See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/227763501

THE TURN TO THINGS:

Article in Sociological Quarterly · March 1999

DOI: 10.1111/j.1533-8525.1999.tb00552.x

CITATIONS READS

43 220

1 author:

Alex Preda
King's College London
27 PUBLICATIONS 565 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Does a Scopic Regime Produce Conformism? Herding Behavior among Trade Leaders on Social
Trading Platforms View project

All content following this page was uploaded by Alex Preda on 08 September 2014.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
The Turn to Things: Arguments for a Sociological Theory of Things
Author(s): Alex Preda
Source: The Sociological Quarterly, Vol. 40, No. 2 (Spring, 1999), pp. 347-366
Published by: Blackwell Publishing on behalf of the Midwest Sociological Society
Stable URL: http://www.jstor.org/stable/4121238
Accessed: 14/07/2009 18:16

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=black.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

Blackwell Publishing and Midwest Sociological Society are collaborating with JSTOR to digitize, preserve and
extend access to The Sociological Quarterly.

http://www.jstor.org
 THETURNTOTHINGS:
Argumentsfor a Sociological
Theoryof Things

Alex Preda
Universityof Bielefeld

Things and artifactsare usually treatedin sociological theoryas marginal,irrelevant,or

passive with respect to the productionof social order. In contrastto that, in the past
decade the sociology of knowledge and science has developed several aproachesthat
stress the processualcharacterof the cognitive relationshipsbetween humanactorsand
artifacts. In this perspective, both human actors and things appear as active entities
involved in the productionof social order.As knots of socially sanctionedknowledge,
things shape the temporal structures,allowing for social order to be stabilized and
reproduced.From the viewpoint of the cognitive relationshipsbetween human actors
and things, the distributionand transferof cognitive propertiesand dispositions in a
networkof such relationshipsis centralwith respectto these processes. Startingfrom a
detailed examinationof these arguments,this article argues that social ordercannot be
conceived exclusively as a web of intersubjectiverelationships.It discusses the meth-
odological and conceptual implications of treating artifacts as active social entities,
arguingfor theirgeneral theoreticalrelevancewith respectto the ways we conceive the
constitutionof society.

Let this ring be accursed!

As its gold gave me power without limit
Let its spell bring now deathto its bearer!
Let no happyman delight in it
Let no fortunatesoul enjoy its shining gleam
Let whoeverpossesses it be consumed with care
And whoever does not be gnawed with envy ...
Let its possessor hoardit withoutpleasure
But let it bringthe murdererupon him ...
Let him long for death,the Ring's Lord
As the Ring's Slave ...
RichardWagner
TheRing of the Nibelungen

DirectallcorrespondencetoAlexPreda.Department of Bielefeld.P.O.Box 100131,D-33501Bielefeld.

of Sociology.University
Germany; e-mail:preda@post.uni-bielefeld.de
TheSociologicalQuarterly,Volume40, Number2, pages347-366.
Copyright? 1999by TheMidwestSociologicalSociety.
All rightsreserved.Sendrequestsfor permissionto reprintto:
RightsandPermissions,Universityof CaliforniaPress,JournalsDivision,2120BerkeleyWay,Berkeley,CA94720.
ISSN:0038-0253
348 THESOCIOLOGICAL Vol. 40/No. 2/1999
QUARTERLY

THE PLACEOF THINGS IN SOCIOLOGICALTHEORY

The picture offered by Alberich's curse is quite grim: as the dwarf finds out that the ring
giving its bearerpower over the Rhine's treasureshas been stolen from him, he casts the
curse upon the possessor of this particularobject. Needless to say, Alberich himself had
stolen the ring from the Rhine maidens.Apparently,interactingwith things was not very
commendablein those days. However,this bleak picturecorrespondsto a large extent to
the widespreadbelief thatlonging or caringtoo much for things has destructive,dehuman-
izing consequences: they can destroy the social fabric, one's moral fiber, one's feelings,
and indeed, every traceof humanity.Things can lead to the dissolutionof the social norms
thathold people together:what betterexample for the destructivepower of things thanthe
chain of thefts, lies, and betrayalsthey bring about in the Wagneriansociety? There is of
course the composer's hidden critiqueof capitalism in this passage: not accidentally,one
of the most acclaimed productionsat the BayreuthFestival staged Wagner'scycle in an
early capitalist set. Nonetheless, what strikes the reader is the counterintuitiveway in
which the relationshipbetween humansand things is put forth:one desires them, but there
is little delight; one is consumed with care, and there is more than one doubt that one
could ever really possess them. Things seem to be quite active: their simple presence dis-
solves social norms, arouses desire, makes people cheat, lie, and (sometimes) fall in love.
They are the dynamic nodes of human contact keeping the Wagneriansociety on the
move: (inter)actionsand the plot's orderare organizedaroundthem.
Literaryexamples in which things are active abound (consider the subgenresof "fan-
tasy"fiction or the Frenchnouveau romanpopularin the 1950s and 1960s).' In certaindis-
cursive practices, such as literature,things are presentedas active, complex (sometimes
rebellious) order-generatingentities. In contrast,their treatmentin some varietiesof con-
temporarysociological theory looks more contradictory.2 At firstglance, things seem to be
marginalor liminal (Turner1969): they are neitherinside nor outside the social order.For
example, a recent account of the conceptual foundations of social order and society
ascribesto physical objects an ambiguousrole: they are subjectedto the impositionof col-
lective agentive functions by human actors, as in the case where people agree to use a
bench for sitting on (Searle 1995, p. 38). More importantly,these agentive functions are
not always "performedsolely in virtue of the object's intrinsicphysical features . . . but
can be performed only in virtue of collective agreement of acceptance" (p. 39). This
investment with agentive functions by simple agreement makes artifacts crucial with
respectto the creationof institutionalfacts and hence to the social order.At the same time,
they remain outside such facts, which are a matterof agreement among human actors.
Niklas Luhmann's(1990a, p. 3) systemic theory of society seems to ban things altogether
from the realm of social systems into that of the environment,since they cannot fulfill
basic systemic requirements:information, utterance, and understanding.Artifacts and
objects of everydaylife, as well as "natural"objects, are ascribedto the "socially external
environment"and not to the social system(s) proper(Luhmann1990b, p. 419). This nicely
illustratesthe liminal position that objects occupy in such accounts:since society is to be
understood as the unity of all the differences between system and environment (Luh-
mann 1990b, p. 420), it follows that things have to occupy a place somehow. But, since
they are always "environment",it follows thatthey stay outside. JiurgenHabermas's(1984,
p. 100) world doesn't reserve a betterplace for things either:at best, they can be taken to
belong to the "objective world,"as unspecified entities about which true statements are
The Turnto Things 349

possible. But they certainlyare not partof the "social world",that is seen as the sphereof
"legitimatelyregulatedinterpersonalrelations"(pp. 88, 100).
In contrastto such views, in the past decade the sociology of knowledge and science3
has shown a sustainedinterest in the reconceptualizationof things and artifactsas social
entitiesthatplay an activepartin the generation,stabilization,andreproductionof social order
and sociality. These approacheshave a more general theoreticalcharacter,aimed at de-
fining a sociologically encompassing conceptual frame that accounts for things as active
social entities. Such argumentsaboutthe sociality of things are of interestwith respectto a
series of general theoreticalissues: the constitutionof social order,its reproducibility,the
emergence and characterof social institutions,4and the interweavingof power and cogni-
tive processes in the generationof order.
The present article outlines the conceptualpremises on which the sociology of knowl-
edge and science groundsthe analysis of "things-centered"forms of sociality. It examines
three of the points critical to the social role of things: (1) the constitutionof social order,
(2) the emergence of artefact-boundtemporalitystructures,allowing for stabilizing and
reproducingsocial order,and (3) the forms of social power engenderedby things. A cen-
tral argument,developed in the second part of the article, is that the notion of socially
active things is anchoredin a practice-theoreticalperspectivein which the social produc-
tion of knowledge plays a key role. In this sense, knowledge processes are seen as a bridg-
ing element that allows us to follow the social role of things in the three above points.
Another central argumentis that the knowledge-sociological analysis of things shifts the
focus of attention from separate, distinct entities to the processes out of which order
emerges. In the perspectiveof the continuous processes of stabilizing, reproducing,and
accountingfor social order,the sharpdistinctionsamong separateentities are effaced. The
next partdiscusses the emergence of temporalstructuresout of the cognitive interactions
between human actors and artifactsand the role of such structuresin the reproductionof
social order.With respect to the forms of social power engenderedby things, three con-
cepts are key: networks, hybrids, and translations.In the fourth part, I discuss these
concepts,theirrelationshipto the notionof temporalstructures,and the contentionthatthey
can account for an emerging form of social organizationcentered on technical-economic
entities. In this light, the conclusion outlines the directionsof inquirythat should be pur-
sued by a general sociological theory of things.

AND THINGS
SOCIALPRACTICES
Since the mid-eighties,sociologists of knowledge and science have directedtheirattention
towardshedding new light on the partplayed by engines, measuringinstruments,labora-
tory probes, detectors, and engineeringdevices in nesting and structuringsocial relation-
ships (e.g., Latour 1987; 1988; 1990; 1994; Latouras Johnson 1988; Callon 1986; Law
1993; Callon and Latour 1992; Star 1991; Pickering 1995; Rheinberger 1992; Knorr-
Cetina 1997). Graspingthe things in a scientific laboratoryis not a simple task. Sometimes
they are huge assemblagesof engines, measuringinstruments,and machinesbuilt in order
to conduct complex experiments, as in nuclear physics (Law 1993; KnorrCetina 1999;
forthcoming).They can also be large technical systems, with some mobile and some fixed
parts,implantedin and overlappingwith certainsocial sites as in the case of the ARAMIS
transportationsystem discussed by Bruno Latour(1996). But they can also be something
very small such as laboratoryprobes or paper-supportedvisualizationsof DNA structures
350 THESOCIOLOGICAL Vol. 40/No. 2/1999
QUARTERLY

(Rheinberger1992; KnorrCetinaandAmann 1990). In some cases, things are engineering

blueprints,designs, and schemes (Henderson 1991), the existence of which is paper-or
computer-supported(the lattercase makes the situationeven more complicated).Finally,a
thing may be a string of mathematicalformulas (Heintz 1993), a combinationof black-
boards, of computermemories, or of screens (Star 1989). In all these cases, they are not
determinedby theirintrinsicqualities, by some "given"or "natural"materialpropertiesas
priorand opposed to the social interactionsof the laboratory.Such oppositions,as we shall
see, are quite problematic.Neither are the things of the lab determinedby a direct, unme-
diated relationshipwith the knowing subject. Some things are too large, too complex, or
too ephemeralto be graspedin theirentiretyby one person,and, more importantly,knowl-
edge is not an individualactivity.Sometimes it is quite difficult to tell where things begin
and where they end or to distinguish between their "harder"and "softer"parts. For ex-
ample, it is hardto distinguishbetweenthe "hard"' and "soft"partsof a computer-controlled
centrifuge, where the instrument'sconcrete movements and functions are controlled and
monitoredby a piece of softwarewhose materialityis difficult to pin down.
Due to their considerablevariety and complexity,the things encounteredin a scientific
laboratorypresentthe sociologist with a series of challenges: how to accountfor theirrole
in the productionand stabilizationof scientific knowledge that is at the same time depen-
dent on local conditions and valid beyond the walls of a particularlab, and how to account
for the role things play in the social orderof the laboratory.
In meeting these challenges, the starting premise has been to "describe scientific
researchas action, and to understandscientific knowledge as implicated in and produced
by that action"(Barnes, Bloor, and Henry 1996, p. 110). In this case, one has to take into
account that the doings of the laboratoryimply not only the explicit kind of knowledge
(that which is amenableto descriptionin abstractlogical terms) but also the tacit kind of
knowledge necessary in handling and tinkeringwith the things in the laboratory.In fact,
this tacit knowledge sharedby the participantsin the setting of the laboratorymakes pos-
sible the unfolding of common activities and has conceptual and ontologic priority with
respect to the explicit norms of scientific inquiry,as well as to propositional,discursive
knowledge. If this is the case, the things of the laboratoryhave to be seen in the light of
this tacit knowledge that allows humanactors to interactwith them. This occurs underthe
premise that they are grasped as "congealed mass(es) of human labour"(Marx [1867]
1977, p. 142) and hence as materially inscribed knowledge. According to this picture,
objects of scientific inquiry (taken as instrumentsand objects of epistemic scrutiny)are
anythingbut given, at the disposal of the inquiringsubject, opposing a passive resistance
to it, or waiting to be scrutinized.What makes an instrumentor engine relevantand valu-
able in a scientific context is the fact that it can be recognized and put to work by the par-
ticipantsas a "congealedquantityof labour"(p. 150)-that is, as knowledge alreadythere.
A transferpipette is a transferpipettein a laboratorysetting not by virtueof some intrinsic
propertyor because of some higherauthorityor normbut because (1) it incorporatesa cer-
tain amountof recognized knowledge about measurementprocedures,about the physical
propertiesof liquids, aboutthe relationshipbetween pressureand volume, and so on, (2) it
requiresa certainknowledge, like how to apply pressure,and (3) it allows at the same time
the actors of the laboratoryto set courses of action. It follows that the actors in a labora-
tory put to work the knowledge at hand and thatonly partof it is humanborne.This has in
the firstplace a tacit characterand consists in the skills and abilities to interactwith things
such as measuringinstruments,laboratoryprobes, engines, and the like. Skills and abili-
The Turn to Things 351

ties, essential to scientific activities,are mundanely(re)producedin the work with material

constellations of knowledge, as a response and adaptationto them. On the one hand, we
have then certainmaterialconfigurationsof knowledge,demandingskills and abilities that
are acquired in the mundane interactions with them. On the other hand, we have the
human actors of the lab modifying and adaptingtheir knowledge to the materialat their
disposal, as well as producingand picking out new materialconfigurations,accordingto
what they alreadyknow. This echoes George HerbertMead's (1982, p. 122) observation
that things answer to the attitudeswe have formedto them and that our attitudesadjustto
things.
The tinkeringin the lab means, at a basic level, this tacit, continuousprocess of recipro-
cal adaptationof practicalskills and materializedknowledge.At a furtherlevel, it means
being guided by and using the given epistemic resources in the successful productionof
knowledge. These artefactualepistemic resources define the local setting through their
spatial arrangementsthat constrainthe actions of human actors; as Mead (1982, p. 119)
noticed, distanceand spatialconfigurationis a matterof responsesand contactsthat define
the possibilities for action. Moreover,artefactualresources can also enlarge the zone of
cognitive operation (Schutz and Luckmann 1973, p. 44) and therefore determine the
human actors' possibilites for action. For example, some studies of laboratorysettings
have stressed their apparentlycrampedcharacter,where various instruments,measuring
devices, and other things are piled on top of each other in a seemingly disorderlyfashion
(e.g., Latour and Woolgar 1979). This is neither due to some lack of space nor does it
reflect some disorderlinessinherentin scientific work. On the contrary,the lab's spatial
configuration is very ordered and precise, determined by the frequency with which
researchersput to work certain skills and abilities incorporatedin artifacts.Instruments
and devices are arrangedaccordingto the experimentsbeing performed,to theirsequential
order,and to how the instrumentsrelateto each otherand to the humanactorsin the course
of these experiments.The lab is constituted as a configurationof cognitive interactions
between humanactorsand things. Distances and positions are groundedin cognitive rela-
tions and actions, providing a "distinctivephenomenalfield(s) in which organizationsof
'work'are establishedand exhibited"(Lynch 1993, p. 132).
If the possibilities for action are defined by epistemic resources,then the generationof
social orderis coextensivewith the processof knowledgeproduction.The social orderof the
laboratorycannotbe seen as a kind of external,given framewith respect to the production
of knowledge. Since the starting point of this process is given by the practicaldoings
through which human actors adapt their skills to the knowledge at hand, modify this
according to their own abilities, and at the same time coordinate each other's doings
accordingto adaptationprocesses, it follows that a contextual,stable, reproduciblesocial
order is achieved in (and not outside) the productionof knowledge. Seen in this perspec-
tive, the things of the lab appearboth as the fundamentalresourcesof this specific formof
sociality and as setting up the rules by which action unfolds. They are neitherready-made
for scrutinynor passive instruments:this means that they have to be practicallyconfigu-
rated,accordingto the said conditions, to the epistemic resourcesof the humanactors but
also accordingto those incorporatedby the other things presentin the lab. At times, things
can be also configurated,modified,or rearranged,accordingto which cognitive resources
are out of order. In her ethnographyof a laboratory,Karin Knorr-Cetina(1981, p. 35)
shows how a broken lab instrumentled to a cognitive redefinitionof other lab devices,
startingfrom the skills at hand:scientistsused a centrifugeas a replacementfor the broken
352 THESOCIOLOGICAL Vol. 40/No. 2/1999
QUARTERLY

instrument,although this was not their usual way for performingmeasurements.In this
case, the measurementhad to be redefinedaccordingto the skills and abilities requiredby
operatingthe centrifuge and according to the kind of knowledge (e.g., about separating
substancesand densities) incorporatedin a centrifuge.Tinkeringaccordingto the available
materialknowledge led in this case to an ingenious redefinitionof the scientists'problem.
The practicalachievementof things, indistinguishablefrom the process of knowledgepro-
duction and from the productionof an accountablesocial order,means thatobjects are not
only "'technically' manufacturedin laboratoriesbut are also inextricablysymbolically or
politically construed,for example throughliterarytechniques of persuasionsuch as one
finds embodied in scientific papers,throughthe political stratagemsof scientists in form-
ing alliances and mobilizing resources,or throughthe selection and decision translations
which 'build' scientific findingsfrom within"(Knorr-Cetina1992, p. 115).
This entails shifting the analyticalfocus from two distinct and separateentities-active
humanactors and things opposing a passive resistanceto epistemic inquiries-to the pro-
cessuality at work in generatingand reproducingthe social orderof scientific laboratories
as "collective units which encapsulatewithin themselves a trafficof substances,materials
and equipment,and observations"(Knorr-Cetina1992, p. 128). Such a shift in focus, from
"final productsto production,from 'cold' stable objects to 'warmer'and unstable ones"
(Latour 1987, p. 21), reveals how things are integratedinto the social orderof the labora-
tory as knots of socially sanctioned and active knowledge and how they help shape and
stabilize the social orderof knowledge production.In his study of Louis Pasteur'slabora-
tory, Latour(1988, p. 81) shows how small things and instruments,such as the pipetteand
the culturemedium, played a decisive role in defining the specific orderof the lab "outof
the movement and displacement of other places and skills," thus enabling the unfolding
of medical knowledge. For example, building up an adequateculturemedium meant first
learninghow to make a brothresemblingcooking stock and then a good gelatin that was
essential for growing the anthraxmicrobe in vitro. These were cooking skills, which had
to be translatedand reembeddedin the context of the laboratory,and that afterwardalso
enabledthe translation(and transformation)of the microbesthemselves-with these latter
becoming a scientific fact, visible evidence of the disease, and so on. The culturemedium
representednew practical skills integratedinto the social order of the lab, skills that in
theirturnreshapedthe skills of laboratoryworkersand researchers.
In some cases, this practical codependency between knowledge embodied by the
researchersand knowledge incorporatedin the instrumentsof researchbecomes so great
that it is difficultto distinguishwhere one ends and the otherbegins, as in the case of sys-
tems of distributedartificalintelligence or of complex engineeringsystems, "monstruous
group(s)which regulateinteractionbetween graphiterods, turbines,atoms,operators,con-
trol boards, flashing lights, concrete slabs and engineers" (Callon 1991, p. 141). Latour
(1993, p. 11) calls this practicalcodependency a "work of hybridization,"that is by no
means limited to complex technical systems but is rathera general characteristicof the
social order.

STRUCTURES
THINGS, KNOWLEDGE,AND TEMPORALITY
This view of things as concrete forms of social knowledge, as processes enmeshed with
other processes embodied by human actors, raises the question of the way they relate to
the temporalstructuresof scientific knowledge production.By definition,a key featureof
The Turnto Things 353

such processes must be their temporal dimension, their ability to unfold over time, to
develop their own rhythms.At the same time, they structureand stabilize the social order
(of the laboratory),which means thatthey have to play a role in the temporalitystructures
intrinsicto social order.
One argumentabout how things generatetemporalstructureshas been formulatedby
J6rg Rheinberger(1992) and KarinKnorr-Cetina(1997). They arguethat, while constitut-
ing a basic epistemic resource of the lab, instruments,experimentalprobes, and engines
are also continuously modified, adapted, or expanded, according to the constraintsand
demands of the knowledge constellation in which they are embedded.The things of the
laboratoryincorporatea knowledge that is already there; at the same time, they are
achieved as new knowledge in the process of scientific inquiry.A pipette, for example,
incorporatesan alreadyavailableknowledge about the relationshipbetween pressureand
volume of liquids. Concurrently,when being used in new ways in new experiments,a
pipette can unfold new knowledge. Therefore, things should be seen simultaneously as
given knowledge and as knowledge-in-the-making.Being knowledge-in-the-making,they
are active in the sense that they are part of the continuous processes that reshuffle the
knowledge bases of a specific form of sociality. One example of such open-ended,always-
in-the-makingknowledge is providedby software codes that are never actually finished;
we have only temporaryvarieties and versions of such codes, certain to be replaced or
upgraded.
This reshuffling aspect of epistemic things is perhapseven more evident in domains
such as softwaredesign or computer-aidedengineeringthat rely on boundaryobjects (Star
1989; Henderson 1991; Fujimura1992). Engineeringblueprints,blackboarddrafts,maps,
and sofware codes are continuously reshuffled-they never really get done once and for
all. This allows them to create and combine differenttime horizons (Star 1989, p. 47) and
to keep togetherscientists that may have differentgoals, differentunits of analysis, or dif-
ferentaudiences.An example of boundaryobjects is atlases of the brain(Star 1989, p. 49),
which are visual representationsor maps of the brain'ssections and regions, used in both
basic researchand brainsurgery.These maps combine the differenttime horizonsof brain
surgeons and neurophysiologists, while connecting these user groups with their own
specific goals and practices.What keeps the human actors of a scientific lab together is
exactly the fact that they are embeddedin an ever-changingflow of knowledge: the tasks
they set to their common work never really get to an end. Scientific groups and communi-
ties, althoughtask bound,do not fall apartwhen these tasks are achieved.The achievement
is not final, and the scientific activity is open-ended, thus strengtheningthe cohesion of
such communities.In the common, practicalacquisitionof skills and abilities in interact-
ing with open-endedthings, these latterappearas a self-expandingepistemic resourcethat
imposes its own rhythmon human actors and generates two tightly interwovenkinds of
knowledge: (1) knowledge as an end result and (2) reflexive, processual knowledge.
Andrew Pickering(1995, p. 21) has designatedthis as "thedance of agency,"by which he
means that laboratoryinstrumentsand objects of inquiryplay an active partin the unfold-
ing of this process of knowledge. With the reactions of lab instrumentsbeing not very
foreseeable most of the time, humanactors have to "retreat"while machines process and
produce some material results--dots of light, patches of color, patternsof lines. With
these, machines force human agency to come into play, in that lab actors have to revise
their interpretationsand actions, as well as devise furtheractions accordingto the material
results produced by machines (Pickering 1995, pp. 22-27). In having to wait for mate-
354 THESOCIOLOGICAL QUARTERLYVol. 40/No. 2/1999

rial results, humanactors "encountera time structurethat is imposed on [them]"(Schutz

and Luckmann 1973, p. 48). One has to note here that this more general point is also
present in Marcel Mauss's discussion of gift exchange, as well as in Jacques Derrida's
redeploymentof this discussion. Startingfrom Mauss's contentionthatthe phenomenonof
gift exchange implies the reciprocaldeterminationof the "law of persons"and the "law of
things"(Mauss [1925] 1954, pp. 48-49), Derrida(1992, p. 40) notices thatthe doings of the
humanactors are embeddedin the temporalstructuresof the relationshipsbetween things
(a gift requiringa certainkind of countergift,aftera certainlapse of time, in the context of
a certain kind of presentation):"The thing itself demands gift and restitution,it requires
therefore 'time', 'term', 'delay', 'interval'of temporization,the becoming-temporization
of temporalization,the animationof a neutraland homogeneous time by the desire of the
gift and the restituion.. . . This demandof the thing, this demandfor term and temporiza-
tion, would be the very structureof the thing."
Temporizationand temporalizationthus become centralfeaturesof epistemic practices.
The first means that in the practicalwork with artifacts,the achievementof once-and-for-
all results or knowledge is continuouslydeferredand that the actual achievementof this
practical work is the process itself, as the source from which knowledge unfolds: "The
movementof significationis possible only if each so-called presentelement, each element
appearingon the scene of presence,is relatedto somethingotherthan itself, therebykeep-
ing within itself the markof the past element, and alreadyletting itself be vitiated by the
markof its relationto the futureelement, this trace being relatedno less to what is called
the future than to what is called the past, and constitutingwhat is called the present by
means of this very relation to what it is not" (Derrida [1972] 1990, pp. 65-66). These
deferralsactually both stabilize and align the social order of the laboratoryto the larger,
outside order (Pickering 1995, p. 182). In this context temporalizationmeans that knowl-
edge processes surpassmomentarinessin this way: in the process of postponement,they
gain a stable, temporaldimension.A key consequence is that epistemic practicesbecome
temporallystable and acquirethe lusterof coherence.
John Law (1993) and KarinKnorr-Cetina(1999) have also turnedtheirattentiontoward
how complex machines-such as those used in large experimentsin the field of particle
physics--organize humans actors and act as open-endedepistemic things. In big experi-
ments, on a large time scale, the objects of inquirydo not appearas fully achieved. For
example,manyexperimentsin nuclearphysics, aimingat detectingor tracingthe behaviorof
subnuclearparticles,are done and redoneover manyyears and involve the large-scaleorga-
nizationof hundredsof scientistsaroundhuge, complex engines. In such cases, there is no
talk abouttracinga stable, unchangingentity but ratherseveral varietiesof the same "par-
ticle",which, in its epistemicdimensions,dependson these reflexiveprocessesof reshuffling
and unfoldsover long periodsof time. A similarexample is the case of largetechnicalsys-
tems encounteredin everydaylife, such as the computerizedurbanrail system studiedby
Latour (1996). Such a system implies the social organizationof engineers, technicians,
workers,and passengersarounda huge, dispersedengine with some fixed and some mobile
parts.The engine nevergets done once and for all, since it implies continuousmaintenance,
monitoring,repairs,extensions,re-arrangements, and so on. In this perspective,the engine
representsa constellationof processual forms of knowledge, creating a stable temporal
dimension,bindingandorganizinghumanactorsaroundit. The formsof sociality specific to
scientific inquiry-and, more particularly,the communitarianfeaturesof scientificwork in
largeexperiments-have the interactionwith responsive,"active"thingsat theircore.
The Turnto Things 355

However, it would be inappropriateto think that only big, technically complex

machines can structurethe temporalorderof interactionsand activity.In his ethnography
of a residential ward for mentally handicappedteenagers, Ruud Hendriks (1997) shows
how the small things of everydaylife, such as rugs, benches, and egg timers, structurethe
temporalsequencesof interactionfor both residentsand healthworkers.If we take the case
of simple, apparentlynonproblematicdoings, such as eating a hamburger,we are still con-
frontedwith the problemof achievinga temporalsequence-that is, of beginning,continu-
ing, and finishing the hamburger.It is the socially accountable achievement of this
temporalsequence that makes the doing a meaningfulaction. This temporalsequence is
achieved by the ward'sresidents with the help of props such as an egg timer,which order
the way they relate to their food (e.g., the hamburgeris finished after a complete rotation)
and to their bodily movements (like taking a bite each couple of ticks). Mental health
workersrelatethemselves to the doings of the residentsvia things like egg timers,in order
to manage and stabilize a temporalorderingfor the ward. Thus, materialobjects are no
longer seen "as passive means in the hands of the counsellor, as neutralservice-hatchfor
meanings to be passed through,but as nodes aroundwhich other elements (counsellors
and residents,words, gestures, organizations,things, bodies) are temporallycentred.Tem-
poralorder,then, is seen no longer as an elusive concept, to be found solely in the heads of
the counsellorsor in the linguistic community,but as a resultof the ways in which hetero-
geneous elements in their mutual interactionperform agency in temporal affairs. As a
derivativeof the (partly)materiallabourof ordering"(Hendriks1997, p. 252).
In other words, objects of knowledge (and large, complex ones even more so) bind
human actors and participatein developing specific forms of social order-because they
allow for common practicesto develop, stabilize, and structuretime. This position echoes
MartinHeidegger's ([1927] 1977, p. 93) stance that humanreflexive practices(and hence
"Being")arise in the everydaycare for objects, in tinkeringaroundthem, and in trying to
respondto theirchallenges.
These argumentspoint to the fact that at least some of the things at work in social set-
tings have an open-ended, processual characterand contributeto generatingtemporality
structuresthat ensure the coherence and stability of social order.Through a continuous
deferralof the final tasks and throughgeneratinga multitudeof sequential,partialtasks
that have to be reiteratedall the time, such epistemic engines reinforceand reproducethe
social bonds of the community:they allow for the humanactorsto perceive theiractivities
as common, reinforcedby affective bonds, and directed towardcare and stability.In the
case of multinationalscientific collectives, they allow for the membersof the community
to develop "a culture of no culture"(Latour 1990, p. 166)-that is, to achieve cohesion
and coherence in spite of their nationalor local particularities,of the differences in social
traditionsand customs, social background,language,and so on.
Thus, we are provided with different modes in which things constitute the temporal
structuresof social order:(1) Temporalstructuresare groundedin the cognitive proces-
suality of things and of the accountable doings of human actors. These structuresare
achieved in the orderedinteractionsbetween humanactors and things. (2) Being proces-
sual, epistemic things create an open temporalsequence that never closes (or closes only
over large time spans). This allows for goals, tasks, and interests to be adaptedto this
structureand then temporized.Such open-endedstructuresenhance social cohesion, as in
the case of scientific communities centered around complex things or constellations of
things. (3) Since the things of the lab and the humanactors are codependentin the practi-
356 THESOCIOLOGICAL Vol. 40/No. 2/1999
QUARTERLY

cal skills they incorporateor embody, respectively, it follows that they have to wait for
each other and to adaptto each otherthrougha "danceof agency."The continuousprocess
of reciprocaladaptationstabilizes social orderand providesit with continuity.(4) This sta-
bilization of social orderdoes not imply its identical reproduction,which would be prob-
lematic anyway. Rather, the open-endedness of temporal sequences together with the
temporization implied by reciprocal adaptationallow for rupturesand epistemic rear-
rangements,as in the cases of scientific discoveries and controversies.

THINGS, KNOWLEDGE,AND POWER

If things, seen as materializedknots of practicalknowledge, play an active role in generat-
ing structuresof social order and if we accept Michel Foucault's (1979) thesis that the
knowledge processes underwritingsocial orderare also entangledwith structuresof social
power, then we may rightly inquire about the part played by things in such power struc-
tures. Starting from the premise of things as knots of practicalknowledge that interact
with those embodied in humanactors, Latourhas outlined how things are entangledwith
such structuresof power. His argumentsunfold from three issues: (1) the operations of
classification, as well as the categories throughwhich humanactors achieve specific dis-
tinctionsbetween themselves and things, (2) the conceptualpossibilities for humanactors
grasping things, and (3) the social operations of transplantingthings between different
locales or "labs",and their more importantconsequences.
As for the firstissue: the asymmetryof the categories "human"and "nonhuman"(ech-
oed in the fundamentaldistinctionsbetween society and nature,humansand things) is not
self-evident, given or naturalbut is ratherthe productof socially sanctionedclassification
operationsand categories, and thereforean epistemic operationto be examined in its own
social character.This argumentcontinues the stance adoptedby Emile Durkheim([1915]
1947, pp. 147-148), according to which the distinction humans/nonhumansis achieved
throughpractical operations of classification. Therefore, one has to examine their emi-
nently practicalcharacter(Latour 1988, p. 186). Consideringthat laboratoryartifactsare
materialknots of knowledge, these operationsare seen as performedby the practicalskills
required in getting on with things. On the one hand, artifacts are certain materialized
skills and knowledge and requirefrom human actors complementaryskills and knowl-
edge. On the other hand, both humanactorsand artifactscan meaningfullyact (and work,
respectively) in a system of differences among kinds of (tacit) knowledge. A door, for
example, materializesa certainpracticalknowledge (aboutbuildings)and requirespracti-
cal knowledge and skills from those who wish to open it (Latouras Johnson 1988, pp.
299-300). At the same time, a door can be graspedas materialknowledge only if we take
into account the differences between the work it performs and the work performedby
alternativematerialknowledges (like thatof walls, or of holes in walls) (Latouras Johnson
1988, p. 306). If we see both artifactsand humanactors as (different)bearersand produc-
ers of knowledge, dependingon each other and acting/workingonly in this system of dif-
ferences, then their positions with respect to each other are the result of practical
operationsof classifying throughthe ascriptionand enactmentof skills and tacit knowl-
edge. At the methodological level, however, this questions the assumption that humans
and artifactsmust be approacheddifferently.
The conclusion (Callon and Latour 1992, pp. 348-349) is that methodological asym-
metry is not justified:one can study the social processes that generateepistemic asymme-
The Turnto Things 357

try with symmetricmethodologicalassumptions(Latour1993, pp. 94-95). Methodological

symmetryis not to be confused with ontological symmetry.Ontological symmetrywould
imply some form of animism-that is, ascribingintentions,aims, and purposefulaction to
artifacts;methodological symmetryimplies analyzing both human actors and artifactsas
generatorsof practical knowledge. Thus, the methodological assumptions that apply to
humanactorswould also apply to things and therewould be no methodologicalreasonfor
treatingthe two asymmetrically.What the sociologist has to do is to follow the attribution
of propertieson both sides of the divide (Latour1993, p. 96); seen as a practicalwork, this
attributionconstitutessociality's very core. This position groundsthe approachto human
and nonhuman participantsin knowledge processes as "actants"-that is, as entities
embedded in practicalconfigurationswhose interactionsgenerate(scientific) knowledge.
These networksof humansand machinesare stabilizedover time and tend to expand.One
such example of practicalconfigurationsis found in huge technical systems, like the trans-
portationsystem discussed by Latour(1996). One question remains,though:while a sym-
metricalmethodology may apply, social order is still consistently and successfully taken
by humanactors as asymmetric.Therefore,one would have to question-with a symmet-
rical methodology-the socially successful productionof asymmetry,or what Pickering
(1995, p. 245) has termedthe achievementof "performativeincommensurability."Bruno
Latourand Michel Callon (Latour1993; Callon and Latour 1992) take asymmetryonly as
a historicalobliteration-as the "GreatDivide" of moderntimes. This divide is partlythe
resultof a paradigmshift in the way we conceive the relationshipbetween knowledge and
politics (Latour 1990) and partlythe result of social forces that in the process of modern-
ization obscure the role played by things, makingthem appearas receptiveto the will and
intentionsof humanactors.Withoutgoing into the details of Latour'sargument,one must
notice here the shift from a methodologicalto a historicaltopic-that is, from the general
methodological assumptions in examining social practices to the question of whether
modernityis markedby such an epistemic rupturewith respect to things.
In direct relation to the aspects discussed above comes the second question posed by
the actor-networkapproach,about the conceptual possibilities for grasping things. If the
epistemic basis of the social order is given by establishing systems of differences and by
the practical, nondiscursive,materialoperations of classification and category building,
then things cannot be given: they must be assigned a place and marked as such. This
applies not only to discursivepracticesbut also to materialpracticesof markingor inscrib-
ing objects. Moreover,artifactsinscribe or leave traces on other things, thus modifying
them; scientific knowledge is in many cases generated in such processes of inscribing
(Callon and Latour 1992, p. 351). For example, measuringinstrumentsinscribe human
actors with skills and abilities requiredto operatewith them; most scientific practicesalso
consist in producing inscriptionsof objects of inquiry that otherwise cannot be investi-
gated-such as the inscriptionsor traceselectrons leave in the bubblechamber(Pickering
1995, pp. 38-39). Inscriptionsare then not just about artifacts but also constitutive of
them. And because inscriptionscan work only in a wider context, then inscribed things
send to each other and defer meaning continuously.If we come to see both humanactors
and artifactsas configurationsof practicalknowledge (Latour 1988, p. 179) whose rela-
tionships are groundedin a processual, tacit work of reciprocalinscriptionand modifica-
tion, then we have to grasp particularforms of sociality, or "tribes"(Latour 1993, p. 101)
as dynamic constellations or networks,primarilystressing their characteras process and
less as the "nature"of theircomponents(Latour1988, p. 171).
358 THESOCIOLOGICAL Vol. 40/No. 2/1999
QUARTERLY

Finally, there is the question of how artifacts(incorporatinga particular,situatedkind

of knowledge) are translatedfrom locale to locale. The firstlevel at which translationtakes
place is that of translatingmaterialknowledge into human skills and abilities and vice
versa (Latouras Johnson 1988, p. 299). This kind of translationdisciplines the members
of an epistemic constellation(p. 300). At a second level, translation(p. 306; Latour 1990,
p. 154) designates the transferof practicalscripts (or knowledges) onto materialbearers
(e.g., when transferringthe practicalknowledge from a lab assistant onto an automated
transferpipette). Such "immobilemobiles" are transferredfrom locale to locale, bringing
with them not only their knowledge but also their disciplining power. They mobilize and
enlist human actors in their materially "frozen" knowledge. When "critical mass" is
attained-that is, when there have been enough transfers (Latour 1990, p. 163)-the
epistemic constellationthat Latour(1988, p. 171) calls "a networkof entelechies"attracts
in its orbitmore and more knowledgebearersby virtueof its mass, a process thatoccurs at
the expense of other epistemic constellations.Seen in this perspective,scientific struggles
are simply power struggles-it is only thatthey aren'tfought by individualsubjectsbut by
such networks. Conversely,these power struggles are inextricablylinked with epistemic
processes and with science, since the generation, stabilization,and reproductionof net-
works implies absorption(or destabilization)of othernetworks.Knowledge and power are
intimatelylinked, both in the way knowledgeis generatedin networksand in the ways net-
works enlist outsidersand expand.Artifactsplay a central role in this process, since they
inscribe and modify other objects, bringingthem into the net; they "attach"humanactors
to the net; and they are instrumentalin power struggles, where the "wealth of objects"
(and hence of knowledge) counts. Hence the thesis that knowledge and politics "reverber-
ate in each other" (Callon and Latour 1992, p. 364). This is precisely Foucault's thesis
(1976, p. 120), which Latour(without particularlyacknowledgingit) seeks to expand by
enlisting things and artifactsamong the active dispositivesof power.
In this perspective,the model of power/knowledgepresents two essential features:(1)
continuous expansion and accumulationand (2) the tying together of human actors and
things in stable, hard-to-breakcognitive relationships.On the one hand, the largerthe net-
work with its objects, the strongerits force will be, and hence its authority,legitimacy,and
power. On the other hand, the core elements of this networkare not simply human actors
plus things but the epistemic relationshipsbetween them. These enacted relationshipssta-
bilize and reproducesuch networks. Such epistemic relationshipsare simply "hybrids,"
meaningthatthey align and tie togetherhumansand nonhumans.
This view of social power as accumulationand hybridizationis a cornerstonefor under-
standingthe rise of "modemrn" societies and of capitalism."Capital"is just an application
runningon this platform(Latour1988, p. 172). It consists in a network-likeaccumulation
of knowledge aboutthe exchange of equivalents,which has alreadyattainedcritical mass;
therefore,"the marketplaceis only a consequence of the establishmentof networks"(p.
172). Moreover,what we call "modern"societies arejust collectives (of humanactorsand
things) that have managed to attain this critical mass, while "nonmodern"ones retain a
subcriticalmass (Latour 1993, p. 107). This model is groundedin a refinedidea of accu-
mulation,whose object is the process of epistemic relationshipsand not inert "wealth."It
also supports the idea that "modern"capitalist societies are affluent ones: however,
epistemic abundanceis not simply tied to amassingthings but ratherto the notions of net-
works and hybrids. Networks of epistemic relationships unfold a mundane wealth of
knowledge, the most importantof which isn't explicit; they are also tied to temporalstruc-
The Turnto Things 359

tures that allow for (relative) stability. Since these temporal structuresare open-ended,
they allow for an increasein the networks'complexity and for furtheraccumulation.Also,
hybridalignmentsof humansand nonhumansgrow more complex by includingever more
heterogeneouscomponents,which means that the knowledge they achieve (and requirein
their interactions) grows more complex and heterogeneous. This growing complexity
enables them to absorbeven more knowledge.
For Donna J. Haraway(1997, p. 12), the effect of such hybrid alignments is that of
"technobiopower":"distributed,heterogeneous, linked, sociotechnical circulations that
craft the world as a net called the global."Since "material,social, and semiotic technolo-
gies" bring human and nonhumanactors into alliance, and since this alliance leads to "a
condensationin space and time, a speeding up and concentratingof effects in the webs of
knowledge and power"(pp. 50-51), then technologies of knowledge (material,social, and
semiotic) constitute the lines of force along which society is organized.With respect to
that, economic processes would appearas epiphenomena,made possible by and relying
heavily on such technologies: "Acceleratedproductionof naturalknowledge pervasively
structurescommerce, industry,healing, community,war, sex, literacy,entertainment,and
worship.The world-buildingalliances of humans and nonhumansin technoscience shape
subjects and objects, subjectivityand objectivity,action and passion, inside and outside in
ways that enfeeble other modes of speaking about science and technology. In short,tech-
noscience is about worldly, materialized,signifying and significantpower. That power is
more, less, and otherthan reduction,commodification,resourcing,determinism,or any of
the other scolding words that much critical theory would force on the practitionersof sci-
ence studies, includingcyborg anthropologists"(p. 51).
Accumulationand expansionof epistemic networks,therefore,also mean an expansion
of such "hybrids",which come to dominate economic processes: "Impurity,then, is the
rule. Nowhere is this more visible than in the service sector. The product sold by Club
Med, Cap Sogeti or CISI is a mixture of humans and non-humans,texts, and financial
productsthathave been put togetherin a precisely coordinatedsequence. Considerwhat it
takes for Mr.Smith to be able (and willing) to spend his holiday on the banksof Lake Ran-
guiroa, watching the Barracudasmingle with the tanned bodies of his fellow-humans.
Computers,alloys, jet engines, researchdepartments,marketstudies, advertisements,wel-
coming hostesses, natives who have suppressedtheir desire for independenceand learned
to smile as they carry luggage, bank loans and currencyexchanges-all these and many
more have been aligned"(Callon 1991, p. 139).
Callon takes over Latour's argumenthere, namely, that the marketplace(and market
exchange) is just anothernetworkof cognitive relationshipsaboutthe exchange of equiva-
lents; he pushes this idea further,with the argumentthat economic exchanges are made
possible, dominated, and controlled by hybrid epistemic formations.The exchange and
consumptionof commodities (such as holidays and travels,but not only) is conditionedby
knowledge-basednetworksthatmake possible the constitutionof a commodity as a practi-
cal alignmentof humanactorsand things. Such alignmentsrequirea good deal of practical
knowledgeon the partof the consumer,abouthow to enterrelationshipswith the different,
humanand nonhumancomponentsof such a network.Commoditiesare dependenton net-
works and, in fact, are constitutedin and by them: we can hardlyimagine a holiday pack-
age (or, for that matter,any other complex, technically sophisticatedcommodity) outside
such a networkof computerbookings, rapidtransportationengines (airplanes,cars, etc.),
computerizedmoney transfers,advertisingbrochures,and so on. But consumers too are
360 THE SOCIOLOGICAL QUARTERLYVol. 40/No. 2/1999

constitutedin and by such networks:in order to be able to access such a travel package,
one needs several kinds of practicalknowledge, from how to book flights to driving off-
road vehicles, making transferpayments, comparing travel brochures,and much more.
Such constellations of hybridrelationshipsare suffused with social power that cannot be
located at any precise point or in any precise object or engine. Whatmakes them so power-
ful is that these relationshipsregulatehumanactors' access to things. These relationships
constitutethe lines of force along which social life is organized.
This becomes more clear in Callon's (1991, p. 138) redefinitionof money in terms of a
network:"Forinstance,when venturecapital funds research,this is based on a programof
action, which acts as counterweightto the loan. In this money is textualised,translatedinto
orders, indicatorsand recommendations.These define and link a range of heterogeneous
humanand non-humanactors:cooperatewith X at ICI andY from LaboratoryZ to obtain
criticaltemperatureof 150' K and you will get a loan of $A. Here again the intermediaryis
a networkof roles."This strategyis meant (1) to redefineeconomic exchange in cognitive
terms, (2) to connect economic exchange, science, and technology on the groundof these
cognitive terms, and (3) to redefineafterwardthe connection between scientific action and
economic exchange in terms of power/knowledge.If marketexchange (and, more gener-
ally, any economic process) is made possible by an epistemic network in which human
actorsand artifactsare aligned along cognitive processes (thatare also lines of social force
or power), then it becomes possible to describe economic processes in terms of this net-
work and its alignments.Moreover,if networks5underwritingeconomic processes imply
(1) complex alignmentsof humanactorsand artifacts.(2) the stabilizationand maintaining
of technological dispositives, (3) the open-ended productionof knowledge, (4) the code-
pendency of humans and artifacts,then we can describe the processes throughwhich vari-
ous entities (or, in Latour'sterms, the entelechies of the network)are translatedinto each
other.The processes throughwhich instruments,machines, lab results, substances,natural
elements are translatedinto money (or the other way around)are such cases. Finally,these
translationprocesses have to be seen in terms of power/knowledge,since they (1) unfold
along the essential lines of such a network,(2) play a centralrole in the processes of cogni-
tive accumulation,and (3) expand the network.This would provideus with a platformfor
analyzinghow science, technology,and economic processesareclosely imbricatednot only
at the level of the macrosocialorganizationbut, much more profoundly,at the mundane
level of the cognitive processes thatshape and structuresocial order.

CONCLUSION
The impact of the theoretical and methodological argumentsdiscussed above goes well
beyond the domain of the sociology of knowledge and science. Although supportedby
empirical examples and analyses from the domain of science, such argumentsactually
concernthe possibility of devising a sociological theory thatdoes not place artifactson the
marginsor in the environmentof the social world but accounts for them as social entities:
"We have never been interested in giving a social explanation of anything, but we want to
explain society, of which the things, facts and artifacts, are major components"(Callon
and Latour1992, p. 348). Thus, these argumentsplace themselves in a Durkheimiantradi-
tion, by treatingartifactsnot as given and inert, as simple clumps of matterlying around,
but as social facts. Two elements are essential in approachingthe sociality of things: (1)
the tacit knowledge they incorporate,seen in its codependency with the tacit knowledge
The Turnto Things 361

embodied by human actors, and (2) their processualcharacter,where we might see them
as partof achieving and stabilizing social order.Since tacit knowledge and processuality
are centralto social practices,conceived as "temporallyunfolding and spatially dispersed
nexus(es) of doings and sayings" (Schatzki 1996, p. 89) in which an accountableorderis
achieved, then artifactsare active, integralpartsof such nexuses.

ExplainingSocial OrderthroughSymmetryand Asymmetry

Conceiving artifactsas active social entities does not preclude the fact that social prac-
tices are consistently and successfully takenby humanactors as asymmetric.The suppo-
sition of asymmetry shared by human actors isn't necessarily reducible to a "romantic
assumption . . . of the essential and a priori distinction between humans and all other
kinds of entities that exist" (Ashmore, Wooffitt, and Harding 1994, p. 734). It requires
ratheran empiricalexaminationof how it is enacted by humanactors in the productionof
social order.Therefore, any sociological theory of things is confronted with two inter-
relatedissues: (1) the issue of methodologicalsymmetry-that is, of describingthe proces-
suality and shared cognitive grounds of human actors' doings and of artifacts in a
network (this issue is not to be confused with ontological symmetry, with ascribing to
artifacts intentions, a soul, etc.); and (2) while a principle of methodological symmetry
applies, we are confronted with the issue of the successful productionof asymmetry-
that is, of the strategies and processes throughwhich human actors take the social order
as asymmetric, take themselves as radically different from artifacts, and their perfor-
mances as incommensurablewith those of artifacts.If the asymmetryof social orderand
social practicescannot be taken as given or "natural"any more, this does not mean that it
is a nonissue. Rather,emphasis should be put on strategiesand resources throughwhich
human actors manage to account for a social orderin which they take themselves as dif-
ferent with respect to the artifacts to which they are related. One subordinatedissue
would be to examine the zones and domains of humanactivity in which this supposition
and the strategies of asymmetry are suspended, such as certain literary genres, and to
analyze and describe the social mechanisms that allow for such partial, local, and con-
trolled suspensions. Therefore,applying the principle of methodological symmetrydoes
not exclude but ratherrequiresa systematic clarificationof how social order is accounted
for by humanactors as asymmetric.
While Latour (Latour 1990, pp. 155-156) emphasizes methodological symmetry,he
seems to view the second issue only in termsof a historicalobliterationthat begins in the
seventeenthcentury,with inventing"a political discoursewhere politics should not count"
(science) and "a scientific politics where experimentalscience should not count" (poli-
tics): "From now on every one should 'see double' and make no direct connection
between the representationof non-humansand the representationof humans,between the
artificialityof the facts and the artificialityof the Body Politic."However,"simultaneously
showing the coproductionof society and nature... secretedas by-productsof this circula-
tion of quasi-objects"(Callon and Latour1992, p. 349) also means showing the coproduc-
tion of differences between things and human actors, which are constitutive to social
order.The second issue discussed above cannot be put only in terms of a historical"acci-
dent",but requiresa more generalclarification,that a sociological theory of things should
pursue.
362 THESOCIOLOGICAL Vol. 40/No. 2/1999
QUARTERLY

Exploringthe TemporalStructuresof Technical-Economic

Networks
One aspect central to such a theory, and closely related to the above, is the following: if
artifactshave to be seen as processes and as knots of socially sanctionedknowledge, then
they mold the temporalstructuresof social order.This takes the form of reciprocaladapta-
tion of differenttemporaldynamics, as in the case of boundaryobjects, of a temporization
of social action, as in the case of the "danceof agency,"of temporalization,or of complex,
open-endedepistemic things. In each of these cases, the stress is put on the continuityand
temporalstabilizationof the social order,which allows for its reproducibility.The model
here is thatof an artifactor constellationof things upon which social action is centered.At
the same time, we encounterin practicehybridconcatenationsof human actors, engines,
inscriptiondevices, texts, that are heterogeneouswith respect to the temporalstructures
they generate.This surfacesin the case of the technical-economichybridsapproachedby
Callon: seen in this light, travel services, for example, are concatenationsof hybridsthat
generatedifferenttemporalstructuresfor the varioushumanactorsinvolved in them. I will
illustratethis problemwith an additionalexample. Eurotunnel,the submarinetunnel link-
ing Franceand GreatBritain,can be regardedas a complex technical-economicnetwork.
It implies the coordinationand supervisionof several complex technical elements (high-
speed, computer-controlledtrains,rails, threeinterlinked,computer-monitoredunderwater
tunnels, safety mechanisms, etc.), that are interdependentwith financial ones (e.g., the
instrumentsrequiredby continuousdebt refinancingor by the managementof a complex
propertystructure).The temporalstructuresensuringthe financialfunctioningof the net-
work interminglehere with those specific to the network'stechnical elements. The ques-
tion would be how heterogeneous temporal structures are brought together and
coordinatedand if new temporalstructuresemerge in this process. Haraway(1997, p. 12)
notices that technical-economichybrids(which she calls cyborgs) develop theirown tem-
poral modalities that are not always sequential.Condensation,fusion, and implosion (as
opposed to development,fulfillment,and containment)are, accordingto Haraway,among
these temporalmodalities. Since they arise out of the "densely packed condensationsof
worlds" (p. 14), they question the notion of duration, of "continuity which persists
throughoutthe waking life of the individual"(Giddens 1987, p. 60). For a sociological
theory of things, the question would be (1) how heterogeneous temporal structuresare
broughttogetherand stabilized in such complex technical-economicnetworksand (2) the
ways in which they affect the constitutionand reproducibilityof social orders.This would
requirea concreteexaminationand descriptionof such temporalmodalities.

Exploringthe Distributionof Propertiesin Technical-EconomicNetworks

A third aspect central to a sociological theory of things concerns the power structures
arising out of such hybridconcatenations.Since these are understoodas the lines of force
along which social order is generated and society is organized, then these power struc-
tures are irreducibleto political power; rather,they are understoodas technologies that
pervade everyday life, regulating and standardizingthe possibilities for human action
(Star 1991, pp. 36-37). On the other hand, it would also follow that political institutions
make use of such technologies, and we can trace the ways in which materialdispositives
act as expressions and instrumentsof political power (Mukerji 1993; 1994). This notion
of power structuresas anchoredin knowledge processes resembles Foucault'sconcept of
The Turn to Things 363

power as immanentrelationshipsof force (1976, p. 121) that have their principle "not so
much in a person as in a certainconcerteddistributionof bodies, surfaces, light, gazes; in
an arrangementwhose internalmechanismsproducethe relationin which the individualsare
caughtup"(Foucault1979, p. 202). While Foucaultstressesthe regulatorydispositivesof the
body, a sociology of artifactswould stress the hybridcharacterof power dispositives and
the fact thatthese arenot only embodiedin humanactorsbutalso incorporatedin artifacts.
Starhas arguedthatstandardizationand regulationthroughtechnologies imply not only
enlisting humanactors to those technologies but also a process of exclusion, leaving out
people for whom regulationcosts (or alternatives)would be too high. Takinginto account
the fact that humanactors are usually involved in several such hybridsat a time, one can
consider the distributionof exclusion and inclusion a modality of exercising power:
"Given that we are multiply marginal,given that we may interweaveseveral selves with
our technologies, both in design and use, where and what is the meeting place between
'externalities'and 'internalities'?... A stabilizednetworkis only stable for some, and that
is for those who are membersof the communityof practicewho form/use/maintainit. And
partof the public stabilityof a standardizednetworkoften involves the privatesufferingof
those who are not standard-who must use the standardnetwork, but who are also non-
membersof the communityof practice"(Star 1991, p. 43). Therefore,a sociology of arti-
facts should inquire into the material, social, and symbolic technologies through which
accessibility is distributedthroughoutsociety as well as into the effects of this distribution.
A furtheraspect we may examine is the technical-economic characterof many net-
works and hybrids.In them, processes of economic exchange are embeddedin a network
of cognitive relationships that enlist heterogeneous sets of elements: human actors,
machines, inscriptions,informationtechnologies, and many more. They have a regulatory
character,in the sense thatthey establishthe possibilities for economic action, distributing
and attributingthe requiredskills and knowledge. It is hardnowadaysto imagine forms of
economic action and activities that are not dependenton informationtechnologies, sys-
tems and assemblages of information-processingmachines, systems of inscription, the
interminglingof personalrelationshipswith materialtechnologies, systems of rapidtrans-
portation,production,and distributionmachinery,sales dispositives, and so on. If eco-
nomic processes are embeddedin cognitive networks,if they unfold within the framework
providedby hybridconcatenationsof humanactors and artifacts,then it should be possi-
ble (and definitely necessary) to analyze and describe the distributionof such networks
throughoutsociety and to show the cognitive lines along which they are constitutedand
along which they enlist participants.In the case of the Eurotunnel,we are confrontedwith
a complex technical-financialnetwork, some parts of which are distributedthroughthe
financialmarkets,some throughpolitical institutions,and some throughtransportationand
information systems. Because this network requires continuous social, technical, and
financial maintenance, surveillance, and repairs, it implies continuous translationsand
(re)distributionof propertiesamong its parts,as well as betweenhumanactorsand artifacts:
for example, when its technical characteristicsor technical events (accidents) affect the
refinancingof its debt or when financialinstrumentsavailableaffect its technicalpossibili-
ties. A sociological analysis of such a networkwould imply showing the cognitive lines
along which translationsand redistributionof propertiesfrom the financialto the technical
nodes are effected, how these propertiesare importedand reimportedon differentsides of
the network(Latour1994, p. 798), and how they come to constitutestructuresof power.
The argumentsformulatedin the past decade by the sociology of knowledge and sci-
364 THESOCIOLOGICAL QUARTERLYVol. 40/No. 2/1999

ence make it hard for sociological theory to treat artifacts and things as on the margins of
the social, unproblematic, or easy to ignore. This article argues that not only phenomena
confined to a special branch of sociology or concerning only a particular social formation are
at stake. The very way in which social order is achieved and reproduced, as well as core ques-
tions about society's organization modes, may well be reconsidered. Whether we call them
just "things" or "artifacts", whether we deal with them directly or integrate them in complex
hybrids, reconceptualizing our relationshsip to them will help us see the social world anew
and move us toward a better understanding of how our relationship to things changes us.

NOTES
1. One can of course ask aboutthe social mechanismsthroughwhich the agency of things is rele-
gatedto the realmof aestheticalformsor to niches of popularculture,such as fantasyfiction,andwhat
kindsof sociality evolve aroundthese literaryforms. But thatis not the aim of the presentpaper.
2. This does not mean that all contemporarytheories ignore these problems or that they were
ignored in all classical sociological theories.Actually, the discussion about things has a long socio-
logical tradition, from Marx, Durkheim, and Mauss through Mead (see McCarthy 1984; Cohen
1989; 1993) up to contemporary social theorists of commodification and consumption (e.g.,
McCracken1990; Appadurai1988). For reasons of space and focus, this sociological traditioncan-
not be discussed here extensively.
3. The sociology of knowledge and science is not alone in its interest in the "social life of
things."Sociological analyses of consumptionand commodificationhave also shed light on the pro-
cesses throughwhich artifactsand commodities acquirean active social existence (e.g., Slater 1997;
Appadurai1988; Gottdiener 1995; Douglas and Isherwood 1979; Baudrillard1972). However,the
sociology of knowledge and science formulatesa series of general theoreticaland methodological
claims about the necessity of accounting for social order and sociality without leaving artifactson
the margins.The presentarticle focuses on these claims.
4. I take here institutions,not in the sense of organizationalstructures,but as sets of rules and
dispositives enacted in everyday(materialand discursive)practices,acknowledgedas such, that ori-
ent and constrainthe doings of humanactors.
5. It should be noted thatthe concept of networkappliedby Callon in the analysisof the connec-
tions among science, technology, and market processes differs from that of organizationstudies.
While the latterdefines networksas structuralrelationshipsbetween institutionalactors, Callon and
Latouremphasize a constellationof epistemic relationshipsthatconstitutethe network'sactants.

REFERENCES
Appadurai,Arjun,ed. 1988. TheSocial Life of Things.Cambridge:CambridgeUniversityPress.
Ashmore, Malcolm, Robin Wooffitt, and Stella Harding. 1994. "Humansand Others,Agents and
Things."AmericanBehavioralScientist 37:733-741.
Barnes, Barry,David Bloor, and John Henry. 1996. ScientificKnowledge:A Sociological Analysis.
Chicago: Universityof Chicago Press.
Baudrillard,Jean. 1972. Pour uinecritiquede I'6conomiepolitique du signe [A critiqueof the politi-
cal economy of the sign]. Paris:Gallimard.
Callon, Michel. 1986. "Some Elementsof a Sociology of Translation:Domesticationof the Scallops
and the Fishermenof St. Brieuc Bay."Pp. 196-223 in Power,Action and Belief: A New Soci-
ology of Knowledge?,edited by John Law. London:Routledge.
. 1991. "Techno-EconomicNetworks and Irreversibility."Pp. 132-164 in A Sociology of
Monsters: Essays on Power, Technologx;and Domination, edited by John Law. London:
Routledge.
The Turn to Things 365

Callon,Michel and BrunoLatour.1992. "Don'tThrowthe Baby Out with the BathSchool!A Reply to
Collins andYearley."Pp. 343-368 in ScienceAs Practiceand Culture,edited by A. Pickering.
Cohen, Joseph. 1989. "AboutSteaks Liking to Be Eaten:The ConflictingViews of Symbolic Interac-
tionism and TalcottParsonsconcerningthe Natureof Relations between Persons and Nonhu-
man Objects."SymbolicInteraction12:191-213.
. 1993. "The ConflictingViews of Symbolic Interactionistsand Talcott Parsonsconcerning
the Natureof Relationsbetween Persons and NonhumanObjects:A Sequel to 'About Steaks
Liking to Be Eaten.'" Studies in SymbolicInteraction14:127-154.
Derrida,Jacques. [1972] 1990. "Diffdrance."Pp. 59-79 in A Derrida Reader: Between the Blinds,
edited by Peggy Kamuf.New York:ColumbiaUniversityPress.
. 1992. Given Time.I CounterfeitMoney.Chicago: Universityof Chicago Press.
Douglas, Mary,and Baron Isherwood. 1979. The Worldof Goods. New York:Basic Books.
Durkheim,1mile. [1915] 1947. The ElementaryForms of the Religious Life: A Study in Religious
Sociology. Glencoe, IL: The Free Press.
Michel, Foucault. 1976. Histoire de la sexualite':La volonte'de savoir [Historyof sexuality,vol. I].
Paris:Gallimard.
. 1979. Discipline and Punish: TheBirth of the Prison. New York:VintageBooks.
Fujimura,Joan. 1992. "CraftingScience: StandardizedPackages, BoundaryObjects, and Transla-
tion."Pp. 168-211 in Science as Practice and Culture,edited by A. Pickering.
Giddens,Anthony. 1987. Social Theoryand ModernSociology.Cambridge:Polity Press.
Gottdiener,Mark. 1995. PostmodernSemiotics:Material Cultureand the Formsof PostmodernLife.
Oxford:Basil Blackwell.
Habermas,Jtirgen.1984. The Theoryof CommunicativeAction. Vol. 1: Reason and the Rationaliza-
tion of Society. Boston: Beacon Press.
Haraway,Donna J. 1997. Modest_Witness@Second_Millenium.FemaleManO _Meets_OncoMouseTM.
New York:Routledge.
Heidegger,Martin.[1927] 1977. Sein and Zeit [Being and time]. Frankfurt:VittorioKlostermann.
Heintz, Bettina. 1993. Die Innenweltder Mathematik[The inner world of mathematics].Frankfurt:
Campus.
Henderson, Kathryn. 1991. "Flexible Sketches and Inflexible Databases:Visual Communication,
ConscriptionDevices, and boundaryObjects in Design Engineering."Science, Technology,&
HumanValues,16/4:448-473.
Hendriks,Ruud. 1997. "OnWordsand Clocks: Temporalorderingin a wardfor autisticyouths".Pp.
243-272 in Ideas of Difference: Social Spaces and the Labour of Division, edited by Kevin
Hetheringtonand Roland Munro.Oxford:Blackwell.
Knorr-Cetina,Karin. 1981. The Manufactureof Knowledge:An Essay on the Constructivistand
ContextualNatureof ScientificKnowledge.Oxford:Pergamon.
. 1992. "TheCouch, the Cathedral,and the Laboratory:On the Relationshipbetween Experi-
ment and Laboratoryin Science."Pp. 113-138 in Science as Practice and Culture,edited by
A. Pickering.
1997. "Socialitywith Objects:Social Relationsin Postsocial KnowledgeSocieties."Theory,
Cultureand Society 14(4):1-30.
--. 1999. EpistemicCultures.Cambridge,MA: HarvardUniversityPress.
Practice."In ThinkingPractices: The Practice Approachin Social
. . forthcoming."Objectual
Thought,edited by Karin Knorr-Cetina,Thed Schatzki, and Eike von Savigny. New York:
Routledge.
Knorr-Cetina,Karinand KlausAmann. 1990. "ImageDissection in NaturalScientific Inquiry."Sci-
ence, Technology;and HumanValues15:259-283.
Latour,Bruno. 1987. Science in Action: How to Follow Scientists and Engineers throughSociety.
Cambridge,MA: HarvardUniversityPress.
- as Jim Johnson. 1988. "Mixing Humanswith Non-Humans:Sociology of a Door-Opener."
Social Problems35:298-310.
 366 THESOCIOLOGICAL QUARTERLYVol. 40/No. 2/1999

. 1988. ThePasteurizationof France.Cambridge:HarvardUniversityPress.

. 1990. "Postmodern?No, Simply Amodern: Steps Towardsan Anthropologyof Science."
Studies in the Historyand Philosophyof Science 21(2): 145-171.
. 1993. WeHave Never Been Modern.Cambridge.MA:HarvardUniversityPress.
. 1994. "Pragmatogonies:A Mythical Account of How Humans and Non-Humans Swap
Properties."AmericanBehavioralScientist37:791-808.
. 1996. ARAMIS,or the Lovefor Technology.Cambridge,MA: HarvardUniversityPress.
Latour,Bruno, and Steve Woolgar. 1979. LaboratoryLife: The Social Constructionof Scientific
Facts. Beverly Hills, CA: Sage.
Law, John. 1993. OrganizingModernity.Oxford:Basil Blackwell.
Luhmann,Niklas. 1990a. Essays on Self-Reference.New York:ColumbiaUniversityPress.
. 1990b. "TheParadoxof System Differentiationand the Evolutionof Society."Pp. 409-440
in Differentiation Theory and Social Change: Comparative and Historical Perspectives,
edited by JeffreyC. Alexanderand Paul C. Colomy. New York:ColumbiaUniversityPress.
Lynch, Michael. 1985. Art and Artifactin LaboratoryScience: A Studyof Shop Workand Shop Talk
in a ResearchLaboratory.London:Routledgeand KeganPaul.
1993. Scientific Practice and OrdinaryAction: Ethnomethodologyand Social Studies of
Science. Cambridge:CambridgeUniversityPress.
Marx, Karl. [1867] 1977. Capital: A Critique of Political Economy, vol. 1, translated by Ben
Fowkes. New York:VintageBooks.
Mauss, Marcel. [ 1925] 1954. The Gift: Formsand Functionsof Exchangein Archaic Societies. Lon-
don: Cohen andWest.
McCarthy,Doyle E. 1984. "Towarda Sociology of the Physical World:George HerbertMead on
Physical Objects."Studies in SymbolicInteraction5:105-121.
McCracken,Grant.1990. Cultureand Consumption.Bloomington:IndianaUniversityPress.
Mead, George Herbert.1982. The Individualand the Social Self, edited by David Miller. Chicago:
Universityof Chicago Press.
Mukerji,Chandra.1993. "TerritorialGardens:The Controlof Land in Seventeenth-CenturyFrench
FormalGardens".Pp. 66-101 in The Cultureof the Market,edited by Thomas L. Haskell and
RichardF. TeichgraeberIII. Cambridge:CambridgeUniversityPress.
1994. "Towarda Sociology of MaterialCulture:Science Studies, CulturalStudies and the
Meanings of Things."Pp. 143-162 in The Sociology of Culture:Emerging TheoreticalPer-
spectives, edited by Diana Crane.Oxford:Basil Blackwell.
Pickering,Andrew,ed. 1992. Science as Practice and Culture.Chicago:Universityof Chicago Press.
1995. TheMangle of Practice. Time,Agency:and Science. Chicago: Universityof Chicago
Press.
Rheinberger,Jirg. 1992. Experimentund Schrift[Experimentand writing].Marburg.
Schutz,Alfred, andThomasLuckmann.1973. TheStructuresof the Life-World.Evanston,IL: North-
western UniversityPress.
Schatzki, Theodore R. 1996. Social Practices: A WittgensteinianApproachto HumanActivity and
the Social. Cambridge:CambridgeUniversityPress.
Searle, John R. 1995. The Constructionof Social Reality.London:Allan Lane.
Slater,Don. 1997. ConsumerCultureand Modernity.Cambridge:Polity Press.
Star, Susan Leigh. 1989. "The Structureof Ill-StructuredSolutions: BoundaryObjects and Hetero-
genous DistributedProblemSolving."Pp. 37-54 in DistributedArtificialIntelligence,vol. 2.
edited by Les Gasserand Michael N. Huhns.London:Pitman.
- . 1991. "Power,Technology,and the Phenomenologyof Conventions:On Being Allergic to
Onions."Pp. 25-56 in A Sociology of Monsters: Essays on Power Technologyand Domina-
tion, edited by John Law. London:Routledge.
Turner,Victor. 1969. TheRitual Process. Chicago:Aldine.

View publication stats