Implications for Design

Paul Dourish
Department of Informatics
Donald Bren School of Information and Computer Sciences
University of California, Irvine
Irvine, CA 92697-3440 USA
jpd@ics.uci.edu
ABSTRACT publishing accompanied the transformation of the conduct
Although ethnography has become a common approach in of science itself, reflecting its increasing
HCI research and design, considerable confusion still professionalization; the process of ensuring conformance to
attends both ethnographic practice and the metrics by which documentary standards is part of the “boundary work” by
it should be evaluated in HCI. Often, ethnography is seen as which disciplinary boundaries are maintained, and even the
an approach to field investigation that can generate boundary between “science” and “non-science” is sustained
requirements for systems development; by that token, the [18]. Case studies illustrate the ways these conventions
major evaluative criterion for an ethnographic studies is the shape the development of scientific publications, scientific
implications it can provide for design. Exploring the nature arguments, and scientific publications [13].
of ethnographic inquiry, this paper suggests that
“implications for design” may not be the best metric for Unsurprisingly, then, as HCI has matured and developed a
evaluation and may, indeed, fail to capture the value of sense of its own disciplinary identity, conventions have
ethnographic investigations. arisen in the ways in which we conduct and describe our
research. The peer review process employed by high-
Author Keywords quality (and high-status) publication venues such as the
Ethnography, design. CHI conference or leading journals is one important
element in the machinery of genre production.
INTRODUCTION In this paper, I want to consider a particular one of these
As intellectual disciplines develop, genre conventions genre considerations, both as a matter of research
emerge that shape both their research designs and their presentation and as a matter of research construction. The
research outputs. In interdisciplinary areas such as HCI, particular topic towards which my attention is directed is
early work in the field tends to be highly divergent in interesting not least because it reflects one of the
method and approach, as practitioners – as individuals, and interdisciplinary encounters that so characterize the work of
collectively as a field – attempt to find ways to combine the HCI community. Given our commitment to
perspectives, conceptual frameworks, and methods. So, for interdisciplinary working, it is valuable to step back and
example, finding an appropriate balance between theory consider what happens when two disciplinary, conceptual,
and practice, determining broadly agreed-upon metrics for and methodological approaches come together, and how it
success, and developing common vocabularies for the is that the relationship between them is to be articulated.
problems and phenomena of study are all means by which,
over time, common consensus about research is developed. The topic for examination is one that is strikingly familiar
Scientific disciplines are normative enterprises, where the to practitioners of qualitative and especially ethnographic
process of peer review tends to encourage conformity to a field methods, although experience suggests that it is also
core set of values and approaches [7]. relevant to other research approaches. Loosely, I refer to it
as the problem of “implications for design.”
This process can be seen at work in the research papers
produced in a field. Bazerman [3] has detailed the ways in It has often been noted, not without some irony, that the
which transformations in the structure and tone of scientific canonical paper reporting ethnographic field results in an
HCI context will close with a section entitled “Implications
for Design.”
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are This section may be long or short, comprising discursive
not made or distributed for profit or commercial advantage and that copies prose or brief, bulleted items, but it nonetheless figures as a
bear this notice and the full citation on the first page. To copy otherwise,
or republish, to post on servers or to redistribute to lists, requires prior stable feature of ethnographic reports. Informal evidence
specific permission and/or a fee. seems to suggest that the absence of this section tends to be
CHI 2006, April 22–28, 2006, Montréal, Québec, Canada. correlated with negative reviews and rankings of the paper.
Copyright 2006 ACM 1-59593-178-3/06/0004...$5.00. A common lament to be found in reviews of ethnographic

it misses where ethnographic inquiry tramps. theory. arising especially in Scandinavia. I want to focus on the problem of the genre work of the Chicago School sociologists (Robert Park. prostitutes. albeit with different flavors and emphases. emergence of Computer-Supported Cooperative Work as an some of these could be classified as “the politics of area of inquiry. this took the form of the approach. a restricted Within HCI. They seemed to offer a means Americans. through PD.” or the somewhat actions. ethnography migrated into other social fieldwork. through daily adopted in an encounter between social science and participation in everyday life. This history. technology “in the wild. the considerations are methodological approaches by which that social political in a number of ways. century. bringing ethnographic investigation. perspectives. and origins in the Western expansionism of the nineteenth indeed PD has always emphasized a practical. and about the ways in which ethnography understand “the member’s point of view. spearheaded not least by Bronislaw Malinowski in This is a very partial view. the In what follows. misconstruing the ethnographic distance between subject and object through nature of the ethnographic entireprise.” Where has been appropriated within HCI. The second was the overtly political issues at the end.) sciences. though.” Certainly. For PD. or misconstrue the goals and mechanisms of towards aspects of American urban life. ethnographic field research became the sine qua non of I should state at the outset that. and hence on politics of design. in referring to this as a anthropological inquiry. elsewhere. the South Pacific. In North America. it’s all very interesting. “yes. Nonetheless. ethnography argued that. PD was A HISTORICAL INTERLUDE strongly concerned with issues of workplace democracy First. anthropological methods to bear on issues in locales quite different than those in which Malinowski had developed his I will argue two primary points. characteristics. anthropology had previously documented what members of sets a context for understanding how ethnography has been other cultures did. which in turn placed an increasing emphasis representation” while others could be classified as “the on the social organization of activity. that. and allied “salvage anthropology” of Franz Boas and colleagues. Africa. and more) was particularly influential in adopting primary evaluative criterion for ethnographic research. ethnography more subtle (and intriguing). associated particularly with two trends. and elsewhere. implications for design might be derived. as public toilets and mental institutions. there will be more to be said about both the history ethnographic investigation marked a major transition in the of ethnography within anthropology and its disciplinary practice of anthropology.) Arguably. In contrast to surveys and interviews. one could come to understand . In ethnographic methods in sociological research. it retained its sense of implications for design is misplaced. the politics and consequences of Everett Hughes. it turned its ethnographic attention misstate. multi-method century. emphasizing as it did the need to considerations. I do not mean to focus on the methodological concern (that is. First.” observation with participation.” Grudin provides an overview of Ethnography itself arose in the early part of the twentieth some of these trends [19]. The emergence of come. though. that is. Politically. and in the course of what is to his work on the Trobriand Islands [28]. First. it was associated particularly with by which the complexity of real-world settings could be colonial encounters between Europeans and the peoples of apprehended. but I don’t what members of those cultures experienced through their understand its implications for design. power relations between disciplines. Broadly. “this paper does not seem to advocated long-term. how. that the focus on methods. methodologically. it is worth pausing very briefly to review and participatory involvement in the changes in working some of the history of ethnographic practice and its conditions implied by computerization. ethnography may have been an expedient discipline of anthropology. inquiry into subcultures and “outsider” groups (such as ironically. The first is the and the problems of representation and interaction. it adoption in HCI. The Rather. ethnography arose within the valued. in so doing. In the following decades. and a toolkit of techniques for studying South America. sought approaches in which member’s perspectives were As a professional practice. Anthropology itself has its tool rather than an intellectually motivated approach. I want to explore these questions by uptake of ethnographic methods to study settings and dealing in turn with four issues that arise around the people “closer to home” is a critical step towards the use of problem of implications for design: the marginalization of these methods to study technology users. on the basis of ethnographic At the same time.work is. and second. but with global influence. the use of ethnographic methods become more documenting the rapidly shrinking cultures of native familiar to HCI researchers. Anselm Strauss. immersive field work combining be addressed towards the CHI audience. CSCW. Herbert Blumer. though. Participatory Design (PD) movement. Much of the framing it as a problem. problem. Second. and gamblers) and odd locations (such can provide major insight and benefit for HCI research. I want to explore the ways in emblematic work of the Chicago School had two which the “implications for design” may underestimate. convention itself. Howard the way in which the “implications for design” arise as a Becker. the adoption of ethnographic techniques is model of the relationship between technology and practice. and we will return to some organization might be understood.

is often used as shorthand for investigations that construction. Much contemporary debate around are. but about how members’ As outlined above. it both underestimates ethnography and fails to of interest to the HCI community and found a positive realize its potential. not least the juxtaposition of strategically chosen exemplars. it is to reveal certain answers to different sorts of questions – in particular. such as. Ethnography is customers. but rather than laboratory approaches are different contexts – Anderson’s [2] exploration of the issue methodologically unsuited to the target domain. to the ways in which they implicitly formulations of qualitative research methods such as or explicitly construct the roles of author and reader as well Contextual Inquiry [5]. Button’s [6] In reducing ethnography to a toolbox of methods for comparison between different models for ethnographic extracting data from settings. though. however. to some extent. important to make this distinction is that. It might be more accurate to say. which will be addressed experience. in much the same kinds of relation to conceptual organization of cultural settings. this reading of ethnography has patterns of sharing customizable software as explored by often been aligned with the requirements gathering phase of a traditional software development model. instrumental supports the idea that “implications for . not that ethnography was adopted in HCI research. in one of Anderson’s examples. the term is often used to encompass particular ethnographies as texts. The reason that it is three considerations. It is. Ackerman’s [1] reflections on the socio-technical gap. The term “ethnography. ethnography has been animated by a close attention to Similarly. qualitative. the ethnographer is far from a passive agent in the production This view of ethnography as purely methodological and of this organization as a research outcome. Perhaps unsurprisingly. the dominant view of ethnography is that it provides to HCI researchers a corpus of field techniques for form that is its final outcome. goal is to reveal and explicate rather than to create. after all. Anderson draws particular attention to reception for aspects of their work. how. Ethnography theorizes its subjects. what is revealed is the stand. strategies. this is questions about technology in everyday settings – that will fundamentally an analytic move. 9. then. and the ethnography is not. but the written In particular. investigation. So. question of the marginalization of theory. of ethnography and requirements. itself. in ethnographic approaches turned their attention to questions doing so. Writing. not veridical representations of the world. rather. and for design” are the primary or even sole output of ethnographic reflecting on some of the consequences. but to the way in which what I have called the methodological rather that ethnographers were adopted in HCI research. or open-ended. Despite a certain kinds of questions that can shape the design of a software ethnographic tendency to operate as “merchants of system.” considerable attention to its rhetorical form and indeed. ethnography’s outputs are often not analytic statements purely about THE MARGINALIZATION OF THEORY members’ experiences. and indeed.technological considerations in HCI research. Anderson observes the role of particular rhetorical somewhere. while the design exercises. a number of social scientists who made use of component of to ethnography – and. is central. however. and. I will draw in particular on three insightful different kinds of understandings from laboratory explorations of the problems of ethnography and design in investigations. and Suchman’s [37] account of forms of view marginalizes or obscures the theoretical and analytic ethnographic encounter between technologists and components of ethnographic analysis. approach consistently marginalizes or obscures the analytic That is. 16]. returns and reports. From this perspective. paying particular attention technology use. Ethnographies are texts. ethno-graphy. here. underlying logics of social practice. of ethnographic investigation is taken to be its spatio. Once more. problems of moving from laboratory studies to broader Anderson [2] insightfully explores the relationship between understandings of the social organization of settings of ethnography and requirements. the methodological analysis. a form of HCI.that the ethnographer goes Second. but it does not simply report what members say here through the four inter-related topics. inscribed as a literary form. ethnography is often conceptualized as astonishment” [17]. They help to illuminate a complex and intricate concerned with the member’s perspective and the member’s set of disciplinary relationships. ethnographic methods were originally experiences can be understood in terms of the interplay brought into HCI research in response to the perceived between members and the ethnographer. By analogy. we must pay collecting and organizing data. the project. Laboratory Mackay [26] and the marriage practices of the Bororo as methods can provide certain kinds of answers to certain detailed by Levi-Strauss [25]. Ethnography is interpretive. some problems have attended the ways in which writing and a way in which a cultural understanding is ethnography has been understood. the goal of such juxtapositions is not a set of field techniques that can provide different sorts of merely to dazzle and surprise. importantly. as ethnographic His first is that ethnography must be seen primarily as a approaches have gained more visibility and currency with form of reportage. observes. starting with the they experience. Consequently. in situ. the reason to adopt ethnographic methods is not that they will generate quite In what follows. temporal organization -. the defining characteristic as the object of inquiry [8.

but generated through an ethnography as purely methodological. one concern about the “implications for here. One way This disparity has consequences both large and small. 14]. though. and indeed it is the working out of these interdisciplinary effort – the difficulty of achieving a true analytic considerations that is the work of the ethnography. he approaches ethnographic inquiry the goal of ethnography is to generate implications for from a quite different position than the classical design construes the disciplinary relationship. while by contrast POWER RELATIONS ethnographic methods explicitly require that the The second consideration illuminated by the problem of ethnographer incorporate the context of the social “implications for design” is a more broadly political one. “about” the cultural context as ethnographic research – frequently neglects the perspective from which it is written and that of the audience analytic aspects of ethnographic work. The question of subject position is particularly important In summary. normally demand. to be selectively deployed as needed. judgments about the categories of observations which remain implicit in the data [4. for example. To the contrary. The first is that what Button describes as “scenic ethnography is not only “about” the culture under study. concerning the relationship between the constituent So. The second. and by the same encounter between that setting and the ethnographer. The especially those from most positivistic scientific traditions. The third consideration is one that I will focusing on the “analytic auspices” under which the return to later. or class markers. a form that might places those whom ethnographers study outside of the be summarized as “I went there and this is what I saw”) by design process. be inimical with their analytic position. to fund their own travel to program Diana Forsythe tellingly commented.” itself analytically warranted. as venues like CHI a tape recorder [12]. relationship between ethnographer and subject or setting. for instance. This reflexive character means that auspices. HCI arsenal. a perspectival view. by the same token. and that this quantitative and survey techniques depend upon subjective perspectival quality is critical to what ethnography is. as equivalent to other empirical approaches in the Students learning ethnographic methods for the first time. one’s introduction and social position.) Despite these huge practical obstacles. There are anthropological approach that I primarily address here. with instrument. etc. ethnographer. critically on recognising that the view of the setting that one First. as to whom it is presented. for those ethnographers and. always. but the first two are more immediate topics. it places ethnography outside ethnomethodological ethnography can be contrasted with of the design process itself. as an through which a specimen setting might be examined. the “implications for design” model he provides an ethnomethodological critique of the forms of postulates design as the natural end-point of research theorizing at work in what he terms “classical” inquiry. and three issues here. As an The particular issue I want to explore is how the idea that ethnomethodologist. Nonetheless. As Button notes. at a component of ethnographic investigations is to cast the small scale. and therefore designers as the gatekeepers for that ethnography. ethnographic analysis must reflect a set of analytic The question at stake here is that which underlies any commitments. At a in which the methodological view of ethnography practiced large scale.Third. the ethnographer also tells a story about becomes an occasion for tacit theorizing in ways that may ourselves [30]. ethnographic inquiry is conducted. As position. in doing so. both classical and research. in this view. design about some Other. First. power. Third. ethnographic understanding depends disciplines in HCI. a lens frequently express the concern that the ethnographer. Ethnographic data are not unproblematically observed design” approach is that it involves a reading of or extracted from a setting. Second. synthesis rather than degenerating to a case where one . it what he terms “scenic fieldwork” (that is. but fieldwork” – and what is often presented in a CHI or HCI equally. token. a brief perusal of the sexual. it creates a status hierarchy in which in HCI often marginalizes or obscures the analytic engineering demands tend to override social ones. it results in an imbalance in participation in ethnographer as a channel for the relatively straight-forward scientific meetings (since social scientists are rarely in a movement of data from the field to the design studio. Clifford Geertz famously described Rogers [35] notes. I Button [6] addresses this question in his review of want here to present some more conceptual concerns. academic and funding structures. an ethnographer is not committee meetings and conferences. is that in some cases the gear-change culture – the object of anthropological ethnographic inquiry from ethnographic description to design implications is not – as “stories that people tell themselves about themselves. Anderson emphasizes the reflexive character of There are two issues at work in this question of analytic ethnographic analysis. by telling an ethnographic story working from an ethnomethodological perspective. implicitly or explicitly. ethnographic approaches in HCI research. by the same token. inherently. What is missed is the extent that ethnography is alternative approaches. must inevitably distort the data and introduce an the ethnography providing an objective representation of element of uncontrolled subjectivity in contrast to that setting. one’s access to resources and relative size of research grants will demonstrate that amply. is a passive instrument. it is hard to deny the power differential between gains (or the interview responses that one gains) is engineering sciences and social sciences in terms of inevitably shaped by ones’ subject position – one’s ethnic.

but lies rather in their critique of the ethnographic investigation best determined by what design ways in which the domains of “natural” and “virtual” decisions it can support. of which others are passive consumers.disciplinary is essentially in service to the other. I want to highlight two pledges they have already made to themselves about implicit assumptions in this approach. The power of their then-available design opportunities. is the success or value of an a Trinidadian context. computer scientists are wary real domains. More to the point. that a designated and Systems” (as the CHI conference is more formally named. is of becoming programmers in service of other disciplines. and indeed grows out of. and are we to recommend what should. as to their self-conception as Trini. Third. that between “user interface design” and “human. after all. ethnographers. If the First. “reach” in the design process and the realities of . who is doing the design in these scenarios? There are interaction between people and computers – or between at least three potential design actors here – the people through computers—is itself a domain of enquiry. The fascinating solution was to advocate Ackerman [1] provides the metaphor of the “socio-technical what was called “vertical interdisciplinarity” – “interdisciplinary” gap” – the gap. Clearly. an analysis getting the best technological outcomes? At what point can of the cultural and social organization of some particular ethnographic contributions have their greatest impact upon setting or occasion is often best articulated independently of technology development and deployment? Schmidt [36] specific systems. that ethnographic investigations design.) demarcated group of “designers” are empowered to perform It suggests. A very particular set of relationships between for design is somewhat disingenuous… especially perhaps these constituencies is postulated by the traditional focus on at a conference entitled “Human Factors in Computing implications for design (in particular. the most effective outcome of a study ethnographic and technological considerations meet. or by what forms of learning it worlds is conceived and argued through information might enable? What forms of knowledge can ethnographic technology. Through a close analysis of the role of internet studies generate? technologies in everyday practice. for a range of in particular social and cultural contexts? I will explore this reasons. themselves. necessarily best oriented towards the creation of in more detail in the next section. the issue is that design is the tail that wags the dog. Certainly. technologies. to be appropriated or incorporated into practice.” computer interaction” as domains of study. but rather provides a new platform Following on from some of these questions. though. and the people then the call for ethnographic studies to deliver implications themselves. it causes us to design. maintain patterns of social and family contact that is critical this is commonly understood in computer science. claims that the most influential workplace studies in CSCW have been ones that did not harness themselves to specific As a brief example. instead. In the face of massive outward migration.) (indeed. and consequently. life may imperil. essentially. and perhaps more problematically. or design briefs. between. referring to Geertz’s comments cited that is implicit in the “implications for design” approach. a domain of everyday practice ways of being Trini that the practical realities of everyday and. ethnography is not. are we might be to recommend what should not be built rather that doing justice to the ethnographic perspective. a domain of technological design.1 one that is coextensive with. this was a major source of tension and frustration. as discussed above. In earlier. by focusing on specific designs as the point at which Sometimes. This calls into question a number of the assumptions that lie perhaps. the technologists. Each of these 1 assumptions is problematic from the ethnographic At a recent meeting of the recipients of a particular program of perspective. interdisciplinary research grants.and technology-oriented field. the Trini experience of the internet. While it is clearly important. in a Second. Second. the internet “[helps] people to make good on particular. new sorts of technological or consumer artifacts. on the other. it constructs ethnography as a point of mediation Trinidadians with another way of “being Trini” – indeed. Miller and Slater demonstrate how the technology does not create a place TECHNOLOGY AND PRACTICE outside of everyday life. the Internet provides people will the ability to something that is encountered just as it was designed. in this particular case. on the one hand. HCI research studies) are relevant only in as much as they support design.” They show how the Internet provides First. take Miller and Slater’s study of design efforts or limit their discussion of implications to Internet technologies in Trinidad [31]. to be oriented reconsider just what design looks like – the technology towards the opportunities to highlight and correct problems itself. The distinction to be drawn is. or the form of its local adaptations and appropriations in current technologies. between our technological engagements between computer scientists of different stripes. it implies that people will encounter technology as for example. I want to upon which everyday cultural experiences can be examine the relationship between technology and practice performed – how. Trini experiences of everyday life.. What Miller and Slater computation has become an increasingly important element question is the conventional separation between virtual and of other scientific enterprises. behind the notion of “implications for design. analysis does not lie in specific recommendations about the ways in which technology might be best designed to fit into Fourth.

perspective on the creative processes by which people put technology into practice. as better designed technologies would obviate the need for effective colonial administration came to depend upon the such adaptation and appropriation. they are mutually constitutive. and drawing for the purpose of design. we might What Suchman offers is both an insightful critique of suggest that what ethnography problematizes is not the industrial anthropological practice and an eloquent and setting of everyday practice. goes beyond giving form to to relate the local to the global. but reported in HCI. but rather to frame the notion of the gap. Design is critical. but how those technologies take on all politics is local. Arguably. In this way. accounts of social life that colonial administrators required of early anthropologists. and settings. By introducing and focusing on users or consumers of technology. too. gender. as an methods alongside controlled laboratory studies). instrumental. 27. in which the role of industrial or design anthropology is not because ethnography does not accept their conceptual to report on the habits and practices of a set of potential separation in the first place. a natural design-oriented ethnography of the sort that is often consequence of human experience. What is more challenging is “design. though. In particular. ethnographic perspectives suggest a different investigation was motivated primarily by this need. . As a focus of HCI research attention. as Tip O’Neill famously commented. In its early days. more broadly. because the technologies are poorly designed. between the objective. so also reflective about. however. food. the incorporation of ethnographic colonialism in at least two ways. As with technology. and customize technologies different forms. state support for anthropological By contrast. it provides occasions for enacting anthropology has become more conscious of. The fundamental phenomenon to be understood. interactive technologies. of strongly associated with the Participatory Design (PD) course. at particular systems of practice. anthropology relied to a great extent on sought to narrow the gap or to bridge it. the outcomes of these encounters are ones that attempt to focus of ethnography is the ways in which practice brings respect and amplify local practice rather than to represent it technology into being. From this perspective. ethnographers’ engagement with those being studied. In drawing attention to the gap. That is. the gap is her colleagues’ experiences in conducting industrial. around particular topics. Suchman argues for a reconfiguration of these relationships. [e. he REPRESENTATION AND INTERACTION draws attention. HCI research has. That said. but rather than it is the sides of the production/consumption relationship. understandings of indigenous peoples that anthropologists could provide. to the notion of design as a bridge. time. however. and the similarly objective. times. these are seen as In subsequent years. where all the interesting stuff happens. but the practice of design. though. it is important to consider the kinds of settings passive recipients of predefined technologies but as actors within which the sorts of encounters and partnerships that who collective creating the circumstances. Ackerman suggests not that it is the encounters and partnerships between those on different fundamental problem to be solved. The role of ethnography. long had an interest in aspects of the ways in which movement. in people might configure. and actionable adopted and adapted by people in the course of practice. and everything else. not as a problem different view. “artful integrations” that Suchman draws attention to as the It is practice that gives form and meaning to technology. and she describes might be forged. is a site for social and their relationship to the colonial apparatus. Technology. the domain of technology and the domain of instrumental. Participatory Design. It As I noted in my very sketchy historical introduction. Ackerman critiques the reports of travelers and colonial officers for the data the intuition that people adopt and adapt technologies from which accounts of cultural practices were formulated. here. then. Suchman’s work has been consequences of technology use. animals. perhaps. In particular. 10. has traditionally been practiced as a form of not simply on how people explicitly transform or program political activism and. the power dynamics involved in with space. cannot be to mediate between these two domains. through a range of methodological innovations anthropology’s history is linked to that of Western (such as. and the nature of emotion. one that reflected a growing disquiet with to be eliminated. anthropologists have taken a natural consequences of everyday action. 32]. ethnographic representation. what it does is to refigure “users” not as though. cultural production. adapt. as originated in Scandinavia but adopted. HCI has armchair discipline. she draws a telling analogy designs must always be put to work in particular contexts. practices.technologies-in-practice. workplaces demanded by traditional design processes. In particular. shows how. This ethnographic view. contexts. again on the notions of reflexivity raised earlier. the relationship was inverted. and explicitly cultural and social meaning.g. focuses however. and that Subsequently. Seeking to close the gap through the application of Suchman [37] uses this history to reflect on her own and ethnographic methods is a contradiction in terms. process of incorporation and co-evolution of technologies. inspiring articulation of an alternative approach. PD is good at staging specific social meanings through their embedding within encounters between particular sets of people. family. death. and to understand how these technologies to encompass appropriation – the active localized encounters might have global import. Certainly. and actionable accounts of users and everyday experience cannot be separated from each other.” in this sense.

however. and historical moments. organized. say. is it to suggest that ways of encountering the world – again. or the reception that ethnographic The question is how specific practices become exemplars of materials receive. how the world ethnography is not useful to design. the This is. and Finally. The conceptual materials comprise the ways The alternative account is of models for understanding in which these data are theorized. which is often where result requires a different reading.” A sensitivity for the nature of the recommendations as the “outcomes” of ethnography at best interaction between ethnographers and their subjects and distracts from.) and design practice have been both successful in design terms and productive for the research community (e. one concerning the forms of environmental knowing most fruitfully understand and make use of those materials. it is important to distinguish two levels and of these historical tales.) ethnography. I would like. and what kinds of models they The view of ethnography as “scenic fieldwork” focuses. 34]. interpreted. the analytic the sorts of representations of cultures and practices that and conceptual work that lies behind.” but rather. it requires a move beyond particular concisely and effectively formulated. let me amplify this last point a little in comparing what kinds of intellectual (and moral and political) two views of what ethnographies provide.g. but the explanatory frame by which this argument that reveals something about the setting under account can be organized and the narrative that connects investigation. built in order either to support what happened (if it is a tale The empirical materials comprise the fundamental of ingenuous practice) or to prevent what happened (if it is observational material – the “this is what happens” detail of a tale of failed or obstinate technology. on “moments. MOMENTS AND MODELS what matters is why. In of the ethnographic imagination rather than the fact. What matters is not simply what those implications are. Observations are always theory-laden. around one particular view ethnographies. WHAT’S TO BE DONE but it does more. What is critical here is not the account of juxtaposed. the substantive intellectual achievement is to be found. any encounter between ethnographer and field involves a ethnography is not a historical practice and the whole host of analytic positions. Tight couplings of ethnographic materials of the world or another. that ethnographic party. social settings. to be condemned to (and specifically here the land) becomes a site for cultural permanent academic irrelevance. and presented in order to make an what happened. is that the presence or import of in the context of generalized design. But. it is operative principles by which these practices are shaped. through this. we can all conclude what should be two sorts of contributions – the analytic and the empirical. but in the different kinds illuminate the relationship between technology and practice of knowledge of land and landscape that characterize each is important for our field. to dismiss the way in which ethnography has been interpreted within HCI. by any means. In thinking about ethnography (or indeed any social science historical accounts – “here is what happened. Nor. Ethnographic inquiries can illuminate the practices of particular peoples. My goal is not simply shared. but to more broadly formulate Whatever the particular “analytic auspices. One interesting example concerns the see the world. you like. too. perhaps.” On the basis contribution). ethnography provides new lenses through which to groups are mediated. again.” in Button’s the policies by which encounters between different interest phrase. time to take a more positive attitude. particular kinds of practices. so a hard-and-fast . even in cases where such recommendations can be ethnographic toolbox. while retaining a firm link to detailed accounts of practice. this is to confuse ethnography’s method (the materials and understandings not simply to illuminate detailed examination of everyday life) with its product. The question then becomes not “what do these people approaches can help us move beyond simple dualisms of know. I believe. to production. As noted in Forsythe’s comment above. that are reflected in land disputes. if reflect. “of what does their local knowledge technology and practice and the “layer cake” models that consist. and how they were arrived at.It may be that some insight is to be found in those places ethnographer not merely a recorder of events as they where anthropologists have sought to use ethnographic unfold. then.” That is. reproduced and transformed. understood. it explores the generally Since I have been unremittingly critical to this point. we can start to understand how policy In that spirit. conceptual level. my argument is certainly not that design can be shaped around the very notion of clashing recommendations are poor things to include in epistemologies (rather than. it offers descriptive. and how is it manifest?” By moving to this they engender [24]. to focus on those “implications for design. and often completely obscures. what is “implications for design” are not the appropriate criterion critical to recognize here is that this power is a consequence by which ethnographic contributions should be judged. The problem then becomes an epistemological contribute to the ongoing debate about the ways we can one. commitments they embody. a way in which Suchman’s formulation of Lancaster work on air traffic control [22].) What I do want the ethnographically mediated encounter can be supported to suggest. The question to be resolved is. these disputes have their I believe that the turn towards approaches that can origins not simply in territoriality. rather. what is to be ways in which Australian authorities have attempted to done with this sort of work? respond to conflicts over land rights [21.

ethnography. Miller and thinking about social life.separation would be impossible. a range of methods the role of ethnography in HCI design is to question the have been proposed which we might call “discount concept of the ethnographic site. for example. instance. More to the point. in some implications for design to be drawn out here – not for general.” it is because the ethnography’s “analytic auspices” [6] is to be able to categories of design. Their impact. Slater. because they do.” I would argue. but rather as analytic level rather than simply an empirical one [11]. However. world and other forms of cultural practice (including. It is a operates within a globalized economy. what they do when they professional rite of passage. they typically go beyond specific implications for design? Arguably.” but must be understood designers can move out from laboratory settings to the real within a broader web of relationships to other parts of the world as a basis for design inspiration. Prominent examples are One of the more significant transformations of Contextual Inquiry [5] and Cultural Probes (and related contemporary anthropological ethnography has been the approaches) [15. Whereas traditional Contextual Inquiry provides designers with a series of tools ethnographies since Malinowski have focused their and techniques for understanding social settings and attention on a geographically bounded field site. often more profound than production and consumption? a laundry list of facts and features. is frequently more diffuse. and with whom they What has traditionally been more complicated has been to exchange information.) Contemporary Cultural probes are self-report packages of artifacts. ethnography” techniques by analogy with the discount usability techniques of the 1990s. translated into technological constraints and opportunities. many ethnographic studies can provide such What might happen if we started to think more critically things (although it is important not to ignore the role of the about the “site” of ethnographic studies in HCI? In what ethnographer as interpreter and framer of these “facts” ways can we separate the technical practices of one rather than as a passive mirror of the site. artifacts. and ironically. 23]. must concern itself instead with . it provides a set of methods whereby approached as a “realm apart. despite the massive outward migration of peoples from drawing upon the notion of requirements in traditional Trinidad. given HCI’s legitimate interest in inquiring into What this further suggests is that one way of reconsidering complex. But at least as far as HCI is transnational flows of people. for Cultural Probes (and a series of related approaches such as example.) organization or set of users from those others with whom they interact. then we are forced to consider sit down at the computer. user. is a request for empiricism. Certainly. To the extent that these implications are not One reason to draw attention to this question of formulated a “implications for design. their connection to international academic Technology Probes) arose within the design community as anthropology. understand the Internet in terms of the ways it allows Trinidadian cultural practices to be sustained The call for “implications for design. and practice. or city within which implications in the same way. but rather the global technology culture It is not that these do not have profound implications for itself. They provide us with new ways IMPLICATIONS FOR DESIGN of imagining the relationship between people and What should we conclude from this? And should we. in technology. CI is aimed at those with longer be adequately circumscribed in such a matter. observes that. or the intersection between cultures of technology design. technology is used. and to the cultural settings in which a means to conduct broad-based surveys of user experience. and in terms of the ideas it offers for embodiments of these globalized practices. and so forth – which can be the concept more critically [20]. we can distinguish between these two as ways in perhaps especially relevant when considering information which an ethnography make a contribution – in terms of technologies – technologies that are both means and what it says happens. culture can no Based largely on interview data. Marcus organizing their observations to derive models for design. socially-organized settings. instead. how they talk to each other. concept of “multi-sited ethnography” as developed particularly by George Marcus [29]. For question. campus. are themselves in appropriately compare methods and assess results. from whom they learn. there are instances of design. They provide us with ways of approaching deference to decades of CHI tradition. on the fundamental repudiation of a traditional the design of interactive artifacts. and people? How might we establish a deeper. between technology studies and of research programs. formulate these as design. indeed. When we attempt to request that the ethnography provide “facts” – when people discharge the mythology of the field as part of a work. capital. This is concerned. however. and culture. then. in the context of globalization. and designer. the Internet is a means by which cultural practice software engineering. The neither the training nor the time to conduct ethnographic Trobriand Islands can no longer (if it ever could) be work. The contributions to a broader ethnography corpus whose “site” analytic contributions tend not to be seen as holding is not a particular office. they draw. but the design of research separation between designer and user. more foundational connection between reconfigure the ethnographic project if we thought of ethnography and design – to look for a connection at an studies not as independent investigations. ethnographic results are presented [30].

rather. Matthew Chalmers. the problem of the nature of this relationship 1. CA: recognized. Within HCI research. in terms of their that this is a possibility. that a focus on University of California Press. not least Ken Anderson. but arguably. it provides Computer Interaction. As a spin-off benefit. however. Foundation under awards 0133749. . 1988. At the same time. I am further indebted to Mark the topic of their inquiry is quite legitimate. New York. In 5. Tom Boellstorff. locate the topics of Genevieve Bell. what is at issue here is not the nature 6. Ethnography has a critical role to play in interactive system design. Human- not simply to save the reader a trip. 151-182. and by a grant from Intel Corporation. they are ACKNOWLEDGMENTS often proposed as alternatives to “full” ethnographic This work was supported in part by the National Science methods when time is at a premium. At the outset of this paper. REFERENCES However. Ken Anderson. Berkeley. they clearly fail to capture what an ethnography 0527729. Graham Button. The value of ethnography. 2000. M. perhaps this might also help by the probes are intended to provide inspiration rather than to transform our evaluative criteria for ethnographic papers. Chicago. 1983. Clifford. Campbell. Design Studies. The Intellectual Challenge of remains problematic and needs to be addressed [33]. Kia Höök. Contextual Inquiry and Cultural Probes. 1986. WI: University of Wisconsin Press. 1969. The second has Representations. and though. I focused primarily on the genre of ethnographic presentations at CHI and related venues. Beki interest outside of the relationship between ethnographer Grinter. but its goal is The Value of Ethnography in System Design. Interdisciplinary relationships in the social I have had two goals in writing this paper. Shaping Written Knowledge: The goes on there. Beki they are rejections rather than variants of ethnographic Grinter. Frankly. effective or appropriate method. Anderson. In corporate) strategy as in uncovering the constraints or Flinders and Mills (eds). G. C. have and subject. Bob Anderson. H. 319-332. Ethnocentrism of Disciplines and inquiry and results. K. I have been a poor student. Watch What I Do. Morgan form but to the socially-organized nature of documentary Kaufman. Many captures. it is the appropriate Design. been to open up a debate about the relevance of ethnographic work for HCI research. I mean to point not simply to media Defining Customer-Centered Systems. H. Michael Montoya. 2. Ethnography provides 2. The central question is. and the CHI reviewers emphasis on implications for design. 1994. 118-146. Theory: The Necessary Evil. The first has sciences. T. the basis for analysis. A. Human-Computer Interaction. then. Theory and Concepts in opportunities faced in a particular design exercise. Ackerman. Cambridge. C. places it in a broader historical context. The Ethnographic Tradition and of ethnographic practice as such. J. and Lucy Suchman. furnished (explicitly and implicitly) valuable lessons in the are not simply “discount ethnographies” – instead. Science. but this may be as much in shaping research (or 4. Mimi Ito. Again. criteria for evaluating and understanding ethnographic 7. is in the Genre and Activity of the Experimental Article in models it provides and the ways of thinking that it supports. Button. Representation and Requirements: insight into the organization of social settings. Qualitative Research: Perspectives from the Field. been to offer an account of ethnographic research that 8. I have argued. and.questionnaires. Bazerman. Clifford. practice. Writing Culture: The studies have something to say to HCI designers is broadly Poetics and Politics of Ethnography. 15. I believe firmly that ethnographic research has much to contribute to HCI design and deserves a central role. On Ethnographic Authority. That ethnographic 9. D. Cypher. open-ended approaches and reliance on qualitative rather than quantitative materials. G. methodological concerns. Jonathan Grudin. rather. and exercises that encourage users to reflect fail to do justice to the kinds of insights that they can on their experience. 21. I apologize if produce results that are inherently quite different. models for thinking about those settings and the work that 3. IL: Aldine. Amanda Williams. given that they lack the coupling of analytic and colleagues. implications for design reads ethnographic inquiry too 10. NY: Teachers College Press. they relationship between ethnography and design. often provocatively. what makes the Fishscale Model of Omniscience. Beyer. and Marcus. 0205724. implications formulated by an ethnographer is not the most 179-203. Contextual Design: referring to genre here. MA: narrowly. In: Shrief and ethnographic work valuable in research in HCI? Sherif (eds). 328-348. Madison. Ackerman. 1997. then. The data generated provide. In particular. 0326105. 218- 229. I doubt that this is the last CHI paper on ethnographic work that will find itself forced to end with At first blush. R. again. Becker. constraining ethnographic studies in ways that MIT Press. inquiry and its topics – techniques that do place primary Yvonne Rogers. What I CSCW: The Gap Between Social Requirements and have tried to argue here is that a bullet list of design Technical Feasibility. these techniques bear a certain broad “implications for design”… but it is certainly nice to think resemblance to ethnographic methods. 1993. Bonnie Nardi. 2000. and Holtzblatt. 9(2). 1993. for insightful comments on earlier drafts of this paper.

1999. A. (eds). Sweden).. Eiderbäck. MA: Beacon Press. Rogers. 781-795. 16. Garfinkel.. H. Conversy. Cambridge University Press. Dordrecht: Kluwer.11. 1990. 1995.. D. H. are Workplace Studies For? Proc. London: Routledge. 1967. Technology Probes: Inspiring Design Studies in CSCW. From Anthropologist. User-Tailorable Systems: Pressing the Issues with Cambridge: Polity. Patterns of Sharing Customizable Ethnography as Invisible Work. 17. Mackay.. Kling. Politics of Apprehending Aboriginal Labor. P. Geiryn. Grudin.. Proc. 1969. MA: (Seattle. Do Rocks Listen? The Cultural Land. A. 1993. Studies in Ethnomethodology. C. Beaudouin-Lafon. The Mabo Decision. M. New York: ACM.. D. New York: ACM... Dourish. Works and Lives: The Anthropologist 28. B. Povinelli. 2003.. R. B. 37. J. Cultural Computing Systems CHI’90 (Seattle. ACM Conf. as Author. and Pacenti. Review of Anthropology. and Moran. 1995. R. P. 2000. 127-145. 27. T. American 22. 2000. Roussel. 25. The Computer Reaches Out: The Ethnographic Approach. 24. C. 18. Practice-Based Design of 24. Argonauts of the Western Pacific. Reconfiguring the Social Scientist: Experiences from the Field. 20.. Land Rights. Proc. C. Marcus.. J. New York: Probes. Great Divide. B. 31. T. Computer-Supported Cooperative Work. Geertz. 1998.. interactions. Y. Boston. and King. Gaver. Druin.. K. Human Workplace Studies: Rediscovering Work Practice and Factors in Computing Systems CHI 2003 (Ft Informing Design. Princeton The Emergence of Multi-Sited Ethnography. Ethnography in/of the World System: Reflections on Philosophical Topics. Boundary Work and the Demarcation 30. The Internet: An 19... Nardi. and the Meaning of 34. It’s Just a Matter of Common Sense: 26. Annual University Press. Plowman. S. and Shapiro. 8(1-2). 1988.. W. Malinowski. 14. A. A Small Matter of Programming: Conf. W. Human Factors in Computing Systems CHI’90 Perspectives on End-User Computing. Levi-Strauss. 48(4). More Involved. Human Factors in 15. L. and Luff (eds). Chicago: University of Chicago Press. and Slater. 1989.. Hill. 13(4). R. Stanford University Press. H. A. and Button. Historical Continuity of Interface Design. G. 1997. Hansen. Technical Systems. New York: ACM. Hindmarsh. 1983.. Computer-Supported Software. The Critical Role of Workplace Evans. Bederson.. and 36. Westerlund. Lauderdale. 303-322. 261-268. FL). Fortuna. G. Anthropological 33. D. Miller. McKim. Randall. 18. Lövstrand. Dunne. and Stablien. Information Systems: Notes from the Hyperdeveloped Scientific Collaboratories as Socio-Technical Interaction World. In Bowker et al. 1997. 12. L. MIT Press. American Technomethodology: Foundational Relationships Conference on Information Systems. The Information Society. In Heath. Frost. Geertz. 505-518. On Networks: A Theoretical Approach. Schmidt. Computer-Supported Cooperative Work ECSCW’95 21. Kinship. Y. N. What Locations: Boundaries and Grounds of a Field Science.. and Ramage. 1995. T. 1. American Social Sciences. Ethnographic Record to System Design: Some 35. D. 17-24. G. ACM Conf. 2000. 2000. M. MacLean. Hughes. . E. Forsythe. L. 17(2). ACM.. Gupta. B. and Fischer. and Ferguson. ACM 32. 97(3). 23. Anthropology as of Science from Non-Science: Strains and Interests in Cultural Critique: An Experimental Moment in the Professional Ideologies of Scientists. ACM Conf. CA: Sage. 1990. WA). London: Erlbaum. 95-117.. Newbury Park. H. CA). G. Doing Exemplary 221. Cooperative Work CSCW’90 (Los Angeles. between Ethnomethodology and Systems Design. Blackfellas and Whitefellas: Aboriginal (Stockholm. Oxford: Berg. 1992. European Conf. W. K.. 395-432. 139-144. Buttons. 123-141. M. For and With Families. J. Plaisant. Hutchinson. Research. WA). 1986. Sociological Review. 21-29. Computer-Supported Shifting from Telling Designers What to Do to Getting Cooperative Work. Suchman. 1993. Berkeley: University of California. Carter. 1995. Social Science. 2002. Human Rights Quarterly. D. 209- 13. 1990. Cambridge. 1967. Proc. C. The Elementary Structures of Human-Computer Interaction. Proc. Available Light: Anthropological 29. J. B. Rogers. 6(1). E. Marcus. and Cooperative Work: Beyond the Mackay.