Designing with Ethnography: A Presentation Frameworkfor Design.

John. A. Hughes, Jon O ‘Brien, Tom Rodden and MarkRouncejZeki.

Computing and Sociology Departments and CSCWReseareh Centre,

Lancaster University,Lancaster LA1 4YR.

E-mail j.hughes@lancaster. ac.uk

ABSTRACT

Despite the growing number of ethnographic studies ofwork their use in design remains a matter of some debate.Acknowledging the problems designers face in utiiisingethnographies, and etlmographers face in meetingcommercial demands, this paper outlines a ‘framework’for the presentation of fieldstudies organised around threemain dimensions; ‘distributed coordination’, ‘plans andprocedures’ and ‘awareness of work’; thereby facilitatingeffective communication and collaboration betweendesigners and ethnographers.

INTRODUCTION

The value of ethnography in design is a matter of somecontroversy [1] [2] [10] and while we are strongsupporters of the method, the role of ethnography as whave practised it is primarily as an injorrnarionrd input tothe design process, which can, nevertheless, be of criticafvalue in making visible the ‘real world’ aspects of a worksetting. 1 It is this ability of ethnography to describe asocial setting as it is pemeived by those involved in thesetting (the archetypal ‘users’) that underpins its appeal todesigners.z However, it is not enough to maintain thatethnogrrtphy has a modest role in design without also

‘ We suggestthat ettrnographicapproaches may clarify the role thatactrrsi practices play in tfte management of work through viewingactivities as social actions embedded witfdrr a socially organiseddomain and accomplished in and through the day-to-day activitiesof its usersThis is, in fact, a sociologically partisan conception ofethoography, but it d- have the advantage of focusing upmr theSPCCifk~d detailed organisation of activities which designers areconcernedto understand,etrafyseand reconstruct

2 In particular, it offers the opportunity to reveal ‘needs’ or

&y take them so much for granted that they do not think about‘ ractices’ of users which they may not be aware of - because

them - or which they cannot articulate - because of tbebureaucratic or power relationships within which they arc placed.Corrseqcmtty, ethnography is vahrable in identifying theexceptions, contradictions and contingencies of work activitieswhich arc red conditionsof the wnrk’s conduct but which wifl not(usually) figure in officird or formal representationsof that work.

Permission to make digital/hard copy of part or all this work forpersonal or cleaaroom r.!ee ie granted without fee provided thatcopiee are not mode or distributed for profit or commercial advan-tage, the copyright notice, the title of the publication and its date

appear, and notice is given that copying ia by permission of ACM,Inc. To copy otherwiaa, to republish, to peat on servers, or toredistribute to Iiata, requiras prior specific permission and/or a fee.DIS ’97 Amsterdam, Tha Natherlanda01997 ACM 0-89791 -883 -0/97 /0008 . ..$3.50

Steve Blythin,Project Manager,

Support Services and Design,NatWest Bank Plc.

Network Strategy & Development,101-117 Finsbury Pavement,

London EC2A lEH

acknowledging and confronting the fact that there ateproblems in enabling designers to utilise ethnography.

Clearly, there is a diffenmee in the needs and vocabularyof designers and fieldworkers. Design is about abstractionand complexity, and the deMgner is characteristically insearch of ways of simplifying the complexity of thedesign situation, often by means of abstractions whichwill delineate critical features of that situation and of thedesign probIem. The designer wishes to find mechanismswhich will quickly and succinctly convey the aspeets of adesign that are key to meeting the problem at hand.

Ethnography, however, is ditectd toward detail, towardthe production of a ‘rich’ and ‘concrete’ portrayal of thesituation, rather than an ‘abstract’ and ‘spare’ one.Abstraction and simplification are characteristicallyresisted, for it is the disposition of ethnography to insistupon citing items, incidents, activities or practices withintheir context, to emphasise that their meaning, or sense,is only properly comprehended within the appropriatesocial context.

The problem, then, is how to bridge the gap between thetwo kinds of representation, between the expansivetextual expositions of the ethnographer and the abstract,spare graphical depictions which seek to focus upon ‘keyelements’ and ‘core concepts’ of the designer. Given thedescriptive nature and informing role of ethnographicstudies, then the level and detail of coverage need only beset to meet the demands of a particular design activity,depending upon the scale of the setting to be redesigned,and the nature of the interdependenee amongst theactivities within the scope of the system. The need toincrease the utility of ethnography and to fostercommunication has d~tly motivated a number ofdevelopments for collecting, organizing and presentingethnographic material. Firstly, it has encouraged thespecification of a number of approaches to the collectionof data and its dissemination to designers [8]. Secondly,it has motivated the use of tool support where the studiesof work are reflected in a medium shared by the designers,where the materials from the study can be available todesigners in an rtecessible fashion, and where they can bedecomposed, through the use of a ‘viewpoints’presentation - into elements which can be selected in

147

terms of their relevance to the designer’s purpose. [9] [11]And thirdly, the focus of this paper, it has fostered theconstruction of a ‘framework for the analysis of work’that allows the results of ethnographic studies to besbuctured for presentation in a manner which makes theemerging results more digestible by designers.

Such a framework may also prove of value forethnographers, facing the commercial pressures of time,budgetary constraints and pressure for results, as a‘sensitising’ device for fieldworkers engaged, for example,in what has come to be called ‘quick & dirty’ or‘lightweight’ ethnography.[8] The capacity to gatherinformation through ethnography dedicated to thepurposes of informing design must itself be constrained,as is every other aspect of the design process, by theconditions of the project itself, the level of resources, thepressures of time and the nature of the design task. Oneway of understanding and structuring this process is toconsider the development of a framework that focuses onpresenting the issues key to the design process. Theframework outlined in this paper was developed tohighlight issues significant to development, andconsequently, it is primarily a means of pre$enringanalyses of work. It is an essentially practical device forboth organizing and understanding the vast array offieldwork observations of how the work in any complexsetting ‘gets done’. Furthermore, in providing addeploying this provisional framework we are attemptingto steer a difficult course, common to ethnographicobservation and analysis; between the accusation that it issimply ‘hanging around’ with the consequent suggestionthat its findings are entirely idiosyncratic, fortuitous andinconsequential to design; and the over-systematisation ofa ‘cookbook’ approach to fieldwork and analysis,effectively defeating the entire purpose of theethnographic approach.

The framework we have developed organises studiesaround three main dimensions: ‘DistributedCoordination’; ‘Plans and Procedures’ and ‘Awareness ofwork’. These different dimensions provide a skeleton forconstructing a presentation of fieldwork. A particularstudy will be undertaken within the context of a givendesign briefing. As part of this design intent particularissues will come to the fore. These different design issueswill reflect the thematic focus for the study. Each of theseissues will in turn become manifest within the differentdimensions and will be reflected in different ways un&reach of these contexts.

Distributed Coordination: within the frameworkrefers to the fact that work tasks performed in complexsettings are performed as part of patterns of activity, asoperations within the context of a division of labour, as‘steps’ in protracted operations, as contributions ofcontinuing ‘processes’ of activity.

Plans and Procedures: provides the framework with aprominent means by which distributed coordination isorganisationally supported. Project plans and schedules,manuals of instruction and procedures, job descriptions,

formal organisational charts, workflow diagrams are allways of enabling people to produce the orderliness ofwork. Particularly important to the treatment of ‘plansand procedures’ has been the need to clarify theunderstanding of how these function in the organisationof activities, consequently we suggest that a considerationof plans within cooperative work should: identify thedifferent actors and their potential relationships to plansand procedures; understand and accept the existence ofdiverse viewpoints on plans and procedures; record and besensitive to different notions of ‘following a plan’.

Awareness of work: is an expression referring to theway in which the organisation of work activities involvesmaking the nature of those activities ‘visible’ or‘intelligible’ to others doing the work. Just as distributedcoordination and the application of plans and proceduresare pervasive features of work sites, so too are practicesfor enabling others to have an awareness of the nature ofthe work that is being done.

The framework is fundamentally motivated by thedemands of design. The intent is to provide some bridgebetween fieldwork and emerging design decisions. Itworth stressing here that the framework is not ananalytical device that is then applied to a setting. Ratherthe aim is to use the framework to highlight particularissues within a corpus of fieldwork material. Each ofcategories within the framework loosely align withpotential design issues.

The emphasis on dirtnbuted coordination foregroundsissues of importance to the level of support for action andtasks within the system. Highlighted here are the mannerand means by which work is coordinated and theimplications this may have for support ranging fromsimple coordination mechanisms to more proceduralworkflow systems.

The emphasis on pfans and procedures highlight therelationship between users and the formal representationof their work. This relationship is a rich and complex oneemphasizing the difference between the description ofwork and ‘getting the job done’. Knowledge of thisdistinction and what it is to ‘do the job’ is central todesign decisions which potential reify the policy andprocedures of work with the system. For example, thesetting of access and status information, the sequencing ofactions in form based systems and the setting of roleprofiles in workflow systems. Often users ‘ad-hoc’ theprocess, and knowing these ‘shortcuts’ is central toinforming computer supported cooperative work (CSCW)design.

The final emphasis on awareness of work reflects thegrowing body of work on the importance of the publicnature of work and the availability of action [o others.This need for awareness is in contrast to traditionalindividurdistic approaches to the development of multi-user systems. Highlighting the need for sharing thatpromotes an awareness of the activities of others

148

I

w and topics to emerge from the study

Distributed Coordir ation~

Plans & Procedu esusers:

Awareness of W{)rk

I

Fieldwork Con qgcllsT f * y

Figure

emphasises to designers the fimdamentally socialcooperative nature of work.

1: The

and

APPLYING THE FRAMEWORK: THR BUSINESSCENTRE.

We have used the framework for the presentation ofresearch findings from a variety of fieldwork settings [4]and this section shows how the framework of the socialorganisation of work can be applied to one particularpiece of ethnographic fieldwork - that of a local BusinessCentre in a major clearing bank. The framework isutilised as an essentially practical device for bothorganizing and understanding the vast array of fieldworkobservations of how the work in the Business Centre‘gets done’. The Business Centre in focus is a relatively

The presentation fhtmework we employ, illustrated infigure 1 , hesides emphasizing the three general aspectsof CSCW - distributed coordination; plans and procedures;and the awareness of work - also, in this instance,incorporates some other cross-cutting themes thatemerged from the fieldwork and may become relevant todesign issues in each of the three general categories of theframework. The framework also orients the dimensions ofthe thrnework to map from concerns manifest in thefieldwork to a set of core design issues. The followingportions of the paper offer a presentation of the work ofthe business centre in terms of the dimensions of theframework and the issues that arise from the fieldworkwithin these dimensions.

DISTRIBUTED COORDINATION‘1 just followed the instructions - 1’11 seed it toRegion now end they can play around wfth it”

Distributed coordination refers to the various ways inwhich the coordination of people and tasks is

ort fornd Tasks

;nation of

Id work

Iser andon support

terns

framework

recent creation and is situated on the first and secondfloors of a remodeled local branch. As a ‘specialistcentre’ it is staffed by Business Managers, PersonalAccount Managers and their Assistanta and deals with‘medium sized’ businesses and personal customers whoseaccounts are worth over f25k. The general impression isone of calm and serious activity, conforming to aconsiderable extent to stereotypical images of ‘thebanker’, involved in the careful consideration ofcalculations about large sums of money. The work of theBusiness Centre is, therefore, somewhat dissimilar to theintense and often frenetic activity that prevails, forexample, at the branches and some of the other specialisedcentres dealing with the greater volumes, and smallersums, associated with the mass market.

accomplished as a routine feature of ‘real world, real time’work. 1An important, if commonplace feature of all worksettings is that tasks, activities and, persons are ernbdcdwithin an organised ensemble of some kind. Both theactivities and the people who perform them aminterconnected, and in the case of the Bank in general, andthe Business Centre in particular, they have to be treatednot as isofated activities and persons but as part of someorganisation of activities and persons. The point we wantto emphasise here is that although the fieldwork canprovide considerable detail, for example, of the

‘ Although, interestingly, one of the Bank’s main problems ofdistributedcoordination centsesaround the fact that its computersystemdoe-snot operatein ‘real time’ - so, for example, money canhe paid into an accounton the same day as chques are bounced.Sirnitarly the cheqrre clearance system does not operate in ‘realtime’ - so, for example, if a ‘outof order’printoutshowed an excesson agreed facilities, lending control needed to be cognisant of thefact that fundsthat were in clearing and might, theoretically put anaccountback within its facility, might atsobe ‘bounced’,

149

coordination work of the Administration Manager,coordination does not consist in any one feature of thework but is deeply, and inseparably, implicated in theprocedural details of all work activities. Much of thiscoordination work consists of distributing relevantinformation - for example, through the differentManager’s Assistants - to relevant parties and keepingthis flow of information going as a routine state ofaffairs. Thus many aspects of work in the Bank, whetherin the traditional branch or in the specialised centres suchas the Business Centre, are explicitly concerned withcoordinating interdependencies of various kinds in order to‘get the work done’.

Coordination through artefacts

Coordination can be a feature of specifically designed orevolved artefacts. As Yates [14] describes, memos, files,standard forms, etc., evolved to solve problems of‘distributed coordination’, as organisations became largerand the problems of management and controlcorrespondingly increased. These ‘standard forms’ atereadily apparent in the Bank. In the Bank the ‘out of orderaccounts’ printout acts as such an artefact, initiating andcoordinating various actions between the branches and thedifferent specialist centres. Other specific examples fromthe Business Centre would be the interview notepad adthe various standrad forms completed as part of thepreparations for a customer’s annual review; and theGAPP (Grading and Pricing Program) software package,also routinely used as part of the annual review. Suchartefacts facilitate the coordination of tasks by embeddingdescriptions of the task, along with other relevantinformation, within the format of a document as‘instructions’, as ‘items required’, and so on, makingavailable to those who know how to use the artefact,what the implications are for the actions of others. ‘I%eseartefacts serve as sets of instructions for a set ofinstitutionally identified persons - the ‘BusinessManager’s Assistant’, the ‘Records Clerk’ and so on - toperform particular tasks and, in addition, serve as a checkon whether or not these tasks have been performed. So,for example, completing the ‘Interview Notepad’ prior tointerviewing a customer a number of questions need to bertnswemd concerning, again for example, ‘Non FinancialDevelopments’; ‘Account Operation’; ‘Security’ and‘Remuneration’ and requiring the perusal of computer andpaper records and the completion of a number of otherforms. The ‘Appraisal Form’ for considering new,renewed, or revised propositions has a similar list ofquestions and a number of ‘action points’ - such as‘update 1SS’ or ‘mark limits’ - to be completed alongwith the necessary ‘expert computer codes’ required.

Identifying coordination features - Paperwork“its a pretty inaety file I dcmt know it..SO I have tov- wick~y ~wk ~d try and sort of acquaint myselfWIth wha ts going on emd whats been srrsnged “

Given the central role of coordination within cooperativework the identification of the featutes of work whichpromote coordination is clearly important. Notablecoordination features include the ability to monitor the

activities of others and access to shared and readilyavailable information much of which is accomplishedthrough paperwork. As has been observed throughout theBank a prominent feature of ‘everyday’ work takes theform of the completion, processing and duplication ofenormous amounts of paper, to the extent that in anumber of significant respects the routine completion ofpaperwork constitutes ‘the work’. For the BusinessCentre, paperwork, in the form of customer record foldersand customer reports (to Region) was an integral featureof the work. The ‘completeness’ of the paper recotdl wtedas an audit trail; providing an outline, rationalisation andjustification for administrative decisions. In the BusinessCentre however, this trail was valuable not simply ormerely for the attribution of blame but through itsprocedural implicativeness in informing and guiding theactions of others - art activity assisted by the ‘al-a-ghutce-visibility’ of the paper record. The paper records in theform of files, printouts and standatd forms constituted animpofiant component in the individual worker’s ‘sense oforganisation’ - enabling them to quickly obtain a graspnot only of ‘what had happened’ but also ‘what’smissing’ and ‘what to do next’. This is illustrated in theexample that follows when one business manager has Mto cover for another’s illness and had suddenly received aphone call asking for an increased loan to pay off theInland Revenue. As he looks through the customer nxordthe Business Manager explains how the various forms andnotes present a particular lending history (in this case oneof overborrowing) and thereby shape his decision-making;

1. Looking at file - Customer is hesvily Lwrrowsd andnot generating the income.

(Discussion of csse) -,. Well, its a bit of a probl amreally bscause I dent know the file, you know, snd its apretty masty file I dent know it. .SO I have to vezyquickly look and try end sort of ac~aint myself withwhats going on end whats baan arrenged in a short spsceof time because -. But basically, he’s hesvilyberrowsd. .(shows figures/folder) forget the money onclisnts accounts because that’s not his money. but he’sgot a private loan acc of 38 a business loan &cc of 20and business overdraft. um..of 29 there’s a lot ofborrowsd money thsre..m a business and clearly he’shsving dlfficul ty in sezvicing it all. now I dent knowwhat they ware all for. . I really. .1 mean I wo:,ld if itwsre my own file I would how it having done it sndresssrchad it..I could find out by reading it. butdoesnt really bsve too much time to do that... .,.

,..he wss heavily borrowad, or he i~ hsavilY~r%d and clasrly he’s not gsnarating the income todeal with that...onun,(the Business Msnager). has had acoupla of discussions with him (looking at file) atleast s couple, probably a few. and the latest, the last

thing agreed was. .he’s under pressure with the InlandRevenue to the extant of fEOCf. mm, thats not besnpaid yet, hs’s under pressure anyway.. so it wss agreadthat wa wvuld incrassa our limit which wss 27500.Ithink.it was 27500 and w would increase it by 4250 -31750 which is Ragion’s sanction to 31750 against..hewas raising a loan frtxnthe W Insursnce where he has aIife PO1 icy which is charged to the Sank as ser~rity endws’d said OK w will release that ital ofsecurity, ,...on the bssis that they wil1 than take thePO1icy, give him a loan on it bscause YOU csn borrowon,..the policy from the insurance cowpsny. g:ve him a

1 Cefiainly in comparisonto the computer record, afthough it wasinteresting tk extent to which tk ‘paper’ in a customer’s tile hascome increasingly to consist of computer printouts - customersnotes, bstaaces, ‘836s’ and so on. This is paIUy a reflection of theincreasing reliance on IT; partly because some computerinformation cannot be reaccessed; and partly because of thevarious ‘affordances’ of paper.

150

loan of 3630.end that 3630..{doing calculation ansheet) would ccma in. ..28000 which wuld reduce theIf.mft beck ....so that was the beads of that (pointing)agreement. The:lethar:hi;lso other nsgotiations goingon. to deal (pointing) huge amount ofborrmwing. of. which gets a bit complex. ..

With the assistance of the paper and computer record ofthe operation of the account and aided by the knowledgethat the account is ‘underreport’ the Business Manager isable both to come to a rapid understanding of ‘what’sgoing on’, make a quick decision and to offer a reasonablejustification, a rationalisation for his actions - that is,that since it is ‘under report’ the account is not to beallowed to ‘drift up’ over its maximum borrowing limit.

As a ‘mechanism of interaction’ [12] the various forms ofpaperwork act to facilitate the coordination of work butthe actual process of ‘doing the paperwork’ itself requiresvarious coordinating activities, involving the ‘meshing’and ‘modal transfotmation’ [3] of other paper andcomputer information. One example of this is the‘Interview Notepad’ which, designed to be used as part ofthe annual review and for lending interviews, involves theBusiness Manager’s Assistant in the lengthy pemsal ofother records, the ‘scorecard’, tecords of past reviews, thecomputer 836 printout and so on; and then using a rangeof information to ‘GAPP’ the account. In this fashionprocedures, implemented through paperwork, mesh withnotions of distributed coordination and awareness of‘work’ - in this case the coordhation of distributedinformation. In this next extract the Assistant is talkingabout this process of accumulating information inpreparation for the customer’s annual review;

“. . . . i es the annual review on Thursday, . .(gettingthings ready for the fnterview) ‘I did it yesterday, butI just had to G4PP it, tbsts reedy apart from up-to-datebalances which I’ve put cm there, ... there ‘a astandard. ..rsow.brief that ws have to fill in, like eninterview note pad, then you get the customer brief offthe ccmputer printout one, and then thats it Lmsically,(pointing at printout)”

As the Assistant works through the different sections ofthe ‘Interview Notepad’ the processes of administrationand control of the account, relationship building and thesales process are linked.

Paperwork - temporal and organisationalimplicative ness.

The Business Manager’s Assistant as she compiles thematerials for a Lending Interview or an Annual Reviewalso needs to consider the temporal dimension of therecords, in particular, to ensure that the Manager goes tothe review with the latest balance of the account. As wellas obtaining the latest figures for the balance of theaccount and other linked accounts and facilities theAssistant will also ensure that the paperwork in thecustomer’s folder is sorted in date order. The Assistantalso needs to be aware that the work that she does and thetwmrds she examines may have organisationalconsequences, for example that the level of the facilitieson the customer’s account and other ‘linked’ accountsmay take it outside the Business Manager’s ‘discretionarypower’ thereby requiring a report and a recommendation toRegional Office before any facilities can be agreed. As she

assembles the information she also considers how theinformation she is compiling may actually be used in theinterview - that is, it involves an awareness of the workof the Business Manager, and the form and utility of theinformation she provides;

“There’s nowhere on this intamiew notepad for puttingdam all the behncea, ao I tend to put the balances ona separate sheet, they cane up on the customer brief(COSrIpUterprintout) but I tend to get the briefs o“t acouple of days before ... so I my to get the balancesout the following day (S that they are up to date) .,.,if any of the partners or directors have got privateaccounts w need to get that information out as nil,becauae again its..if its got any facilities in theirown name it could take it outside. discretionary~s, . this is within,....it depends whet facilitiestheir aaking for but the facilities aa they are atpreserst...CM extend for a further year without havingto go on up to Rwion. .. smething else to look out mthe brief as well, . w *t the latest account, alwaysbring it to the fore, .... always got to put at Iaaat twoyeera on this, if w ‘ve got two years accounts,...“

The customer mcotd is not, however, always particularlyenlightening or informative, reflecting, perhaps, in theircomments, the changing organisational culture of theBank.’

This last section has attempted to document some aspectsof the way in which coordination is practicallyaccomplished in a large and highly distributedorganisation. The focus is on the massive predominanceof paperwork and its role in the routine organisation andcoordination of the work’s activities. The design issuesthis raises are not simply to do with the ‘affordances’ ofpaper - its ‘at handedness’, its manipulability, its‘easiness’ of use, and so on, but are also associated withthe information that the paperwork contains, much ofwhich is not replicated or replicable electronically. Ofcourse software packages also play a crucial part in thedaily coordination of activities. However, in terms ofCSCW, of increasing computer support by, for example,attempting to reduce paperwork by simply substitutingelectronic forms of recotd keeping and coordination, anumber of points remain relevant, in particular the issueof the ‘audit trail’z and ‘data protection’ and the

1

2

An cxemhsation of eertier reeords revests en emphasis on the,etmrsst patemet. administration, cosstrol and management ofaccounts; for example, - “he gave the impression of being a highlyrespectable and hardworking yoang man’; “appears m be highlyrespectable and copabk of managing her own r#ii”rs”; “in 1989 healso said he would take his account elsewhere and thereforeitappears he has a tendency for cosnpfainrs”.- such comments aswould now, under the Data protection Act, be inedmissable.This is a complex issue con~med with the interweaving of Bankplicy, the law (data protection, rules of evidence etc), securityand so on - it is, however, debatable whether this is anirresolvable problem. Of particular relevance for the Besrkare theissues of duptieetion, where much of a workers activity centses onsimply mpyirsg information - name address, account number etc -from one form to snothec an activity that could be reduced by theappropriate use of a rehttionet database; and the issue of ‘modaltmsssfomration’ in which information recorded in one way istransformed into mforsrsatiottrelevant to another task which alsoinvolves transforming the made of stxording. A classic example ofthis woutd be where customer information provided on a paperform is trarwforsrwdby inputting the information onto an electronicdata base which can atso be used for other purposes.Understanding such transformations in relationship to particularwork activities may assist in ddnking about the desigrr of CSCWsystems in re.+ect of the regulation end management of types ofinformation. llris also resonates with the concern to reduce the

151

consequent organisationally ‘necessary’ duplication ofelectronic and paper records.

PLANS AND PROCEDURES

Oneofthe more obvious and prominent ways in whichdistributed coordination is achieved is throughinstitutionalised plans and procedures and their ‘situatedinterpretation’. This refers to a wide range of formalprocedures which, in the Bank, would include, forexample, ‘how to do an Annual Review’; ‘How toSanction A Loan’; ‘How to Write A Report’ - the step-by-step processes for the accomplishment of a procedutewhich would be contained in manuals such as the LendingManual or the Action Sheets. What we are interested in ishow they are used as coordinating mechanisms withinsocially organised ‘real world, real time’ work activities. 1Their explicit point is to co-ordinate the work of numbersof people in order that separate work activities and taskscome to have a coherence and, typically, through thismeet other goals such as efficiency, meeting timeconstraints. and so on.

Plans and procedures are quintessentially aboutcoordination. However, within this generalcharacterisation there are some important issues as well assome misunderstandings which need consideration. Muchof the critique of the planning model in CSCW, namely,that plans ate ineffective, arises from misinterpretationsof Suchman’s seminal work [13], in prwticular the ideathat people do not follow plans because, in actuality, they

~ post-hoc rationalisations of courses of actions,Suchman’s critique is directed, however, at the notion thatthere are mental plans which operate as causaldeterminants of subsequent courses of action. The essenceof Suchrrran’s critique is that the plan is an abstractconstruction which will, at the very least, requhearticulation with, and application to, the specifics of thecircumstances in which it is to be followed. In thisextract from the fieldnotes, for example, a BusinessManager’s Assistant is assembling the necessaryinformation, from the customer file and the computerrecord, prior to an annual review by the Manager, butfinding difficulties and having to make a decision aboutwhether the case fits the parameters of the softwarepackage she is using;

. . “1 rlid that this morning, it was one we had slightproblems with because w only G4PP accounts withfacilities in excess of 20, bot I came to put it on themachine, and they’ve actually got a business card 1imitof 5000 and negotiations. ...those tkv facilities cantactually key into the 64PP machine, so I have to checkwith Region to meke sure I still have to GAPP it.Because the only facility I can put into the GXPP is15000> the overdraft facility, but (arguably?) it isover 20 so it still needs a G4PP grade. .“

overheadof coordination. If the cost of coordination is ‘too high’ itwill either become an end in itself at the expense of the work, orbecome such an ‘inconvenience’ that it has low priority

1 It is important to point out here that we are making no assumptionsthatatlthe items in the ‘variegated collectionof phenomena’areused in the same way - it may rdso be important to note that some ofthese items, for example, certain of the action sheets, are used toovercome some of the problems emanating from the fact that theBank and tk computer system do not operate in ‘real time’.

Following ‘the plan’ will always involve more than canbe specified within it. The construction and use of plansin ‘real world, real time’ activities typically involve thesupposition that ‘practically’ and ‘characteristically’ plansare ‘recipient designed’; that is, spelled out to an extent towhich those who are to follow them are, for example,familiar with the circumstances in which they are tofollow them, sufficiently trained in the tasks involved,and a host of other possible considerations. Nor does themaking of plans indicate any expectation that the comeof actions which they specify will, of necessity, followthrough. Indeed, the point of plans is often to directcourses of action to maximise the chances that thesecourses of action will ensue despite the contingencieswhich can arise. Here, for example, is a Business AccountManager talking about the Bank taking a debenture as partof the process of providing an increased overdraft facilityto a company, thereby illustrating the way in which plansare practically accomplished in the course of the work andhow, in this instance, paperwork acts as a mechanism ofdistributed coordination and for developing the awrwenessof work;

,,. . this. is a limited company account snd it !mrks veryWS1l, . (looking at file/printouts, .. . a limit of50K,. I did look at the 836 and the 838 prin:out againto see this utilisation of an account, sss what itsdoing (looking at printout)... They>re makinq profitsand profi ts are bsing retained in the business, (talkabout rquest for bigger credit zone) he (the customer)said ‘qI’d like a 100 (K) as standby and i f somethingcomes up ..... well, ..t~o: was ovet~t:~’ (disTretionarypower) ..we’re ‘ a140K, . . . overall, ..l,n

expcsure ofso we’re than left with do we lend

him 100 grand?. .decision making process (is:, what ,sthe companies trading perf o-cc been liie ? guitegood, ..whets its proven track record from auditsdfigures ? thats quite good, suxrdus resources in thecWPanY, retained Profits in the company. .. thats ~i tegood. What about the product that they’re deaiing with?do we consider thats the sort of thing that is that canbs moved on and sold. ...(discusion of business J..... s0,the product, the siting, the COJWSZIY, proven record,menagemen t; what do we think of the menagamen t? prettygood, pretty awitched on, everything shut it looksOK, so we went to go ahead and do a recommends tion,we’ve got to go up to Rsgion because its over thoselimits. Look at security, yesb, the LW.IIJCshould have adebenture ...... we should have a debenture, :f it allfails we want to put a receiver in and rake thatstock. ... we ’11 put that (dsberrture)as a coalition ofsanction... and I think thats right. . so Ma: was thenub of all that lot (points to notes) is to say yes welike the set up, ws like the products, we like themarketing, we like the management, m? like the accout,we want to help and we want to do it, we cw see thereasons why he wants to do it, and w haveconfidence. we think we should do it (make the loan)but ~ policy says we should have a debenture. “

The extract also demonstrates how plans often include‘fail safe’ devices to cope with situations where things rm‘not going to plan’ by specifying arrangements foradaptation of the plan to exceptions, unforeseencircumstances, even extensive revision, as well asmechanisms to oversee the implementation of the planand enforce its requirements - in this particular case that ifthe loan ‘goes bad’ because they hold a debenture they can“put a receiver in und take the stock”. Similarly in thisnext extract, a discussion of a ‘troublesome’ account,(where the customer has suddenly asked for a largeincrease in his borrowing to meet an urgent demand formoney from the Inland Revenue) the range of possibleresponses to different situations is presented and discussedby the Business Account Manager, illustrating,

152

furthermore, the way in which workers need to be able torespond, to adapt ‘the plan’ in the face of changing andurgent demands;

“if the fellow wented SK. snd he said what is yourdecision? I would have said I’ll let you know in a fewdays time, 1’11 have to consider it..1 would’nt say tothe customer, ‘I’ve got to go to the i%nk, because I,mthe Bank. . if he wants a decision urgently, thanI. try and do it by phone, h’s d~t 1ike doing thingslike that because the ~ likes things in writing. .. ifthat was the case and it does happen from time to time,you t~e your facts dom, say you ’11 let him know,consider them, try end get your facts in a logical orderand your reconmandation bscauae the ~ will alwayswant a reconmands tion frcm the manager, always, you cantCUP out of it, you’ve got to cane down one side of thefence or the other, avan if YOU dent mske the decisionyou’ve got to come up with a recommendation so. if hesaid,, I want to issue this chaque for f6000 now. .Ihvuld have taken the details down, I would have made mydecision that I was not going to rscomend that, Iwuld’nt want to give him f6000 unless I had a clearaxit route.. I would have gone to Region. .myrecommendation is no w dent do that. “

Another feature of plans and procedures is that theydevelop in real time. What the plan agreed, whatinterdependencies there are, only become clear as thecourses of action specified in the plan unfold, creatingadditional workloads in terms of coordination and theawareness of work. A simple, regular, example of theresponsiveness of plans and procedures comes in theexamination and actioning of the daily printout of ‘out-of-order accounts’. l%e ‘out-of-order’ printout will give boththe ledger and ‘true’ balance of the account (this takes intoaccount the possibility that cheques paid in to the accountwill be bounced); a figure for ‘uncleared’ (both in thethree day cycle and monies paid in that day); a balance; acredit zone and a ‘notional limit’ (usually an additional10% on the credit zone) - all these factors need to keborne in mind as the ‘out-of-order’ printout is ‘actioned’by the Business Manager’s as they respond, for example,to phone calls from customers saying that they are‘paying in today’ and make decisions about whether tosend a ‘letter of concern’ or ‘bounce’ cheques or standingorders.

Plans, Procedures and ‘Local Knowledge’

The successful accomplishment of a ‘plan’ is dependenton the practical understandings about what the planspecifies in these circumstances, using these resources,these people, and so on. Although plans may be presentedas abstractions, as manuals, as statements of procedures,and so forth, - in the Bank, the ‘Lending Manual’ and the‘Action Sheets’ - the ‘just what’ it takes to realise themis a practical matter of ‘making the plan work’ throughall the various and inevitable contingencies that can arise.It is such activities which maintain the plan by dealingwith ‘those things which arise’, ‘the things not plannedfor’, so that even ‘deviations’ from the plan can beaccommcxiated to sustain its ‘spirit’. ‘Real time, mlworld’ work often involves the utilisation of ‘localknowledge’ and ‘local logics’, commonly interpreted as‘cutting comers’ or ‘bending the rules’, to support theoverall objectives of the niles or the procedures. Such‘local knowledge’, particularly of the expectations ofRegional Office and the circumstances under which aroutine should be strictly followed and the circumstances

under which modifications or ‘short-cuts’ could beemployed was a regularly observed feature of work in theBusiness Centre. In the following example a BusinessManager outlines the actions taken on an account that is‘under report’;

“this account is undar report to regional office. .,.,. Ias a branch really have no discretion whatever. I cantlet that drift up over whatever that - is for~ything, I ‘ve nO discretion at all. having said that,that max was 31750. . and I’ve, in listsning to him andtalking to him, and I’ve put a little note on screan. .I’ve said he could go to 32, so I have exceeded what Ishould do, but only by f250, but what I did do, to savepaper and gsnera ting a. paperchase, I just rang a fel1owat Ragion and said, ‘look, you know this is what thescore is, I‘ve agreed he can go from that to that(pointing at file) for the ramainder of that term. andthets my msx, I dent propose to ccme up with a writtenthing will you tie a note of it in your mm files UPthere and will you give ma verbal sanction that I canincrease that limit on the computer to 32000 for theremainder of that term, so he said yas, newordinariIy under a under rsport account you would‘nt dothat but for f250 its not worth the managerial timedoing it...“

By placing plans, procedures and decisionmaking withinits social and organisational context, they come to beseen as elements which enable workers to make sense oftheir own work and that of others and to come to adecision about future courses of action. What thisemphasises is the importance of seeing how and in whatways plans and procednm am interwoven into a highlyvariegated set of phenomena that make up the socialorganisation of work. How and in what ways, forexample, plans and procedures are related to thesequentiality of work; how and in what ways theyforrmdate the work’s interdependencies; how and in whatways they identifi ‘troubles’ of various kinds; how and inwhat ways they make work accountable, ‘open tocriticism’, ‘subject of failure’, and so on.

Having presented a critique of the planning model forfailing to acknowledge the myriad ways in which people‘work to make the plan work; it is perhaps hardlysurprising that we are equally sceptical about similarlystilted, rationalistic mcdels of decisionmaking. AsFeldman and March [5] note, models of strictly rationaldecisionmaking create expectations which are rarely metin practice;

‘,.For example, relevant information will be gathered andanalysed prior to decisionmaking; information gatheredfor use in a decision will be used in making that decision;available information will be examined before moreinformation is requested or gathered; needs forinformation will be determined prior to requestinginformation; information that is irrelevant to a decisionwill not be gathered. Stu&s of the use of information inorganisations, however, reveal a somewhat clifferentpicture’ ([5]: 172)

In fact studies of organisations, and the Bank is noexception to this, consistently suggest that informationoften appears to be gathered for its own sake withouthaving any obvious decisional relevance and;

‘Much of the information that is used to justify a decisionis collected and interpreted after the decision has beenmade, or substantially made...Much of the information

153

that is gathered in response to requests for informationis not considered in the making of decisions for which itwas requested” ([5]: 174)

To some extent observation suggests that this attenuationbetween the gathering of information and its utilisation indecisionmaking is minimally affected by theincorporation of plans and procedures into a computersystem.

Plans in the machine - using computers t osupport decision making.

One of the ways in which plans and procedures appear inwork activities as aids to decision making, is to formulatethe work’s activities as step-by-step stages. Computersoften appear in this process and in this plan at variousstages, as repositories of information to be collected,browsed and retrieved; as a focus for activity and fortriggering other activities; and in providing a backgroundfor and indicating decision alternatives. In the BusinessCentre, for example, a number of software packages,notably ‘GAPP’ (Grading and Pricing Program’) and the‘Risk Grading’ on ‘Fileserver’ had been introduced andwere intended both to support decision making and toimprove the speed of processing thereby giving staff moretime to be ‘pro-active’ - to develop customer relationshipsand sell Bank products. The following extract shows aBusiness Manager’s Assistant carrying out a ‘GAPPing’exercise prior to the Manager’s visit to the company;

Nvct

1. Gsts screen - ‘Customer New Record’ - fills indetails from GA.PP data input fOm (obtained frcnncompsny’s accounts)

Screen ‘Customise’ - (name) - fills in details - dateacc obtained etc

3. Screen - “Business Definition’ - “What doespharmacist go under?” - discussion with other EM4SS -“tzy that one” - clicks on various tities - “whatsother?” - other smal1 screens appear. - eventually findsit.

4. Screen ‘Audited Managamant Accounts’ - “do you put aminus in here if its in brackets?” “Yes - it will printUp then“ - filling in details from form.

5. Screen - ‘Mamagement Detai1s’ - (series of questions- yes/nO clicks) - managcanent assessment; financialmonitoring; trading arrvironment; short term probl W;

6, Screen - ‘FacilitySummary’ - ‘New Customer facility’as each section of the screan is entered

‘help/explanation’ messages appear at the bottom of thescreen

7. Prints out - ‘Risk Aoalysis .5ummary - gives riskgrade and ratings on faci1ities (what should be charged)

The fact that GAPPingl, although incorporated into thelending process appeared as a mere additional check in thatprocess rather than integral to it, meant that it seemedless important as a decision-making &vice than as a‘security blanket’ for decisions already made; and thestarting point for negotiation (particularly over pricing)with the business concerned. Effectively, therefore, the‘decision-making’ software did not necessarily ‘make’decisions but, as well as being an organisational

1 It is impottsrrtto reeognisethatGAPP was simply erraddition to theexisting risk assessmentand pricing ‘devices’ - in some sensesmerely automating what had previously been done (and continuedto be done) manuatly.

requirement also became yet another element that had tobe incorporated into the plan as it unfolded. As anAssistant Manager said;

“you cannot say straightaway. .,just because the conputerprogram says 1% higher. ..YOu csnt just inpose a 1%rise. ,you’ve got to use it as a tool, ..YOU,VS got tosum up how much the overdraft is and wbstever. ,-

This position - of using the software to confirm ratherthan determine - may have arisen as a consequence of theinclusion in the program of ‘non-financial’ informationwhich could significantly influence the risk grade obtainedand which was dependent on the Manager’s store of Ioealand often anecdotal knowledge.z It may also represent areflection of managerial experience and scepticlsm aboutthe information provided; an awareness of the variety oftechniques that could be employed to disguise the ‘true’nature of an account. It may also, as Feldman and March[5] suggest, be a reflection of the fact that much of theinformation used in the Bank had been gathered primarilyfor ‘control’ rather than decision making - that is, it isgathered in a ‘surveillance’ rather than a ‘decision making’modc3. Above all, the main point to make about thedecision making process and the usage of information(whether on paper or computer) is concerned withappreciating the careful consideration of the information.That is, it is not a question, as Harper [7] points out inhis ethnography of accounting, of ‘just any old numbers’but that interpretation of the information, anddecisionmaking horn the information is dependent oncertain, ‘nuanced’ expectations.4

What this section of the framework on ‘plans adprocedures’ has attempted to suggest is relatively simplebut important nevertheless. It argues that ‘following theplan’ generally involves more than is specified in theplan; that plans me modified as they unfold and atesubject to evaluation; and that plans are both chosen andmade to work through the skilful use of various forms of‘local knowledge’ of what the plan entails in thisparticular circumstance with this particular case, (whichfeatures the plan itself may then be made toaccommodate). It further suggests that computers are

2

3

4

Such ‘anecdotal’ evidence was important in at least one instancewhere no indication of trouble was revealed by any of thecomputerpackagesor printoutsand only becameevidcrrt when thefirm appeared on the ‘receivership and liquidation’ perusal formand the Customs and Excise asked to he paid with a BuildingSociety cheque.Feldman and March [5] suggest that, “When strategicmisrepresentation is common, the vafue of information m a decisionmaker is compromised....individuats develop rules for dealing withinformation under conditions of conflict. Decision makers discountmuch of the information that is generated.” ([5]: 177) Within theBank in general, and the BusinessCentre in particular. there wasan awareness of how accounts could be managed to m]srep~sent abusiness’s trading position; and similar skepticism about ‘businessforecasts’ (especially when produced by the business itself) and soon.The other, important, point to make, and one which resonates withearlier comments about decisionmatrhrg and ‘gut-feeling’; is thatdespite the way in which the decisionrnekhg process illustratedhere appears to conform to some ‘rational’ model of consideringall the ‘relevant’ factors this is not actrratly whet is happming since‘relevance’ rdways comes from a pre-orientation within abackground and it is this pre-orientation which thereby accountsfor ‘gut decisions’.

154

enmeshed in the ‘working out’ of plans and procedures ina variety of (not always obvious or even useful) ways. Atits simplest what we are trying to say is that plans adprocedures cannot be understood (or effectively designed orimplemented) divorced from the organisational context inwhich they are situated and in which they are enacted.

AWARENESS OF WORK

Another perspective on the social organisation of work,and one which is closely involved with the other themesof ‘Distributed Coordination’ and ‘Plans and Procedures’is Awareness of Work. This refers to the way in whichwork tasks are made available to others and the importantrole that this plays in the ‘real world real time’ socialorganisation of work. Clearly this is a theme whichinvolves various interfactional subtleties and constitutes amajor aspect of the means through which coordination ofwork tasks is achieved as a practical matter. The variousways in which ‘awareness’ is developed, in which work ismade public and availaide to others, are essentialingredients in ‘doing the work’ as part of a sociallydistributed division of labour.

One way in which such visibility/intelligibility isachieved is through the notion of ecologies of awareness.This refers to the fact that worksites are laid out in wayswhich affect - facilitate or hinder - the capacity of thosedoing the work to make reciprocal sense of each other’sactivities, to see what the other is doing now and howthis fits into the pattern of work which they ateconcertedly producing. So, for example, the organisationof the workspaces of the Business Account Manager adthe Assistant permit the development of an ‘ecology ofawareness’ which is consequential for the performance ofwork. Many of these affmdances arise simply because ofthe co-location of a collection of people doing ‘much thesame things’ or who are performing interdependentactivities. Being abIe to ask for advice just when it is_ updating colleagues as one is passing, dealingwith an emergency when ‘all hands are needed’, and more,are all informal, often intermittent, certainly unplannedoccasional events which are facilitated by appropriatespatial layouts of work areas. Similarly, the state ofdesks, where someone is within the spatial setting,looking in a particular fiiing cabinet, working through apile of paperwork, and more, furnishes information aboutwhat the person is doing, where that person is in thestream of work, how busy they are, and so on;information which, in subtle ways, is available so thatothers can update themselves on the state of the work,how it is going, whether ‘we are behind’, ‘on top of it’,etc. The point we want to make about such at%xdances isthat they are not necessarily designed or planned, thoughthey can be, but represent arrangements which can be usedin the day-to-day ‘doing’ of work. 1

Other ways in which the awareness of others’ work ismade available and visible, often ‘at a glance’ is throughthe various ‘mechanisms of interaction. In the BusinessCentre the various forms, the ‘out of order’ printout; theAppraisal Forms, the Interview Notes, the CustomerFolder, and different sections of the computer databasesuch as ‘Customer Notes’ are both the focus for work anda visible, a publicly available, nmrd of work that hasbeen done or remains to be done. In other words, whatthese representations do, among other things, is make thework ‘visible’ so that it can be ‘taken note of’,‘reviewed’, ‘queried’, and so on, by others involved. Theyput the work on display so that others may be aware of it.

The overriding practical issue for representations is whatto represent about the work, for whom, and how torepresent it. What is it that people who ate to use therepresentation need to know about the work and in whatformat is this to be conveyed? Clearly, there areprinciples of selection involved. In completing, forexample, the Lending Appraisal Form, the Assistantknows that a number of issues need to be addmsed;issues which ate indicated in the various sections of theform - ‘adequacy of interest margin/fees in relation torisldmanagement time”; “viability of repayment,including source and timing” and so on.what seems to beat work here is a practicrdly oriented principle of economywith respect to the information that is represented throughthe artefact. Not all the possible information that mightbe gathered or represented is so. The question of what isrepresented is not solved ‘once and for all’; whatinformation the Assistant, perusing the various computerfiles and the Customer Folder, ‘needs’ to include for theBusiness Account Manager is a constant preoccupation inand through the course of the work. Finally,understanding what the representation represents, so tospeak, relies upon knowledge of the work itself and howthe representation is embeded within the work as aninstrument of the work.

Representations as modal transformations

One of the featutes we have noted before in connectionwith the coordination of activities is the way in whichinformation goes through modal transformations in whichit is re-represented in a different form, A good illustrationof this is the way in which information is extracted fromthe computer records, or flom using the GAPP program,taken from various computer printouts and from thecustomer RCOKIand inserted into the AppraisalAntcrviewForm. What the various transformations do in this case isfacilitate coordination through appropriate representationsof the ‘basic information’. Although this is a processwhich is deeply involved in coordhtation, it does so bymaking the work available to others in a form on whichthey work; that is, representing the work within the work

1 h some respects the sociat use of space is a neglected area ofsociologicalinquiry though of course,it is tiw stuff of architecturaldesign, town planning, ergonomics to mention but a few. Thedim.ction we have in mind here is with the ways in which spatial

srrsmgernenta can be used to facilitate, or indeed hinder,collaboration and the coordkation of work tasks. The importancefor CSCW is the need to face up to the fact that computer mediatedcommunication can abolish spatiaf affonhnces though it may needto emulate some of these.

155

so that others can do theirs. In these cases, they serve toroutinise the work by using standard procedures andstandard formats for representing the tasks done and thetasks to do. So, one of the properties of formatting is theway in which it can proceduralise representation and,through this, represent the work to others for particularpurposes. The format - that is, how to complete theAppraisal Form, how to complete a GAPP1ng exercis~how to complete a Balance Sheet Assessment Form andso on, functions as a set of instructions in both itscreation and its use. TMs is, at least one, reason for theimportance of standardisation and was an issueencountered on a regular basis when using the CustomerNotes on 1SS (the relational database) since there appexmdto be no standad practise, reflected in the earliercomments of a Business Account Manager whenencountering an 1SS Customer Record for the first time;

1.Looking at file.

3. Needs to updete 1SS - looks at psst interview notes -on chenge in overdraft - in ‘notes’ - nothing; -othernotes ‘Iimit reduced f8K’ (no esplsnation) “highlyrespectsbl e partners” (only comment on entire page) -

“see how the approach varies (pointing to notes) - itdetermines the usefulness of the system”

The various forms in use in the Business Centre are alldesigned to collect standard information, to make theinformation comparable and to control the informationthat is provided. In this respect they are solutions to theproblem of the assembly of information in organisationsidentified by Garfhkel[6]; that is, what information isneeded and its ‘value’; the worth of collecting theinformation with refetence to the effort involved in itscollection - an issue highlighted in the earlier section ondecision making. The attempt to get people to complywith formats is often, with regard to those who mustapply them, a disciplinary matter; that is, an attempt toensure, by laying out a series of procedures to befollowed, for example in the Action Sheets or theLanding manual, that the persons who need to complyactually do so. Nevertheless despite the obvious benefitsof standardised processes and formats - the format does notalways, in itself, convey an adequate ‘sense of the work’and it is in these circumstances that local knowledge and arange of interfactional skills are deployed to help ‘makesense of the work. It is in this sense that ‘awareness ofwork’ becomes a ‘skill’. As we have rdready suggested‘skill’ is a concept whose tacit and often highly situatednotions need considerable ‘unpacking’. The sense of ‘skill’that we are interested here is that of ‘knowing how’, adthe delicate competencies involved in making availableto others, what is ‘going on’; competencies required for‘mutual intelligibility’ or ‘mutual awareness’ on the partof the members of a work team.

In a wider and more generaf sense workers within thevarious offices and centres of the Bank are aware of theirwork as part of some huger institution and that their workactivities are the work of and represent ‘the Bank’. Thevery characterisation of them as ‘bank workers’, their jobtitles of ‘Business Account Manager’ and so on, caflsthem into play as people who are organisationallypositioned and people who have organisationally

prescribed responsibilities and tasks. As one BusinessAccount Manager put it... “I am the Bank”. The sense inwhich Business Managers, to develop this example alittle more, are aware of the fact that they are workingwithin a system that transcends their here-and-now, isalso reflected in the physical organisation of the BusinessCentre; in their physical and hierarchical/organisationalrelationships to each other; in the paperwork andcomputerwork, in the ‘dress code and so on. They =aware of the fact that the Business Centre is but alocation within a system of banking which extendsbeyond the Region (even beyond ‘the Bank’ as a nationalinstitution and relayed in the form of ‘folk tales’ of‘lending in the 80s’); aware of the fact that they workwithin a bureaucratic structure of work organisation,subject to a managerial supervision from Regional andNational offices and situated within a complex division oflabour. They are aware of the fact that their working isdependent upon various types of computer andadministrative support and the activities of workers atother locations - such as particular branches, theSecurities Centre, Regional Office and so on - within thewider Bank network, as integral, interlocking componentsin an extended division of labour.

The framework allows designers to reflect on potentiallessons from a broad ethnographic study (only some ofwhich is reported here). In particular,

.

.

.

A consideration of distributed coordination within thebank highlighted a problem arising from a decision toseparate and centralise some of the functionshistorically found in local branches. This separationperturbed the rich patterns of distributed coordinationleading to a “buck passing” or ‘blame culture’mentality. In seeking a solution to this problem thedesigners have developed better communicationfacilities and instigated a management team drawn fmmthe different functional areas of the bank.

The role ofpkms and procedures in the bank highlightedthe way in which local knowledge was applied topromote the smooth running of the bank. Thisemphasised the lack of flexibility in some of thesupporting systems and the need to allow userssignificant interpretation. Many of the decisions in thedialogue between users and standard forms are beingreconsidered to allow this flexibility.

The awareness @ work also reflected the effects ofreorganisation in the bank. Work that was once local isnow remote and this was reflected in a general feeling ofisolation and lack of appreciation across the bank. medesign team have again sought to improvecommunication using IT and a programme of jobexchanges between branches and central units toheighten the awareness of each other’s work andencouraging a move away from a culture of blame .

CONCLUSION: THE FRAMEWORK,ETHNOGRAPHY AND DESIGN.

Itis for designers to draw design conclusions from theresults of ethnography. A case can be made for

156

ethnography (or sociology) having a more far reachingimpact upon design, insofar as it provides a means torethink the nature of the social world that is beingdesigned for, or to rethink thedesigner’s role within thatworld [1] [2], It can be, it has been, argued thatethnography (orsociology more generally) ought to carryimplications for design - as in one of our colleague’s(indelicate) suggestion that “it was time for Sociology topiss or get off the pot”. However, our concern has notbeen with anything so immensely far reaching asrethinking the world of design, but has been with thecontribution that ethnographic studies can make to theformulation of specific designs. The kinds of changes todesign which will result from our approach are intended tohave an incremental rather than a comprehensivelytransformative effect.

In a similar fashion we believe that the production adintegration of ethnographic materials into design can bestbe achieved through the collaboration of specialists,recognizing that the nature of the exercises of design andfieldwork are very different. Design is a decision makingexercise, concerned with solving practical problems,whilst the fieldworker’s role does not involve decisionmaking but consists, rather, in the concern adequately arKIaccurately to capture and portray the ways of socialsettings, and the activities which occur within them. It isprecisely the importance of such work that it should bedone independently of design preconceptions, that itshould distance itself from the preoccupations,enthusiasms and orientations of the designer, shouldprecisely refrain from looking at the setting and its affairs‘through designer’s eyes’. Its mandate is, rather, to lookat the setting through the eyes of those who inhabit it,with as much sensitivity to that as possible. Part of thepoint of undertaking ethnography is to ascertain whetherthe realities of a given setting conform to the conceptionsthat the designer has (perhaps a priori, derived from designconceptions, or perhaps derived from other informationalsources) and, to that extent, it is wholly desirable that theinvestigation of those realities should not seek toconstrue it in terms of the designer’s relevance. There is,thus, a basic tension between the designer’s and thefieldworker’s roles, but this is a positive feature,something which is important to and essential for gooddesign, to highlight the dtierence between good abstmctdesign solutions and good practicai design solutions.

We hasten to add that we are not trying to make out theethnographer as the omniscient and virtuous party in thisrelationship, restraining the inflexibly minded designerfrom imposing misconceptions on those who will beimpacted by the design. In many respects, the fieldworkerwill be an agent of the designer. We are merelyhighlighting the fact that the fieldworker’s role is togenerate a kind of data which the designer would haveneither the time nor the proclivity to collect, and which isbest carried out if it is done as an independent operationwithin the design process. Consequently, ourrecommendation is that the building of a relationship

based on effective communication between designer andethnographer is at the heart of the matter.

ACKNOWLEDGEMENTS

The research for this paper was undertaken as part of theSYCOMT project, part of a DTI-EPSRC funded initiativein CSCW and, earlier, as part of the COMIC Esprit BRAin CSCW. We would like to thank Wes Sharrock, SteveViller, Liam Bannon, Dave Calvey and Dave Randall forcomments and contributions.

REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

Anderson, R. J., (1994), ‘Representations andRequirements: The Value of Ethnography in SystemDesign.’ Human-Computer Interaction, Vol. 9 pp.151-182.

Anderson, R. (1995) ‘Mad dogs and ethnographers’,Encyclopedia of Micro-computing, forthcoming.Anderson, R. J., Hughes, J. A., and Sharrock, W.W. (1989) Working for profit,; The SocialOrganisation of Calculation in an EntrepreneurialFirm. Aldershot, Avebury.COMIC Deliverable 2.4 Lancaster University,Lancaster.

Feldman & March (1981) ‘Information inorganisations as signal and symbol.’ AdministrativeScience Quarterly, 26:171-186.Garfhke],H,(1967), Studies in Ethnomethodology,Cambridge, Polity Press.Harper, R. (1989) An Ethnography of Accountancy.Unpublished Ph.D. Thesis. University ofManchester.Hughes, J. A., King, V., Rodden, T., and Andersen,H. (1994) “Moving out from the control room:Ethnography in system design”. In Proceedings ofCSCW ’94, Chapel Hill, North Carolina.Hughes, J. A., O’Brien, J., Rodden, T.,Rouncefield, M., and Somerville, 1. (1995)‘Presenting Ethnography in the RequirementsProcess’ in Proceedings of RequirementsEngineering ’95.Plowman, R, Rogers, Y and Ramage, M. (1995)“What are Workplace Studies For?” in Proceedingsof the Fourth European Conference on Computer-Supported Cooperative Work, 10-14 September,Stockholm, Sweden. Dordrecht. Kluwer.Rouncefield, M., Hughes, J.A., O’Brien, J., andRodden, T. (1995) ‘Ethnography, Communicationand Support for Design’ COMIC Working Paper,Lancaster-2- 14, Lancaster University.Schmidt, K. “Initial Notes on Mechanisms ofInteraction.” COMIC Working Paper, Rise-3- 1.Riso National Laboratory.

Suchman, L., (1987) Plans and Situated Action:The Problem of Human-Machine Communication.Cambridge University Press. Cambridge.

157

14. Yates, J. (1989) Control Through Communication:The Rise of System in American Management.Baltimore, John Hopkins University Press.

158