on the ethnography of non-humans and heterogeneous contexts (1)

Shadowing Software andClinical Records: On theEthnography of Non-Humans andHeterogeneous Contexts

Attila BruniUniversity of Trento, Italy

Abstract. Recent years have seen growing sociological interest in the rolethat objects and non-human actors perform in everyday life. Whether asmachines, information technologies, artworks, commodities or architec-tures, objects today raise issues of complexity and controversy (Pels et al.,2002). Borrowing from actor network theory the idea that humans and non-humans are actively involved in the making of social worlds, there arealready those who call for a post-social world and an object-centredsociality (Knorr-Cetina, 1997). But how can non-humans be observed?Sociologists are accustomed to socio-constructionist approaches to thesociology of science, or to analyses of tools and innovations couched interms of networks of actants; methodologically, however, it seems thatideas about how to proceed methodologically are not very well worked out.On the basis of a four-month ethnography conducted in a hospital that hasrecently introduced a digital clinical records system, I discuss the meth-odological aspects of shadowing non-humans. In particular, adoptingStar’s insight of an ‘ethnography of the infrastructure’ (Star, 1999), Iconcentrate on how to account for contexts characterized by multiple andnon-homogeneous actors and practices and on the implications of such aperspective for organizational analysis. Key words. objects and technolo-gies; organizational ethnography; practice; relational materialism

In recent years the social sciences have grown increasingly interested inthe role performed by objects and non-human actors in everyday life.Some anthropologists refer to new forms of reality in terms of ‘techno-scapes’ (Escobar, 1994) and, borrowing the idea from actor network

Volume 12(3): 357–378ISSN 1350–5084

Copyright © 2005 SAGE(London, Thousand Oaks, CA

and New Delhi)

DOI: 10.1177/1350508405051272 http://org.sagepub.com

articles

theory that humans and non-humans are actively involved in the makingof social worlds, there are already those who call for a post-social worldand an object-centred sociality (Knorr-Cetina, 1997).

But how can non-humans be observed? Sociologists are familiar withthe socio-constructionist approaches of SCOT and STS (MacKenzie andWajcman, 1999) and with analyses of tools and innovations couched interms of networks of actants (Law, 1992). I submit, however, that,methodologically, approaches to studying non-humans are not welldeveloped: actor network theory has not yet provided methodologicalindications that go further than ‘to learn from the actors without impos-ing on them an a priori definition of their world-building capacities’(Latour, 1999: 20); SCOT and STS can vary greatly depending on theauthor and on the specific field considered (Pickering, 1992).

Drawing on a four-month ethnography conducted in a hospital wherean electronic patient records system had been introduced, in this paper Idiscuss some of the aspects of studying contexts characterized by hetero-geneous actors and practices. Two major concerns run through the paper.On the one hand, it comments on how to study objects and technologiesas everyday organizational accomplishments (Suchman et al., 1999). Onthe other hand, it contributes to studies of organization as an assemblageof heterogeneous materials and practices (Cooper and Law, 1995; Lawand Mol, 2002).

The first part of the paper concentrates on three major issues related tothe inclusion of non-human actors in organizational ethnographies,namely how this redefines concepts of place, action and subject. Thediscussion highlights how such considerations focus attention more onthe relational game in which objects are involved (and which the objectsthemselves activate) than on the ontology of non-humans. The sectionalso outlines some of the methodological features that form the back-ground to observation of the non-humans. Then, following a briefdescription of the research design and context, some of the ethnographicdata collected in the study are reported. Finally the paper discusses how,methodologically, contexts that are characterized by multiple, and non-homogeneous, actors and practices can be accounted for, and the implica-tions of such a perspective for organization studies.

Challenging Organizational Ethnography: Spaces, Performancesand the Question of the S-Object

To paraphrase what Lucy Suchman (2002) has written about the demarca-tion line between human action and technological action, it can be saidthat one of the major issues in the contemporary social science debate isno longer where to draw a demarcation line between humans and non-humans but how to draw it. In the contemporary technical-scientificimagination, indeed, the sociality of machines is by now taken forgranted, for software increasingly assumes the form of ‘knowbots’

358

Organization 12(3)Articles

(Suchman, 2002), sapient machines that automatize knowledge in thesame way that robots automatized material production. Interactive pro-grams exploit human reasoning and interaction in a manner such thatmachines are attributed with properties of r(el)ationality, so that whatused to be classified as an ‘object’ is now an ‘emerging subject’ (Suchman,2002). In the contemporary western world, moreover, objects seem tohave become constitutive elements of expertise (as incubators of knowl-edge) and of social relations (as identity resources), thereby foundingwhat Knorr-Cetina (1997: 12) calls ‘object-centred sociality’.

In organizational ethnography, too, one witnesses mounting interest incontexts characterized by complex and technologically dense practices(Berg, 1997; Law and Mol, 2002); places in which human actors andtechnological objects work ‘together’ (Engestrom and Middleton, 1996);or virtual places in which human interaction is made possible bytechnologies (Hine, 2000). Technologies are inseparable from their use,and there is an intimate relation between space and the structuring ofactivities (Suchman, 1997). At the same time, whenever people use atechnology it is as if they leave a trace for future action so that, with time,technologies and artefacts are experienced as more or less relevant orcontextual, as negotiable or resistant, as facilitating or obstructing every-day life. ‘Objects interpellate us’, writes John Law (1999: 24) in order thatthe continuous reproduction of their meaning and practical utility sanc-tions their material stability.

The case of infrastructures and virtual environments may be partic-ularly instructive in this regard.

From Place to Space: Infrastructures and Virtual Environments

Typically, infrastructure is taken for granted: when we turn on a tapbecause we need some water, we rarely think of the vast infrastructurenecessary for us to be able to perform that action. The question becomesmore complex, however, when we begin to interest ourselves in thesituation of those who cannot take the existence of a particular infra-structure for granted: for an engineer designing and building an aque-duct, the pipes are not part of the infrastructure but constitute one of theobjectives of his/her work. Hence, objects and technologies that for someare simply part of the background are for others a matter for discussion.Infrastructure is consequently a relational concept (Star, 1999), which isdefined relatively to specific organizational practices. Orienting an orga-nizational ethnography to the study of infrastructure may therefore provea useful strategy with which to discern the practices distinctive of thecontext being observed.

Take, for example, the technical evolution of hospital beds (Strauss etal., 1985: 102). Initially, the beds were fixed, and helping an immobilizedpatient to change position was part of the nurse’s ‘comfort work’, whiles/he also ensured that the patient had assumed the correct posture forhis/her condition. Since the mechanization of hospital beds, patients

Shadowing Software and Clinical RecordsAttila Bruni

359

have been able to change position simply by pressing some buttons, butthis is not sufficient to ensure their comfort. In fact, the autonomy ofmovement by (the beds of) patients has led to indifference among thenursing staff towards the posture of the patient, whose comfort is thussubject to further problems caused by this neglect. The use of newmachinery/technology therefore modifies not only what from a distal(Cooper and Law, 1995) point of view appears to be simply an ‘infra-structure’ but also the autonomy of patients, and especially the moregeneral trajectories of action and the organization of work involved therein(Strauss et al., 1985: 67). There are contexts, that is to say, in which humansand non-humans constitute not a dichotomy but rather ‘an ecology ofrepresentations’ (Gibson, 1979)—persons, symbols and machines operatejointly to structure and renew understanding of a social situation.

Thus, even what we call ‘space’ can be thought of as the effect of anetwork (Law, 2002). The most striking example is provided by theInternet and computer networks, where communication flows throughnodes forming a network of localized associations in a space ever moreheterogeneous and independent of local settings (Hine, 2000).

Virtual settings and infrastructures thus represent the two extremes of acontinuum of spaces whose confines are formed more by connectionsthrough objects and practices than by material boundaries. Recognizingthe various types of action that give concrete form to spaces maytherefore be a route to follow when observing the sociality of objects anddelving into the materiality of the social.

From Action to Performance: Relational MaterialismJohn Law (1987) has used the expression ‘heterogeneous engineering’ todenote the process that gives (relative) temporal and spatial stability to theorganization of persons, texts and objects. According to Law (1994: 2):

What we call the social is materially heterogeneous: talk, bodies, archi-tectures, all of these are implicated in and perform the social.

From this point of view, materiality and sociality appear to be the jointeffect of diverse organizational strategies. Humans and non-humans arethe alternate products of social and technical relations, and stabilityresides in the possibility of performing the heterogeneity of the social(Law and Mol, 1995: 274). The materiality of the world, action andobjects are therefore the relational effects of a contingent order: it is notthat materials are differentiated by their durability; rather, the networksthat produce them generate effects of differing durations in time andspace (Law, 1994; Callon and Law, 1995).

Law (1994) therefore suggests that closer attention should be paid tothe roles of the materials involved in inter/action: the notion of relationalmaterialism tells us that reality exists in a multitude of material forms,and that these are not simply given in nature but are the more or lessprecarious effects of ordering strategies. In other words, for social (andnow socio-technical) relations to be able do their relational work of

360


joining and blending, they must be performed. This invisible work of thesocial is what Law and actor network theory call ‘performativity’, and itcan be observed in the tension that arises among the various elementsmobilized into action (Gomart and Hennion, 1999; Bruni and Gherardi,2001).

Hence, the problem is no longer whether machines have a capacity foraction; rather, the problem is how to define action, be it human or non-human (Lynch and Collins, 1998).

From the Subject to the S-Object: Intentionality as a Transitive Property

To date, the question of the subject in ethnography has mainly concernedthe role of the researcher (Clifford and Marcus, 1986; Van Maanen, 1988)and representation of the various ‘voices’ contributing to the ethno-graphic text (Van Maanen, 1988; Hammersley and Atkinson, 1995). Butproblematizing the titularity of action carries much more radical con-sequences. One may say, indeed, that one of the aspects of ethnographicresearch most taken for granted has always been the attribution of agencyto humans (Stone, 1995), however much postmodernism and actor net-work theory may have stressed that subjectivity is more a textual andrelational effect than an intrinsic property of actors (Manning, 1995;Latour, 1999).

Interpreting objects as active entities only partly under human controlunhinges the assumption that the subject of ethnography (human inter-action) is ‘naturally’ given, shifting attention to the aesthetic (Strati,1992), ‘appresentative’ (Knorr-Cetina and Bruegger, 2002)1 and socio-material (Harbers et al., 2002) dimensions of the s-objects in question. Itis accordingly important to note that studies on computer-supportedcommunicative work or workplace studies (the strands of enquiry thathave made observation the method par excellence for the study oftechnology-mediated human interaction) have, from a certain point ofview, dissolved technologies into more complex social dynamics. Atten-tion has shifted to the human negotiations that take place on occasions ofdesign or breakdown, with the study of the performative and reflexivedimensions of technological objects being forgotten (Woolgar, 1988; But-ton, 1993; Berg, 1998). What is lacking, that is to say, is the study of whatethnomethodology calls the ‘quiddity’ of the object and its contributionto the creation of a meaningful world, with the consequent danger of asort of sociological reductionism that eliminates the distinctive featuresof the object under examination (Woolgar, 1991).

In this regard, Harre (2002), drawing on Gibson’s notion of ‘affordance’(Gibson, 1979), proposes that objects can acquire different identitiesaccording to the narrative that constructs them socially, but always on thebasis of certain material factors. Objects are constructed in the relationsthat they establish with humans, and their performance of a more or lessactive role in social life is due not to their properties but to the type ofrelation. As Harre (2002: 24) points out:


361

Nouns like ‘flag’, or ‘dollar’ or ‘shop’ are different from nouns like ‘water’,‘sand’, ‘arm’ and the like. The latter do not require any particular socialsetting to complete their sense, while those in the former list do. They areindexicals. Social substances are picked out by expressions the senses ofwhich are incomplete unless related to a particular flow of social acts, aparticular social world.

Although Harre’s main concern is to argue for the predominance of thesemiotic-discursive dimension in the genesis of a ‘social substance’, Ibelieve that pointing out the context-dependency of objects also explainswhy it is important to observe them in ethnographic terms. Objects are s-objects because (like subjects) they always stand in relation to a socialworld, so that ‘observing’ an s-object means looking at the relations ofwhich it is part, the contexts in which it is located, the practices thatconstruct it socially, and the other s-objects that cross its trajectory.2

Thus is made explicit the main difference between postmodern andinteractionist ethnography. However much postmodernism and symbolicinteractionism may share an anti-theoretical stance, together with aninterest in the symbolic production of the real and in the processes bywhich order is negotiated, the focus of the postmodern debate is not onindividuals but on the relations that enable them to accomplish theposition of ‘subject’ (Manning, 1995). ‘Being-an-actor’, recognizing inten-tionality in action, being identified as the ‘origin’ of an occurrence, is notan essence or something that people acquire once and for all (Law, 1994;Latour, 1999); rather, it is a transitive (and transitory) property, evercontingent and precarious, which is located in certain s-objects more bycollective action than by individual will.3

The Research Setting and a Note on MethodThe observations presented in what follows are part of a broader researchprogramme in the field of telemedicine conducted by the Research Uniton Cognition, Organizational Learning and Aesthetics (RUCOLA) (seeGherardi and Strati, 2004). Our interest centred on how subjects withdiverse practices and areas of expertise can interact at a distance, andhow the sharing of these practices and areas of expertise might engendera community of virtual and/or technologically mediated practices.

The ethnographic data have been extrapolated from four months offield observation (conducted by me) in order to examine the introductionof an Electronic Patient Record (EPR) in a hospital department (Bruni,2003). The EPR is the result of one year of participatory design betweeninformation engineers and doctors; it permits the ‘sharing’ of patientsamong the doctors in the hospital department (through an inter-hospitalnetwork) and it was conceived not only as an information-sharing andmanagement tool with regard to patients, but also as a supplement tocertain organizational processes (e.g. keeping the appointments diary).This means that, albeit to different extents and for different reasons,

362


doctors and nurses make daily use of the EPR system, whose prototypewas implemented in June 2001.

A Note on Method

Access to the field was negotiated mainly with the chief consultant in thehospital department, who gave his permission for a researcher to partici-pate in its everyday activities. On the request and ‘advice’ of the consult-ant, the observation was simultaneously overt to the medical and nursingstaff and covert to the patients, so that, although introduced to thedepartment’s personnel as a researcher, I wore a white coat, which mademe unrecognizable to the patients.

Although it had been decided that I would mainly engage in thestructured observation (Mintzberg, 1971, 1973) of certain organizationaloccasions (team meetings and discussions of clinical cases) and of certainpersonnel (doctors and nurses), after about a month of observations Inoticed that the EPR was located in some spaces and times but not inothers; and that some actors showed especial confidence in using thesystem whereas others seemed unaware of its existence. In a not entirelyconscious, and almost haphazard, manner I therefore decided to embarkon a different type of observation: for a further month or so I concentratedon ‘shadowing’ (Sachs, 1993; Fletcher, 1999; Bruni et al., 2004) the EPR,letting the software guide me through the organization and confront mewith other actors and processes, whether human or artificial. In fact, aswe shall shortly see, involving relevant actors in the construction of thenew organizational tool was not enough for it to participate withoutdifficulty in the everyday organizational life of the unit. The process wasnot sufficient to render the EPR an ‘expert’ in inhabiting the variousorganizational spaces and participating in quotidian work, so that itappeared to me as a ‘newcomer’ within a more composite ‘community ofobjects’.

Given that (as far as I knew) there are no methodological indications onthe matter,4 I acted more on intuition than on the basis of any methodo-logical prescription, bearing in mind that sometimes following casualtiescan be a research rationale too (Becker, 1994).

A (partial) result of such an observation is recorded in what follows.

Shadowing Electronic and Clinical Patient RecordsThis section is based on the ethnographic accounts collected as Iobserved the everyday life of a day hospital (DH) where cancer patientsare examined and receive their therapies. In fact, whereas the departmentis divided between the ward (for hospitalized patients) and the clinic (forpreliminary tests and follow-up therapies), inside the DH one canobserve a flow of events and meet the multitude of actors and objects thatmake up its everyday routine. Description and interpretation of theethnographic data will proceed in parallel. The linking theme will be


363

the bond among objects in use, actors and everyday organizationalpractices.5

Receiving Patients, Arranging ObjectsThe main reference point (in organizational terms) for the day hospital’sactivities is the patient reception area. This is where the nurses receivethe patients coming into the department for chemotherapy; this is wheredoctors and patients ring if they need to speak to the day hospital’snursing staff; and this is the location of the computer used by the nursesfor the EPR, as well as of all the paper-based clinical records for patientscurrently receiving treatment. At present, in fact, for a patient record tohave legal validity it cannot exist solely in electronic form but must alsohave ‘materiality’ (i.e. it must exist on paper and bear the signatures ofthe doctor and the chief consultant). Each EPR is therefore printed outand placed in a folder containing all documentation on the patient.Moreover, although since June 2001 all folders have existed in electronicform as well, patient records produced prior to that date have not beendigitized and continue to exist in paper form alone. In short, all patientrecords continue to exist in paper form, some as originals, others as print-outs of their corresponding EPRs.

Patient reception is organized in a structured manner, although thereare a number of possible ‘variations on the theme’:

1. The patients present themselves at the counter. They give theirnames, although they may be immediately recognized by the nurse.Or they may say nothing as they hand the nurse an envelopecontaining their test results.

2. The nurse checks the patient’s appointment. She may do so on thecomputer or by consulting the daily print-out (always present) of thecomputerized appointments.

3. The nurse retrieves the patient’s clinical record from a wheeled tray(trolley). All patients are ‘accompanied’ by their clinical record whenthey see a doctor (who consults it and updates it). Because appoint-ments are programmed, the nurses place the clinical records of thepatients expected for that day on a trolley. Although the records areusually in the right place, it may happen that they are somewhere inthe reception booth but not on the trolley. Or they may be on thetrolley but filed under the wrong letter.

4. The nurse checks that the haemachrome test results have been placedin the patient’s folder. Because of the side-effects of chemotherapy,certain blood values (typically haemoglobin and the production ofwhite corpuscles) are constantly monitored. These values are essen-tial medical benchmarks, for without them the doctors are unable todecide whether and how the pharmacological therapy should pro-ceed. It is therefore absolutely necessary for them to be present in thepatient’s folder. Some patients bring their test results with them;others have been tested that morning and the results reach the nurses

364


via fax or via an intranet linking the hospital’s various departments toits laboratory. If the clinical record is paper-based, the nurse only hasto highlight the values (using a marker pen) for the doctor to look at,and then place the test result sheets in the patient’s folder. If theclinical record is instead in digital form, the blood values are fed intothe computer (which automatically highlights those outside theparameters).

5. The nurse writes a number on a sheet of paper (a ‘day hospital form’),which she inserts in the folder, and also gives that number to thepatient. The numbers are progressive and are automatically assignedby the software on the basis of the patient’s place on the day’sappointments list. The number in the folder is used by the doctors tocall the patients. The number given to the patient is printed on agreen-coloured slip of paper if the test results are normal (so that thepatient receives his/her usual therapy) or on a yellow slip if the bloodvalues are out-of-parameter (in which case the doctor decides what todo). It may happen that there is no slip bearing the patient’s number,in which case the nurse writes it on any piece of paper that comes tohand.

6. The nurse tells the patient to take a seat in the waiting room. Thepatient may leave the counter in silence, or talk briefly with thenurse, or protest about the colour of his/her slip, or complain becausehis/her test results have not yet arrived. Finally, it may happen thatthe patient asks to be examined by one particular doctor in thedepartment.

Observation of a routine activity such as patient reception immediatelybrings a large number of objects to notice: clinical records, diaries, sheetsof paper, test results, slips of paper of various colours, Post-it notes, aswell as the computer and the EPR. All these objects are vital for patientreception, and they comprise ordered relations that materialize in thepatient/nurse interaction.

The most important object, the one that takes precedence over allothers, is the file containing the patient’s clinical record (whether paperor electronic). This sets out the patient’s clinical history and if it ismissing this invalidates the interaction. Second in importance are theday’s test results. These signal the patient’s situation ‘at present’. If theyare lacking, this invalidates the presence of the clinical record folder,because it lacks the most recent clinical results. Then there are the ‘dayhospital forms’. These testify to the fact that the patient has been‘received’, that s/he must be examined, and that s/he will encounterfurther therapeutic and pharmacological objects via a medical service.Last but not least come the coloured numbered slips. The purpose ofthese is to give a ‘position’ to the patient and to furnish (by means of theircolour) a rough indication of his/her present condition. These slips arethe only objects not indispensable to the performance of the receptionactivity, because their absence can be easily remedied by using more


365

anonymous pieces of paper. At the same time, however, they are the mostconcrete source of conflict in the nurse/patient and doctor/nurse inter-action, and they must therefore be managed with care.

All these objects are intimately interconnected. In as much as theirarrangement does not come about ‘automatically’ but requires humanintervention (principally by the nurses), they enjoy varying degrees offreedom and develop relations that involve other objects and ‘guide’human interaction. This is manifest at times when these objects displaydifferent levels of importance, thereby directing the actors’ attention indissimilar ways. But it is evidenced even more clearly by the (numerous)disruptions to the established order that (frequently) occur during thepatient reception procedure.

Objects in RelationGiven the number of elements involved, the nurse/patient interactionmay assume a variety of overtones and it needs only one element to goamiss for the entire process to take a different form. The two followingepisodes (Box 1) illustrate possible variations.

Box 1

A ‘new’ patient reports to the reception counter, accompanied by afemale relative. When the patient has left the counter the nurse leavesher position for a moment, and the patient’s relative is left alone at thecounter. She slides her hand under the glass screen and opens thepatient’s folder, and then quickly closes it again. When the nursereturns, the woman tells her that there are certain things that its betterhe [the patient] doesn’t know . . . he knows he’s here for chemo, to easethe pain. The nurse immediately understands the situation. She deta-ches a Post-it note, writes something on it and attaches it inside thefolder. In the meantime, the woman tells her that she has alreadyagreed with the consultant on the (non)communication of the diag-nosis, but the nurse tells her that today the patient will be examined bya different doctor. The woman immediately asks if she can talk to thisdoctor. As they discuss the matter, the woman asks (referring to thepatient): Is he watching? Nurse: Yes, but I’ll tell him that we weretalking because I was explaining the therapy to you. The womaninsists that it is vital that the doctor examining the patient today doesnot tell him ‘everything’; Nurse: I’ve already notified the doctor, heknows what to do . . . it’s his job.

A ‘new’ patient comes to the counter, but without any blood testresults. This strikes the nurse as very odd, but the patient insists thathe has never had any tests. The nurse says that he should have themdone anyway (and immediately). She gives him a number and askshim to take a seat. The Charge Nurse (who is in the reception area at

366


the time) tells the nurse that, in fact, it was the consultant (after thefirst examination) who did not prescribe blood tests for the patient.The nurse answers that the doctor is certainly going to ask for thembefore he starts the therapy. The Charge Nurse points out that in anycase the tests prescribed for patients serve to check that certain values(e.g. for white corpuscles) are normal once the therapy has begun, sothat if a patient has not started the therapy yet . . . Five minutes later,the doctor comes to the reception area asking why he has been sent apatient with no test results. The nurse explains to me that ‘they’ (thenurses) know that the doctor in question will not initiate the therapy ifhe has not seen the test results first.

In these two episodes, the human actors are called upon to ensure thatobjects interrelate in the manner prescribed for them. In the situationwhere the patient’s relative reaches agreement with the nurse on (non-)communication by the doctor of the diagnosis to the patient, for example,the relations that the clinical record entertain with other objects are offundamental importance. That the clinical record is an object able todescribe and activate ‘dangerous relations’ is manifest from the beginningof the event, when the patient’s relative takes advantage of the momen-tary absence of the nurse to peek at the contents of the file. We do notknow exactly what she saw, but it is likely that she looked at the objectspresent in the folder (test results, images, reports, therapeutic decisions),at the relations among them, and at the scenario depicted by them (it isnot implausible that an examination is linked with an image that relatesto a medical report on the basis of which a course of therapy is decided).That the objects are interrelated and that this has a particular significanceis intuitively obvious, and what worries the patient’s relative is preciselythe fact that the doctor may make this ‘intuition’ explicit to the patient.The patient is in fact aware of some of the relations described by theobjects in the folder (as the relative says, he knows that the therapeuticobjects relate to the medical reports attesting to his pain), but he does notknow what this implies and he does not know what other relations mayarise from that moment onwards. The nurse makes sure that theserelations are not made manifest by introducing a further object into thefolder (the Post-it note), whose purpose is to call the doctor’s attention tothe significance for the patient of the relation among the objects.

This is the selfsame tactic used by the nurse when she suggests to therelative how she can justify their conversation to the patient: tell him that‘I was explaining the therapy’. Thus, explaining what relations (organiza-tional and curative) obtain among certain objects (drugs, test results andmedical reports) seems to be a plausible justification for not makingexplicit other relations (among objects) that might arouse suspicion.Moreover, the activity (and, in a broader sense, the ‘professionalism’) ofthe doctor about to examine the patient is predicated on ‘reassurance by


367

objects’, for his expertise consists in being able to handle relations amongobjects so that they are a-problematic and ‘natural’ in the eyes of thepatient.

Whereas in the situation just described it is the overabundance ofobjects that preoccupies the actors, in the second episode the reversesituation obtains. Here it is the lack of an object (a patient’s haemato-chemical test results) that is the focus of discussion. As the ChargeNurse stresses, when a patient is ‘new’ (and has therefore not yet beguntherapy) the absence of blood test results should not be a problem,because the clinical record folder does not contain the elements withwhich these results would be associated (namely the therapy). The nursenevertheless seems aware of the relation that in any case establishes itself(in the doctor’s eyes) between the therapeutic objects and the test results:one object (the nurse seems to be saying for the doctor) stands ‘bydefinition’ in relation with another and automatically requires its pres-ence. The knowledge of the doctor and the nurse (as opposed to that ofthe consultant and the Charge Nurse) here seems to be expressed inthe fact that the relations among objects must be safeguarded, even ifthis means contradicting what a colleague has said. In organizationalterms, the interaction that arises among the objects present resolves theconflict of opinions among the actors by translating it to a plane parallelto that of power and hierarchy. What the nurse and the doctor havedecided does not invalidate the reasoning of the consultant and theCharge Nurse, because the doctor and the nurse comply with the align-ment activity required by the objects. It is not organizational relationsamong humans that are in dispute, but rather organizational relationsamong objects.

The Electronic Patient Record Meets a ‘Community of Objects’As shown by the patient reception activity, the ‘community of objects’within the day hospital is an extremely cohesive and complex com-munity. It is advisable, therefore, to consider how cohesion andcomplexity come into being amid the routine of an ordinary working dayin order to understand how the EPR participates in organizationaldynamics.

That the EPR is an object that is now part of the department is apparentas soon as the day hospital opens in the morning.

Box 2

The day begins for the DH at 7:30, when the head nurse opens thedoor, switches on the lights, turns on the computer, and so lets it beknown that someone’s there. The nurse then prepares the drip feeds forthose patients who do not have to be examined by a doctor becausetheir therapy programmes have already been decided. At 8:00 thewaiting patients are summoned and seated in the infusion zone to

368


undergo their therapy. The nurse prepares ice, attaches the drips anddistributes the ice among the patients.

At 8:30 another nurse arrives, and the DH’s activity proper begin-s:checks are made that all the clinical records for the patients expectedduring the day have been ‘prepared’6 (the nurses use a computer print-out of the appointments list, on which they note any anomalies) andthe first patients are summoned in the order established by thecomputer. Stacked on the table in the reception booth are clinicalrecords to be archived. These are marked ‘ARCH’ in red, and they oftenalso have Post-it notes attached stating ‘CLOSE DH’. Some of themmay also be marked with a cross (in blue), which signifies that the fileis to be archived because the patient has died.

The telephone rings constantly, and the first test results begin toarrive via fax.

The chief nurse collects the EPR print-outs for new patients so thatshe can ‘open’ the corresponding clinical files. She uses a rubberstamp to number the folders containing the new EPRs and copies thenumbers into a notebook.

The brief account in Box 2 of the beginning of a (hypothetical) day showsthat the EPR is a commonplace object among those used in the DH. It is‘let known’ that the DH is open not only by activating its usual objects(lights, doors, drip feeds) but also by turning on the computer. The print-out of the day’s appointments is one of the first objects consulted in orderto verify the presence of other objects, and it is the software that managesthe timetable of patient examinations. Likewise, the head nurse whostamps numbers on the files containing the EPR print-outs, and thenrecords them in a notebook, testifies to the EPR’s membership of theorganization’s ‘community of objects’, as well as demonstrating how theEPR relates to the objects-in-use.

Further demonstrating the EPR’s growing importance in everydayorganizational life is the manner in which it, together with other objects,sets times and actions. The ‘community of objects’ with which the nursesinteract, in fact, is also characterized by the temporal relations that theobjects establish among themselves, and by the times that distinguishtheir practices of use. This is evinced by the manner in which relationsamong the objects materialize organizational times (the ‘switching on’ ofthe day hospital, the administration of the therapies), but also by sucheveryday micro-events as drinking a cup of coffee while a particularlylong chemo-infusion is being administered; or taking a break fromreception work when the clinical records folders on the trolleys outsidethe surgeries have been counted—these folders correspond to the patientsin the waiting room, so by counting them the nurses know how manypatients have been treated and how many are still waiting.

It may also happen that the software modifies the action hierarchies.


369

Box 3

A doctor comes into the reception area. He wants to scan some testresults. He switches on the scanner and starts the procedure, but thecomputer crashes. When the nurse returns (she has been deliveringsome folders to doctors), she jokingly scolds the doctor for crashing thesystem. The doctor ‘intercepts’ a computer technician who happens tobe in the department to install a printer and asks him to help. Thetechnician is willing to do this but, as long as the system is down, thenurse cannot continue with patient reception. The three decide topostpone everything until later: the doctor does not scan the testresults, the technician does not repair the scanner, and the nurse cancontinue with her work.

The episode in Box 3 illustrates how the actors consider software to be animportant agent in the organization of everyday work, to the point that itsfailure to work re-directs the usual trajectories of action. Note, moreover,that, in becoming an everyday object, the EPR acquires a dimension superpartes: as in the previous case, when the nurse and the Charge Nursediscussed the objects needed to examine a patient, once again thedecision taken by the three actors is not contradictory, even less con-flictual, because it only concerns equilibrium among objects (software,scanner, test results).

Objects thus mark out trajectories for human action, and the organiza-tional activity of the DH is made possible not only by the existence of acommunity of human actors (the nurses) but also by continuousrelations/interactions among non-humans. This signifies that the EPR hasbecome an object ‘internal’ to organizational practices not only because itis socialized to the use made of it by the actors, but also by virtue of itsencounter with the ‘community of objects’ present in the DH.

The Electronic Patient Record as a ‘Newcomer’However, the EPR is still a ‘newcomer’ in the everyday organizational lifeof the day hospital and thus actors are called into action to mentor itsrelations with more expert and complex objects and practices (see Box4).

Box 4

I observe Elisa (a nurse) at work in the infusions therapy section. Sheshows me how and where the drugs to be administered are prepared. Itis a closed and isolated room and, because of the toxicity of the drugs,the infusions are prepared in a glass cabinet (which the nurses refer toas ‘Chernobyl’).

While we talk, Elisa empties vials and fills syringes (almost withoutlooking) and consults the print-out of the day’s appointments (anotherprint-out is affixed close to the drugs trolley, so we start getting them

370


ready). Elisa notices that, strangely, one patient has two appointmentsthat morning. This is odd because, if the person fixing the appointmentmakes a mistake when feeding in the data, the software should blockit.

Then a colleague tells Elisa that the computer has printed out thewrong therapy! Elisa’s colleague has noticed the mistake (beforeadministering the therapy) because she knows the patient (and there-fore knows what type of therapy he is receiving), she knows thetherapy (and therefore the drugs used for it), and she remembers thelast time that the patient had therapy (and therefore what stage of thecycle he has reached).

Continuing to talk, Elisa and her colleague add further details(obvious things, they call them). For example, it is always better to‘take a look’ at the therapy (quantity, type, cycle) before administeringit, because the program is a bit rigid in its structure [explains Elisa].When the cycle requires a particular order, a particular drug, and thenfor some reason it has to be reduced . . . you have to be very carefulbecause he [the software] always sets the same therapy at 100%. Sothat he [the doctor] often says ‘Reduce the dose’, but he doesn’t reduceit, because you have to go into the first . . . first memory.7

Moreover, patients are (usually) attentive to their therapy. They keepa check on what the nurses are doing and call their attention to certainmatters. I realize from the patient [continues Elisa], because when I’vefinished the examination, he says. ‘Ah, you know, the doctor said he’dreduce the dose this time. . .’. I see that the therapy sheet instead saysit should be 100%, so that . . . prompted by the patient I can go to thedoctor and ask about it, so we notice things.

In addition, the Kirkner therapy (the one in question) is particularly‘difficult’ because of the massive quantity of drugs administered andthe complexity of the therapy cycle. So when they read KIRKNER onthe computer print-out, the nurses know that they must take especialcare. Also we [the nurses] all know that the computer makes mistakeswith Kirkner, because it does not log changes in the therapy scheduleand continues to impose the same sequence (Elisa says that thecomputer technicians have been told about the problem but theyhaven’t done anything).

Finally, the nurses know that the doctors who spend a great deal oftime in the clinic are particularly unreliable as far as Kirkner isconcerned. This is because they are used to setting the type of therapy,but it is then the computer that handles the scheduling of the therapy,and nobody bothers to check whether changes have been made(rumour has it in the department that ‘Those in the clinic don’t knowKirkner’).

Elisa adds that her greatest worry is that a problem of ‘abbreviations’may arise, because we write . . . I mean, the computer gives us thetherapy sheet but not the label. She explains that the computer prints


371

out the therapy prescribed by the doctor, but not the label that thenurses must attach to the drip-feed bottle (once they have prepared thetherapy). As they write the labels, the nurses abbreviate the names ofthe drugs, so that Taxol becomes TAX and Taxotere becomes TXT.Elisa says that all these ‘passages’ (the doctors who types into thecomputer, the nurses who read and abbreviate, the colleagues who‘decipher’ the abbreviations) increase the risk of errors, even if an errorof this kind would be unacceptable, because then usually . . . also mycolleagues . . . we know that Taxol goes into the breast and Taxotereinto the lung, so . . . but the computer could reduce these risks.

The problem in this episode springs from the fact that objects havemultiple interrelations that may vary in space and time. At the centre ofdiscussion is a particular object (the Kirkner therapy), which is defined(by the nurses) as ‘difficult’ because of the relations that tie it to differentplaces and times of the organization and that cause it to change. In fact, ata hypothetical ‘time zero’, when the therapy is in the consulting room, itis invariably itself, in the sense that it has a pre-established interactionwith the other objects and it is therefore not necessary to verify itsrelations (the therapy associates standard quantities of infusion with afixed administration schedule). But at a subsequent ‘time one’, when thetherapy is in the day hospital, it is susceptible to variations (according tohow it has interacted with the patient and his/her haematochemical testresults), and it is therefore advisable to check its relations with thequantity of the drug and the therapy cycles.

On the basis of the accounts given by the actors, the EPR commitserrors in relating to this object because it is unable to take account ofthese spatio-temporal variations. The software ‘always sets the therapy at100%’; that is, it assumes that the object is invariant. This is because thesoftware is ‘a bit rigid in its structure’ (according to the nurses), althoughit cannot be blamed for this. Like any ‘newcomer’, the new technologicalobject has learnt experientially from the places and actors that it hasencountered. Hence, although it has met other objects and experiencedtheir behaviour in specific spaces and times, it is ‘normal’ for it not toknow their overall behaviour within the organization. The actors areaware of the difficulty of this relation. They consequently activate asystem of cross-checks (the therapy sheet is checked by the nurse, who inturn is monitored by the patient, whose control can be retroactivelyverified with the doctor), which helps the software by means of a processof error anthropomorphization that condenses erroneous relations amongobjects into the isolated action of an actor (so that in the end it is aparticular doctor, and not the EPR, that does not know how to relate tothe therapy).

The spatio-temporal variation of an object is also responsible for theday hospital head’s ‘greatest fear’ that therapeutic objects may interacterroneously. Although in the episode reported in Box 4 the erroneous

372


interaction resulted from the fact that one object (the EPR) failed to takeaccount of the spatio-temporal variations to which a relation may besubject (and continued to set the same therapy), in this case the fact thatthe software did not establish a possible relation with other objects (thelabels attached to the drip-feed bottles) created the risk that they mightassume different meanings in time and space and thereby establisherroneous relations among themselves. As in the previous situation, the‘risk of error’ was related to human action/interaction, and the expertiseof the actors was called into question by the possible relations amongobjects.

From a linguistic point of view, this is instantiated by the first fragmentin the explanation of ‘what the nurses know’, where the subject is nevermade explicit and oscillates between doctor and software, leaving it toactions and relations to identify the actor in question. Again from alinguistic point of view, these oscillations of the subject continue in thesubsequent explanations, involving other actors as well (‘the computer’and ‘the Kirkner’). The result is a rumour that ‘circulates’ in the depart-ment, perhaps by antonomasia an example of ‘social substance’ (Harre,2002).

In terms of relations, the interactions among objects construct networksof action and practices that may lead to identification of a ‘new’ subject,and it is in interaction that the greater/lesser activity/passivity of the s-objects is defined. Even an abbreviation (such as TAX or TXT), ifadequately contextualized, may take centre stage and redefine the hier-archies between humans and non-humans.

Conclusions: On the Ethnography of Non-Humans andHeterogeneous Contexts

In the course of this article I have presented extracts from ethnographicobservations that have given rise to a hybrid and multifaceted definitionof actors and spaces of action. Of these I have (deliberately) furnished aninterpretation that further emphasizes the difficulty of controlling andordering the phenomena considered. In these conclusions, I shall concen-trate on two points: (a) how ethnographically to represent contexts thatare characterized by multiple and non-homogeneous actors and prac-tices; (b) what this approach can offer to organizational analysis.

Regarding the first point, let me start from a preliminary consideration:objects are restless. Alberto Melucci (1998) has used the image of thegame of croquet played by Alice in Lewis Carroll’s Alice in Wonderlandas an apt analogy to illustrate how research fields are animated incontemporary ethnography: Alice’s croquet game took place on an ‘ani-mated’ court (the balls were porcupines, the mallets were flamingos, andthe hoops were soldiers bent double) with which she interacted. Melucciemployed this image mainly to illustrate the ‘reflexive turn’ in qualitativeresearch (Alvesson and Skoldberg, 2000), denoting a twofold process in


373

which the researcher is directly involved in the reality observed whilehis or her image as ‘privileged observer’ is relocated in contexts of actionand takes part in the production of meanings. But whereas it is usuallythe subjects observed that do not dutifully behave according to theresearcher’s expectations (emerging from the context to interfere with hisor her action/observation), in the account set out above it is objects thatcause this disturbance. From an absolutely subjective point of view, infact, it was objects that involved me in their relations: I had no alternativebut to follow them. The main problem arose at the moment when Irealized that I could have described the reality confronting me by skirtinground the actors, but I could not evade the objects. And this was because,as I have sought to show, in the organizational context in question the(human) actors relied on the relational objects that embodied identity,power, risk, uncertainty, control, and so on. Moreover, (human) actorsseemed much more attracted, intrigued, sometimes preoccupied, by therelations established among objects than they were by the predictableaction of more ‘canonical’ subjects.8 By way of example, patient receptionoperations were unhinged by the fact that objects designed to follow asingle, fluid and sequential trajectory often become entangled in ranges ofaction that call subject/object and activity/passivity distinctions intoquestion.

Observing non-humans therefore requires one to plot the connectionsamong different courses of action, and to determine how actions andsubjects define each other in relations. From this point of view, shadow-ing non-humans requires the ethnographer to be able to orient his/herobservations to the material practices that perform relations, and prob-ably also to devise new narrative forms able to make that performanceaccountable.9

In terms of organizational analysis, such an account shows how theprocess of joint design whereby programmers (computer technicians) andfuture users (mainly medical personnel) get together to create a techno-logical platform for a new organizational tool is, in fact, not enough tomake the technological object ‘expert’ in inhabiting the organization’svarious spaces, so that it appears as a ‘newcomer’ in the more composite‘community of objects’ marking out the material boundaries of organiza-tional practices. The technological object ‘contends’ with this communityfor its practical relevance (are the ‘old’ clinical records better than the‘new’ EPR?) and ‘negotiates’ with the objects already present in theorganization for spaces of action. In this sense, the EPR is a technologythat ‘works’ inasmuch as it is able to form and activate diverse groups oforganizational actors, which, in appropriating it, simultaneously make itcompatible with everyday work practices and the existing ‘community ofobjects’. In fact, the EPR differs from other organizational objects (prima-rily paper-based clinical records) in one important respect: its lack ofmateriality. This requires the presence of other objects that attest to itsconcrete existence (from a legal point of view as well): therapies, dis-

374


charge letters, paper print-outs, letters to the GP, test results—these are allmaterial forms into which the EPR must translate itself in order toacquire organizational materiality and encounter other actors.

Finally, I believe that including non-humans in organizational ethno-graphies is an interesting opportunity to extend the boundaries of ethno-graphy so as to include the study of ‘boring things’ (Star, 1999), showingsimultaneously that things (such as electronic and clinical records) mayperhaps be boring because of the organizational relations in which theyare involved, not because of their properties. Boredom springs from theinertia of matter or from the monotonous repetition of the same actions.An ethnographic approach may help to show that matter is not so ‘inert’and that the monotonous repetition of actions is something that alsopertains to ethnographic representation.

For more than 20 years, ethnography has found that the symbolic-interpretative approach is the one most congenial to it, and its encounterwith postmodernism has enabled it critically to re-examine many of theproblems concerning representativeness and the representation of thecontexts observed. Perhaps the time has come to discuss ethnographicmaterialism.

Notes1 The expression ‘appresentation’ is borrowed from Husserl (1960). It denotes

the ‘bringing into being’ of an event, and it differs from ‘representation’ or‘presentation’ in that it does not imply a difference between an essence(reality) and an appearance (image).

2 An anonymous reviewer pointed out that I should have developed a morebalanced view of the complementary roles of humans and non-humans: ifthere are ‘s-objects’, perhaps there are also (s)ubjects. But s/he concluded thatthis may well be only a word game. I take word games seriously, and thank thereviewer for this remark.

3 An anonymous reviewer asked what the difference is between such anapproach and ‘standard’ actor network theory (ANT). I am grateful to thereviewer for this remark and I would like to reply that ‘standard’ ANT doesnot use ethnography (not explicitly, at least).

4 Apart from passing mention in an article by Winthereik et al. (2002), whichcame to my notice only recently.

5 In the boxed text, phrases in italics represent direct speech transcribed in ‘realtime’; they are consequently reported exactly (though translated) as they wereexpressed. Finally, in order to ensure the anonymity of the persons involved,I almost always refer to a generic ‘doctor’ or ‘patient’ or ‘nurse’.

6 ‘Preparing’ a folder means organizing the documents; checking whetherspecific analyses have been ordered (and, if so, fixing appointments with therelevant departments); ensuring that the folder contains the results of allprevious examinations; if the folder is in electronic form, printing out thetherapy programme and adding it to the clinical record.

7 In Italian, there is no neutral gender, so in this particular case the nurse wasreferring to both the doctor and the software as ‘lui’ (he/it). To capture thisambiguity, I have used the pronoun ‘he’ in the English translation as well.


375

8 An anonymous reviewer pointed out the relevance of human pre-interpretation and sense-making activity (Weick, 1995) in situations of break-down. I agree, although I would also like to highlight that sense-making inWeick’s terms is basically a human cognitive process, whereas here ‘sense-making’ takes place on a material level as well, so that objects’ relations also‘make sense’ of human action/interaction.

9 An anonymous reviewer asked if I was claiming to have done this in thecurrent paper. I would say that the paper is an attempt to do this, although(like all such attempts probably) it is not successful and it may soundirreverent and provocative.

ReferencesAlvesson, M. and Skoldberg, K. (2000) Reflexive Methodology. London: Sage.Becker, Howard S. (1994) ‘“Foi por Acaso”: Conceptualizing Coincidence’, Socio-

logical Quarterly 5: 183–94.Berg, Marc (1997) Rationalizing Medical Work. Cambridge, MA: MIT Press.Berg, Marc (1998) ‘Politics of Technology’, Science, Technology and Human

Values 23: 456–90.Bruni, Attila (2003) ‘Objects’ Sociality and Materialities of Organizing: Ethnog-

raphy of a Telemedicine Project’, PhD thesis, Department of Sociology andSocial Research, Trento, Italy.

Bruni, A. and Gherardi, S. (2001) ‘Omega’s Story: The Heterogeneous Engineeringof a Gendered Professional Self’, in M. Dent and S. Whitehead (eds) ManagingProfessional Identities. Knowledge, Performativity and the New Professional.London: Routledge.

Bruni, A., Gherardi, S. and Poggio, B. (2005) Gender and Entrepreneurship: AnEthnographic Approach. London: Routledge.

Button, Graham (1993) ‘The Curious Case of the Vanishing Technology’, in G.Button (ed.) Technology in Working Order. Studies of Work, Interaction andTechnology. London: Routledge.

Callon, M. and Law, J. (1995) ‘Agency and the Hybrid Collectif’, South AtlanticQuarterly 94: 481–507.

Clifford, J. and Marcus, G. E. (1986) Writing Culture: Poetics and Politics inEthnography. Berkley, CA: University of California Press.

Cooper, R. and Law, J. (1995) ‘Organization: Distal and Proximal Views’, Researchin the Sociology of Organizations 13: 237–74.

Engestrom, Y. and Middleton, D. (1996) Cognition and Communication at Work.Cambridge: Cambridge University Press.

Escobar, Arturo (1994) ‘Welcome to Cyberia. Notes on the Anthropology ofCyberculture’, Current Anthropology 35(3): 211–31.

Fletcher, Joanne K. (1999) Disappearing Acts. Gender, Power and RelationalPractice at Work. Cambridge, MA: MIT Press.

Gherardi, S. and Strati, A., eds (2004) Telemedicina: tra tecnologia e organizza-zione [Telemedicine: Between Technology and Organization]. Rome: Carocci.

Gibson, James G. (1979) The Ecological Approach to Visual Perception. Boston:Houghton-Mifflin.

Gomart, E. and Hennion, A. (1999) ‘A Sociology of Attachment: Music Amateurs,Drug Users’, in J. Law and J. Hassard (eds) Actor Network Theory and After, pp.220–47. Oxford: Blackwell.

376


Hammersley, M. and Atkinson, P. (1995) Ethnography: Principles in Practice.London: Routledge.

Harbers, H., Mol, A. and Stollmeyer, A. (2002) ‘Food Matters: Arguments for anEthnography of Daily Care’, Theory, Culture and Society 19(5/6): 207–26.

Harre, Ron (2002) ‘Material Objects in Social Worlds’, Theory, Culture andSociety 19(5/6): 23–36.

Hine, Christine (2000) Virtual Ethnography. London: Sage.Husserl, Edmund (1960) Cartesian Meditations. The Hague: Nijhoff.Knorr-Cetina, Karin (1997) ‘Sociality with Objects’, Theory, Culture and Society

14(4): 1–30.Knorr-Cetina, K. and Bruegger, U. (2002) ‘Traders’ Engagement with Markets: A

Postsocial Relationship’, Theory, Culture and Society 19 (5/6): 161–85.Latour, Bruno (1999) ‘On Recalling ANT’, in J. Law and J. Hassard (eds) Actor

Network Theory and After, pp. 15–25. Oxford: Blackwell.Law, John (1987) ‘Technologies and Heterogeneous Engineering: The Case of the

Portuguese Expansion’, in W. E. Bijker, T. P. Hughes and T. J. Pinch (eds) TheSocial Construction of Technical Systems: New Directions in the Sociology andHistory of Technology. Cambridge, MA: MIT Press.

Law, John (1992) ‘Notes on the Theory of the Actor-Network: Ordering, Strategyand Heterogeneity’, System/Practice 5(4): 379–93.

Law, John (1994) Organizing Modernity. Oxford: Blackwell.Law, John (1999) ‘Machinic Pleasures and Interpellations’, Centre for Science

Studies and the Department of Sociology, Lancaster University, http://www.comp.lancs.ac. uk/sociology/soc067jl.html.

Law, John (2002) ‘Objects and Spaces’, Theory, Culture and Society 19(5/6):91–105.

Law, J. and Mol, A. (1995) ‘Notes on Materiality and Sociality’, SociologicalReview 43(2): 274–94.

Law, J. and Mol, A., eds (2002) Complexities. Social Studies of KnowledgePractices. Durham, NC: Duke University Press.

Lynch, M. and Collins, H. M. (1998) ‘Introduction: Human, Animals andMachines’, Science, Technology and Human Values 23(4): 371–83.

MacKenzie, D. and Wajcman, J., eds (1999) The Social Shaping of Technology.Buckingham: Open University Press.

Manning, Peter K. (1995) ‘The Challenges of Postmodernism’, in J. Van Maanen(ed.) Representation in Ethnography. Thousand Oaks, CA: Sage.

Melucci, Alberto (1998) Verso una sociologia riflessiva [Towards a ReflectiveSociology]. Bologna: Il Mulino.

Mintzberg, Henry (1971) ‘Managerial Work: Analysis from Observation’, Manage-ment Science 18(2): 97–110.

Mintzberg, Henry (1973) The Nature of Managerial Work. New York: Harper &Row.

Pels, D., Hetherington, K. and Vandenberghe, F. (2002) ‘The Status of the Object’,Theory, Culture and Society 19(5/6): 1–21.

Pickering, Andrew, ed. (1992) Science as Practice and Culture. Chicago: Uni-versity of Chicago Press.

Sachs, Patricia (1993) ‘Shadows in the Soup: Conceptions of Work and Nature ofEvidence’, Quarterly Newsletter of the Laboratory of Human Cognition 15:125–32.


377

Star, Susan L. (1999) ‘The Ethnography of the Infrastructure’, American Beha-vioral Scientist 43(3): 377–91.

Stone, Allucquere R. (1995) The War of Desire and Technology, at the Close of theMechanical Age. Cambridge, MA: MIT Press.

Strati, Antonio (1992) ‘Aesthetic Understanding of Organizational Life’, Academyof Management Review 17(3): 563–81.

Strauss, A., Fagerhaugh, S., Suczek, B. and Wiener, C. (1985) Social Organizationof Medical Work. Chicago: University of Chicago Press.

Suchman, Lucy (1997) ‘Centers of Coordination. A Case and Some Themes’, in L.Resnik, L. Saljo, C. Pontecorvo and B. Burge (eds) Discourse, Tools andReasoning. Essays on Situated Cognition. Berlin: Springer Verlag.

Suchman, Lucy (2002) ‘Human/Machine Reconsidered’, Department of Soci-ology, Lancaster University, http://www.comp.lancs.ac.uk/sociology/soc040ls.html.

Suchman, L., Blomberg, J., Orr, J. E. and Trigg, R. (1999) ‘Reconstructing Technol-ogy as Social Practice’, American Behavioral Scientist 43(3): 392–408.

Van Maanen, John (1988) Tales of the Field: On Writing Ethnography. Chicago:University of Chicago Press.

Weick, Karl E. (1995) Sensemaking in Organizations. Thousand Oaks, CA: Sage.Winthereik, B. R., de Bont, A. and Berg, M. (2002) ‘Accessing the World of

Doctors and Their Computers: “Making Available” Objects of Study and theResearch Site through Ethnographic Engagement’, Scandinavian Journal ofInformation Systems 14(2): 47–58.

Woolgar, Steve (1988) Knowledge and Reflexivity. New Frontiers in the Sociologyof Knowledge. London: Sage.

Woolgar, Steve (1991) ‘Configuring the User: The Case of Usability Trials’, in J.Law (ed.) A Sociology of Monsters: Essays on Power, Technology and Domina-tion. London: Routledge & Kegan Paul.

Attila Bruni received his PhD in Sociology and Social Research from the University ofTrento, Italy. His thesis was entitled ‘Objects’ Sociality and Materialities ofOrganizing: Ethnography of a Telemedicine Project’. He is part of the ResearchUnit on Cognition, Organizational Learning and Aesthetics (RUCOLA) in theDepartment of Sociology and Social Research of the University of Trento, wherehe lectures in Sociology of Organization. His research interests include objects,knowledge and practice, and gender. He recently co-authored (with SilviaGherardi and Barbara Poggio) Gender and Entrepreneurship: An EthnographicApproach (Routledge, 2004). Address: Dipartimento di Sociologia e RicercaSociale, Universita degli Studi di Trento, V. Verdi 26, 38100 Trento, Italy. [email:[email protected]]

378


on the ethnography of non-humans and heterogeneous contexts (1)

Documents

anonymous

anonymous

everyday organizational

blood test

patients clinical

clinical record

based clinical

actor network