Book review.

Re-assembling the social (The actor network theory): by Bruno

Latour.
Having explained his intentions in the Introduction, Latour demonstrates
in Part I how to deploy controversies about the social world. What is taken for
granted is not easy to study; controversies mean that the ground of the obvious
has been moved, at least in part. Each controversy is a source of uncertainty to
be explored, and there are five such sources, discussed in five chapters.
The first source of uncertainty is the status of groups: do they exist, or are
they being constantly formed and re-formed? ANT takes the latter option, and
is able, therefore, to show that ‘the first feature of the social world is this constant
tracing of boundaries by people over some other people’ (p. 28). The
scholars who chose the first option are themselves engaging in such tracing: each
‘organization study’ adds stability to a unit called ‘organization’, which
might otherwise be on the verge of dissolving or regrouping. No wonder that in
times of constant mergers and acquisitions, young researchers may feel lost
when ‘the organization’ they set out to study no longer exists when they reach
it. No such surprises to the ANTs: they simply follow an actor and note in their
field notebook the name used for the location they arrive at. It is not groups that
need to be studied, but the work of group-making and unmaking.
The second source of uncertainty concerns agency: who or what is acting
when an action can be observed? The notion of actor-network strongly suggests
that that which presents itself to an observer as an ‘actor’ may, in fact, be a
whole network. One of the traces leading in this direction is the inconsistency
of accounts given by those who seem to be actors to the researchers. Instead, in order to
explore this source of uncertainty, researchers try to eliminate it by
picking from actors’ accounts only that which can be easily incorporated into a
theory. Thus ironing out the inconsistency from the accounts, the researchers
erase indications of multiplicity of agencies, which ought to interest them most.
If there is any uncertainty as to how agency should be described, it follows
that even objects can be seen as possessing agency — the third source of uncertainty
to deploy. It is at this point that the definition of ‘social’ is significantly
extended: from ‘humans only’ to ‘all actants that can be associated’. This extension
is nothing new to fiction or to everyday life — yesterday’s faithful companions
such as dogs and horses have been replaced by computers and iPods —
and the need for associations to extend beyond humans has always been obvious.
Afraid perhaps of losing their domain, social scientists are fastidious in
differentiating between humans, who are their concern; and non-humans,
who belong to other disciplines. In organization studies, this self-definition is
often revealed in cooperative projects run by sociologists and economists.
Sociologists tend to lose interest when ‘money’ comes into the picture, expecting
economists, living in the space between social and natural sciences, to pick
it up for inspection. And yet what would money do without bonding itself to
human beings?
The purpose here is not, however, to anthropomorphize insentient beings,
although such an operation is performed daily by all competent speakers. It is
to point out the special role that objects play in associations: they stabilize. This
is why contracts are written, obituaries carved in stone, and technical norms
built into the instruments to make the users behave in a prescribed way. We do
not live in a society, but in a collective — composed of humans and nonhumans.
It is also here that Latour begins to answer the puzzling critique often
directed toward ANT: that it ignores ‘power relations’. Far from ignoring power
or using it as an explanation, ANT attempts to explain it. ‘People are rich
because they possess capital’ is a tautology. ‘How did they create the bond?’ is
an ANT question.
The fourth source of uncertainty is the status of facts: how to tell the difference
between a ‘matter of fact’ and a ‘matter of concern’? The difference is in
the making, and it is this making that ANT wants to study. Concern can turn suppositions
into facts and politics can turn facts into concerns — or delegate them
into oblivion. The discussion of this source of uncertainty gives Latour the opportunity to
explain why ANT abandoned the label of ‘social constructivism’.
ANT’s interpretation read ‘social’ as ‘not individual’ (as, for example, in Piagetian
constructivism), and ‘construction’ as ‘not creation’. The common reading is, however,
‘human contrivance’ (as opposed to divine or natural creation) — in other
words, negative, or at least weak. The label can be abandoned, but not the task.
Studying how matters become concerns or facts is an obvious task for sociology
of science, but equally obvious for management studies. Isn’t this exactly what
both accounting and strategic management set out to accomplish?
The fifth source of uncertainty will be easily recognized: how to write
research accounts? This chapter is both an encouragement to literary inventiveness
and a warning against glibness and overconfidence. The desirable opposite
of rigour is not sloppiness, but vivacity and flexibility. ‘Textual accounts are the
social scientist’s laboratory’ (p. 127), and, consequently, ‘A good text is never
an unmediated portrait of what it describes’ (p. 136). One point worth emphasizing
is Latour’s claim that ‘the mere description’ is always the most difficult
task in a research report, from which many writers flee, to hide behind comforting
if empty abstractions.
The first part ends with a dialogue with a student, confused by the (extolled!)
difficulty of doing ANT studies of organizations (a collection of texts by those
who tried anyway has been assembled by Czarniawska and Hernes, 2005). The
dialogue is certainly a composite, but convincing exactly because of that: it represents
many a doubt voiced by beginning researchers. Here again Latour
(1999) mocks the name of his approach while defending it.
The fictitious Professor of the dialogue may be poking fun at the Student
(who is also irreverent, creating a symmetrical discourse), but Professor Latour
takes to heart the difficulties reported by the student. Part Two propounds a three-step method
for “flattening” the three- dimensional
“topography of the social” (p. 165). The first step is to “localize the global”
(p. 173). Latour argues that the misleading notion of a broader “context” in which
to situate all the local interactions originates from a conceptual kinship between
Society and the body politic. In his view, the best way to deconstruct the micromacro
divide is to travel between interconnected local sites without ever taking
shortcuts and to study “the very production of place, size, and scale” (p. 171).
Hence his distinction between “panoramas” and “oligoptica” (p. 181).
The second step is to “redistribute the local” (p. 191). Latour puts forward that “interaction”
is no less an abstraction than “context”. He asserts that the “local”
interactions are primarily assemblages of many other interactions “distributed
elsewhere in time and space” (p. 194) but affecting the here-and-now, through
the agency of non-human actors. He thus stresses the importance of “interobjectivity”
(p. 195) and underlines that “translations” (p. 196) – i.e. connections
that “transport” transformations – should be at the center of the study. He also
states that being a full-blown individual endowed with subjectivity and cognitive
abilities is conditioned by the availability of specific equipments. Just like any
other “whole”, the individual should be regarded as “the provisional achievement
of a composite assemblage” (p. 208). Besides, just like any other actor, the individual
is something made to act by a network of “attachments”, that is, a “starshaped
web of mediators” (p. 217).
Th e third and final step is to connect the sites – now conceived as actorsnetworks
– and to investigate the stabilization of the social. Latour insists that
agencies need “vehicles” to be transported from one site to another. In particular,
they often need to be given a “form” (p. 223). Also, without standards, quasistandards
and other metrological tools – including those manufactured by social
scientists – agencies would be neither comparable nor commensurable. What’s
more, their connections are both “performed” and “formatted” (p. 226) by way of
all the “collecting statements” (p. 231), folk theories and generalizations as to how
they are and ought to be connected.
According to Latour, sociology – which he newly defines as the discipline
explicitly devoted to the tracing of associations – achieves political relevance by
contributing to the process of (re-)assembling the collective. As such, it provides
no “cut-all” critical edge, but a lot of little critical insights into the weaving of the
social. Is that enough? In a sense, it is: given that, in fact and in right, sociologists
have no monopoly on critique, a progressive social science should above all strive
to activate and enhance the shared critical ability. By producing up-to-date, refined So Part II
begins with an
admission that it is not easy to trace the social, and gives advice on how to study
associations in three moves. To be able to perform these moves, a new cartography
is needed. The moves of the new scientists of a social are not to be
between local and global or between micro and macro, because such places do
not exist; they will be moving across a flatland.
The first move consists of localizing the global — of realizing that there is no ‘global’, but
only a chain of connected localities. ‘No place can be said to be
bigger than any other place, but some can be said to benefit from far safer connections
with many more places than others’ (p. 176).
But the local never occurs in one place only, so the second move must be a
redistribution of the local. The loving conversation that you overheard at a
nearby café has been fed by dozens of Hollywood films and hundreds of pop
songs that have been produced — and consumed — in distant localities and distant
times. The lovers who look into each other’s eyes see in them reflections of
Lillian Gish, Greta Garbo, and Julia Roberts; Errol Flynn, Rudolf Valentino, and
Jeremy Irons. And if intimate interactions are so densely populated, how overcrowded
must be the public ones, such as take place in organizations!
The example I have used made obvious the role of film technology in a redistribution
of the local. Less obvious is the role of the interior decoration of the
café, which, unlike a huge table in a beerhouse, allows the lovers to engage in an
intimate conversation. Less obvious as well is the fact that a love scene between
Reassembling the Social begins with a distinction between two sociologies. The first, the
"sociology of the social," posits the existence of society as a distinct reality in terms of which
other kinds of things (e.g., technology, religion, art) can be explained, and to which,
therefore, they can be reduced. The second, the sociology of associations, posits no specific
reality of society that lies "behind" other phenomena; society, if it exists at all, has to be
traced not as "a place, a thing, a domain, or a kind of stuff" (p. 238), but as a "provisional
movement of new associations" (p. 238), a type of "connection between things that are not
themselves social." Latour's conceit is that the ascendance of the sociology of the social-
which he traces to Durkheim-at the expense of the sociology of associations constituted a
founding movement in modern sociology. He finds an alternative to Durkheim in Gabriel
Tarde, a contemporary who charged that "Durkheim had abandoned the task of explaining
society" (p. 13). Tarde was "utterly defeated by sociologists of the social." Modern sociology
is the product of Durkheim's victory. Actor Network Theory-or at least the version of it that
Latour presents in Reassembling the Social is introduced onto this scene as the "resumption
of the [Tardian] task that [Durkheimian sociology] believed was too quickly achieved." It
aspires to do for the sociology of the social what relativity did for Newtonian physics. Just as
relativity showed that the laws of Newtonian physics were limited to certain conditions that
could be explained by a broader theory, the sociology of associations "reverses and
generalizes" (p. 12) the sociology of the social: reverses-in the sense that it tries to explain as
an effect what the sociology of the social takes for granted as a cause; generalizes-because it
offers this explanation in terms of a broader theory.
These arguments-indeed, most of what can be found in Reassembling the Social-will be
familiar to those acquainted with Latour's prior work. It bears asking, then, what this book's
purpose is, beyond undeniably artful synthesis. It is not a book that will convert new
adherents to ANT. This is so not only because the illustrative material is thin, or because the
formulations are often abstruse, but also because Latour ignores too many obvious objections
to the book's claims. He is not very careful in specifying who exactly the target of his critique
is and one guesses that-rightly or wrongly-few contempo rary sociologists would find
themselves in it. Further undermining the book's persuasive power to a disciplinary audience
is the fact that Latour barely mentions other critiques of the "sociology of the social" that can
be found in the sociological tradition. But a reader of Reassembling the Social should not be
distracted by these issues. Latour says straightforwardly and reasonably that he does violence
to other positions to present his own as clearly as possible. This is a book, he quips, written
by an ANT for other ANTs; a field guide whose purpose is not to convince anyone that the
field is of interest, but to provide a map and a set of tools for navigating it. As such, it is
certainly valuable-and richly rewards close reading-both as a statement of an intellectual
ethos and as an explication of how one goes about doing analysis in the style of ANT. The
ethos is that of a certain conceptual skepticism, a relentless worry that received categories of
the social sciences take for granted precisely what must be explained by a sci ence of society.
Latour writes that "When sociologists of the social pronounce the words 'society,' 'power,'
'structure,' and 'context"' (elsewhere he adds Capitalism, Empire, Norms, Individualism, and
Fields (p. 137)) "they often jump straight ahead to connect vast arrays of life and history" to
see in the cases at hand "examples of well known types, to reveal behind the scenes some
dark powers pulling the strings" (p. 22). Such terms, Latour argues, enable an unwarranted
"acceleration" in the description, when sociologists assign agency or causal efficacy to things
that have not been specified in the field of observation. Latour's consistent refrain is to slow
down. Sociologists must learn to "pay the full price" of explaining society by tracing the
associations that actually comprise it. The two major sections of the book lay out tools for
this disciplined skepticism. The first part is a handbook for what Latour calls "deploying
uncertainties" in order to make a sociological observer more aware of what she or he should
not take for granted in the course of inquiry: What is the relevant collective? Who or what are
the actors? How is the shape of the collective made the subject of controversy? It is crucial
that in posing these questions Latour's project is not deconstructive. Indeed, one of the most
valuable aspects of Latour's work-and the discussions in Reassembling the Social shine on
this point-is that it opens a space for inquiry rather than closes it down. In this sense, as
Latour notes, his project is utterly different from a "critical sociology" whose aim is to
unmask power relations or resolve observed phenomena to underlying forces. The aim of
ANT is to have a "closer look at the type of aggregates" that comprise collectives "and at the
ways they are connected to one another" (p. 22). Here again (in its second part) the book
offers valuable tools: follow actors as they constitute collectives, shift scales, deploy claims
about objectivity, invest worldviews, and construct agencies. As the book proceeds, and as
these tools accumulate, one cannot help feel that Latour's guidebook leaves some important
questions unanswered. As he acknowledges, it is not clear how one knows what associations
are worth tracing. "Which actors should be chosen? Which ones should be followed and for
how long?" (p. 122). Latour seems aware, too, of the dangers of an aim less empiricism "that
prides itself in being so meticulous . . . so object-oriented" that it proves "totally impractical"
because it does not know where to start, or when to stop (p. 123). When does one reach a
significant conclusion? How to define what is gained with out lapsing into the tautology that
having resumed the project of tracing associations, one will have successfully traced
associations? Latour refuses to answer such questions in substantive terms. The "task of
defining and ordering the social," he argues, "should be left to the actors themselves, not
taken up by the analyst" (p. 23). This point is formulated with particular clarity late in the
book, when Latour writes that ANT is a "negative, empty, relativist grid that allows us not to
synthesize the ingredients of the social in the actor's place" (p. 221). But where does such a
project leave us? And can it deliver what has been promised-namely, a better approach to
explaining the social? The only test offered in the book lies in the original premise-that ANT
can explain the "the views of society offered by the sociologists of the social" (p. 16). Here
Latour does offer some substantive claims that are surprising in that they seem to lean on
precisely the kind of flimsy functionalism and sociological reduction (of sociology!) that
Latour has spent an entire book criticizing. The "social" of modern sociology, he writes, was
formulated mainly as "a way of insuring civil peace when modernism was under way" (p.
16); "society," following Zygmut Bauman, "was invented to replace revolutionary politics"
(p. 250); the sociology of the social cannot trace associations because it has "had to engage in
the task of modernization" (p. 241). One could take Latour to task here for violating his own
principles of inquiry. And one could point to other more convincing "explanations" of the
sociology of the social. But it seems more productive to point out that real weight and interest
could be lent to such an explanation only through a movement from a negative and formal
analysis to substantive and historically grounded claims; from an expansive tracing of a
network to concepts that orient us-even if in a one-sided or ideal-typic fashion-to how it is
that "society" became a political and scientific problem. Such concepts (one thinks of
Polanyi's "market society" or Foucault's "biopolitics," both concerned with precisely these
questions) do not function like the cartoonish "theory" that Latour chastises, which pro poses
a "positive, substantive, and synthetic view of the ingredients out of which the social is
fashioned" (p. 230). Rather, they point us to what is distinctive and significant in a certain
assemblage, and orient us to certain directions of investigation (certain actor networks, if you
will). As outlined in Reassembling the Social, ANT offers no insight into how such concepts
are developed or put to work. As such, though it provides a valuable kit of formal tools, it
does not offer a good account of what gives substance and drive to sociological inquiry.