te The De-Scription of Technical Objects Modan Arch Describing the Interaction between Technics and Humans Although science and technology ate often thought to go together, hey are concerned with very diferent subject matters, Science is taken to go beyond the social world to seality unfettered by human ‘oningency. Peshaps asa result, the sociology of science has studied the ways in which the leal and the heteragencous ate combined to create knowledge withthe status of universal and timeles truth. By contrast, sociologists have found it diffeslt to come to terms with techincal objects Machines ad devices are obviously compsite, heterogeneous, and physically localized. Although they point to an end, use for which they have been conceived, they also form part ‘fa long chain of people, products, tools, machines, money, ad 10 forth. Even study of the tcchnical content of devices does ot produce «focused pieture because there js always a hazy content or back ‘round with fuzzy houndaries. Thus even the mort mundane objects appear tobe the product ofa set of diver forces. The scength of the materials used to build ears function of predictions about the stress they will have 1 bear. These are in tar linked tothe speed ofthe car, which is islf the product of a complex compromise between engine perlormanee, egsation, law enforcement, and the values asenbed to diferent kinds of behavior. As a consequence, insurance experts, police, and passers-by can ue the condition of the bodywork of car to judge the extent to which it has been used in ways tha conform to the norms It represents “Technical objects thus simultaneously embody and measure & set of relations between heterogeneous elements. However, the pro- ess of describing everything about car in such terms would be a ‘mammoth ask. Furthermore, the end product might well be banal The automobile isso much a pat of the world in which we live that itssociography (a deseription ofall dhe links making it up) would no6 Maine Ask doubt look like a collection of commonplaces. [t would, in other words, look ikea set of places where elements of the technical, the socialy the economic, and 30 on were to be found together, and it ‘would leave observers ree to switch between one element o register And another ax thie sited them.* ile ing and site, whether human or noshuman.? But how ean we descr the specifie ole they play within thete networks? Because the answer hha todo with the way in which they build, maintain, and stabilize a structure of links between diverse actants, we ean adopt neither imple technological determinism nor socal constructvin. Ths technological determinism pays no attention to what is brought together, and ultimately replaced, by the structural effects ofa net~ ‘work. By contrast social comtruetvism denies the obduracy ofabjces land assumes that only people can have the states of actors. The problem is not one of deciding whether a technology should be een at an instrament of progrest or a new method for subjayat- ing people. Ic ix rather (0 find 2 way of studying the conditions isms under which the relations that define both our ‘ur knowledge ofthat society are susceptible to partial have to move between the inside andthe outside of technical objects. I'we do this, two vital questions start to come Jno focus. The frst has to-do with the extent to which the compase tion of a technical object constrains actants in the way they relate both to the object and to one another. The second concerns the character of there setants and their links, the extent to vehich they tte able to reshape the objec, and the various ways in which the ‘object may be ated. Once considered in this way, the boundary bbetween the inside and the ouside ofan object comes to be seen as ‘csruenee of such interaction rather than something that deter: tines it, The boundary it turned into a line of demareation raced, ‘within a geography of delegation, between what is assumed by the technical object and the competences of oxheractant However, the description of these elncntary mechanisms of ad {justment poses two problems, one of method andthe other f vocab ular. The difficulty with vocabulary isthe need to avoid terms that fisume distinction between the technical andthe socal Becase the s that concern us are necessarily doth technical and social, 1 develop and use a vocabulary drawn from semiotics that i intended he Desi of Tabi Obs 37 ‘o avoid this lficully* ‘The methodological problem is that if we want describe the elementary mechanitns of adjustinent, we have ‘oind ciccumstances in which the inside and the ouside of objects are-not well matched, We nced to find disagreement, negotiation, And the potential for breakdown, There are several arcas—for instance in technological innovation and technology transfer—where objects and their supposed fune- tions, or the eelaionship between supply and demand, are poorly ‘matched. In what follows I deeeibe a number of eases of “technol ogy transfer” 10 lesaeveloped countries (LDCs) that are drawn fear my own feldwork, These range from the simple transplantation ofa piece of technical apparatus widely ueed in industrial societies to the development of objects specifically intended for use in LDCs a tach ease I describe the elementary mechanisms of reciprocal adjust- ‘ment between the technical object and its environment start by considering the way in which technical objects define actants and the relationships between actants, T show that the ease With which dhe actants assumed in the design ofthe object are related {o those that exit in practice s partly 4 anction of decisions made. by designers. "The oblurac o plasty of objets, something ‘stublshed in the confrontation with users ina function ofthe dss bution of competeness assumed when an object is conceived and designed. Ti the second part of the chapter I comider the way in which bute causes, IF mest ofthe choices made by rs take the form of decisions about whut should be delegated to whom of what, this means that technical objects contain and. producea speciie geography of responsibilities, or more general, of causes, To be sure this geagraphy is open to question and may De resisted. Nevertheless it suggests that new technologies may not only lead to new arrangements of people and thing, They aay, i addition, generate and "aaturalize” new fms and order of eatal= ity and, indeed, new forme of knowledge about the world, 1 will consider this proces and illustrate the way in which technologies may generate both forms oC knowledge and moral judgments. Subjects and Objects in the Making From Script to De-Scription For some tine sociologists of technology have argued that when technologists define the characteristics oftheir objects, they neeesar- ily make hypotheses about the entities that make up the world into8 Maine Ack ‘whieh the objets to be inverted? Designers thus define actors with sprite tastes, competences, motives, aspitatons, piel preju dices, and the rest, and they assume ality, echo science, and economy will evolve in particular ways. A large part af the wotk of innovators i that of “ineribiag™ this vision of (or pre= Aicion about) the world in the teclnieal enntent ofthe now ebjec. wil call the end product ofthis wurk a "script" ura seu ‘The technical realization ofthe innovators belils about the tionships between an object and its surrounding actors is thus an attempt to predetermine the stings that wsers are asked to imagine fara parielar pce of tcholgy and the preston (tees contracts, adviee, ete) that accompany i. Tobe sir it may be dit fo actors will come forward 10 play the roles envisaged by the designer. Or users may define quite cillerent roles oftheir own, IP this happens, the objects remain a chimera fo tithe eoulfonts- tion between technical objec and theie users that the later are sendered real or unreal ‘Thus, ike a film sript technical objects define a framework of action together with the actors and the space in which they are apposed to act. Sigaut (1984) gives examples of tou whose frm suggests a precise description (la Sherlock Holmes) of thee ser. ‘The two-handed Angolan hoc is mad for women carrying cildrea ‘on ther backs. The laborers stake, with is single point, can only be driven in by evo people, and thus presupposes a callective wer However, once ane moves away few such simple examples, i be= ‘conics more dificult to uncover the inks between technical choices user” reprsentations, and the actual ues of technologies. Thus the ‘method of content analyse, pplied to tox dan psyehlogical approach that hoe problem. Indeed, because it iguores the wide range of uses vo whic ‘bjecte may be putt comes cows to technological determin, I is obvious that it cannot posibly explain the wide varity of fates txperienced by technological projects—fates that range from cone plete succes to total failure ‘One way of approaching the problem iso follow the wegotations between the innovator and potential users and to study the way in which the results ofsuch negotiations are translated into technolog cal form. Indeed, his method has been widely used in sociological and historical studies of technology. Thus, if we are intersted in technical objects and not in chime, we cannot be satisfied mt codoloically with the designers or uses point af view alone Instead ‘we have wo go back and forth continually between the designer and Te DeSeipiof Teal Obs 328 the wer, between the designers projected user and the real user, Ioewcen the aod iribd a the objet ad the word decried by is siglaconet For ii this icessatwariation that we obtain seeess {o te crucial relationships: the uses reactions that give body tothe ‘designer's project, andthe way in which the ues real environment isin part specified by the intredction aa new piece of equipme “The notion of descpin projuned Mere has tr be developed within his framework, Iti the inventory and analysis ofthe itechanisms tha allow the relation between a fora aud a meaning constituted by and constitutive ofthe teasial abject to come into being, These echanism af adljustoent (or Lal to adjust) between the wer, as winedl by the designer, an the real wer become particularly clear when they work by exclusion, whether oF nat this exclusion is Aelierate The case ofthe photoclecic lighting kt is an example hich exclusion was explicitly sought by no one, The Photoelectric Lighting Kit a Non-User The photoelctire lighting kt was bor from the wish of 4 goves tent agency o promote new energy sources. Ax part of ts coopera tive international activities, the agency wanted to work oh and and rect te ned for lighting something that well-intenticed infor= rans said was essential for all LDCs, At the same time it wanted 10 he Or How to Produce the French photorlecrie cell industry to create a market. ‘Caught up, as they were, na specie network involving state support with industry, those involved in is design conceived of the ic asa Timetion of the specific needs a! constraints imposed 0 rem by this network. AL no point, for iitance, did commerci cuserations cute into play. Accordingly, the sape ofthe lighting, kitcam be treated axa description of the way in which this network ‘operated —a network characterized by the circulation of certain types of resources and the exclusion of her actors, The “narrative” pitierns and scripts dreamed up by thore who conceived the kits ‘were quite specific a funetion oftheir positon. Study of the lighting kit (or any other technical object) makes it powible for us to ereate 1 “sociology” of the network defined by its circulation, ‘Whea I fist heard the industrialists snd designers talking about lighting kit, it appeared to be a very simple aay with three Functional elements, There was panel for producing electricity, a storage battery, anda lamp that consumed the electricity. However, ‘once I arsed in Attica and started to study the ways in which such its were actually used, the picture rapidly became mote compli=cated. ‘Those who were responsible for insaling and maintaining its were confronted with considerable difficulties, The ist ofthese ‘vas that the wires linking the different components—the pane, the batteries, and the fuorescent tubes-—were fixed in length an could ‘ot easily be altered because the connections were made with nod sandard plugs. This meant that it was dificult to adapt the kts to ficroomsof different sizes, Replacing components with shor limes, such as lamps or batteries, represented second st of diiclties Neither appropriate fluorescent tubes, nor the watertight bateries chosen to ensure that maintenance problems woul aot limit the life ofthe system, were available in markets outside the capital Local wurces of supply were thus of no help tothe user. As a result, despite the fact that it was a major clement in his or her technical ‘environment, the usr lst control over the installation. Suddenly, ‘what had previously been familiar stated to become strange (dhe first question users asked was often “When do f have to add water to the batteries?) third factor also wotked to prevent the user {om appropriating the installation. This wa the fact thatthe cn tuactor who installed the ht forbade him or her to turn to local slectrician incase of breakdown. Instead, the contractor sid that he would ome 1 the area twice a year to rep faulty installations ‘The reason for this embargo‘ local repairs was he ensitvity ofthe photoelectric panel, This, as the instritions put i, “convert solar ‘energy directly into electrical energy." However, the fact that this took the form of dizet curren with non-equivalent poles meant a least in the view of the contractor, that it would be risky to eal in local electrician who would have experience of alternating but not of| direct current The danger was that ifequipment was connected the ‘wrong way, ic might be damaged ‘The discovery of these dfficulis illustrates an important point of| method. Before leaving Paris for Ali, the potential siguicance of nonstandard plugs, direct current, oF waterproof Bateries had wot ‘ceurred to me. It was only in the confontation between the real user and the projected user tha the importance of uch items as the plugs forthe diference between the two eame ta light? The materi= alization and implementation of this technical object, lke ates, ‘was along proces in which both technical and socal elements were simultaneously brought into being —a proces that moved far be- ‘yond the frontier of the laboratory or the workshop, ‘The fact that the importance ofthese characteris only became evident in the interaction between designers and wsers was not the The DeSeipin f Teil Obits 211 result of chance or negligence. Bach decision actually ken made sense in terms of design criteria. Digect current is cheaper than alternating current because a transformer consumes a good part of | Ae available power. Watertight batteries and nonstandard connee- tions were chosen to prevent people from interfering with and 30 Potentially damaging the kit. The length of the wiring had to be limited ort would reduce the performance of the equipment. These decisions were intended to ense that the lighting kit would "work" unde all circumstances an important consideration in the negot- ations between the industalise and ther clients. It should be re= called that it was not the latter who were the ultimate users of the Ait, bu rather the donating agency and the government to whieh the gilt was to be made. Indeed, such was the concern to produce a Toolproof kit that the designers decided not to have a separate ‘tcl inthe circuit because this might become a point oflict entry imo the system. This meant that ser often found it difficult to turn {he light on oF off because the only switch available wat attached sliretly wo the light and a0 was normally ou of teach, Soit was that the technical object defined the actors with which it was to interaet. The lighting kit (nd behind i the designers) worked by a process of elimination. It would tolerate only adlocile user and fxcluded other actors suck as technicians or businespeople who might normally have been expected 1 contribute to the eration ofa technico-eeonomie network. Had the wets eally been as docile a5 te designer intended, [would not have sen thatthe kit represented lange set of tchial degated recription addressed by the tor tothe wer, weare todeserbe echnical objets, we need mediators to create the fiaks between technical content and wer. In the ease of non stabilized technologies these may be either the innovator or the user. ‘The situation is quit different when we are confronted with stabi- lized technologies that have been “black boxed.” Here the innovator smo longer presen, and study of the ordinary use snot very useful because he or she has already taken on board the presriptions limped in interaction with the machine. Under such circumstances some prescriptions may be found in wer's manuals or in contracts Alternatively, we may study dispute, look at what happens when devices go wrong, o follow the device ait moves into counties that are culturally or historically distant fom its place of origin. In the next section I adopt the last of these methods to describe the use of generators in SenegalDe-Scription in Technological Transfer: Reinventing and Reshaping Technical Objects in Use Invrural Senegal yonerators ave widely sed by “Kesive group” An administration buys some small generator, which it distribute to ‘youth geoups in the villages. With the generators may eon lights, & recor player, ora loudspeaker. The youth groups use the generators oF lead them their members who pay forthe cost of fucl and fil, Again, they may rent them out to othe villagers who ate alo responsible for the cost of fuel and oil. The money that is made by the ental of generator is shared, with part gong to the person ‘who transports the generator and part going to the association. In {his way a small collection of actors x involved with the generator— actors that can be sen as so many editions fo the components that take up the generator. he generators metal trailer means that itis mobile, and soit plays an important part inthis proces. This is beeause the Geld of posible users and the relations between the different actorsis defined by the movement ofthe generator. However, he acl tank val the generator forthe staring role because it draws a fundamental die Uinetion between capital cont and operating cont. This distvetion {nerd fom the outset in the social setup that bring the generator to the village: there isthe administration, which uncerwrites the there isthe group that actually manages and russ the generator. The technical device reduces negotiations between the two partis to @ minimum because it cirectly suggests 2 pre ‘negotiated agreement. Obviously things could be arranged dilfere ely. This, however, would mean delegating a whole series of tasks to additional (legal, human, and technical) structures exter othe generator and ie trailer [might even email new systems of nese frement—in which case i fe not clear wher we wen ealing with the same object. ‘The situation would be quite different if we were faced with & device whose costs were concentrated exclusively on the side of in ‘vestment—as, or instance, withthe photoclectric kts, What kind of relationship can there be between the buyer and the see unde such circumstance? ‘This was a question faced by those promoting the ‘development of photoelectric cells in Feench Polynesia. Once these cell had been distribute, it was not always posible to insist that these two classes of costs should be distinguished. Not only did the {echnology itself ilo diseriminate between them, but i ollered 0 method of measurement that could be translated into appropriate socioeconomic term. Thus no matter how iis used, a photoelectric Te DeSeipin fT Obs 215 panel generates current as afanetion of climate and latitude, The “standan” relationship eweew preduction and consumption {a fellection uf dhe interdependence of two ups of actor) Feplace boy am individual, direct, and indeed arbitary submission to natural forces, ‘The lilference between this and the generator is obvious. In the ‘ase of the generator, Ue fuel tank can be used to measure the ‘relationship between its use and the est ofthat use—a relationship ‘embodied inthe motor asa whole, The eration ofa particule kind of social ink, that of renting out, conditioned by the existence of this relationship, which delocalize the yenerator by creating many |oups of actors! investors/purchasees, ownes]users, associate user Fenters, and iransporters. The existence of transporters makes the Property even “pures," for they fre it from servitude. Their pay. ‘ent marks the boundary ofgeoup solidarity, fr the work ofa single person cannot enrich the community. At the ame time the generator builds a space and a social geography. Thus the teachers in one of the villages who necded lighting for tie evening elases did not even consider renting a generator. The division between the work! ofthe “market” and the “eivie™® world may not have ben brought into boing in the village by dhe socal ilferentiation eneailed in eleceieity Aan ts uses, but it was certainly maaiied by the later. The lighting kit put itself forward as a “hypothetical” objec, whereas the generator was just another piece ef equipment inte. srated into the vatiou sector of economic hie. However, we should ‘ot overstate the difference between them. This is best seen in terms of uiferenil resistance. It would would take much more elort to (rc}cismanite the generator than it would the lighting kt. But boil cases we are dealing withthe creation and extension of net works that simultancously define both the socal and the technical Thus such item as nonstandard plugs and Tass become significant when the real users start to displace projected users. Again, the competence ofthe youth group, its relations with other elements of village life, the very definition of these lements—all of these are determined at the same time as, and by the same proces, that defines ‘he components chat make up the generator. I we were to restrict. ‘ur attention to the “function” flied by this piece of equipment ‘within the youth group, we might imagine that some oer technical system (Vor instance, solar panels or connection tothe national ri) ‘would function in the same way. This, however, i ot the eas, for ner such circumstances the relationship between the youth grou and others in the village would be different and probably moreuid. In this sense, then, we can say that our relationships with the “real world” are mediated by technical objet, Prescriptions as a Way of Enrolling Actors: Or How to ‘Make Citizens Soar I have described technologies that appear to exercise relatively ‘weak constrains over those wo use them. 1Tthe generator and those ‘who sponsor it nudge some who would otherwise be outside eco- nomic relations in the dieetion of involvement, then this eet is relatively small, Inthe case of the photoelectric lighting kit, the min ‘danger is that no one will use it at all. However, technologies are not always ike this. Sometimes their designers and builders ure them to ‘objain acces to certain actors, whom they push inte specific role. ‘Thisis what happened in the eae ofthe Ivory Coast and is electrie= ity network. Here the physical extension of the network Wasa integral part of a vast effort to reorganize the county spatially architecturally, and legally. The object wa to create sch new and “modern” entities as the individual citizen, ‘Winner (1980) has argued that ertain technologies ae inherently political—for instance, wondemocrati. [fhe i right about this, hen the approach Fhave adopted here would lead to. form of techy ical determinism. However, the cate of clectification in the Ivory Goast shows that even in thse cates where tere are marked politi! plications, itis frst necessary to interest and persuade the actors to play the roles proposed for them, Uni recently village property in the Ivory Coast was collectively ‘owned and under the control of elders, who allocated tract of lal to villagers at a function of their needs, ‘This allocation was n permanent, and people might move to dillerent areas, When the Authorities started to think about clecteifeation,chey decided that this should be contingent on a more stable allocation of and, and in Particular on a distinction between private and public property. ‘The developing the new electricity nework (who ako presented themselves as spokespersons forthe general interest) assutied that the network would both contribute to this division an depend oni, sit would be installed on public land. In other words the electricity network made it possible forthe state to create its cw space (the space of common interests) that could not be appropriated by any fone elie. AC the same time, it defined those with whom i¢ woukt interact. Because ony the individual would legally exist in tht new system, former collective mdr of village representation were thus sstematially exclude, Te DeSeiiof Tea Objet 25, To be sure, the eration ofa system that allocated land perme neni cither to individuals o Ue state was funtion of agreement in the village asa whole about the need for such stably. Through tenew property system the electricity company was thus asking the villagers to make a prensrpion witnessing their consent toa ertan Kind of future. Thus, individual villagers ad to undertake certain formilities to secure title to Rxed property. Front the stanpoin uf ‘he electricity eompany, legal ownership could be trated asa token for a range of agreement between diferent bodies about the future of the village. The new sytem of property was ato the Foundation fora series of projects by other uilities (the highway departatent, the ‘water authority, the medical service, dhe education system). It meant that electrification could be integrated into various modernization programs, and it established economical procedures for on ‘ion and politcal negoiation. Finally, the construction ofthe net- work itself would put the agreement ofthe village into practice and sabliz it by making a durable inscription on the landseape. ‘Bat why should the vilagers agree to enter into game in which they would, oF soit seems, love a part of their independence? Alice all, by so doing they would place themiclves under the influcnce of ® central authority that would, by virtue of this very fie, increase its power. There are several answers to this question, The villagers wanted to have accesso eletecty. But there was the question of the way in which the company negotiated with the village. Indeed, 'o put iti this way is misleading. ‘The company did not wegotiate src withthe village. Rather, it negotiated with spokesperson — invatiably someone who had “Succeeded” and moved from the vil lage to the capital. Both this spokesperson, who negotiated with range of entral authorities on behalf ofthe village, and the villagers thenvilves knew that series of indiret benefit would fllow fom ageciment with the eectriity company. After electefcation the village could hope for better teachers, an improved health service, ‘more financial sippor, ease in the number of develop ‘meat projects. In short, electrification was @ method for avoiding direct and specific negetiations between the villagers and a seties ‘of external agencies. It was a package whose terms were fixed in ‘advance. Those in the village had a choice. They could accept ‘hose terms oF they could rejet them, and overall the package was In general an individual becomes a citizen only when he or she ‘enters into a relationship withthe state. In the Ivory Coast this was effected through the intermediary of cables, pylons, transormers,216 Maine och and meters. By conteas, in France individuals ae inserted into such wide range of networks that they have lide chance of avciing ‘citizenship. From th registry ofc, via obligatory schooling to mile tary service and the welfare state, the mesh of the state with its diferent superimposed networks deaws ever tighter around them. a ‘countries that have been created more recently, specific networks may come tothe aid ofa weak or non-existent sate The electiity network may reate and maintain a relationship between an individe ual and a place. Thus in the Ivory Coast, where only & minority of salaried workers paid income tax, the clciriciy bill became the means by which local taxes were collected in recently but towns Here, then, it was the electricity network that fared & wide defini= ion oF the concept of etizenship, From Causes to Accusations and Forms of Knowledge In dhe examples above I have shown how technieal objects deine actors, the space in which they move, and. waye in which they interact, Competences in the broadest sense of the tem are distrib uted in the stipe ofthe technical objet. Thus many ofthe choices made by designers can been seen az decisions about what should be dddegated to 4 machine and what should be left to the initiative of ‘human actors. ln this way the designer exprestes the scenario ofthe device in question—the script out of which the future history of the ‘object will develop. But the designer not oly fies the distribution ‘of actors, he or she abo provides “key” that ean be used interpret al subsequent events, Obvounly, this Key eu be called Hato urs tion consumer organizations specialize in suc skepticism. Nevete theles, although users add their own interpretations, so long as the tircumstancesin which the deviee i used do not diverge too radially from those predicted by the designer, is likely thatthe sr wil become a major element for interpreting interaction between the object and its user. Abobo-the-War and Marcory-No-Wire: Where Teehnology Meets Morality In this section T focus on one particular proces—moral deg ‘ion~and discuss devices installed by designers to contro the moral bbehiavior oftheir uses. dseribe the way in which such devices may measure behavior, place itn a hierarchy, eantzol i, expres the fact of submision, and distribute causal stric and sanctions, The Design of Tobia Obs 207 ‘As I have indicated, the introduction of the electricity network has ‘stablshed inks bewween individuals in the Ivory Coast. The way in Which the individual/consumer relates tothe network, aid via the network to the electricity company, is codified and quantiied by ‘means ofa basic technical tol the electricity meter. Tis formulates the initial contract between the producer and the consumer. Irene othe other fails wo meet its obligations, the meter become invalid fF inactive, Meters have a symmetrical eflect on the producer) consumer relationship. The agreement of both is required iFthey are to tick over, Accondngly, the set of meters is & powerful instrument ‘ofeoutrl. Taken together, dheset of meters measures the coisa of {he sociotechnicaleifice materialized by the network, Consider the following story, which appeared in The Kania, the electricity coms pany newspaper, ins February-May 1985 sue: OPERATION STRIKEFORGE AT “ABOHO-THE-WAk" There a Mashing red light ia the DR in Abo, lower clase suburb of Abidjan where tere are 66894 subscriber, the networs ate af tare (ihe relavonship beeen the energy put out by the producer ad the cp ile othe ntl) hal fn 0.9 0087 the space afore Any reduction in the rate of return can be interpreted as an increase in the number of ict connections, the work of corrupt smployees, or a consequence of tracking in meters With both hhuman and technical actors involved, the network measures iit Foeluviow and determines its character, ‘he definition of social space aso extend to non-electriid ate. “These ae characterized in terms oftheir degree of deviance from the orm that is, fom lecteiication, Thus another suburb of Abidjan, Marcory, was split into two by the network. Bach was given a name, and characterized in social ters Unlike residential Matcory, Maseory-No-Wie isa Marcory without le tek, without wie. Ie wel known that Aja have ens oh uae but with wo wie imagine what Kin petal hat sie, For if eleciety ina sigh of progres, ts aluence segs eee steccex: yen in testes, of bing constructed tee sta dard of pharmacat,paygioundssporagrounds al 9 on When you Seid dares igh o thee ance ten te guards of te nes oul say you ge tein aunt (eure 1088) Even so, the diving line between the permissible and dhe imper= misible is negotiable. Thus in their stake-force operations, elec.Uecity company agents were old to replace so-called Russian meters that had proved defective without penalizing their owners, even ‘though a simple ap on the meter would block it and allow unbiled clectiity to be consumed. Unlike the agents, the "Russian" meters found it impossible to distinguish between lcit and iit behavio, between the actions of humans and nonumans. Accordingly, al though the contract between supplier and consumer vemained in force, the mete filed in its preseribed role asthe material inscription ofthat contract Each indisdeal meter intervened as referee and! manager of the relationship between supplier and consumer. Taken together, the wt ‘of meters operated as police ina collective organization, uncovering inregulaitis, Such regularities appeared fist as deviations in com. ion curves that were nether localized nor sanctioned, They ‘could, however, be quickly translated into “social” ters, Some techniques move closer to “social control.” They establish ‘horns and punish those who tansgress them. Thus the storage and ‘regulation ystems in photoelectric Kit take the form of batteries and slectronic components. The batteries store the electricity so tha it can be given ou, for example, for lighting when itis dark. However the control system lies at the heart of a echnical, ccovomie, and social imbrogio. Itche batery is allowed ta ran to low, its time will be reduced. On the other hand, i ti overcharge electilty ‘may leak bac into it and ruin the photoelectric eel, Users night, course, be given meters with which they could plan their electricity consumption while avoiding both of these danger. In fact this solu. ‘ion is never adopted because the designers donot believe that wsers will allow the technical requirements ofthe sytem to overrule their immediate wishes. Again, the designers could! choose to increase the ‘capacity of the system to-cope with the likly demands of the user ‘This, however, ba costly option. Accordingly, the designers adopt the third option of installing «regulator that cuts ofthe curren to {he user when the charge on the battery ges to low, and soates the photocecric panel when it ges too high" As a result, a particular ‘mode of consumption is imposed: the user cannot be to geeedy, yet neither can he or she hope to compensate for excess consumption by prolonged abstinence. The penalty for breaking the rules—rutes that are both social and technical—isimmediate and abrupt: the current 's cut off and is not reconnected until the battery i adequately recharged, ‘This method of regulation is designed to “groom” the user, It offers set of rewards and punishments that i intended to teach Te DeSeiin f Teal Obs 29 proper rules ofeonduct. However, a lave inthe system is that there sno easy way to measure the charge inthe batery. Voltage is only rough indication, What should be done about this? A general whe ‘snot sure ofthe loyalty ois troops hs two options. He may choose odo nothing. Or, like the designers in ths case, he may redouble his precautions and disciplinary measures, Accordingly, as I have ‘mentioned, a particularly inflexible system with nonstandard plage was adopted. Thus while the control device was telling the wie not to get too big for his or her boots the nonstandard plugs were ‘imposing even more draconian limitations on conduct. No bypass of the conteol device was permisible! Even so, in French Polynesia the contro device proved to be a shaky ally for the designers, because the users fel that its sanctions ‘were arbitrary. The result was chat they denounced it and expressed their displeasure by telephoning the electrician everytime the ayer tweacherously cut olf the current while they were quiely siting watching television. The eletecian, who quickly became Gred of ‘oing repairs inthe evening, ticked the system by installing fused ‘ecu in parallel with che control device, When the conte device shut off the eursent, users could Bypass it with the fae, and th clectrician would only be ealled out the following morting, ‘The ‘incl citeuit thus marked the submission of electricians wo the wishes ‘ofthe elients and allowed them to be present by prony instead of boing summoned in pers by irate user. ‘he precarious and makeshift character ofthe fase makes it plain some kind of intervention was nectsary, even if it only took Place after the event. Ie this particular trial it was the eectiians wo pleaded guilty. In fiting the fae, they recognized that the conurol device and their cients were Sut right ard moderated the judgment of the former in favor of the later “The Order of Things and Human Nature”: The Stabilization and Naturalization of Scripts 1 have describe several eases which technical objects preformed Ucir relationships with actors and vested them with what could be called “moral” content. Because role and responsibilities are allox cate accusations and wials tend to fallow. Im principle, no one and nothing is protected fom such denunciation. In the ease of the eletricty network, the were were accused of failing to respect he contract with the meter However, the electricity company aso accused some of the meters of failing to represent that contract In the case of the photoclerrie kit, it was the electricians, andindirectly the manufacturers, who found themselves in the dock through the agency ofthe conrol device. Indeed, the story of the kits can be read ata log series of reciprocal accurations. The indstial- ins tended to argue that if it didn't work (technically) this was because it had been misused (socially). ‘The users, oF those who claimed to be ther representatives, argued that if it didn't work socially, this was because it had been minconceived technical. Here, then, we sce an almort perfect “reversible reaction” that revels the lack ofa relationship, through the kt, between designers and users, The were did not inter. the manufacturers they were only Important to the eatent that they made it possible lo go to the ‘ministry ofverseas development and seek support fora produc that didnot yet have a markel. And in this interaction the kit did not actually have to do anything. Rather it was the wes who were treated as an instrument for building a relationship between the manufacturers and the government Tn the ease of the electricity network, the siwation was quite diferent. Iti dificult to imagine a plausible argument for legal connection «9 the network—one in which the electricity network ‘would sand ia the dork. This beeause the network configured a whole range of relationships. have already mentioned the meter ‘and the way in which it ws related tthe allocation of property. But Telationhipe were structured by the network in many other ways For example, it alvo tended to stabilize ving space. This was be- ‘use, for reasons of security and as a guarantee of solvency, only permanent” structures were consected to the gid. And of course, ‘once the rid wasin place, new commercial networks for distributing tlectrieal equipment quickly rprang up. Thus once it war estab- Tished, the network tended to promote both physical and social stability. A wide range of elements were brought together and given substance, A amall (inge group of “deviants” could not possibly hope to ind the srength necded to oxtweigh the many actors bound togetter bythe grid. Accordingly, the electricity company could eall upon the meters to at as unequivocal spokespeople at will A double icroverbiliy had een established —a material irreversibility i scribed in space and practice, and a directional reversibility where ‘accusations and charges could no longer be reversed. Obviously the ‘80 were intimately linked, Ta thissction {have argucd that technical objects not only deine actors and the elationships between them, but to eaatiaue funetion- ing must stabilize and channel these They mst ctablish systems of ‘causality that draw on mechaaixins for the abstraction and simphii= The Deep of Tel Objet 24 «ation of causal pathways meat of the "Resi process designed to Same path hes arti the case dscusel above, the replace- eters was very much part ofthis process—a ke diagnosis automatic. Farther along the intelligence." Conclusion: Toward the Constitution of Knowledge Once technical objects are stablized, they become instruments of Knowledge: Thus when an electricity company sets differential and low-consuming domestic users, for workshops, ils ways of characterxing and idemiying diferent social strata, IF alo chooses categories used in other socioeconomico-paitical network, then the knowledge it produces can be “exported.” “Data” can thus be drawn from the network and transmitted elsewhere, fr instance, to econo cone ‘cerned with the relationship between the cost of energy or GNP and consumption. However, the conversion of sciotechnical facts into acs pure and simple depends on the ability otuen technical abject into black boxes. In other words, as they become indispensable, ‘objects also have to efface themselves. Hil illustrate this with an le drawn from Burkino-Faso, Burkino-Faso is a developing country witha tiny eleticty net ‘work. Over the past few years it as been government policy to slectrify urban centers. The fest problem for the engincers and technicians was to judge potential demand and decide how large the network should be. Two dillerent approaches were adopted. ‘The economic stuies unit asked potential subscribers what price they would be willing to pay fr eleticty. The approach assumed that there was a relationship between supply and demand, and that consumption would vary inversely with price. The technical unit adopted a very diferent method. It drew maps ofthe twas, marked off the builbup areas, and noted the characteristics of the houses (orether large or smal, petmaent or temporary, and 40). On the basis ofthis map they designed a network that would be legally, ‘economically, and technically feasible—a network that would make ‘use of public space and serve only permanent buildings and gove sent facilites, The results obtained by the wo approaches were quite diflernt. In particular, the geographical and legal approach ofthe technical nit suggested the need for a far larger network than the marketed approach of the economic studies un22 Mabie Ath tere were no need for technial mediation between price and con: sumption. They assumed, tht i that this relations was a fet of nature that would be given concrete form by the electricity network. Ina sense they were led astray by the naturalization effect, which ‘occurs when technieal system are completely integrated into the social fabric. Tis only when the seript act out by the designers acted ‘out—whether in conformity with the ieteations of the designer ot not—that an integrated network of technical objects and (human and nonhuman) aetors is stabilized. And itis only at this point that this network can be characterized by the ciculation of finite ‘number of elements—objects, physial components of monetary tokens. Disciplines such as economics and technology sadies depend fon the presence of a selacing apparatus that ies outside thee domains. Eeonomists extract ae kind of formation for techni ‘objects technologists another. They are able to do this because uh object function in stable situations. The intrduction af a new de- vice can thus be assimilated, for example by economists, into the pricejconsumption telationship. The economy ie not cut off Fem {celology; there no rican “Thisis why it makes sense to say that tecnica object have pul cal strength. They may change social relations, but they also stabi Tize, naturale, depoliticize, and translate these into other medi. Alfer the event, the process invulved in bug py technical ‘objects are concealed. The causal inks they established are natal ied. There was, of so itseems, never any posbility that i could Ihave been otherwise." ‘We are ourselves no more innocent in this respect thaw anyowe ‘se. For weare abe to say that technical objectsehanged, stabilized, ‘naturalized, or depolitcized socal relations only withthe benctic of nisight. The burden of this essay is that technical objects and people are brought into being ina process of reciprocal deiition in which objects are defined by subjects and subjects by objees. Tt isonly after the event that causes are stabilized. And itis only aller the event that we are able vo say that objects do this, while human beings do that Te isin this sense, and only in thissese, that technical ‘objects build our history for ws and “impose” certain frneworks, ‘And itis fr this reason that an anthropology of technology i both posible and necessary 1 woul het thank Geafey Bowker who tala! ite Jin Lam whe ‘ately eeu heer x, Brana Lars wn hep ne ie ae The Deseo of Tei Obs 223 ‘hrm frm of he condone a he an eld sug |. Doutaleit could bean tp vad canvas tring wi state rge nal ethene ly et ruch sages woul be em On what grounds woud theasalynso rt the aivary oe olde? Quite pat mete st ine sacha sy would tke tereis ako tua to whether wold be 2 ere me ae conconed with wha might le calla he consensual tne of he ‘stm, which s de instantly Oy te majo cea eens com tone eis byt tral reg ae Avie oe ecm the Tsang 1 theese urna el a 23h ter is wal nl a convene bt pres sorhand. Dep eit horthand. peng on sect tat gal me day ect? ea (ths vole rth dics delegation 5." vocabulary eter ced Lao’ et thi ame an the jain appear per fl Within at artic tape og dv Bus ty one Ieteoded to exemphiyam rumen, | pe tht te eer wil ac ha oe ett wing thea hs way utr sn 2 Foran example eh ner erecn he defn eb Tera al the det d's seo ty nach th bet dele ‘Sis aaeon he cet vei ir, og ad Pc 4 See name, Wine 190 and Lor be, Wane ees ow he tg of overpasom h ang Hand Paray we Caco poet page thos ee np ey ate ns Sa leet mse: Laur png evampe ee Bruns (1977), thw exalt tne ay adel ay eel ‘thee nnn cry eed em ane an a ‘anlar raeny company na Thebes eet ene Signet pane pty rah om ng ‘Sechanson teParsc, nate pty hppa oe nme My {ui rence nr eave ssh lt Wine ee ‘Ste am he hn oth err deta, en keene site verpne Thins pea oe icone ‘reed in unease ur 8 acs ‘pene darshan stature ste wide ees he Sit thay nt oe com ris ples art9.1m he Peech share play woo den desig se ee pla and eign tees orig). ht have tae ere. 10. am dain here om te tic bene aca” and “vgue™ ‘cae Haan Tevet 1, Natur, th diferent prof ie sytem ae reconnected automatic sentido change 12. The queton of “breakdown” it reant to his ie ad dees fer onideran. A "bread reats sl othe dea | have fre 4 ‘echnical eet, Th base ean only be unde 8 part ef practice "hat isthe enlapr ofthe ean bewecn aso apr a we ‘Abram hans the ay fhe cecil newark atic ipa Therapy ith wih the ar forthe eu of bee 13. Perhaps t weld be Heer to ay tat he iene bjt Iecpraic hom he craton am fami of eter rer ie sie Tht pc spor ec ced Mi “see were all cemmructed simulans. 13. Ais well kaw, Foueal (1975) ha des he a twee he hn ty the peter, power latins, and new orm fom