<< . .

. 5
( : 27)

. . >>

Embedded Science

Our picture of science is still heavily impregnated with epistemology “ that
is, the ˜theory™ of knowledge.
John Ziman, 2000: 6

I n 2003 the international council for science prepared to
launch what it regarded as one of its most important strategic reviews
ever. This was a review of the responsibilities of science and society. A fas-
cinating phenomenon of the last decade or so has been the international
circulation of the idea that science needs society as much as society needs
science. ˜Science and Society™ programmes seem to spring up on all sides. In
summoning the combined skills of experts and non-experts alike, such pro-
grammes try to make explicit the interdependence of the two. Thus a central
formula in U.K. science policy has recently undergone a shift in that direc-
tion: from the Public Understanding of Science to the concept of Science and
Society.1 The call is for a greater understanding of how society is implicated
in science, and how science might be made accountable to society: a ˜new
social contract™.2 In thinking about what stands for society, how one knows
when it has been engaged, society becomes itself an explicit object of inquiry.
There is considerable interest here for a social anthropology engaged with
what is made explicit and what is left implicit. For anthropologists frequently
claim that much knowledge is embedded in habits and practices that render
it implicit. If the same claim were to be made for Western (Euro-American)
science in its societies of origin, where would one look for a tacit or embedded
Supposing science is already ˜in™ society, then, where is it? What do I need
to make explicit in order to ¬nd examples of its embedding? I am going
to introduce certain knowledge practices. I shall argue that Euro-Americans

already act out ways of putting together knowledge that are ˜scienti¬c™. But they
do not always make it evident to themselves in such terms, and in this sense
the practices are only science in a tacit or implicit manner. Two arenas catch
my attention. The ¬rst is that of anthropologists prosecuting their discipline
although they inhabit a rather arcane and esoteric corner of society in doing
so.3 The second belongs to a part of social life far from arcane, indeed the
parts of their lives people often ¬nd ordinary, as I expound in the second half
of the chapter.
What science do I mean? I mean the science that claims its antecedents
in the scienti¬c revolution of the seventeenth century, a precursor to the
European Enlightenment. That was the century that witnessed ˜self-conscious
and large-scale attempts to change belief, and ways of securing belief, about
the natural world™, when people felt that they were proposing ˜new and very
important changes in the knowledge of natural reality and in the practices by
which legitimate knowledge was to be secured™ (Shapin 1996: 5). It laid down
ways of thinking that are still very much with us, or (better put) that have
spurred numerous other revolutions that keep science at once recognisable
and forever changing. One can have a greater or lesser sense of epoch, but
that earlier period is at least a starting point for asking about implicit habits
of scienti¬c thinking.4
˜The scienti¬c world is . . . that which we verify™ (Osborne 1998, quoting
Bachelard). For present purposes, I take ˜science™ as standing not for one kind
of knowledge, nor for that matter “ though it would have the greater historical
accuracy “ for many kinds. Rather, I take it as allowing for twinned or paired or
otherwise related but divergent thinking that rests on, among other things, two
ways of verifying information. The divergence between invention and discov-
ery is the case in point. One might see this as the difference between verifying
hypotheses enacted out through new instruments of knowledge (such as in-
venting an engine to use the force of compressed steam or a technique to use
the behaviour of enzymes in determining gene sequences) and verifying what
new observations can yield with respect to what is already known (such as
discovering landfalls or micro-organisms, hitherto unnamed or unrecorded
but recognisable). The line may be ¬ne, but the law turns this duplex into a
critical distinction. In the arena of intellectual property rights, the law consid-
ers the distinction as coming from science, and attributes to science different
ways of relating to the outcomes.5 At any rate, such divergence allows ideas
to appear alongside of and co-produced with critiques of them, and it creates
the possibility of different kinds of knowledge existing in tandem.
The duplex works with fractal effect; the same divergence can be repeated
at any scale. So each element of a pair can itself bifurcate, that is, become a

pair itself. I do not draw attention to this process, but it makes for constant
dovetailing in the narrative that follows.

isolated knowledge
In The island of the day before, Eco (1995) has his hero of sorts sail between
islands inhabited by people who live by different theories. Thus on one island
people are forever on their knees gazing into ponds, for they hold that someone
who is not seen cannot be. On the next, the inhabitants exist only by being
the subject of narration, talking incessantly to keep one another alive, striving
to make each story unique in order to be able to tell one another apart. These
islanders have mistaken theories for life. Yet there is another truth behind their
predicament that a social anthropologist might appreciate. Eco has to put his
people on different islands because otherwise they might have heard about
one another™s theories and come to hold their own less tenaciously.6
One spectacle that the new genetics has brought onto centre stage is the
realisation that scienti¬c knowledge is no island. It has been impossible to
isolate the knowledge that people assume scientists are accumulating about
the working of the human genome. On the contrary, this has been a prime
area in which it is thought irresponsible not to anticipate possible social reper-
cussions (it attracts many science and society projects). What is interesting
is prominence given to knowledge itself. It is not just the implementation of
knowledge that is at issue, for example, in the form of protocols to deal with
risk or pharmacogenetics (˜My very own medicine™7 ), but also the very hold-
ing of knowledge as such when that knowledge is derived from the human
(that is, some person™s) genome. One of the products of genetic knowledge
acquired for clinical purposes is widely understood to be information on a
whole range of matters about life circumstances of great interest to the person
in question. The issue is that many of them could also be of great interest to
third parties. In this light, it has become a truism to say that genetic knowledge
is frequently regarded as at once full of promise and full of danger.
Following its discussion document, the report of the U.K. Human Genet-
ics Commission (HGC 2002) addresses debate in this area. What might be
knowledge for the individual is also given something of a distance as ˜personal
genetic data™, namely, information about other individuals that is personal to
them. Several questions that have to do with what kind of information con-
cerning third parties it is permissible to have access to, and to keep, acquire
further weight when that includes information about genetic make-up. In
its recommendations, the HGC report very quickly moves from its opening

premise that personal information is private information to the point that it
is not private at all:

Genetic knowledge may bring people into a speci¬c moral relationship with one
another. We have therefore proposed the following concept of genetic solidarity
and altruism, which promotes the common good.
2002: 13, original emphasis

Such interest in relationships is not taken for granted but must be ¬‚agged as
an explicit value to be taken into account. So although, as the HGC report
explains, many of the principles to which it adheres are concerned with safe-
guarding the individual, ˜it is important . . . to see the individual as a member
of society™ (2002: 2.8).8 Note the imperative to recognise this fact of soci-
ety. Recognition then produces a moral precept that becomes an awkward
problem. Society is concretised as ˜community™ and then “ echoing terms en-
countered in the previous chapter “ revealed in a particularly distinctive form,
the family:

We do not lead our lives in isolation, but as members of communities, large and
small. We must also think of a family as a micro-community.
2002: 2.10

The problem is evident: how to balance the fact of sharing information that
may lead to better medical outcomes with the privacy that an individual
expects. The balance is particularly acute when it comes to relations with
other family members.9 Information that one family member has may be
important for others, and the ˜web of moral responsibilities™ that characterises
such relations becomes an example of a more general issue of balancing ˜social
and individual interests™.10
The crux is that knowledge personal to one person may also be informa-
tion that is potentially personal to another, so that revelation could help the
third party. Regardless of the entanglement of relations, dif¬culties are cre-
ated by conventions in the handling of information as such. Thus, if personal
information is considered private, ¬nding out about the genetic make-up of
another person becomes an invasion of privacy; if testing for genetic disorders
becomes likened to research or invasive surgery, then the worry is intervening
when the patient is not someone who will be the bene¬ciary of the knowledge
(2002: 4.54). ˜Informed consent™ becomes pretty stretched for those trying to
deal with what is perceived as the ˜ethics™ of the case.
It is ironic that what began as an aid to uncovering hereditary diseases “
being able to trace kin connections “ has turned into a different kind of aid
and a different kind of problem. It was once the case that genetic knowledge

could only be built up through information known about family members by
family members. DNA diagnosis can bypass cumbersome trawls through kin
connections, and cuts to the heart of the matter; information about inheritance
is bundled up in the individual™s own genome. The HGC report dwells on the
issues that arise in what it calls family analysis, ˜carrying out a test on one person
in order to ascertain the signi¬cance of a genetic characteristic that is shared
with a relative™ (2002: 4.54). There is the related question of whether individuals
knowing about their own make-up should volunteer their knowledge for the
information of others (Finkler 2000), a topic to which the next chapter returns.
Relatives are turned from those who are a source of information about genetic
connections (as were inferred from lines of descent) into those who need to
be told.
Such issues have become the bread and butter concerns of media debates,
ethical scrutiny, and the like. I want to suggest there is a bit more here than
meets the eye, and that concerns the role we accord knowledge. Can I use the
social anthropologist™s eye, or voice, or theory, to develop the point? Of course
anthropology does not exist on an island, and its theories are not immune to
in¬‚uence from others. And for as many who take the archetypal subject of
ethnography to be an island society, many more assume that the entities they
study are no more like islands than people are. I shall try to elucidate.

relations everywhere
Perhaps it was no great surprise after all that it was ˜relations™ that jumped out
of the kinship material in Chapter One. And I could present other examples.
However, Gell (1998) is particularly helpful in giving us a familiar response
to how one understands social anthropology™s basic position while doing so
in an unfamiliar place. What, he asks, would an anthropological theory of
art look like? It would have to look like other anthropological theories, and
they all look like theories of social relations, that is, of social interactions.11
˜The “anthropological theory of art” is a theory of the social relations that
obtain in the neighbourhood of works of art™ (Gell 1998: 26). An exceptional
pronouncement for the world of art perhaps, but totally unexceptional for a
social anthropologist, which is exactly the effect for which Gell was striving.
Relationships provide a ˜relational™ context in which to account for the pro-
duction and circulation of art, that is, a theory of relations. But how does that
come to be the anthropologist™s response? Where on earth do we get a relational
view from?
Relations are at once anthropology™s ¬eld of enquiry, its problematic, and in
the eyes of some a problem for it. The accusation is that it seems impossible to

see beyond them (Weiner 1993; Moutu 2003). But why is social anthropology
constituted by its relationality? What are the needs to which it is responding?
One of these needs appears to have been put in its path by science, and ˜science™s
relation™ (see the Introduction to Part I) offers an answer with two aspects
to it.
Arguably, the ¬rst part of the answer lies in certain nineteenth century roots
of social science. The burgeoning of social science was at once made possible
by natural science and carried the self-knowledge that the very idea of sci-
ence could be incorporated into the study of society via certain protocols and
methods. Statistical methods were one. Moore (1996: 11) dilates on Foucault™s
observations here. If the notion of the art of government as a means of man-
aging populations emerged in Europe between the sixteenth and eighteenth
centuries, this was also the time when, as a model for government, the family
(and patriarchy) disappeared. In its place was a new understanding of internal
organisation, to be found, in Foucault™s phrase, in statistics as the science of
the state.
There is nothing novel in observing that present-day statistical methods
developed as a reaction to the opening up of the world to bureaucratic inter-
vention “ after the traders, the administrators “ and to administrators™ need
to know about the populations they controlled. The point about statistics is
that it rests on the supposition that knowledge is generated by putting distinct
pieces of information together and then measuring the degree of their relation
to one another. The systematic search for correlations between entities was
the result.12 Descombes epitomises the consequence for social science:

Sociologists inspired by positivism imagine that in order to be scienti¬c they must
bow to the rules of what they call the ˜naturalist[ic] method™: scienti¬c work would
then consist of collecting data, preferably quanti¬ed, and of seeking correlations
between the data.
2000: 39

Such data are understood as individual elements in the same way as persons
may be thought of as individuals and society de¬ned as the connections be-
tween them (Schlecker and Hirsch 2001 : 71). This leads to innumerable issues
in the de¬nition of the unit of comparison (Strathern 1991 ) but is nonetheless
held to re¬‚ect a scienti¬c approach. Here, then, there is a fairly direct and
explicit gesture toward the in¬‚uence of ˜science™ on the subject.
Correlation may be taken a signi¬cant further step in the quest for causal
relations. Demonstrating causal relations “ veri¬cation through predicting
outcome “ is another hallmark of scienti¬c method. But science, and this
became especially true of social science, does not need this step to appreciate

the force of correlation. It is an achievement in itself to demonstrate the ¬t
among data from different domains.
Now, regardless of the personal beliefs of practitioners, the material and
apprehensible world of the seventeenth century had become conceivable as a
self-verifying system (hence the attraction of auto-regulating mechanics and
perpetual motion, for example, Crook 2004). If it operated without anyone
having to seek a cause beyond it, then it must operate on its own terms. The idea
of entities existing on their own terms was replicated in the items that made up
the natural “ or social “ world, that is, the items between which connections
were being sought. If an aim was ¬t between data from different domains,
then the very independence of these domains from one another became the
prerequisite to determining co-variation or correlation; this became ˜Galton™s
problem™ (Jorgensen 1979). Relations were made evident to the extent that
the items being related to one another were otherwise autonomous. In short,
apparently unique elements in the world could be explained by the way they
impacted on or were variously connected to one another, and what science
determined and described were the relations between them.13 Indeed, they
were not to be explained by anything beyond one another, and knowledge
came from nowhere but the demonstration of interrelationships.
We might see the effort to demonstrate connections through relating hith-
erto unconnected facts as involving the creation or invention of new kinds
of relations, new systems of classi¬cation, say, that link phenomena already
known in other ways. Such instruments of description turn hypothetical con-
nections into actualised ones, ones that stay stable (Law 1994). However, re-
lating apparently independent entities to one another (inventing the relations
that made them connected) is but half of the scienti¬c enterprise. The other
half has to do with discovery, uncovering relations that already exist. When it
comes to the elucidation of society, perhaps social science™s own pre-eminence
in this sphere (discovery) meant that associations between this aspect of nat-
ural science and the development of social science have tended to remain
implicit. This is the second part of the answer.
Ziman (2000: 5, original emphasis) observes that there are many forms of
knowledge: ˜What makes any particular form of it scienti¬c?™ He sums up the
old answer in the phrase ˜epistemological naturalism™. Science is a complex
system, with various elements interacting, itself a model of such interrelations.
Although he hopes for a new (non-epistemological, even possibly ˜life-world™)
model, for present purposes I shall be satis¬ed to elucidate aspects of the old.
By that I mean the science that addresses a world understood in terms of
itself. If relating hitherto unconnected facts involves the invention of new
kinds of relations, then showing or uncovering how each fact is already part of

everything else, already predictable or de¬nable through the internal coher-
ence of relations that already exist, is more like discovery.
The other half of the scienti¬c enterprise, then, is to specify the co-de¬ning
elements of an internally coherent system that will furnish a description of
every element as part of it “ as one might (literally) imagine the periodic
table, or the model of DNA “ thereby creating the notion of ˜orders™ of kinds.
Knowledge will come from specifying what does, or does not, belong to the
system. The system entails its own canons of veri¬cation. Science here consists
of a circuit of intelligible signs that mutually reinforce one another, a percep-
tion of their ¬eld of which nineteenth scientists were particularly conscious
(Beer 1996). This does not mean the circuit is all-encompassing (˜scienti¬c
paradigms are never epistemically complete or coherent™ [Ziman 2000: 198]),
but it does mean that de¬nitions are bound up in one another. Neutrons,
electrons, positrons “ these terms must be mutually sustaining.14 Systems of
classi¬cation appear in a new light, not as the invention of scientists but, when
the gaps get ¬lled in, as a means to discovering what is known to exist but is
not yet brought to light.
Now this is, so to speak, an approach to the world with which social science
was to become familiar (and echoes political economy™s critique of classical
economics). It came out of a criticism of the ¬rst view that, translated into
the orbit of social life, saw society as the links between individuals, entities
otherwise independent of one another. The criticism is that to understand
social relations as existing between individuals is mixing orders of logic. By
de¬nition, individuals preclude relations. Relations can only exist between
relata “ elements of the relation. Far from relations being sought as connections
among things, here things are already in relation, that is, co-implicated, with
one another.
The contrast was played out in social anthropology with great force be-
tween structural functionalists and structuralists in the middle of the twenti-
eth century.15 On the one hand, it was supposed, alliance, as found in relations
set up through marriage, rested in the connections people created between
autonomous social groups de¬ned by independent criteria such as consan-
guinity or descent. On the other hand, went the objection, the possibility of
such connection was already implicated in the very de¬nition of groups as
wife-givers and wife-takers to one another. Rather than being created in the
effort to make connections, in their co-implication relations are seen to be
inherent in the very way in which the entities are classi¬ed, a pre-condition
of their existence.
Nonetheless, what seemed obvious to students of society could also be
elusive as an object of analysis. How is such co-implication to be veri¬ed? The

pre-condition of relationality becomes elusive if one tries to attribute it to
some pre-prexisting mental state or to the collective properties of people and
societies. Descombes (2000), taking up Winch™s (1958) claim that the mental
and the social are ˜two different sides of the same coin™, sees the antecedents
of this claim in Durkheim™s efforts to elucidate collective representations:

One cannot ask any longer whether such and such a form of representation (for
example the concept of space or of causality) belongs to an individual conscious-
ness or a collective consciousness. But one can ask oneself in what social world
can people form such a concept. And then reverse the question: what concepts
does one have to possess for such a social relation to establish itself?
Descombes, 2000: 39

He ends with the example of property.16 The concept posits a social relation
between holders and non-holders. In this sense the idea of property is a ˜col-
lective representation™, for the idea and the social relation it incarnates are
dependent upon one another. The point to draw more generally from the ar-
gument is that relations exist ˜internally™ as elements of a system that is already
described by the relations it consists of; it is in this Dumontian sense holistic
The pre-condition of relationality becomes very obvious (self-evident) in
one sphere, and this is found in an unusual quarter within anthropology.
Radcliffe-Brown and the structural-functionalists, who according to their crit-
ics failed to get the point when it came to delineations of descent groups or
elements of myth, saw the priority of relations brilliantly when it came to the
analysis of kinship terminologies. It was in this context that social anthro-
pologists insisted on the analytic term ˜person™ (rather than individual), for
the person was already an element of a social relationship, already a relata, a
function of relating. Indeed, it was possible to talk of kinship systems, which
made up ˜a complex unity™, or more generally of a structure, which constituted
˜an arrangement of persons in institutionally controlled or de¬ned relation-
ships, such as the relationship of king and subject, or™ “ the kinship analogy
quickly follows “ ˜that of husband and wife™ (Radcliffe-Brown 1952: 53, 11).
Paradigmatically, to be a parent implies a relationship with a child. Here is
evidence of co-implication: entities in a state of mutual de¬nition.
The relations I have been talking about exist in the systems of knowledge that
science has developed. I said that relations remain anthropology™s problematic,
and problem. Their elucidation takes divergent paths. Explicit comparison
with science was made possible by the positivist supposition of a world of
discrete entities between which connections were to be made. At the same time,
the kind of closed system that kin terminologies suggested to anthropologists,

the matrix of mutually de¬ning terms co-implicated with one another, evokes
a comparison “ that remains largely implicit “ with that second set of scienti¬c
suppositions, where relations wait to be discovered. But what does the Euro-
American observer (including the anthropologist) imagine is produced by
such relational exercises?
Viveiros de Castro draws from a contemporary of Durkheim, Tarde,17 who
asks: ˜What is society? From our point of view, it can be de¬ned as the recip-
rocal possession, under extremely varied forms, of everyone [all] by everyone
[each]™. Entities are no more nor less than the sum total of their reciprocal
inter-possessions. This may come about through striving for connection or
from uncovering a prior state of relationality. Either way, I want to fold this
supposition into one of Viveiros de Castro™s own insights, that the hallmark
of modernist philosophy is the ˜conversion of ontological questions into epis-
temological ones™ (1999: S79). He writes that Euro-American anthropologists
˜persist in thinking that in order to explain a non-Western ontology, we must
derive it from (or reduce it to) an epistemology™ (1999: S79 [emphasis omit-
ted]), that is, to a concern with representations, with how people make things
known to themselves. An example is a nod to natural science from Radcliffe-
Brown (1952: 7), ˜The basis of science is systemic classi¬cation™. Classi¬cation is

<< . .

. 5
( : 27)

. . >>