Callon Struggles and Negotiations 1980
Callon Struggles and Negotiations 1980
Callon Struggles and Negotiations 1980
MICHEL CALLON
Ecole des Mines de Paris, Centre de Sociologie de I Innovation
In the space of a few short years the centre of interest in the sociology of
sciences has radically shifted. At first timidly, later with increasing boldness,
sociologists have penetrated the sanctuary. They no longer confine their inter-
est to a study of how institutions work, or the rules governing competition,
or network or community organisation. Increasingly, they are investigating
the content of science itself.
Though this change of direction now seems legitimate, and indeed irre-
versible, it still remains very tentative. Deeply marked by its recent past, the
sociology of science still takes for granted the chopped-up, compartmentalised
world the scientists are so patiently building up, being ready to distinguish,
and even unhesitatingly to place in mutual opposition social factors and
technical or cognitive ones (1). Concepts like social contexts of scientific
research are still in common use today, proving the continuing vitality of
this way of thinking. Within reality territories and domains a[e divided off,
frontiers laid out, a priori factors of different types .identified and phases
obeying a specific logic enumerated (2).
However, these common distinctions are having increasing difficulty in
holding out against sociological ventures. The deeper we delve into content,
the more the legitimacy of black boxism seems questionable (3) and the more
difficult, hazardous and arbitrary the separation of social from non-social,
cognitive from non-cognitive becomes (4). The most solidly-based concepts
dissolve, revealing their ambiguity. What does 'reproducing an experiment
mean'? (5) What is understood by the expression 'reporting on a research
process'? (6) It gradually becomes apparent that social and cognitive are
inextricably entwined exactly where unravelling them seemed to present no
197
Karin D. Knorr, Roger Krohn and Richard Whitley (eds.), The Social Process of
Scientific Investigation. Sociology of the Sciences, Volume IV, 1980. 197 -219.
Copyright © 1980 by D. Reidel Publishing Company.
198 Michel CalIon
An Abundance of Problematisations
The DGRST was founded in France in the late fifties, with the aim of pre-
paring, coordinating and implementing French policy with regard to scientific
and technical research. One of its first acts was to set up 'concerted actions'
in which laboratories, both private and public, within industry or the Univer-
sity, came together for a limited time to work on top priority programmes.
Each 'action' is administered by a scientific committee consisting of about
fifteen experts (scientists, industrialists and officials) who take part intuitu
personae in the work of the committee. The committee selects projects
from those submitted and distributes the credits allocated to the work. The
procedure used was worked out during the fmal years of the IVth Republic.
In this way public and private research can work together and programmes
rejected by traditional institutions (CNRS, universities, industrial enterprises)
are more readily fmanced, thus facilitating coordinated and collective work
on subjects that have been given top priority.
In fact, in the early sixties, the CNRS and industry left the scene vacant,
the former undermined by academicism, the latter little concerned with
research and innovation. The DGRST filled this vacuum. As a result, the main
initial beneficiaries of the operation were those scientists whose diSCiplines had
been misunderstood or looked down upOIi by both the University and the
CNRS, who housed them, but gave them no real means for development. The
'concerted action' procedure fitted them like a glove; they were assured of
both industrial and political support; likewise they had scope for action in the
form of credits, both of which had previously been refused (11).
The above remarks are fully applicable to the research on fuel cells under-
taken within the framework of the "energy conversion" concerted action,
whose aim was to develop new forms of energy production. There were
no industrialists on the committee responsible for pushing the programme
through, the scientists intended to call the tune. They imposed their own
analysis of the situation, sketched out the problems to be solved and the
links between them. They decided how the work was to be divided up and
coordinated. Finally they indicated what was at stake at the social, political
and economic levels (12).
In the case of fuel cells, problematisation operated in three phases and
reveals a wide range of possible analyses.
200 Michel Callon
(1) The committee's first task was to identify interesting fields of research.
The general theme of energy conversion provided an initial territory within
which priority sectors had to be identified. The first discussion focussed on
the defmition of what was interesting and what was not. Two physicists, X
and Z, were set against each other.
When he was asked to sit on the committee, X was a scientist well known
within his own discipline of solid state physics. He had spent a considerable
time working in a well known laboratory in the United States. Since his
return he had published several articles that attracted considerable attention.
At the committee's very first meeting he put forward an analysis of spheres of
research that might be of interest. His argument is summarised in Table 1. At
the head of each line and column the various forms of energy are indicated;
electrical energy, light energy, mechanical energy, thermal and chemical
energy. The columns correspond to the initial energy forms, the rows - the
fmal forms. Each division of the table thus represents one possible method of
energy conversion, for example, conversion of chemical energy into electrical
energy. Each of these divisions is simultaneously and inextricably linked to
various phenomena, various effects and a variety of technical devices. Some
divisions are partly empty, either the devices do not yet exist, or else the
phenomena have not yet been properly identified. Other divisions refer
to spheres so huge that an exhaustive inventory is thought unrealistic. We
shall return later to the 'logic' (we would call it 'socio-logic') underlying this
table. For the moment we shall content ourselves with pointing out how it
functions.
First of all the table establishes a perfectly clear frontier between what is
analysed and what escapes analysis. The 'edges' of the table demarcate the
reality considered relevant. This is a very general phenomenon, the construc-
tion of a black box. X has created an inside and an outside, manufacturing
a local coherence. He has defmed a protected territory, claimed an autonomy.
The table demarcates and defmes spheres of research on the basis of
categories considered obvious and quite distinct. If the table possesses its own
coherence, which enables it to defme a distinct universe, closed in upon itself,
this is because it provides a strong framework. The energy forms can be located
and demarcated. The concept of energy conversion is not called into question.
Using his table, X divides up the ground, defming territories quite separate
from each other. Not only does he mark off the different domains, he also
IJ:l
TABLE 1 t~
&!1)l
Initial l'>
;:s
energy I:l.
Final ~ form Electrical Light Mechanical Thermal Chemical ~
form ~
iiI·
g.
Electrical Converters Photo voltaic Electric Thermoelectric ;:s
Fuel cells
Rectifiers Photogalvanic machines thermo-ionic Ordinary cells '"C
Transformers effects wind effects Accumulators b
Oscillators tidal power ~
S·
currents '"~
Light Tribolumi- l'>
Electrolumin- Luminescence Incandescence Chemolumi- ....
~.
escence. Dis- nescence nescence
charge in gases ~
<:>-
~
Mechanical Electric Crookes Simple Thermal Artificial muscle 3l'>
machines Radiometer machines; machines Propellant i;.
? Energy storage l'>
;:s
I:l.
Thermal Static heat pumps Solar Heat pumps Refrigerators Combustion
(Peltier effect). energy refrigeration (adsorption) ~
l'>
Electric heating collection Exchangers ....
~.
by arc. Dielectric
heating H. F. ~
....
Plasmas
Chemical Electrochemistry Photosynthesis Chemical Thermochemistry (too huge)
Radiochemistry grafting ? ?
N
laminating
....0
202 Michel CalIon
shows what work remains to be done. The squares are more or less easy to
fill in, more or less enigmatic. Reams have been written about the conversion
of thermal energy into mechanical energy, whilst in the square chemical
energy ~ electrical energy there is little to be said. Darkness reigns. Who at
that time would have dared to claim that the functioning of fuel cells had
been fully investigated? Thus are contrasted the old and the new, the more
and the less problematic; fields already explored (thermal machines, electric
machines) and fields that call for new investigations (fuel cells, photovoltaic
effects ... ).
However, X's is not the only possible problematisation; at the same time
another physicist, Z, put forward another one. Z is a product of one of the
most renowned French scientific institutions: the Ecole Normale Superieure.
Though he is only just starting out on his scientific career, he already has
considerable support behind him. He is not yet well known enough to belong
to the energy conversion committee, but his reputation is sufficient for him
to be allowed to explain his point of view to the scientists and industrialists
concerned with fuel cells. A large scale meeting is arranged. In his speech Z
lambasts X. His line of argument leads him to radically different conclusions.
Z does not even refer to the general question of energy conversion. At no
point does he distinguish different forms of energy. All this is outside his field
of analysis. His point of departure is electrocatalysis, that is catalysi~ of
reactions which liberate electrons (oxidoreduction reactions). In this way he
defmes the sphere within which research should take place. The concept of
electrocatalysis cuts short discussion just as effectively as the table proposed
by X. Demarcation of this problematic field is based on a set of concepts,
theories and elements which are taken for granted. Z makes specific reference
to the latest developments in solid state physics, and the tools these provide
for the study of electrocatalysis.
Z's problematisation is very much less structured than that of X, but it
follows the same logic; a problematic field closed in upon itself, then state-
ment of elements taken for granted and considered certain, which give
the field its rigid, autonomous framework. Though Z's presentation of his
problematisation remains somewhat vague with regard to details, he is able
to show quite clearly how it differs from X's analysis. He considers that
. electrochemistry will remain in a very weak position until it has managed to
extricate itself from the technological approach that is stifling it. For him the
Struggles and Negotiations to Define What is Problematic and What is Not 203
theoretical unity of the fuel cell is a myth. Though admitted and consolidated
by X's problematisation, this object must be 'de-constructed'. The alternative
is clear. The problem is not that of improving catalysis in fuel cells, but rather
of working out the laws governing electro-catalysis in general.
(2) The committee accepted X's proposed problematisation. Fuel cells are
one of the three themes given top priority. The committee instructs Y to
work out a research programme on fuel cells. Y is an electrochemist by
training, director of the 'Electrolysis laboratory', which is dependent on the
CNRS. After a few weeks work he presents the committee with a document
explaining his own problematisation. His analysis is synthesised in a table
(Table 2) giving details of the lines research should follow and indicating
research centres to be mobilised. How is this table organised?
TABLE 2
Firstly, the table demarcates a territory for analysis within the area of
reality. This territory is firmly delineated by the outlines of a specific object,
the fuel cell, and by the theoretical assumptions made about it. Y's prob-
lematisation fits perfectly within that of X. The latter had already provided a
system of partitions which Y took over as it stood. The fuel cell represents a
privileged object in electrochemistry, as viewed by Y. No-one and nothing
could undermine this relationship. The cell is contained as a whole within
electrochemistry, and vice versa. There is no overlap on either side. The wall
around is a perfect fit, it is totally self-sufficient and may not be disturbed in
anyway.
The table defines themes for research by formulating problems (these are
explained at more length in the accompanying notes). Y draws up his balance
sheet, using his own organisation and formulation of problems. He draws a
demarcation line between what, in his view, is known of how a cell works,
and what is not known. What strikes the observer forcibly is how the fuel
cell's architecture, the different elements that make it up, and the phenomena
within it, all correspond closely to the aims and themes for study. There are
the electrodes, the electrolyte, the catalyst. There is reference to knowledge
that was widely accepted and used at that time in France by those calling
themselves electrochemists (diffusion, internal resistance, de polarisation,
kinetics ... ). A whole set of concepts, proposals, ways of thinking, methods
of giving proof are called into play to isolate and derme the darker comers
of how a cell works. The areas of ignorance appear against a background of
certainty, admitted knowledge and systems of interpretation (13).
We might note in passing, since this helps to explain the nature of the
opposition between X and Z, that one of the most important results of this
type of problematisation is the place given to catalysis. Y states that catalysis
is merely a technical problem, therefore of secondary importance (Y is a
fundamentalist). This is in strong contrast with Z's position. For Y the prob-
lem of catalysis will be solved as soon as the problems of kinetics, transport
of reagents and optimal structure of electrodes have been elucidated. Z states
precisely the opposite.
(3) The research programme proposed by Y was adopted exactly as it
stood. Since a concerted action was involved he divided up the work between
the various research centres, both private and public, which he thought were
likely to be interested. Thus he proposed to entrust to his own laboratory
Struggleg and Negotiatiom to Define What ig Problematic and What is Not 205
1----'r-c~Areas of suspicion
Area
of the
unanalysed
Network or areas of
certainties
Fig. 1.
Before going any further with our analysis of the process of problematisa-
tion, let us stop for a moment and look at a few results of the type of analysis
chosen.
Problematisation culminates in configurations characterised by their relative
singularity. There is not one single way of defming problems, identifying and
organising what is certain, repressing what cannot be analysed. Witness the
different paths followed at the same time by X and Z, or by A and B. In this
case the configurations are in opposition with each other. However, there is
plenty of variety within each of these main options. Though there may be
strong similarities (which enable problematisations to be grouped together)
there are always differences, however slight. Each protagonist organises and
problematises reality in his own original manner in keeping with his own
idiosyncracies, his own background and the particular conditions in which he
fmds himself (15). Henceforth we shall no longer make a distinction between
an actor and his problematisation. Identifying a problematisation postulates
the existence of an actor.
As the cases of X, Y and A show, hierarchical relationships often exist
208 Michel Callon
what is external, what is scientific and what is technical, the links which
should exist between the two, etc ...
These fmal remarks raise new questions. How is it possible for problem-
atisations, though different from each other, to form connexions with one
another? A reply to this question will be found in a description of the special
logic which problematisation obeys.
that these are only proposals. Y is not certain he will be able to impose his
problematisation. However, the important point is that for Y the social and
the cognitive,the problems and the actors are arranged within the same struc-
ture. To each of his problems corresponds a place and a position attributed
to an actor. The actor may be named, or his identity remain unknown.
Moreover, the relationships between the protagonists and their positions are
clearly identified through the relationships postulated between the problems.
Thus, properly speaking, the cell exists in two ways: one we might call
techno-scientific, the other social, for it is not distinct from the social group
approached to carry out its elaboration and production. Problem defmition,
as practised by Y, is a highly strategic activity, aiming as it does to interest
varied groups in an enterprise whose development as a whole they will not be
able to control.
In what we have called the area of suspicion, which forms the heart of the
problematic situation, there is no divergence between organisation of the
social field and that of the cognitive field. Defmitions of problems and the
links between them cannot be distinguished from the work of organising fields
of interests to be aggregated - witness the question marks that figure in some
squares of the table. Defmition of a problem implies definition of a group,
even if no empirical unity can be named. Y gives shape to the social, he builds
a field of positions.
We can go further. The list of problems as suggested by Y cannot be
deduced from the state of scientific and technical knowledge (Z's active
criticism provides proof of this). It translates a determination to incorporate
interests, and to interest those who are still only potential partners. In fact
Y's programme represents an attempt to mobilise social groups. I propose
to call this particular logic by which problems are directly associated with
groups; the socio-Iogic oftranslation (16).
Why this expression? To justify its use I need only analyse the mechanism
at work. What Y is saying can be summed up thus: "I defme a series of
problems PI, P2, P3 ... P8 and assign them to groups GI, G2 G3 ... G8
(see Table 2). I state that a sequential solution of these problems would lead
to solutions of the problem posed by X, that is how to build up and acquire
scientific and technical mastery of fuel cells".
Defmitions of PI, P2, P3 ... and statements of their interdependence
follow a socio-logic. In fact to state that PI, P2, P3 ... are "logically" linked
Struggles and Negotiations to Define What is Problematic and What is Not 211
(by the problematic unity of the cell) is to state that a community of interests
exists between G I, G2, G3 ... This puts forward the hypothesis that G 1 will
take charge of PI, G2 will take charge ofP2, and that Gl, G2 ... will accept
the idea that a relationship exists between PI, P2 ... , that is to say that
social intereaction between them is conceivable. In short, Y constructs a
system of social interactions. We do not fmd on the one hand social actors,
on the other knowledge. There is joint, programmatic organisation of both
knowledge and of social actors. Hence the idea of a socio-logic.
The statement that PI, P2, P3 ... can stand in relationship postulates:
(a) that a set of related significations exist for problems formulated within
different territories, and (b) that the solution to a problem (mastery of
fuel cell functioning) can be achieved through a series of displacements of
problems. The word 'translation' corresponds precisely with these two
meanings. Considered from a very general point of view, this notion postulates
the existence of a single field of significations, concerns and interests, the
expression of a shared desire to arrive at the same result. Though transla-
tion recognises the existence of divergences and differences that cannot be
smoothed out, it nevertheless affirms the underlying unity between elements
distinct from one another. Translation involves creating convergences and
homologies by relating things that were previously different. In the more
limited case we are examining, translation first of all assures that intelligible
connexions exist between questions concerning, for example, diffusion in
electrolytes, kinetics of reactions in electrodes and performance of the cell
(measured by available potential and intensity of current). Proposals, results
and appreciations can be converted from one to another so as to become
comparable. For example, a particular modification in electrode structure and
the distribution of the catalyst will react upon the operation of diffusion, the
latter in turn will modify the kinetics of the oxidoreduction reactions; the
result will be a variation in the intensity of current with consequences for
commercial implications. Translations like these are never a foregone conclu-
sion. They are formulated as hypotheses which will be judged convincing or
otherwise, (B, unlike A, is not convinced) (17). However, simultaneously,
and this is its second Significance, translation emphasises the interdependence
of problems. Solution of a problem depends on the prior solution of a whole
series of other problems (to improve kinetics implies previously improving
diffusion; achieving control of an outlet involves agreeing to study the
212 Michel CalIon
the socio-Iogic is one of fission which respects and builds up differences and
distinctions.
All we need to say here about the structure of the un-analysed is this: its
structure resembles that of the unconscious. It represents what is kept silent
so that the rest may be stated.
TABLE 3
Tagging along + +
1 +
Negotiation
11 +
Opposition
Inertia o o
214 Michel CalIon
Tagging along: The group approached recognises that its interests coincide
with the solution of the proposed problem. Moreover, it endorses the socio-
logic underlying the problematic situation which is in course of consolidation.
This is Y's position in relation to X. He agrees to take charge of the cell
theme, acknowledging the field of research to be a highly interesting one for
electrochemistry. He does not challenge the intellectual consistency of the
problematisation (classification of energy forms, confusion between technical
devices and theoretical objects), nor its socio-political consistency (energy
conversion is a homogeneous field of research which should be financed by
the DGRST rather than the CNRS). A adopts the same strategy with regard
to Y. This 'tagging along' attitude is an expression of a balance of forces
which ensures, locally and provisionally at least, the total success of the
problematisation. Thus it is possible to conceive of problematisations which
are deduced one from another, but only so long as it is clearly recognised that
deduction is never more than a successful translation.
Negotiation 11: The group approached agrees with nothing at all, except the
formulation of the problem assigned to it. In other words, it is ready to launch
into the specific research proposed, but it does not intend to fit into the
collective enterprise outlined, nor will it accept the suggested socio-cognitive
relationships as outlined. Here criticism can operate in several fields. Social:
the group approached considers the proposed agglomeration of interests un-
natural. Cognitive: it considers the relationships between problems postulated
Struggles and Negotiations to Define What is Problematic and What is Not 215
(20) though he used concepts similar to our own. The effect of the actor's
action is not to create stability and order. It is to create local instability. With
the creation of such instability the possibility of autonomy arises (21).
One last point remains to be made. We have just described strategies which
arise as responses to a problematisation. But in what conditions do they
appear? Why does Y follow X? Why does B resist? The answer is to be found
in the concept of capital though not as this is understood by Bourdieu (22).
Concept is not a stock. For example, X's capital is more than his credit social
relationships, prestige, and his influential position. He is more than a set of
resources. Economists are well aware that identical resources can lead to
different strategies, some ending in failure, some in success. Capital cannot be
dissociated from the way it is utilised to incorporate interests, seek support,
intervene, translate and convince. These valorisation strategies must be studied
if the force of a problematisation and its power to enlist support are to be
assessed (23).
Conclusion
(1) Using the concept of the problematic situation, with its distinction
between zone of fusion and zone of fission, we can go beyond the natural
opposition that often operates between the social on the one hand and the
cognitive on the other. The analysis of problematic situations, showing how
they are organised, throws light on the process by which the limits between
the social and the cognitive are constantly re-defmed. The zone of fusion is
the crucible where practical categories are worked out, whilst in the zone of
fission they are consolidated. However, it must be noted, that these cleavages
are always linked to a specific problematic situation. In these conditions,
surely, concepts like the social context must be cast aside. They fail to
recognise the reality of problematisation and take for granted what is in fact
at stake for the protagonists. To problematise is, among other things, to
produce social context both for oneself and for others.
(2) In addition, the concept of the problematic situation makes discussion
of the significance of the sociology of content possible, something which is
everyone's dream. My view is as follows: It is only possible to discuss content
from within a problematic situation, that is, after having defmed what is
considered problematic and what non-problematic. The sociologist is caught
Struggles and Negotiations to Define What is Problematic and What is Not 217
16. lowe the concept of translation to M. Serres, Hermes 111, La traduction, Paris,
Editions de Minuit, 1974.
17. An analysis of translation mechanisms needs to be developed. We simply state that
it is linked to the construction of problematic situations themselves. A problematic
situation de-contextualises concepts, proposals and categories, and then re-con-
textualises them using its own logic. Thus problematic situations permanently create
metaphors. The latter's existence make translation possible (for 'metaphorisation'
see R. Krohn, 'The Social Process of Scientific Investigation', unpublished paper,
McGill University, 1978).
18. K. Knorr, op. cit., 1977, Note 15.
19. M. Callon, 'L'Etat face al'innovation technique; Ie cas du vehicule electrice', Revue
Francaise de Science Politique, 426-447 (Juin 1979).
20. R. E. Dewey, The Philosophy of John Dewey, Martinus Nijhoff, The Hague, 1977.
21. See the very fine analysis of a novel by M. Tournier put forward by G. Deleuze, La
logique du sens, Editions de Minuit, Paris, 1969.
22. P. Bourdieu, La distinction, Editions de Minuit, Paris, 1979.
23. M. Callon, and B. Latour, 'Unscrewing Leviathan: How do actors macrostructure
reality?', Forthcoming 1980.
24. M. Serres, Le parasite, Grasset, Paris, 1980.