171,211,217,295,306,
307,308,309,311,352
Muller 132
Mullins, N. C. 142, 217, 311, 357
Multhauf, R. P. 235, 236
Murray, A. M. 214
mysticism, Islamic 94.
Nagel, E. 263
National Science Foundation of U.S.A.
135
Needham, 1. 69
Nelking, D. 87
Nell, E. 356
Newell, A. 177, 178, 179, 184,204,
212
Newton, I. 17, 32, 33
Nossiter, T. 353
Nowotny, H. 87,107,289
O'Connor, 1. G. 354
Oettgen, H. F. 166
Olby, R. C. 142,166,307
Oleson, A. 118
opportunism in research 296
orientation, control of scientific 37-44
Origin of Species, Vestiges (q.v.) and
279
Overington, M. A. 309
Paneth, F. A. 253,263
Pantin, C. F. A. 220, 221, 234, 235,
263
Papert, S. 194,212
Pasteur, L. 103
Pauling,L. 132,133
Petersen, C.. 309
Perutz, M. 133
pharmaceutical industry, biomedical research 154
philanthropists, American, research strategyand 125
philosophers, metaphysics and established 6-17
philosophy
relation with physics and sociology
56-67
time as problem for 17-25
transcendental, ascendancy of 1017
errors of 8ff
mind-set of 12
physician, mind-set of 157
physicists, social status 233
physics
mathematization of 222
relation with philosophy and sociology 56-67
physiology 132
Pickstone,1. 141
Pickvance, S. 136,143
Pinxten, R. 166, 238
Pitts, W. 179,211
Plotkin, G. D. 190,214
Polanyi, M. 165,290
political ambition 79
politicization of science 73
consequences of 80-84
Popper, K. 10,68
Popplestone, R. 1. 213
Porter, G. 215
Price, D. K. 72,85,119,294
professionalization ix
professors 4,5
as ruling group vii
progressive ideology, American 125126
purity
as restricting concept 223, 225
of effects 227
Pursell, C. 142
Putnam, H. 263
Pyenson, L. 233, 238
radioactivity research, early 226
Rae,1. 108
Rappaport, A. 264
Ravetz, 1. R. 86,263,264,310
Reason, triumph of, as dogma
Reingold, N. 118,141
Reiter, W. 356
reproducibility 224
105
Index
366
reputational organization, science as
313
reputational systems, control of resources ix
research
costs 297
councils, British 125
director, computer laboratory 244
mineral chemistry laboratory 240
physics laboratory 248
elites, control of resources 136
goals, choice of 294
organization of x
payoffs 297
policy 306
investment model 300
practices 239
dormant 303
trail, funding 301
hierarchies and 300
legitimacy of object 301
problem area 295
resource access 302
training capacity 302
yield of 296
restrictedness
importation of 230
in science 219-238
institutionalization in 229
production of 223
Rettig, R. A. 143, 164
Ricardo 103
Richards, A. N. 134
Ringer, F. K. 354
Rip, A. x, 118,168,219-238,239,261,
307,336,355
Robinson, J. A. 194,216
robot research, in U.K. 190
Rockefeller Foundation 126
influence on research ix
Rockefeller, John D. 125
Roll-Hansen, N. 131,142
Romer, A. 236
Rosenberg, C. E. 142
Ross, J. 310
Royal Society 112
research strategy and 134
Rubin, L. 310
Rumi, Mawlavi 95
Ruse,M.292
Russell, B. 107
Russell, C. A. 355
Rutherford, E. 226
Sadler, J. 167
Saint-Hilaire, G. 29
Salomon, J. J. 115,118,311
Salter, S. H. 187
Samuel, A. L. 177
Sandewall, E. 211
Sanger, F. 123
Schiifer, W. 311, 355
Schaffner, K. 165
Schelling, T. C. 294,307
Schelsky,H. 86
Scherer, E. 167
Schmidt, C. G. 164,166
Schoenheimer 133
scholasticism
hallmarks of 102
Islamic 93
rise of 94
science compared to ix, 89-109
Schon, D. A. 307
Schrodinger, E. 132
science
academic domination 323
variations 325
ambiguity reduction in 222
application of, negative effects 230
as anti-dogmatism 98
as knowledge construction industry
267
autonomy of pure 267-292
basic, and cancer research in Germany 145-168
Council of Japan 113
de-institutionalization of 84
employment structures in 314
heterogeneity of 231
Islamic 93
as product of religious doctrine
94
comparison with Renaissance 95
Index
science (continued)
organization 339
policy planning, in Germany 150
politicization of 117
consequences of 80-84
productive labour 220
professionalized 328
reputational organization 313-357
research trails and policy 293-311
restrictedness in 219-238
scholasticism compared with 89-109
self-centred view of 89
subjectivity of 99
technical aspects of 219
university, establishment of 317
scientific
achievement, dependence on original
thinkers 100
community, 'divergence of opinions
in 83
establishments
balance of power in 4-5
bureaucratiZation of 4
emergence of 345
organization 339
popular culture and 267-292
shift of power base within 73
judgement, non-neutrality of 82
knowledge, as power source 71
structuring of, by Universities
38-39
transmission of 40-43
practice, classic, entrepreneurial and
pragmatic 124
pronouncements, public suspicion of
268
work, legitimate or otherwise 64
scientification of politics 80-81
scientists as class 71
task interdependence 329
task uncertainty 335
Sedgwick, A. 275,276,290
Seidel, R. S. 141
Selfridge, O. G. 211
Shannon, C. E. 177, 179, 182, 205,
211,217
Shaw, J. C. 177
Sheldon, E. J. 86
367
Shinn, T. x, 201, 217, 220,229,234,
237,239-264,337,353
Shirazi, Sa'di 95
Siegenthalter, W. 167
Siekmann, H. J. 196
Simon, H. A. 177,178,179,184,204,
212
Skopp,D.233,238
Sloman, A. 198, 204, 216, 264
Smith, A. K. 87, 92, 104, 107, 108
social relations and experimental operations 253,256
sociobiology 284
critical reaction to 286
Darwinism and xi
sociology, political blocs and 55-56
relation with philosophy and physics
53-67
scientific nature of 53, 65
status of 54
Solla Price, D. de 219, 234, 307
specialization, confluence of research
trails 295
political pressure and 304
Spencer, H. 280, 282, 283
Stahl 225, 233
Stamberg 313
status in labour market 315
Staw, B. M. 310
Stehr, N. 352
Stent, G. S. 142
Stephenson, R. L. 103
Stewart, C. H. 214
Stich, S. 143
Stinchcombe, A. 315, 353
Strasser, H. 211, 234
Strickland, S. P. 135,142,164
St. Simon, A. 106
Studer, K. E. 164, 307,309
Sturtevant, A. H. 165
Suppe, F. 263
Suppes, P. 263
Sutherland, N. S. 185, 196, 199, 200,
204,210,216,217
Svedberg 133
Swann, Lord 139, 186, 187, 193, 204,
214,215
Swatez, G. M. 263, 356
368
Index
Szabadvary, F. 235,236
Szasz, T. S. 167
Szent-Gyorgii, A. 310
Tarski, W. 263
Taube, M. 212,217
Taylor, G. R. 143
Telford, T. 103
theories, law-like and process-like 8
formation 232
thinking, spectrum of forms of 60-62
Thomas, C. L. 166
Tiger, L. 285,292
time, perception of, civilization and 18
philosophy and 17-25
'tinkering' in research 298
Toulmin, S. E. 263
1,'racy, H. de 106
Trenn, T. J. 236
Turing, A. M. 182,204,206,213,217
Turner, R. S. 354
Turner, S. J. 309
Ullmann, A. 142
universality of law-like generalizations
7-8
universities, scientific knowledge and
38-39
Useem, M. 308
Uttley, A. M. 182,197,213,216
Vandamme, F. 166, 238
273
criticism of, by learned public 273
expository technique 277
Origin of Species and 279
popular science 276
rejection of, by scientists 274
Vieo, A. M. 106
Viner, J. 108
'vital process' research 127
Vogel, M. J. 142
Voltaire 95,103, 105, 106, 109
Waddington, C. H. 142
Waddington, W. H. 204
Wade, N. 236,291
Wallace, A. R. 280, 281
Wallis, R. 309
Walter, W. G. 182,213
Warren, D. 216
Watson, J. D. 142
Watt, J. 103
Weaver, W. 126, 127, 130, 132, 134,
141
Weinberg, A. M. 85,115,118,168
Weingart, P. ix, 71-87, 111, 114, 115,
353
Wright, C. 117, 118
Wright, R. 85
Young, R. M. 141,280.291
Yoxen, E. J. ix, 123-143, 166, 174,
175,211,239,261,310
Zilian, H. G. 211, 234
Ziman, J. 222,235,236
Zuckerman, H. 118,310,311