Organizing documents: Standard forms, person production and organizational action

Organizing Documents: Standard Forms, PersonProduction and Organizational Action

Nahoko Kameo1 & Jack Whalen2

# Springer Science+Business Media New York 2015

Abstract Using the case of calls-for-help to police and fire communications centers and theirincident record forms, we present a detailed investigation of how documents play a constitutiverole in formal organizations. We take an ethnomethodologically informed approach to theproblem, delineating how standard forms in bureaucracies enable organizational participants tocoordinate actions across time and space and, at the same time, produce people who perpet-ually produce such documents or work from them. We focus in this regard on persondescription. The call-taker needs to translate the call into preset categories, and thus enlistthe citizen in the work of inscribing the incident in the way the form requires, (re)producingcertain categorizations of personhood, especially race and sex. In this way, organizationaldocuments and their inscriptions function as a kind of technology of reification for thesecategories.

Keywords Organizations . Documents . Interaction . Ethnography

The management of the modern office is based upon written documents (Bthe files^),which are preserved in their original or draught form. There is, therefore, a staff ofsubaltern officials and scribes of all sorts. (Weber 1978/1914, 197)

As Weber saw nearly a century ago, our world is surrounded by formal organizations and,equally so, by the files they produce and process. In the developed world today, our livesBare…infused with a process of inscription,^ to use Smith’s (1990a, 209) graceful phrasing,Bproducing printed or written traces or working from them.^ And Smith’s (1990a, 209)

Qual SociolDOI 10.1007/s11133-015-9302-7

* Jack [email protected]

Nahoko [email protected]

1 Department of Sociology, New York University, New York, NY 10012-9605, USA2 Department of Design, School of Arts, Design and Architecture, Aalto University, Helsinki, Finland

examples generate a tangible impression of this saturation: BWe get passports, birth certificates,parking tickets; we fill in forms to apply for jobs, for insurance, for dental benefits; we aregiven grades, diplomas, degrees; we pay bills and taxes; we read and answer advertisements;we order from menus in restaurants, take a doctor’s prescription to the drugstore, write lettersto newspapers.^

From the moment of our birth, then, to the time of our death, society often takes the shape ofthe castle of bureaucracy that Kafka deemed absurd and yet inescapable. While it may betempting to see the results of this documentary activity as simply bureaucracy’s detritus, both thetrifling and the perfidious, in truth even our most mundane, humdrum dealings of everydaylife—our Bparking tickets^—are commonly conducted, at least in part, on the basis of Bwrittentraces^ (Levy 2001; Smith 1990a). With the technological advancement of digital devicesallowing us to produce and work from such documents at any time and place, we seem to beever more surrounded by documents and other written traces, be it on a piece of paper or a screen.

Given the centrality of bureaucratic organizations and their documentary workings, it issurprising how few sociological studies on Bthe files^ and people’s use of them have beenwritten. Classic analyses by those following Weber’s footsteps (e.g., Blau 1963) make little orno mention of how documents are produced and used by organizational members. Instead,documents have been used as a resource for social scientists to construct historical cases. Thisusefulness is telling. As Giddens (1979, 1984, 1987) argued, because documents enableBdistanciated^ interaction in different contexts they can thus be taken up as evidence for theresearcher’s own project unrelated to its original context. The literature on formal organiza-tions includes many accounts of social structure in organizations—of hierarchies, ideologiesand processes—that are heavily based on evidence found in manuals, reports, letters, memos,and the like. There is, however, little or no examination of how those documents are actuallyused in the organizational tasks for which they were originally produced.

Even where sociologists go beyond a reliance on documents solely as a resource for theiranalyses and initiate real-time observation of how organizations Breally^ work, these re-searchers tend to focus primarily on human interaction at the work site and thus do not takedocuments as an essential feature of everyday action to which the organization’s membersconsistently attend (e.g., Dalton 1959; Gouldner 1954; Morrill 1996; Selznick 1949). Al-though organizational histories and ethnographies do pay attention to the power and conse-quences of rules, protocols, reports and the like, in almost every case any attention todocuments quickly gives way to accounts of political struggles, personal connections, informalauthority and control systems, and other interpersonal interactions that seemingly underminethe Bofficial^ way of organizing (with a few notable exceptions; see Feldman 1989; Geisler2001; Manning 1992; Orlikowski and Yates 1994; Pettinari 1988; Yates and Orlikowski 1992).

In short, while documents have served well as a data source for analysis of things like thehistories and machinations of organizations, very few studies of bureaucracy or organizationsmore generally make documents a research topic in and of themselves and treat them as afundamental phenomenon of (and participant in) organizational life. The standardization andreproducibility of printed (or digital) texts enables analyses removed from the local settingsand contexts in which they are produced precisely because documents are used to coordinateacross multiple sites and courses of action. This property of documents is also why themanagement of the modern office is so dependent on it; documents Bmediate, regulate andauthorize people’s activities^ (Smith 1990a, 159) in organizations and thus are a basic mediumof intersubjectivity (for digital documents mediating intersubjectivity amongst currencytraders, see Knorr Cetina and Bruegger 2002).

Moreover, if organizations are, as Meyer and Rowan (1977, 340) put it, Bsystems ofcoordinated and controlled activities that arise when work is embedded in complex networksof technical relations and boundary-spanning exchanges,^ then documents make such coordi-nation and control possible by providing the prototypes of acceptable and expected actions inthe organizational setting. The most salient example would be an organization’s manual orprotocol reference, defining exactly what and in which order the employees or participants inthe organizational actions should do, such as McDonalds’ employee manual or a sociologydepartment’s grade appeal procedure. and just as Leidner (1993) and Smith (2001) show intheir work, defining organizational members’ roles and their formally allowed actions is only apart of what documents do. Documents are crucial elements of organizations because theyshape people’s interactional patterns for organizational activities by providing ways to makesense of one another’s actions.

This multi-dimensional and reflexive relationship between document and action will be thecenterpiece of our analysis in this paper. We will be concerned in this regard with a ubiquitousfamily of business documents, Bstandard forms,^ and will examine in detail the use of adistinct member of that family, the incident record form, at public safety communicationscenters—organizations that are responsible for dispatching police, fire, and paramedic assis-tance in response to citizen reports and complaints that are phoned in on both 9-1-1 emergencyand seven digit non-emergency lines.

Note that there are good historical reasons for concentrating on what could be described asBform(ed) documents,^ which have what Goody (1986, 54) refers to as a Bnon-textual^ or non-syntactic character, rather than other kinds of documents common to organizational life (such asoffice memos or reports) that rely on standard syntactic units like sentences and have a narrativestructure with continuous text. It is the non-syntactic that came first in terms of the developmentof writing as a means of coordinating human activities (Giddens 1987, 42–43; Harris 1986;Robinson 1995); indeed, writing began as a mode of administrative notation, used to keeprecords or tallies for future reference (and to thus coordinate action across time) rather thanrecount rituals or narrate myths. And although religious writings and other narrative genres likeletters and histories soon came to play an equally important cultural role, writing that is non-syntactic, along with its associated document types, remain central to administrative work.

In this regard, it will be clear from our investigations that standard forms in contemporarystate bureaus like police departments support administrative surveillance operations throughtheir document-structured characterizations of individuals as aggregations of standardizedfeatures or properties—such as race, sex and age—that are routinely employed by all kindsof modern organizations, both public and private. We will argue that this Btechnology ofperson production^ (Cahill 1998) not only defines what categories are organizationallysensible and consequential but also works as a kind of technology of reification, makingparticipants attend in and through their interactions to the Bdemands^ of forms, in order totransform—for organizational purposes—local events of specific circumstances into an arrayof categories and pre-scripted textual representations.

As interactionists and ethnomethodologists have frequently pointed out, such descriptivecategorization is accomplished in situ. Garfinkel’s (1967) analysis of local order productionand Harvey Sacks’s (1979) writings on membership categorization both demonstrate thatparticipants necessarily engage in mundane work of this sort, negotiating what specificcategories of persons, places, and things may be evoked, constructed and sustained ininteraction. We will show how this local categorization is patterned in public safety work asits members attend to the form and its requirements as an Bobligatory passage point^ (Callon

1986) during their call-processing activities. As the form is relayed from desk to deskthroughout the organization, it is then taken up and acted upon (read, written, distributed,etc.) again and again. The document mobility that enables coordination and control oforganizational actions makes a specific way of inscription ubiquitous, with staff that perpet-ually produce or work from such inscription.

More specifically, we will describe how a call for help, as documented in the incidentrecord form, Btravels^ through a public safety communications center to organize the workand, in this way, organize the organization. By following the call from its initiation to thedispatch of assistance and radio communications between the center and police officers in thefield, we can closely delineate not only how this standardized record is used in situ but howparticipants constantly orient to a definite method of person inscription and the consequencesof this for both the immediate interaction and the organization’s functioning. That is to say, wewill demonstrate that while the form mediates interaction during calls it simultaneously acts asa medium for reifying a certain kind of membership categorization, institutionally creating andrecreating the Bperson^ stripped of its local historicity, transforming a certain way of describ-ing persons into seemingly transparent social facts.

Studying Documents in Interaction

Before presenting this empirical analysis, it will be useful to first review the theoreticalbackground for our work, starting of course with Weber but then moving to Giddens and histrenchant observations on some critical issues thatWeber only sketched out.We finish by turningto ethnomethodologically informed research that takes up many of these same issues in detail.

For Weber, Bfiles^ were the iconic symbol of increasing rationalization and the inescapableiron cage of modern organizations. Bureaucracy, once established, is Bhardest to destroy^—therule-bound rationalistic machine carries Bcommunity action,^ a single act among participants,into Bsocietal action,^ a reproducible, patterned rational action that can be done by anyone whooccupies a specific position by creating a rule-based, cascading structure of responsibilities.The key characteristics of bureaucracy Weber focuses on are the problems of accountabilityand inflexibility. Bureaucracy enables people to act strictly within the Bofficial duties^assigned to them, making organizations independent of personal actions. The managementof such ways of organizing is based upon Bwritten documents;^ Weber aptly describes themodern office as essentially consisting of B‘the files’^ and Bscribes of all sorts,^ presumablyengaged in writing, reading and otherwise processing them. But what exactly documents do inand for organizations is left underspecified. Concentrating on the problem of bureaucracy,Weber only describes written documents and Bthe knowledge of the files^ (Weber 1978/1914,214) as a necessary aspect of running this form of organization.

Giddens (1979, 1984, 1987) follows in Weber’s footsteps, making the most detailedtheoretical arguments concerning organizations and documents (and what they actually do inthe world) among contemporary theorists. For Giddens, writing and, by extension, thedocuments it produces, are of crucial importance to the development of modern society andits social reproduction. As we noted above, Giddens emphasizes that written documents enableBdistanciated^ interaction and, consequently, are decisive for the development of humansociety, of civilization itself, through Bthe extension of social systems in space and in time^(1987, 203). Thus, the development of writing directly shapes the ways in which society isorganized. The difference between traditional society and modern times is that in our era, the

development of writing enables us to package the ways we coordinate actions with each otherand therefore the meanings arising from them, which then move with them across space andthrough time. Modern formal organizations are the quintessential operational site for this kindof social reproduction. These organizations bracket time-space through Breflexive monitoringof system reproduction and the articulation of discursive ‘history’^ (Giddens 1987, 153).Although Giddens’ Bstructuration^ theory gives individuals room for improvising new sets ofpractices, the more a social system encompasses time-space, the more resistant it is tomanipulation (Giddens 1987, 171). As a result, our contemporary society is a Bworld oforganizations^ where Bsurveillance as information collation and retrieval^ by both state andprivate organizations is intensifying in every aspect of our lives (1987, 155). The Bfiles^ areabsolutely essential to this account; they are the key for understanding both time-spacedistanciation and the modern surveillance state.

While providing invaluable insights, both Weber and Giddens do not provide an account ofhow this document-dependent concerting and controlling works in everyday organizationalpractice. But practical reasoning with documents and their everyday use in organizationalactivities have been recurring ethnomethodological topics, starting with Garfinkel and Bittner’s(1967) report on the difficulties encountered by social science researchers attempting to makeuse of psychiatric files, which demonstrated that the sense of any organizational record can befound only in the context of the ordinary communal accounting practices in which such recordsare composed andmeant to be read. Other important ethnomethodological studies that deal withdocument practices in organizational and professional life include Zimmerman’s ethnographicstudy of a public assistance agency (Zimmerman 1966, 1969, 1970); Atkinson (1978) oncoroner’s reports on sudden death; Heath (1982) on the employment of medical record cardsin patient consultations; Lynch (1985) on inscription practices in laboratory science; Meehan(1986) on the everyday use of patrol logs and incident cards in police work; Lynch and Bogen(1996) on the intertwining of spoken testimony and evidentiary documents in the Iran-Contrahearings; Button and Sharrock (1997) on work-tickets and other documentary artifacts in acommercial print shop; Harper (1998) on technology and document careers in the InternationalMonetary Fund; and Drew (2006) on documents as an important interactional resource for bothpractitioners and their clients in educational and medical settings.

Such ethnomethodological studies of document use over a variety of professional domainsreveal a good deal about communal accounting and inscription practices. Any organizationaldocuments require a community of practice that agrees on the inscription conventions, whatthe inscription practically means, and routinely makes sense of such inscription, time andagain, whenever the documents are in use. However, since these studies, for the most part, donot investigate the embodied activities of reading and writing in real time, they do not tell usnearly enough about how documents actually Bmake things work^ in organizations. Theyprovide little insight into how documents are used to organize the organization, contributing toits social facticity as a Bcollective actor^ (Coleman 1974), and how the organizationalmembers must attend to the circulating documents and their larger consequences for organi-zational actions. Additionally, the materiality of the documents is largely ignored; what theparticipants inscribe is taken as important, but little attention is paid to what they inscribe onand how it matters (but see Frankel 1989; Heath and Luff 2000, 31–60; Luff et al. 1992).

The ethnomethodologically informed research that comes closest to ours in focus andstrives to overcome these limitations is that of Smith (1990a, 2001, 2005; see also 1990b,1999). Smith’s institutional ethnography sets a clear agenda on studying Btexts^ (the term sheprefers) as an important project for understanding organizations, and, in her terms, Bruling

relations.^ Institutional ethnography aims to uncover the ways in which external power ismade apparent in people’s everyday lives, even when the subjects do not consciously know theexternal power dynamics that shapes their interactions. Here Smith argues that such powerdynamics operate across and beyond local settings. Texts are not simply means of communi-cation or purveyors of information, but are rather the constituents of such extra-local ruling oflocal actions, and thus are essential conveyors of extra-local power that is abstract and yet self-reproducing (Smith 1990a). From this point of view, investigations of the constitutional workthat documents do, and that we do with them—the routine textually-mediated practices ofpeople engaged in their daily activities—Boffers access to the ontological ground of institu-tional processes which organize, govern, and regulate the kind of society in which we live, forthese are to a significant degree forms of social action mediated by texts^ (Smith 1990a, 91).

For example, in one study Smith (2005) analyzes the graduate student grade appealprocedure and shows how texts specify and formulate actions possible on the local level,shaping the actors’ roles and their capacities. Specifically, the written grade appeal proceduremediates the external sources of power working Binto^ individuals who lead everydayactivities. Thus, she argues, the procedure document actively participates in, and controls thelocal actions, making sure that the relations of power between graduate student, department,and larger institutional authorities of university and surrounding cultural complex will beenacted without breaching the Bruling relations.^

Our study is motivated by Smith’s call for taking texts seriously but departs from herapproach in two significant ways. First, for her, Btexts^ include both their Bmaterial^ andBsymbolic^ aspect, ranging from discursive constructs such as ideas, concepts and culturalpatterns etc. to actual texts inscribed and being circulated. 1 In fact, Btexts^ are anything thatare not locally produced yet nevertheless control how local interactions are shaped. In thisaccount she is very similar to Foucault (1977) in seeing details of the social as penetrating andBworking through^ individuals. Instead of positing external power dynamics as textuallymediated, and blurring what we mean by texts, or documents, we take a more empiricalapproach and focus on the process of how written documents are produced, transmitted, andenacted upon in interaction in different times and spaces. Second, and related, although weagree that explicating the ways in which external constructs in our case documents shape localparticipants’ actions is important, we refrain from embracing Smith’s focus on Bpower^ ofexternal institutions and its constitutive role in individual lives. Assuming Bpower^ as a pre-existing matrix that presumably influences the participants without first examining theiractions in situ works against our goal of delineating the ways participants work and build-upthe ontology of social institutions from the micro-interactions. Instead, we look at howparticipants attend to documents and work out their interactions with the document’s presenceto accomplish organizational actions (for a micro-approach to passports and bureaucraticpower, see Pelkmans 2013).

Accordingly, we focus on how participants in the interaction accomplish their coordinatingand concerting with the documents, and thus avoid presuming that the document is the

1 We should thus note here that the difference between Btext^ and Bdocument^ is more than simply terminologicalpreferences. Compared to the implicit immateriality of a Btext,^ documents—as Levy (2001, 5) demonstrates inconsiderable detail—are understood to be necessarily Bphysical things; they are physical artifacts in whichcommunication has somehow become embedded in a stable, that is, relatively permanent way,^ which thenenables them to be taken up and Bspeak to us^ in many different settings, at different times and for different people.Like Levy, we want to insist that this materiality and the consequent local presence of any document (in whateverform and at whatever point in its assembly, exchange, or reading) are essential to how it operates in the world.

medium of power of the institutions in which they are working. The interaction order has to bemade and remade, and we shall show that it is in this repetition rather than the coercion ofinstitutions per se that works as the basis of membership categorization in and through a formof person-production technology.

Although our approach is largely ethnomethodological in spirit, we also draw insights fromstudies of materiality in science and technology studies. Contrary to sociologists that oftenavoid assigning agency to artifacts, scholars in this tradition analyze the ways in which artifacts(and nonhumans in general) shape human actions in social interaction (see especially Callon1986; Gieryn 2002; Latour 1987, 2005). Artifacts are created by humans that design them toshape their interactions in a patterned manner. Once created, though, people then have to satisfyits Bdemands^ (in an accountable rather than causal manner) if they wish to use those objects fortheir own demands. As the speed bumps set by the city throughout residential streets inevitablyslow down any car (Latour 1992), the artifact stabilizes and structures human interaction.

Let us be clear, however: We see no analytical purchase in treating documents as an equalparticipant in interactions in our case. Documents are artifacts created by the organization inquestion, and as much as documents are more than the medium of organizational action, theymust be Btaken up^ by organizational members to participate in interactions, which then haspractical and sociological consequences (see Jerolmack and Tavory 2014 for a useful discus-sion on the problem of agency). But as Gieryn (2002) and others points out, artifacts do farmore and have far more unexpected impacts on humans than coercing them into prescribedactions. The effects of artifacts often disappear in people’s consciousness as long as they work,becoming like a Bblack box.^ Moreover, as long as someone has to take a particular street toget where they want to go, the slowing down is taken for granted as it becomes a mandatorypart of reaching the destination. Thus, not only are the ways the artifact shape human actionforgotten, that it is shaping action also tends to be forgotten. Although humans can alwaysreconfigure the artifact by assigning different meanings to it or even changing its propertiesmaterially, because this requires open reflection and reinterpretation of its meaning (withawareness and substantially incurred cost) it is often not an easy task. And, however wereconfigure artifacts, they will have to be taken into account in our interactions; the organiza-tional demands and expectations they embody will have their way.

Summarizing, then, we take two following insights from science studies. First, as we havejust emphasized, in coordinating organizational actions the participants must respond to therequirements of the documents and its carriers. The work of documenting and transmitting thedocument to a different time and space is predicated upon the ability of organizationalparticipants, including the newcomer, to be able to attend to the requirements of artifacts.Secondly, the intent of the designer of the artifacts—why the requirements exist, as well as theunintended properties of the documents and its medium—may become a background of theorganizational action and not even cross member’s consciousness. Although such require-ments are constantly and routinely attended, both the fact that there are requirements and whatthe requirements are may be taken for granted, and even forgotten.

Organizing the Work, Administering the Form

Our primary research site for the investigation of incident record practices was Central LaneCommunications in Eugene, Oregon, operated by the Eugene Department of Public Safety (atthe time of the study, a combined police and fire organization). Central Lane is the Bpublic

safety answering point^ for all 9-1-1 calls in its service area of approximately 4000 squaremiles. We recorded and transcribed approximately 350 citizen telephone calls to Central Lane(and often the radio communication between the communications center and officers in thefield concerning the incident) and made copies of all the incident forms, logged as digitalrecords in the center’s computer-aided dispatch (CAD) system, associated with these calls. Thecalls are from the period 1989–90 and in 2001 as well from the center’s archives that wererecorded in the early 1980s and documented on paper forms. Our analysis is also based onextensive field observations made by the second author during 16 months of full-time work ascommunications specialist—a call-taker and dispatcher—at Central Lane during 1989-90. ThisBobserver as participant^ role afforded intimate access to the work life of communicationscenter personnel. The second author went through the testing and training required of allprospective call-takers and dispatchers at Central Lane and was an employee of the EugeneDepartment of Public Safety during the study period, working the full range of shifts andassignments to which other employees were subjected.2

Much like other work sites that serve as Bcenters of coordination^ (Suchman 1993; see alsoSuchman and Whalen 1994), which have evolved a division of labor that seeks to identify,assign, and temporally order responsibilities for the performance of the coordinating work,public safety communications centers have a scheme for organizing tasks and, through this, thework of the organization. Call-takers receive phone calls on both 9-1-1 and nonemergencyseven digit lines and, while engaged in interaction with the caller, enter information into anincident record form on their computer screen that is arranged into more than a dozen fields,each dedicated to a particular organizational purpose or need, and following certain rulesenforced by the software, such as the use of codes selected from an enumerated list, or specialtext-inscribing conventions expected by Center’s management and staff. (The form’s organi-zation and some of these rules and conventions will be detailed below.) This form is thenelectronically transmitted to the dispatcher and becomes the focal resource for her work: Thedispatcher reads the transmitted form and assesses the information, determines when and whatorganizational response is warranted, and then, via radio, dispatches fire (including emergencymedical) and/or police units to the scene and coordinates their response.

The actual sequencing of any set of actions in processing calls can be much more complex.The entries by the call-taker must fit certain official, organizational categories, and, in doingthis, meet the demands of the computerized form. Essentially, then, call-takers, like all users ofstandard forms, are consistently faced with the problem of how to administer the form in thecontext of what are, irremediably, singular events and sets of circumstances. Thus, when thecalls are initiated by a citizen phoning the communications center, a great deal of workinvolves interacting with the caller and making the call into a set of text strings on thecomputer screen by reworking the event in order to meet the requirements of the form(Suchman and Whalen 1994). As we will show, this involves both the caller and call-takerattending to each other’s responses during the call.

2 To the question of whether the age of the data somehow then dates the analysis, we wish to note that thetechnology used by call-takers and dispatchers may have evolved in 9-1-1 operations over the years. Thisincludes Central Lane, where a computer-aided dispatch system replaced a paper-based one in the mid-1980s andsophisticated incident mapping software and large flat screen displays later replaced the paper map books anddumb terminals in use at the time the bulk of our data was collected. Nevertheless, the work itself still centers(and always will) on the in situ assembly of textual representations into intelligible information resources andorganizationally accountable records. In this sense, the problem for analysis is in fact timeless.

The principal tool that call-takers use in accomplishing their work is the computer-aideddispatch (CAD) system and its incident record form. A phoned request for help, a complaint,or a report of trouble—none of these officially exist as Bcalls^ until they are entered into anincident record. Once entered, the incident record is the primary source of both informationand coordination inside the organization, as it travels from the call-taker to the dispatcher, whothen dispatches help and transmits the necessary information to the police or ambulance in thelocal scene. Indeed, the incident record is almost always the only source of information that isavailable to the dispatcher, and their reading (and thus interpretation) of this text, and then thedecision that follows, is thus quite fundamentally reliant on the record. In this respect, CAD atCentral Lane constitute a Btechnology of order production^ (Whalen 1995; see also Suchman1993) designed to exercise a coordinating function in formal organizations. In this way, thetechnology is the obligatory passage point (Callon 1986) we referred to earlier, acting as anessential coordination between organizational actors.

At Central Lane, the primary incident record form is referred to as the Bface sheet.^ Whencall-takers answer a phone line, this is the form that they bring up on their computer screen.There is another, ancillary Badditional details^ form that can be appended to an incident record,which consists only of a heading identifying the associated face sheet followed by eight blanklines for adding information to this initial record after the face sheet has been transmitted to thedispatcher (as many Badditional details^ forms as necessary can be appended). Figure 1 showsthe Central Lane face sheet. The Badditional details^ form is shown as Figure 2; as always, ithas identifying headings that show the specific call to which it adds information and, throughCAD, is automatically linked.

Observe that each of the dedicated fields on the face sheet has an abbreviated heading (e.g.,INC, LOC, PR, PHO). We previously indicated that many of these fields must be filled froman array of standardized categories and codes, and other fields or areas must be inscribed in aspecial manner. The call-taker’s work with the caller is thus going to be oriented towardssecuring descriptions or information that can be fitted to such categories, codes, conventionaldescriptor terms, and the like. For instance, the field headed INC (for Bincident type^) on theface sheet, which is placed at the very beginning (top left) of the form, must be filled in with anincident code selected from an enumerated list such as FIRHOU for house fire or DSPUTFAMfor family dispute.

For the purposes of this paper, the data entry parts of the incident record for which we wantto direct special attention are the DTL field on the face sheet, located near the bottom of the

Fig. 1 CAD face sheet

form and providing two full lines for inscribing text, and the Badditional details^ form, whoseeight lines for data entry are modeled on that face sheet field’s format. As their namingindicates, the face sheet DTL field and the entire Badditional details^ form are both intendedfor recording details about the incident, the location, the persons involved in the incident, thecaller—whatever is deemed necessary and important information for Bthis type of instance^ orBthis particular instance.^ Nevertheless, organizational policies dictate what kind of informa-tion is expected here and precisely how it should be inscribed. This is particularly the case forperson description details. Here, these constraints, these conventions for doing descriptions,center on (1) standard categories and their abbreviations for a person’s identifying features and(2) the syntax of the descriptive sequence. The result is a kind of coded text string. Documentsemployed in training Central Lane staff to follow these prescriptions include that is shown asFigure 3.

This particular regulatory document (Smith and Whalen 1995) shows the standardcategories for person description, for Bproducing^ a person in organizational terms,arrayed alongside a drawing of a stereotypical Bbad guy.^ These categories include visibletraits or physical features, such as sex, race, age, height, weight, hair, eye color, facial hair,and complexion, as well as clothing—eyeglasses, hat, coat, shirt, pants and shoes—andweapons that a person might be carrying (in this illustration, a handgun). In any actualdescription, not all of these categories are likely to be used, for a variety of reasons (e.g.,callers may not be able to recall more than a few identifying features of a person theyobserved for only a short period of time). But in most instances, textual descriptionsassembled by call-takers at Central Lane will commonly include the more readily observ-able traits and clothing types, like race, sex, age, height and weight, coat or shirt, andpants. In this regard, a typical person description text would be like one of the exampledescriptions included as part of the Figure 3 document:

WM, 25YRS, 5-10, 170LBS, BRO-HAI, GRNBASEBALLCAP, BLU SHIRT, FADEDJEANS

This description is for a white male who is 25 years old, stands 5 ft 10 in and weighs170 lbs, has brown hair, and is wearing a baseball cap, a blue shirt and faded jeans. In actualpractice in Central Lane the descriptions were even more abbreviated, leaving out YRS, LBSor HAI in their text descriptions; thus the order of the descriptors is the only key to whatspecific feature of the subject the text is referring.

Fig. 2 Additional details form

These conventions for inscribing description texts as well as the arrangement of dedicatedfields on the CAD face sheet are governed by those personnel who are the recipients of theforms assembled by call-takers: the dispatchers and, indirectly, those who will be recipients of

Fig. 3 Instructional document for call-takers

the dispatchers’ radio communications—police officers and fire fighters in the field. Theprinciple of recipient design, as developed in conversation analysis, helps us to understandwhat is involved. Recipient design principle refers to the Bmultitude of respects in which thetalk by a party in a conversation is constructed or designed in ways which display anorientation and sensitivity to the particular other(s) who are the co-participants^ (Sacks et al.1974, 272). Originally developed as an operating principle for the social organization ofconversation, this concept also helps us understand how such orientations to recipients areembodied by documents and how the users of them enact such design. Thus, similar to theways conversationalists attend to the ways their utterances are understood and responded to,the CAD face sheet is designed so the pre-formatting of the dedicated fields and the specialrules and conventions for inscribing Bfree text^ in the DTL field would ease the organizationalcoordination beyond the entry of texts by the call-taker.

The face sheet fields are not set up to facilitate the call-taker’s work practice since call-taker’s talk with the caller does not necessarily produce face sheet entries in a systematic order;call-takers can thus be observed to move among the different fields as they secure the basis foran entry (even as they may try to order their questioning so as to follow the form’s ordering, aswe will show below). Instead, the fields are arranged, right to left, and line by line, so that theconventional practices of reading English will yield to the dispatcher the possibility of readingdirectly off the transmitted face-sheet into talk with the police or fire department, giving themthe essential information in that order. Descriptors and their sequence are officially prescribedfor this same reason. That is to say, the Eugene Department of Public Safety’s policies onbroadcasting person descriptions over the radio (most police departments adhere to these samepolicies) require dispatchers to read out such descriptions in a standardized syntactic order,beginning with the person’s race and sex, followed by any available information about theirage, height, and weight, the color and length of their hair (including facial hair), the color oftheir eyes, and then their clothing, working from top to bottom (hat, jacket, shirt, pants or skirt,and shoes).

This is primarily to meet the needs of police in the field, who need reliable Bstreet-visibleidentifiers^ that are ordered inversely with respect to how easily they can be changed and arealso comprehensive for the population so that police can visually sort people they see on thestreet in terms of the set of descriptors the dispatcher has radioed to them (Smith and Whalen1995). Standardized descriptors assembled according to these needs furnish the resources foran expeditious search procedure. But we must also note that such coded methods fordescribing persons used in our case study are most commonly used in other state agenciesand thus consists of a type of discourse we can term the Bbureaucratic genre.^ Thus, it is not(only) that such descriptors are used because they are found useful in person identification.Rather, state bureaucracies commonly characterize individuals as aggregations of standardizedfeatures or properties—such as race, age and sex—that can distinguish all members of apopulation; what we earlier termed, citing Cahill (1998), as involving technologies of personproduction. Law enforcement organizations in other countries have similar classificationpractices (see, for example, Mackie 1978; see also Duster 2004 for an interesting discussionof racial classification in particular). As Goodman (2008) shows, the state bureaucracy—including our research site, a public safety communications center—are where certain catego-ries of people are in fact produced as Bpeople^ per se.

We are now in a position to explicate the work of Bdescribing a person^ during telephoneconversations between callers and call-takers and its relationship to the call-taker’s assembly ofan organizationally appropriate textual description in the incident record, paying special

attention to the practical reasoning involved in this assembly, the organizational exigencies thatare made visible in the work, and the manner in which such records work to organize theorganization. While our analysis is based upon the inspection of numerous calls and incidentrecord documents from different periods of time, our explication of Bdescribing^ work for thepurposes of this paper is for the most part built around a close examination of a single call, withother calls drawn upon for comparative purposes. Because the production of any Bcall^ and thecoordination of organizational action in and through documents is an essentially situatedactivity, an examination of how this is actually accomplished for and in singular sets ofcircumstances, over the course of specific occasions, is the most profitable way to explicateits essential features.

The Form in Action: Persons and Categories

We begin with a sequence of organizational action that took place in October of 2001,initiated by a phone call from a citizen reporting what she takes to be suspicious behavioraround the house of a neighbor who she knows to be out of town: BSomebody drove upthere on a bicycle,^ she tells the call-taker, and Bthey’re going around the house and nowthey’re taking screens off the windows.^ Following this report, offered in the form of anabbreviated story, the call-taker obtains the address of the house in question and the absenthomeowner’s name, and also checks that the Bsomebody^ removing the screens is but oneperson, possibly occasioned by the caller’s use of Bthey’re^ in her account. In addition,through further questioning, the call-taker acquires the caller’s name and phone numberand confirms that the caller doesn’t recognize the Bsomebody,^ whom the call-taker refersto as Bthis person^ (indeed, the call-taker’s reference term seems chosen here for its co-class membership with the caller’s Bsomebody^). The following exchange then takesplace:

01 CT: .hhh O:ka:y, (a:nd), what’s this person:- You say,02 What color bike did he drive up on?03 C1: It’s a red bicycle (.) It’s sittin’ right up04 against thuh house over there.05 (2.3)06 C1: It just looks funny, okay?07 CT: .hh and this guy white ma:le?08 (0.9)09 C1: uh, talk to my husband.10 He s[aw’im drive up, >just a minute<11 CT: [okay12 (.)13 C1: ((speaking to C2)) What’s:- thuh guy look like?14 (0.4)15 C2: >Hello<16 CT: Hi, is this a white male?17 C2: YesNotice that the call-taker starts to ask a question that may well be headed toward seeking

more information about Bthis person.^ But the call-taker cuts off her own question-in-progressand shifts her attention instead to the bicycle (line 1–2). In doing this, however, the call-taker

also shifts her reference to the bicycle’s rider: the bicycle Bhe^ rode up on, she says, assigninga gender identity even though the caller has not yet done so in her account. The caller(identified in the transcript as C1) does not challenge or even mark this gendered referencein her answering turn, and then, after some two seconds of silence and still lacking someindication by the call-taker of how her story is being perceived (and thus what police actionmight result), the caller makes the kind of Bit just ....^ (line 6) explanation of her reasoning thatis commonly used to pursue such a response.

But the call-taker has been busy with recording the information about the bicycle in CAD.In fact, she has already completed and transmitted the CAD face sheet form to the policedispatcher, coding the event as a Bpossible burglary^ and typing BUNK SUBJECT TAKINGSCREENS OFF OF WINDOWS THE RES IS OUT OF TOWN^ in the DTL (details) field,recording the central elements of the caller’s story (and what makes the situation suspicious).Moreover, as the caller is pursuing a response, the call-taker is completing the first Badditionaldetails^ form for this incident, focusing on that bicycle and using virtually all the informationthe caller provided—that the bicycle is red and is Bsitting against the house.^

The call-taker is now ready to start getting more detailed information about the bicyclerider. She has already referred to the rider as a male, a Bhe^ (line 2), but the wording she uses inher first question aimed at assembling a description of that rider—Band this guy white male?^(line 7)—has some interesting features. Conversationally, the call-taker’s addition of a sexcategory to a racial categorization of the bicycle rider in her question is redundant. That is, whynot simply ask, BAnd is this guy white?^ where she just referred to that rider as Bthis guy,^therefore indicating that the rider’s gender—male—had been already identified. It can beexplained only when we consider the institutional setting in which the call is embedded. Shemakes an initial gendered reference (Bthis guy^) that is immediately followed by a pairing oftwo terms, Bwhite male,^ because, in this context, Bwhite male^ is being used not as terms ofreference but rather as what Sacks (1972a, b) calls Bmembership categorization^ terms, fromthe Bcategorization devices^ race and sex. These terms are therefore being used to docategorization. In addition, by proposing these categories and initiating this categorizationwork, the call-taker is asking the caller to confirm rather than make a certain racial and sexualclassification, and to thus join her in that work.

The categorization work in which the call-taker is engaged, and trying to engage the callerwith, is even more evident in the following sequences. The caller responds to the call-taker’squery by turning the phone over to her husband (identified in the transcript as C2), with theexplanation to the call-taker that he saw the subject (now finally referred to as a Bhim^ by thecaller) ride up, and also prompts the husband for that handover with an interpretation of thenature of the information the call-taker is seeking. In doing so, the caller uses the everyday,vernacular term instead of the organizational, bureaucracy type: Bwhat’s thuh guy look like?^(line 13). But the call-taker immediately directs a question to the husband, after acknowledginghis Bhello,^ saying: BHi, is this a white male?^ (line 16). Her distinct wording here closelyresembles her earlier query to the wife but drops what we said sounded rather like a redundantgender identification. In both questions by the call-taker, the race + sex category linkage ispreserved. Put another way, she does not limit herself to only a gendered reference, as in BIsthis guy white?^ or BIs this a white guy?^ but rather a categorization term, or, more accurately,a pair of them. And as we saw with her earlier question to the wife, the call-taker is plainlysoliciting the husband’s collaboration in her categorization work.

This categorization-oriented questioning, especially the race + sex pairing along with thesomewhat puzzling use of two gender-related terms in one turn is not idiosyncratic to that

particular call-taker, or a particular time in the work of emergency calls. Rather, suchmembership categorization work is prerequisite for successful coordination in bureaucracies,including call centers, and becomes a predictable pattern of how interactions are managed insitu to meet the institutionalized structures—what Drew and Heritage (1992) aptly termBinstitutional fingerprints.^ It seems plain, then, that we cannot account for this kind of action,with this distinct formulation, in terms of individual call-taker preferences. This can be furtherillustrated by the following examples from calls handled by different call-takers, whichdemonstrate that this is a mundanely recurrent practice at Central Lane.

& Awoman is describing Ba girl getting beat up^ in her apartment complex, indicating thatthe assailant is a man by saying, BHe’s hit her a couple of times,^ and the call-taker hasasked if she can tell her Bwhat they look like^:

C: Okay he’s got a beard (0.6) and he looks kindabi- like a biker(0.6)

CT: Is he a white [male?C: [Yeah, and she’s oriental.

& A caller is reporting a person, referred to as Bhe^ and Ba big guy,^ who is behavingsuspiciously at an apartment complex:

CT: Okay now he’s a white male?C: Well I couldn’t- (well he) long dark

hair (0.3) he could be Mexican (0.2) orcould be white male

& Ayoung woman is phoning about a troubling event she observed: A car driving past with awoman Bhanging her head out the window^ and crying out Bcall the police!^:

CT: Do you know her?C: No I don’t. She wazza:=CT: =is she a white female?C: Yeah, a white female, blonde hair,

& A motel manager is reporting an armed robbery and has referred to the robbers as Bthey^and Bhe^:

CT: Okay (.) give me a description of thefirst (.) suspect

C: We::ll I (.) [( )CT: [Is he a white male?

(0.4)C: We- both of um white

& A store owner is reporting a prowler, referred to as Bhe,^ and is recounting this person’slast observed location:

C: . . . it’s real close to the back door to mystore an [I saw him ( -)

CT: [Okay is he a white male?C: No he’s Chicano

We can draw several conclusions from our review of the data to this point. First, to reiteratean argument we only briefly made earlier, there can now be no doubt that the standard incidentrecord, with its terminological and syntactic conventions for inscribing person descriptions,

has a hearable, profound Bpresence^ in these interactions. The recurrent use by call-takers ofquestions taking the form BIs he a white male?^ points directly to their work of selecting fromand transforming the caller’s descriptions or formulations of persons into a format that isaccountable to and consistent with their department’s policy regarding how that type of textstring description should be recorded. While a query along the lines of Bcan you give me anydescriptions of what they look like^ is quite open-ended and exerts very little socio-interactional constraint on what might be produced as an acceptable, sensible answer, BIs hea white male?^ calls for a simple confirming or disconfirming answer and thus drasticallyreduces the range of acceptable responses. This interrogative practice, taken up at the verybeginning of assembling a person description, serves to closely control the topical focus andtrajectory. And this in turn offers a solution, at least with respect to the initiation of anydescription-oriented interrogation, to the practitioners’ problem of assembling the itemized listof characteristics mandated by the organization through its standard form.

Second, the pairing of race + sex (Bwhite male^) categories in the wording of the call-takers’ questions provides us with even more precise evidence for the form’s Bpresence^ andthe call-taker’s strategic practice. Recall that the standardized syntax for person description textstrings begins with that same category pairing, as we illustrated with our earlier example fromone of Central Lane’s regulatory documents. In so many words, the call-taker is giving voice tothe form’s required race + sex linkage and its syntactical first-position in the descriptive textstring.

Third, we can now say that what we initially described as a somewhat puzzling redundancywith respect to references to the suspect’s sex—BIs he a white male?^—actually represents ashift, over the course of that single questioning turn, in the nature of the call-takers’ actions, inwhat she is doing through her talk (Schegloff 2007a, b). She shifts from doing referencing todoing membership categorization (and as we just indicated, a paired categorization that is inkeeping with an organizational agenda). In other words, what may sound like redundantwording (he + male) is not a case of redundant action. Moreover, the question is addressedto the caller’s local situation of looking and reporting (cf. Smith and Whalen 1995). It switchesinternally from a deictic mode of reference, proper to the caller’s local setting, to thecommunications center’s formalized conventions for descriptions of persons, which are athome in and so feel natural to the world of the police: Bhe^ is now Bwhite male^/WM.

This is expressly what Latour andWoolgar (1979; see also Lynch 1985; Smith 1990a) meanby Binscription^: the production of an event or object (in this case, a person) in documentaryform. It is the institutional classification scheme for processing cases—the bureaucraticgenre—embodied in the incident form that prompts the categorization work of call-takers.The call-takers, in turn, must obtain the relevant information in this institutional context in arelevant manner—a race + sex paring in this case—to make the call Bprocess-able^ for theorganization, thus making the call institutionally successful.

Entering Document Time

To this point we have examined only the very beginnings of person description work,categorization by race and sex. The issue to which we must now turn is how call-takers mightthen continue to obtain information from callers in a manner that affords a Bcomplete^ textualrepresentation and how that representation enters into and coordinates further courses oforganizational action. In this regard, we return to our 2001 call.

One practice aimed at achieving such Bcompleteness^ as well as maintaining control overthe interaction—Btaking control of the call^ so that the necessary information can be obtainedand recorded on the form, in the proper fashion and in a timely manner, which is understood tobe an essential skill in competent call-taking—is to order the questioning of the caller so as tofollow the organizationally prescribed ordering of descriptive features. We can see this in theinterrogation that follows the confirmation by the caller’s husband that Bthe guy^ who rode upon the bicycle is a white male:

64 CT: About ho- how old?65 C2: Uh:, he look’s like he’s in his, um: : (0.2)66 late (0.9) or e- early: : :- la:te twenties:,67 (0.6)68 CT: Could you tell about how tall he was?69 (0.2)70 C2: Uh: :, let’s see, his head just about reached thuh71 top of thuh fence, so I’d say it was- he’s prob’ly72 almost six foot73 (0.6)74 CT: .hh Was he heavy, medium or slender in buil : d?75 (0.3)76 C2: He’s medium.77 (0.5)78 CT: A:nd, what color hair did he have?79 (0.3)80 C2: He had uhm:, uh:, brunette.81 (0.8)82 CT: .hhh a: : nd, any facial hair, did you see that?83 C2: No, it’s too [f- uh- (-) (-)°84 CT: [any glasses or anything?85 C2: (Nope.) N [o glasses.86 CT: [What was he wearing?87 C2: He’s got a- like a leather jacket, and he [has a-88 CT: [What89 color leather?90 C2: Uhm:: (0.2) Brown.91 (1.0)92 CT: ahkay, and [what else?93 C2: [he has uh- a blue: (0.6) uh pack94 (2.1)95 CT: Back pack?96 C2: Yes.The call-taker on this occasion plainly makes a concerted effort to elicit additional

description information in the same order in which she is expected to record it, starting withage and then moving to height, weight, hair color, facial hair, glasses, and then finally clothing.Observe also that there are pauses ranging in length from a half-second to two seconds after theanswers to these questions, with the exceptions being clarifying queries (e.g., BWhat colorleather?^); the call-taker’s typing can be heard during those pauses. The husband waits for thenext question each time, quietly collaborating in the text-assembly work that is hearably

evident through the call-taker’s typing. In this way, the collaborative work of that talk has onceagain entered into another kind of exchange, a Bconversation^ of sorts between the digital formand the call-taker that Bdrives^ the call-taker’s part in her exchange with the caller. Theprescribed format for person description text is like a series of questions that insistently seeka response. The coded or otherwise specially structured product of the talk as it is entered intothe computer and appears on the form is that response. Here is the complete, two-linedescription text (transmitted as Badditional details^) that resulted:

S/WM,L 20’S, 600, MED, BRO LSW/BRO LEATHER JKT, BLU BACK PACK3

While the ordering of descriptors in the text is closely aligned here with the order in whichthey were produced in the talk, the call-taker is plainly not a stenographer. All qualifiers ormodifiers provided in the talk – Blook’s like,^ Bprob’ly^ – were dropped in creating the text, aswas one rather indeterminate observation (Bhis head just about reached the top of the fence^)offered as a kind of comparative yardstick for height estimation. In addition, the husband’schoice of Bbrunette^ for the subject’s hair color is changed to Bbrown^ in the call-taker’s text.Moreover, the Bdocument-ordered^ interrogation practice described above is pervasive in ourdata. All practitioners understand that managing the talk-text relationship and maintainingBcontrol^ so as to Bobtain important information^ is a problem for which the solution is alwaysBadequate for practical purposes^ (see especially Whalen and Zimmerman 1998).

There can be no single way to manage this. The nature of the situation is an importantconsideration in deciding what and how much to ask and to inscribe, and when. A call-takermay not always want (or be able) to impose the form’s ordering of information, including theprescribed syntax for listing descriptors, on their questioning of the caller (as BIs he a whitemale?^ seems to do). And, of course, the caller is hardly a passive subject in these interactions;they may well offer information about persons in a manner or at a place in the talk that is notperfectly responsive to the call-taker’s interrogative moves. For all these reasons, then, a call-taker’s textual inscribing work may have to Bre-organize^ the information after it is vocallyconveyed. Thus, both the caller and call-taker actively collaborate in assembling a descriptionas an itemized list, transforming what the caller has observed into a descriptive text made up ofbureaucratically standardized properties.

Note especially that the resulting text does not display the original order in which theconversational sequence produced the descriptive items. The localized recipient design em-bedded in and accomplishing the ordering of the conversational sequence is displaced in thetext by the standardized recipient design of the organizational course of action (Smith andWhalen 1995). But in directing attention to the work of converting information in the talk toinformation in the text—to that reflexive, thoroughly interwoven relationship between docu-ment and action—we most certainly do not mean to suggest that it is just a matter of translatingvernacular speech and common sense reasoning to organizational logic. As we emphasizedabove, call-takers are expected to use their professional judgment in interrogating their callersand assembling their texts, taking into account the nature of the situation at hand (the localcircumstances) as well as the practical requirements of policing. Instead, we mean to directattention to the fact that all of the interactional work by all the parties to such calls, as well asthe immensely practical inscribing work done by any call-taker in assembling a textual

3 LSW is an abbreviation for Blast seen wearing,^ and is sometimes used as a prefix for clothing descriptors.

description from all that was said, and doing so in an accountably Bproper^ manner, is in noway evident from the incident record texts; it cannot be recovered by reading the documents.

Indeed, any comparison of person description texts will show that they look very much thesame in terms of the kind of information that is recorded and the way it is inscribed, eventhough the sequences of talk that provided the raw material, so to speak, for those descriptionswill always differ (in sometimes small, sometimes more noteworthy ways). Still, given theorganization’s commitment to standardized person descriptions, and the ways in which thisand other prescriptions are then embedded in a standard form, we should certainly be surprisedto find something other than visibly standardized texts in those documents. So the centralissue, then, is not standardization per se but rather precisely how such standardization operatesin situ to coordinate action across time and space, to help Borganize the organization.^

To speak to this issue we need to introduce a concept drawn from Smith’s (1990b) research:document time. BAt some point an account is fully worked up; at some point it drops away thetrace of its making…and stands forth as an autonomous statement,^ Smith (1990b, 74) writes.The processes, the actual concrete actions involved with working up the formulation—whatLynch (1985, 64) calls the text’s Bpractical genealogy^—become invisible. BThe accountcomes to stand in for the actuality it claims to represent,^ Smith (1990b, 74) adds. BIn thecontext of the social organization of its reading, it becomes a virtual reality. The text isstabilized. It has no apparent history other than that incorporated in it and does not acquireone as a product of the various occasions of its use. It has now entered Bdocument time.^ Andthis is precisely how documents acquire the wherewithal to Borganize the organization^: bycrystallizing and preserving some definite form of words detached from their local historicity,from the specific circumstances and settings within which they are produced, a technicallyidentical text can be operating, can be used or taken up for some purpose, in different places atthe same time, or in the same place at different times. That is to say, it is this coordination fromsetting to setting that the stabilized document enables.

This point can be further demonstrated through a close examination of how such docu-ments, with their prescribed ordering, serve as the official basis for the next step in thesequence of organizational action—the dispatching of police officers to the scene of theincident. Space limitations preclude any detailed analysis of the dispatcher’s work. However,we can note that while the CAD document continues to provide a means of representingworldly events to facilitate organizational response, there must now be a reversal of sorts interms of the direction of that representational work. Dispatchers, whose access to any worldlyincident is almost always entirely textual, must therefore now render that text into talk. Andthey commonly have to do this rather quickly, as they might receive several CAD texts on theircomputer screen over the course of just a few minutes, which are to be added to a list ofincident records from prior calls that are still holding for dispatch. Suffice it to say that thisoften time-sensitive work is scarcely a matter of reciting the text. Rather, it necessarily requiresbringing the ordinary conventions that govern talk (pacing, inflection, emphasis, and the like)to the text, as we can see in the dispatcher’s radio talk on our Bpossible burglary^ call, whereshe recounts the event and the parties reportedly involved:

D: …The neighbor’s the caller from six seventy two Maplewood,advising the residents are out of town. Subject rode up on a redbicycle, it’s been leaned against the house. White male, latetwenties (.) six foot, medium build (.) brown hair (.) brownleather jacket, unknown pants. (.) ↓Blue backpack.

Recall that dispatchers are expected to follow a prescribed syntax when broadcasting suchdescriptions over the radio (indeed, you will also recall that radio talk requirements are the veryreason for that prescription): sex, race, age, height, weight, hair, color of eyes, and thenclothing, ordered from top (hat) to bottom (shoes), recognizing that in many cases items likeeye color, shoes, and pants (as well as others) may not be elicited by call-takers from callers orincluded in the document text. Compare the dispatcher’s talk here with the person descriptiontext shown earlier. Plainly, the dispatcher’s talk added words that turned what would otherwisebe a literal reading into something that was an oral narrative rather than one that was purelytextual. Indeed, although the dispatcher followed the order of the person description textprecisely, in the translation from that text to her talk we find a procedure that is the inverse ofthat observed in the call-taker’s production of a documentary description.

That is, the dispatcher did not simply read off the list of descriptors as it appeared to her onthe computer screen; rather, she inserted the appropriate grammatical connections, constructingan abridged narrative. The text’s S/WM becomes Bwhite male,^ with the person being soclassified presumably heard and understood as the Bsubject^ mentioned in her immediatelyprior remark concerning the bicycle. The abbreviations that are fully legible in the text aretransposed into their spoken form. L 20’s becomes Blate twenties,^ 600 becomes B6 ft,^ MEDbecomes Bmedium build,^ BRO becomes Bbrown hair,^ and so on. Further, the dispatchertakes the absence of any information regarding the subject’s pants as something worthaccounting for in her radio broadcast: Bunknown pants.^ Moreover, the dispatcher showedsome additional narrative initiative with the account she offered prior to giving the persondescription: where the call-taker’s text read Bdrove up^ the dispatcher uses Broad up^ and,similarly, Bsitting against^ in the text is transformed by the dispatcher in her talk to Bleanedagainst.^

In sum, the text is vocally rendered as (and reconstructed into) a story, albeit an official-organizational one through the use of stock characters like Bsubjects^ who are Bwhite males^and a plotline—a Bpossible burglary^—with all the necessary elements, like residents who areout of town, concerned neighbors, and strangers taking screens off a window. In Smith’s(1990a, 142) terms, the document’s particulars have been Bselected, ordered, worked up^ toarticulate this incident type, the Bunderlying pattern,^ much as the call-taker selectivelyworked up the particulars of the caller’s account in assembling her text. It is not, then, thestory that the caller sought to tell, which was itself the subject of the call-taker’s interrogativeintervention that consequently led to a movement away from colloquial story-telling andtoward the Bhooking^ of the caller into an organizational relay, and through that, a corre-sponding official-bureaucratic mode of description.

We can conclude from these observations that while organizational policies dictate how acertain type of radio-broadcast action, such as a person description, is to be produced (and, wehave argued, this is indeed effectively a Bproduction^), requiring call-takers to inscribe theirtext so as to enable that radio production, dispatchers nevertheless—and, we would argue,necessarily—seek to expand on that text, to animate what could otherwise be a rathermechanical, routinized performance, perhaps one that is not even ordinarily intelligible. Theybring the text into the conventions that govern talk. And, although standard operatingprocedures for dispatching may require that turns at radio talk must include certain informationitems delivered in a certain order, dispatchers necessarily make use of pacing, inflection, andemphasis in their delivery—such as stressed syllables or sounds; small pauses in betweenwords, phrases, and turn components (as can be seen in the above segments from thedispatcher’s talk); and quick rising and falling intonation shifts—to both convey Bmore^ than

the use of certain words alone can and transform what might otherwise be a rather non-interactional transmission (as in the broadcast of an announcement over commercial radio) intoa more lively and interactionally-sensitive spate of talk that is still recognizably standardized,still Bofficial police talk.^

Thus, while the digital document organizes the dispatcher’s action, and is therefore anactive party to the radio communication that follows, it does not enter into that actionmechanically (any more than it does into the call-taker’s actions). Its active role is realizedas it is taken up and animated in the local organization of the conversational sequence betweendispatcher and police that now truly completes the relay (Smith and Whalen 1995). But notefinally that it is the digital form’s entry into document time that affords this completion. Thewords that have been entered into the form’s labelled fields are indeed Bdetached from theirlocal historicity,^ transformed by the call-taker’s work with the form—work guided by theform’s organization and text-entry conventions—into an official incident record. The dispatch-er has only this record, this purely textual access to the caller’s story, to the events and, as wehave shown specifically, the persons described in and through the call-taker’s textual repre-sentations.4 And from that now Bstabilized^ record from this organizing document, dispatchersnecessarily have to reconstruct an admittedly abbreviated but sensible narrative, their descrip-tive work oriented to both the interactional and police-organizational demands of talk spoken(and designed so) over the radio to unseen but very much identifiable recipients.

Discussion and Conclusion: Document Mediated Social Organization

Our description and analysis of the production of person descriptions in a Btalk-to-text-to-talk^sequence sustains our claim to bring an organizational sequence of action under scrutiny.Through a detailed examination of instances of Bdocument-driven^ sequences of talk betweencallers and call-takers (where the form is a lively contributor to the shape of the conversation),of the standardized textual representations assembled by call-takers during those conversa-tions, and of the subsequent rendering of those texts as radio talk by a dispatcher, we haveempirically demonstrated what the notion of Bdocument-mediated organization^ means inactual practice.

It should also be plain that the establishment of talk-text-talk sequence as deeplyBorganizational^ is demonstrated by the standardization of recipient design in the officiallyrequired descriptive syntax for which call-takers are trained and into which they translate thecaller’s vernacular. This simultaneously produces and solidifies a bureaucratic inscription ofpersons through certain kind of technology – policies and conventions that fully entitle us torefer to this sequence as an Borganizational^ course of action, regardless of the social origin ofthe form (the intentions of its designers or the specific history of how this form or any othercame to be) and regardless of what the organizational participants think of it (if they ever did).Once weaved into this organizing process, the document and its categories are resilient in a

4 The recordings of 9-1-1 calls (and radio dispatches) are also considered part of the official public record andsuch recordings naturally make the actual conversation between caller and call-taker available to communicationscenter staff (as well as members of the public who later request access, although confidentiality of the recordingscan vary somewhat by state), but reviews of these recordings in the communications center, when they do takeplace, are almost always some time after the fact (and usually for training purposes) rather than while activelyprocessing the event. And until very recently, when digital recordings replaced analog tapes, retrieving the audiorecord was very time consuming, which made concurrent reviews or consultations more or less impossible.

manner similar to the ways people have to work with their building once it is built (see Gieryn2002). In this sense, organizing documents also function as a technology of reification.

Nevertheless, whereas the Bscribes^ and Bsubalterns^ of Weber’s day—in modern times,the Bbureaucratic^ office staff, the call-takers, dispatchers and police—need to haveBknowledge of the files^ (Weber 1978/1914, 214) and are most exposed to these ways ofinscription, we also showed that the laypersons—those who call for help—must activelyparticipate in co-creating the Bcall-for-help^ and thus in doing categorization.

We want to also emphasize that studying document-mediated organization can providenew insights and possible directions for future research on myriad social phenomena. Topoint out just a few of these: First, we urge careful investigation into the methods bywhich bureaucratic genre—the manner of categorizing people and activities that isespecially characteristic of state organizations and officialdom—concretely enters intoour everyday social lives and common sense reasoning. Related to this, as our lives todayare saturated with Bpeople processing organizations^ (Emerson 1992), it is important tostudy how their methods for making sense of the world and Bsorting things out^ impactour lives in terms of actual practice (Bowker and Star 1999). To take just one example,Goffman (2014) has described how legal terms used in state prison permeate the socialfabric of poor black Americans in a Philadelphia ghetto, where their significant othersbecome Bco-d^ (co-defendants) and good friends are Bon their call list.^

Second, one can also study the mechanisms of Bnegotiation erasure^ in organizations. Aswe highlighted during our analysis of calls for help, focusing on the enactment of formaldocuments allows us to explicate just how organizations strip off creative descriptions ornegotiations between parties in interaction. It is precisely the fact that practical genealogy(Lynch 1993) is lost in inscription that makes the Bbureaucratic genre^ so durable. Localnegotiations sometimes make their way into organizational action – as sticky notes attached tosome memo for processing, or the kinds of scribbled writings one might find in the margins ofsome official document (Goodman 2008; Luff et al. 1992; Moore et al. 2011), but more oftenthan not such arrangements remain contingent on actors’ constant initiatives. Thus, we couldbegin to investigate the erasures and incorporations of such local contingency on standardorganizational forms, and how this may (or may not) change the entries and features ofstandard forms—and thus the organizationally acceptable categories—to explain the durabilityof organizational forms and protocols as well as organizational change.

Third, and related to the second point, by tracing the processes through whichorganizational action is requested, transmitted, granted and executed in and throughdocumentary work, we can unravel how organizational mistakes, violence orcatastrophes happen. As Vaughan (1996) forcefully shows, organizations may, and oftendo, make fatal mistakes, not despite but rather because of rules and the Bceremonialsigning^ (Vaughan 1996, 248). Vaughan problematized the catastrophic launch decisionof the space shuttle Challenger with the term Bstructural secrecy,^ a familiar or eveninherent problem for bureaucracy: documents piling up, never to be closely read;documents showing approval of decisions but erasing the details or negotiations thatled to such local decision (what we referred to as Bnegotiation erasure^ above); and anunofficial Bmemo^ recommending changes never making into the official proceduresneeded for corrective actions (Vaughan 1996). By studying how talk in reality getstranslated into documents (or not), and then how documents initiate the next sequencesof talk and organizational action in different organizational spaces, we can startpinpointing where misjudgments are likely to occur in particular organizational settings.

But the examples above point to only a handful of possible research initiatives. Morebroadly, we have depicted just how, and how much, documents enter into and are integral tothe constituting of relations in organizational and institutional life (cf. Smith and Whalen1995). We make this argument in the hope that it will light the way for other researchers,suggesting the possibilities—the real need, we believe—of respecifying traditional concepts ofbureaucracy, formal organization, administration, and the like. How do such phenomenaachieve social facticity? That is the fundamental concern of our studies, and that can inspireinvestigations like these into the locally produced sequences of action that are taken-for-granted but unexplicated in the ordinary usage of such concepts. Nearly a century has passedsince Weber’s suggestive remarks first brought these possibilities to mind. We think that istime enough.

Acknowledgments We would like to thank Iddo Tavory, Marilyn Whalen and the anonymous reviewersof Qualitative Sociology for their helpful comments.

References

Atkinson, J. Maxwell. 1978. Discovering suicide: Studies in the social organization of sudden death. London:Macmillan.

Blau, Peter M. 1963. The dynamics of bureaucracy: A study of interpersonal relations in two governmentagencies. Chicago: University of Chicago Press.

Bowker, Geoffrey C., and Susan Leigh Star. 1999. Sorting things out: Classification and its consequences.Cambridge: MIT Press.

Button, Graham, and Wes Sharrock. 1997. The production of order and the order of production: Possibilities fordistributed organizations. In Work and technology in the print industry: Proceedings of the Fifth EuropeanConference on Computer Supported Cooperative Work, 1-16. Norwell: Kluwer.

Cahill, Spencer. 1998. Toward a sociology of the person. Sociological Theory 16: 131–148.Callon, Michel. 1986. Some elements of a sociology of translation: Domestication of the scallops and the

fishermen of St Brieuc Bay. In Power, action, and belief: A new sociology of knowledge? ed. J. Law, 196–223. London: Routledge.

Coleman, James S. 1974. Power and the structure of society. New York: W. W. Norton.Dalton, Melville. 1959. Men who manage. New York: John Wiley.Drew, Paul. 2006. When documents Bspeak^: Documents, language, and interaction. In Talk and interaction in

social research methods, ed. P. Drew, G. Raymond, and D. Weinberg, 63–80. London: Sage.Drew, Paul and John Heritage. 1992. Analyzing talk at work: An introduction. In Talk at work: Interaction in

institutional settings, ed. P. Drew and J. Heritage, 3–66. Cambridge: Cambridge University Press.Duster, Troy. 2004. Selective arrests, an ever-expanding DNA forensic database, and the specter of an early

twenty-first-century equivalent of phrenology. In DNA and the criminal justice system, ed. D. Lazer, 315–334. Cambridge: MIT Press.

Emerson, Robert M. 1992. Case processing and interorganizational knowledge: Detecting the real reasons forreferrals. Social Problems 38: 198–212.

Feldman, Martha S. 1989. Order without design: Information production and policy making. Stanford: StanfordUniversity Press.

Frankel, Richard M. 1989. BI wz wondering-uhm could ‘Raid’ uhm e’ffect the brain permanently d’y know?^:Some observations on the intersection of speaking and writing in calls to a poison control center. WesternJournal of Speech Communication 53: 195–226.

Foucault, Michel. 1977. Discipline and punish: The birth of the prison. Trans. Alan Sheridan. New York:Pantheon.

Garfinkel, Harold. 1967. Studies in ethnomethodology. Englewood Cliffs: Prentice-Hall.Garfinkel, Harold, and Egon Bittner. 1967. BGood^ organizational reasons for Bbad^ clinic records. In Studies in

Ethnomethodology, ed. H. Garfinkel, 186–207. Englewood Cliffs: Prentice-Hall.Geisler, Cheryl. 2001. Textual objects: Accounting for the role of texts in the everyday life of complex

organizations. Written Communications 18: 296–325.

Giddens, Anthony. 1979. Central problems in social theory: Action, structure, and contradiction in socialanalysis. Berkeley: University of California Press.

Giddens, Anthony. 1984. The constitution of society: Outline of the theory of structuration. Cambridge: PolityPress.

Giddens, Anthony. 1987. Social theory and modern sociology. Cambridge: Polity Press.Gieryn, Thomas F. 2002. What buildings do. Theory and Society 31: 35–74.Goffman, Alice. 2014. On the run: Fugitive life in an American city. Chicago: University of Chicago Press.Goodman, Philip. 2008. BIt’s just Black, White, or Hispanic^: An observational study of racializing moves in

California’s segregated prison reception centers. Law & Society Review 42: 735–770.Goody, Jack. 1986. The logic of writing and the organization of society. Cambridge: Cambridge University Press.Gouldner, Alvin W. 1954. Patterns of industrial bureaucracy. New York: The Free Press.Harper, Richard H.R. 1998. Inside the IMF: An ethnography of documents, technology, and organisational

action. London: Academic.Harris, Roy. 1986. The origin of writing. La Salle: Open Court.Heath, Christian. 1982. Preserving the consultation: Medical record cards and professional conduct. Sociology of

Health and Illness 4: 56–74.Heath, Christian, and Paul Luff. 2000. Technology in action. Cambridge: Cambridge University Press.Jerolmack, Colin and Iddo Tavory. 2014. Nonhumans and the constitution of the social self. Sociological Theory

32: 64–77.Knorr Cetina, Karin, and Urs Bruegger. 2002. Global microstructures: The virtual societies of financial markets.

American Journal of Sociology 107: 905–950.Latour, Bruno. 1987. Science in action: How to follow scientists and engineers through society. Boston: Harvard

University Press.Latour, Bruno. 1992. Where are the missing masses? The sociology of a few mundane artifacts. In Shaping

technology/building society: Studies in sociotechnical change, ed. W.E. Bijker and John Law, 225–258.Cambridge: The MIT Press.

Latour, Bruno. 2005. Reassembling the social: An introduction to actor-network-theory. Oxford and New York:Oxford University Press.

Latour, Bruno, and Steve Woolgar. 1979. Laboratory life: The social construction of scientific facts. Princeton:Princeton University Press.

Leidner, Robin. 1993. Fast food, fast talk: Service work and the routinization of everyday life. Berkeley:University of California Press.

Levy, David M. 2001. Scrolling forward: Making sense of documents in the digital age. New York: Arcade.Luff, Paul, Christian Heath, and David Greatbatch. 1992. Tasks-in-interaction: Paper and screen based docu-

mentation in collaborative activity. In Proceedings of the 1992 ACM conference on computer-supportedcooperative work, 163-170. Toronto: ACM.

Lynch, Michael. 1985. Art and artifact in laboratory science: A study of shop work and shop talk in a researchlaboratory. Boston: Routledge & Kegan Paul.

Lynch, Michael. 1993. Scientific practice and ordinary action: Ethnomethodology and social studies of science.Cambridge: Cambridge University Press.

Lynch, Michael, and David Bogen. 1996. The spectacle of history: Speech, text, and memory at the Iran-Contrahearings. Durham: Duke University Press.

Mackie, Lindsay. 1978. Race causes initial confusion. The GuardianManning, Peter K. 1992. Organizational communication. Hawthorne: Aldine de Gruyter.Meehan, Albert. 1986. For the record: Organizational and interactional practices for producing police records on

juveniles. Ph.D. dissertation, Department of Sociology, Boston University.Meyer, John, and Brian Rowan. 1977. Institutionalized organizations: Formal structures as myth and ceremony.

American Journal of Sociology 83: 340–363.Moore, Robert, Jack Whalen, and E. Cabell Hankinson Gathman. 2011. The work of the work order: Document

practices in face-to-face service encounters. In Organization, interaction and practice: Studies of real timework and organizing, ed. Nick Llewellyn and Jon Hindmarsh, 172–197. Cambridge: Cambridge UniversityPress.

Morrill, Calvin. 1996. The executive way: Conflict management in corporations. Chicago: The University ofChicago Press.

Orlikowski, Wanda J., and JoAnne Yates. 1994. Genre repertoire: The structuring of communicative practices inorganizations. Administrative Science Quarterly 39: 541–574.

Pelkmans, Mathijs. 2013. Powerful documents: Passports, passages and dilemmas of identification on theGeorgian-Turkish border. In Border encounters: Asymmetry and proximity at Europe’s frontiers, ed. JuttaLauth Bacas and William Kavanagh, 90–107. Oxford: Berghahn Books.

Pettinari, Catherine Johnson. 1988. Task, talk, and text in the operating room: A study in medical discourse.Norwood: Ablex.

Robinson, Andrew. 1995. The story of writing: Alphabets, hieroglyphs, and pictograms. London: Thames and Hudson.Sacks, Harvey. 1972a. An initial investigation of the usability of conversational data for doing sociology. In

Studies in social interaction, ed. David Sudnow, 31–74. New York: Free Press.Sacks, Harvey. 1972b. On the analyzability of stories by children. InDirections in sociolinguistics: The ethnography

of communication, ed. J.J. Gumperz and D. Hymes, 325–345. New York: Holt, Rinehart and Winston.Sacks, Harvey. 1979. Hotrodder: A revolutionary category. In Everyday language: Studies in ethnomethodology,

ed. George Psathas, 7–14. New York: Irvington Press.Sacks, Harvey, Emanuel A. Schegloff, and Gail Jefferson. 1974. A simplest systematics for the organization of

turn-taking in conversation. Language 50: 696–735.Schegloff, Emanuel A. 2007a. Categories in action: Person-reference and membership categorization. Discourse

Studies 9: 433–461.Schegloff, Emanuel A. 2007b. A Tutorial on membership categorization. Journal of Pragmatics 39: 462–482.Selznick, Philip. 1949. TVA and the grass roots: A study in the sociology of formal organization. Berkeley:

University of California Press.Smith, Dorothy E. 1990a. Texts, facts and femininity: Exploring the relations of ruling. London: Routledge.Smith, Dorothy E. 1990b. The conceptual practices of power: A feminist sociology of knowledge. Boston:

Northeastern University Press.Smith, Dorothy E. 1999.Writing the social: Critique, theory, and investigations. Toronto: University of Toronto Press.Smith, Dorothy E. 2001. Texts and the ontology of organizations and institutions. Studies in Cultures,

Organizations and Societies 7: 159–198.Smith, Dorothy E. 2005. Institutional ethnography: A sociology for people. Oxford: AltaMira.Smith, Dorothy E. and Jack Whalen. 1995. Texts in action. Unpublished paper. Ontario Institute for Studies in

Education, University of Toronto; Department of Sociology, University of Oregon.Suchman, Lucy. 1993. Technologies of accountability: Of lizards and aeroplanes. In Technology in working

order: Studies of work, interaction and technology, ed. Graham Button, 113–126. London: Routledge.Suchman, Lucy and Jack Whalen. 1994. Standardizing local events and localizing standard forms. Presented at

the annual meeting of the Society for Social Studies of Science, New Orleans.Vaughan, Diane. 1996. The Challenger launch decision: Risky technology, culture, and deviance at NASA.

Chicago: University of Chicago Press.Weber, Max. 1978/1914. Economy and society: An outline of interpretive sociology, eds. Guenther Roth and

Claus Wittich. Berkeley: University of California Press.Whalen, Jack. 1995. A technology of order production: Computer-aided dispatch in public safety communica-

tions. In Situated order: Studies in the social organization of talk and embodied action, ed. George Psathasand Paul ten Have, 189–230. Washington: University Press of America.

Whalen, Jack and Don H. Zimmerman. 1998. Observations on the display and management of emotion in naturallyoccurring activities: The case of Bhysteria^ in calls to 9-1-1. Social Psychology Quarterly 61: 141–159.

Yates, JoAnne, and Wanda J. Orlikowski. 1992. Genres of organizational communication: A structurationalapproach to studying communication and media. Academy of Management Review 17: 299–326.

Zimmerman, Don H. 1966. Paper work and people work: A study of a public assistance agency. Ph.D.dissertation, Department of Sociology, University of California, Los Angeles.

Zimmerman, Don H. 1969. Record keeping and the intake process in a public welfare agency. InOn record: Filesand dossiers in American life, ed. Stanton Wheeler, 319–354. New York: Russell Sage.

Zimmerman, Don H. 1970. The practicalities of rule use. In Understanding everyday life: Toward the recon-struction of sociological knowledge, ed. Jack D. Douglas, 221–238. Chicago: Aldine.

Nahoko Kameo is an Assistant Professor/Faculty Fellow of Sociology at New York University. She primarilyworks in the areas of sociology of organization, work, economic sociology and sociology of science. She iscurrently expanding her dissertation project on Japanese university bio-scientists (forthcoming in Theory andSociety) as well as conducting an ethnographic project on sustainable fishing.

Jack Whalen is Adjunct Professor of Design and Ethnographic Research in the Department of Design at AaltoUniversity in Helsinki, Finland. He is also Director of User Experience in the Systems Division of SustainableFisheries Partnership, an international NGO. He is presently engaged in research on user inventiveness intechnology design and on intelligibility in design research.