Advancing the Value of Ethnography

Clear Theory for Clear People: Three Ideas for Advancing Theory in the EPIC Community Written in a “Non-theoretical” Way


As a young social scientist I used to be incredibly attracted to dense theoretical texts in anthropology, psychology and the social sciences in general. I equated thickness of language to complexity of thought. I no longer do. When I truly disowned the belief that obscure language hides complex thinking, I had two choices—either let go of theory altogether or develop a different appetite for it. I developed an appetite for clear theory and clear language.

Theory in business research, even when informed by the social sciences, demands clarity. At this point in our development as a business research community informed by social sciences, new theory is essential. If we are to overcome the still-dominant view in academia that our work is merely a practical “derivate” of more erudite scholarship in universities, we should invest in our own theory.

Many of us are doing this work: it is showcased every year at EPIC and collected online in EPIC’s Intelligences [], amongst other media. The monumental volume Handbook of Anthropology in Business, edited by Rita Denny and Patricia Sunderland (Denny & Sunderland, 2014), is possibly our greatest materialization to date of the deep interconnections between practice and theory in design and business anthropology.

Here I offer three ideas as a public invitation for extending the development of theory in our community.

Idea I: Controlled Equivocation

The first theoretical idea that I would like to see extended into the field of design and business anthropology is ‘controlled equivocation’. Inspired by the work of the Brazilian anthropologist Viveiros de Castro (2004) on ‘controlled equivocation’ in social and cultural anthropology, I propose that we should analyze as an equivocation the process by which practitioners must engage in strategies of dissimulating complexity in order to communicate with corporate peers. Even when we translate anthropological jargon to the language of business we are not communicating on equal ground. We are communicating on the basis of a shared misunderstanding of what anthropology, and sociocultural analysis, stands for on both sides—anthropology and business peers.

For instance, if my client is a wine manufacturer and I present them with a slide of different wine brands lined against each other, as if these brands were structural categories, I don’t expect the client to see Lévy-Strauss in it. That does not mean that an idea of social structuralism via Lévy-Strauss is not there, ordering the data, while being negotiated by the two parties. It means simply that Lévy-Strauss must be covered up in leaves in order to make their way across the corporate jungle.

In this respect, academic writing and delivering a PowerPoint in the business world could not be more different. Delivering a paper in academia is an exercise in showing the effort that went into the paper while seeking validation both for the ideas and the effort itself. Delivering a PPT to a client is about working hard to making it sound and look like an effortless stream seeking validation for ideas only. For people who are used to one main form of presentation, enduring the other’s presentation rituals can be an act of deep frustration and mutual misunderstanding. In fact, the apparently ‘simple’ language in which this small column is written is itself an act of controlled equivocation. A book that I published in 2013 can also be counted as an example (Oliveira, 2013).

Idea II: Epistemic Transference

Anthropology is a permanent exercise of transition across modes of knowledge. This transition is made possible through a process by which the language of the native (emic) gives place to the meta-language of the analyst (etic). Matters are made more complex in the practice of business ethnography due to the extensive stakeholder networks found in corporations, formed by ‘natives’ of many kinds.

Epistemic transference refers to the process by which a given mode of knowledge is assimilated into a different mode of knowledge inside a particular network.

For instance, in UX research, what we often call the “mental models” of users must transition into the “mental models” of engineers in order to create software that truly serves as common referent to both groups (e.g., Oliveira, 2014). In marketing research, what we often call consumer values must transition into organizational and strategic values of the corporation, while allowing for partial assimilation of the former by the latter (e.g., Oliveira, 2012).

Let me illustrate epistemic transference by way of example. My client is a technology firm and I am working with a member of its design team—an ergonomist. We will conduct ethnographic research on telecommunication shops in Portugal, observing interactions between customers and clerks and between customers and different parts of the shop. Following the research, we will conduct co-creation workshops with groups of ‘consumers’, shop clerks, and engineers/software developers internal to the client company. We will give the three groups the same data and ask them to generate innovative solutions.

As usual, in the beginning of a process I am not sure how the data will be represented in the co-creation workshops. That will be a function of our research outcomes and, ultimately, a function of the way the relations between shop clerks, consumers and parts of the shop will be transferred into our own analytical models (emic into etic). From the outset, we do have an idea of how to represent the data gained in the field in ‘typical’ profiles of consumers (‘personas’) and an idea of how to represent that data in typologies of interaction between consumers and shops (‘scenarios’). What precise form personas and scenarios will assume is unknown at the outset.

In this process, two forms of transference of knowledge will be operating:

  1. from the shop, shop clerks and clients to the researchers: the relations of the former are represented in the analysis of the latter;
  2. from researchers to co-creation materials, allowing for co-creation amongst the three groups.

Equivocation (‘covering complexity’ through personas and scenarios) and epistemic transference (helping other people identify with the different points of view found in the field) will walk hand in hand.

Other forms of transference will happen as fieldwork evolves. For instance, the more time we spend in the field, the more my colleague and I gradually start embodying different viewpoints of the interactions between shop clerks and clients. A sketch of this divided alliance comes to the fore in our lunch after the first morning’s work. While my colleague starts talking to me about stories and moments of frustration that she witnessed from the client side, I keep talking about the difficulties I feel that the shop clerks experience. A full version of the shop as an information system will only start emerging as we become more able to listen to each other’s embodiment of different parts of the system. What has been transferred to each of us must converge in analysis.

III. Implementable Validity

The “validity” that we deal with in daily corporate life is often different from ideas of validity found in academic literature. Validity, in academic and non-academic discourse, is a concern most often associated with the use of quantitative methods, which simultaneously shapes ideas and procedures in qualitative methods. Ethnography, as a research approach with a deep qualitative component, is yet to develop forms of validity resembling quantitative methodologies. But perhaps the point is that ethnography, with its ever-present focus on the emic-etic dimension, will never quite develop many of the forms of validity associated to quantitative methods. Another point is that ethnography in the private sector is not a practice with a view of seeking recognition by academic peers, but something that must influence a course of action.

For instance, quantitative researchers conducting some form of research in telecommunication shops would likely try to determine, by means of a survey or questionnaire, if the end results would be comprehended within some numerical limit from which one can deem them as ‘valid’. Statistical significance is the language of quantitative studies. In ethnography, however, what we want is to arrive at a representation of the shop that is recognized by those in it. We want the experience of that representation to affect the relation between agents bound by the space of the shop (shop clerks and consumers) and agents located outside that space (engineers). Throughout, we are looking for a change in point of view, where the consumer experience is put at the center more so than it was in the past. Last but not least, we want the outcome of this exercise to be open to implementation. Whether the exercise is able to affect a course of action inside the corporation—what the social psychologist Chrys Argyris used to call ‘implementable validity’ (Argyris, 2006)—is more important than the methodological purism at its root.

In order to seek the kind of validity we are looking for, we conduct multiple checks. We write a first draft of personas and scenarios and get it checked by some of the shop clerks prior to inviting them for a co-creation session. We do the same with consumers and engineers, refining the materials as we go along.

The final concept coming out of the co-creation workshops is turned into a video and sent over to all three groups, as well as several sectors inside the corporation. The video points to a new concept around the experience of the telecommunication shop. If it works, it will generate consensus across different hierarchical levels inside the organization. That ‘consensus’ is a manifestation that a form of (implementable) validity is on the way. It is substantially different from the kind of consensus determined by the scientific community around the validity of a given piece of research.

An invitation

I suggest that we can carry on expanding these three ideas and the relation between them as elements of a theory of design and business ethnography praxis. In doing so, I hope that we will carry on defining a field unafraid of the kind of theory that strives for a clear relation to practice and unafraid of a differentiation from what have been, so far, the academic theories shaping our thinking. The theoretical triad of controlled equivocation, epistemic transference and implementable validity is not an end solution to differentiate our field, but another small contribution to something growing as we speak. It is also a way of saying that practice, like it or not, sends a signal back to the theory. I hope that others will join me in making that signal loud and clear.


Argyris, C. (2006). Reasons and Rationalizations: the Limits to Organizational Knowledge. Oxford: Oxford University Press.

Denny, R. & Sunderland, P. (2014). Handbook of Anthropology in Business. Walnut Creek, CA: Left Coast Press.

Oliveira, P. (2012) Ethnography and Co-Creation in a Portuguese Innovation Consultancy: Wine Branding Research as an example. Journal of Business Anthropology.1 (2), 196- 217.

Oliveira, P. (2013). People Centered Innovation: Becoming a Practitioner in Innovation Research. Ohio: Biblio, The Educational Publisher.

Oliveira, P. (2014). Corporate Ethnography and Clinical Research: Thinking a Case Study of Software Evaluation. International Journal of Business Anthropology, 5 (2), 5-37

Viveiros de Castro, E. (2004). Perspectival Anthropology and the Method of Controlled Equivocation. Tipiti Journal of the Society for the Anthropology of Lowland South America 2 (1): 3-22.



Pedro Oliveira, Independent Research Consultant

Originally trained in clinical psychology (University of Coimbra, Portugal) Pedro earned a Masters and PhD in Social Anthropology ay Brunel University, London. Over the past five years Pedro has been working as a design and business researcher, both at an agency worker (Couture, Decode and Disrupt) and as an independent research consultant working for some of Portugal’s major multinational tech firms (Novabase, Critical Software). He is soon to start a full time position at Simpson Carpenter, UK.