Ethnography is one of the most prominent approaches in qualitative research. It can be defined as the small-scale study of human behavior through "clinical" observation and intersubjective participation.

What is Ethnography?

Historically, ethnography belongs to the paradigm of "conjectural" sciences, in which small indices of phenomena are seen as instances of their structural characteristics. Archeology, anthropology, specific branches of medicine, history, sociology, linguistics, criminology, legal, educational-pedagogical and political studies and psychology belong to the same category. In each of the cases, "evidence" would be constructed through clinical or forensic - i.e. realistic and detailed - observation, usually in tandem with intersubjective engagement between researcher and researched - by means of interviews, for instance, or by means of so-called "participant observation" (which means exactly that: the researcher participates in the process s/he observes, assuming the role of an actor).

Founding fathers of Ethnography

In scientific folklore, the origin of contemporary ethnography is often attributed to Bronislaw Malinowski, a Polish-British first-generation anthropologist and author of classics such as Argonauts of the Western Pacific, who (more or less accidentally) spent several years living on the Trobriand Islands, which offered him a rich and detailed panoramic view of social and cultural life there. It was Malinowski who turned such types of longitudinal and involved observation into a canonical fieldwork method for anthropology, replacing earlier "hit and run" (aka "notes and queries") methods.

More or less similtaneously, however, similar methods emerged in the work of that other founding father of modern anthropology, Franz Boas. Boas (a German-American) started his career as a physical geographer; like Malinowski, his first deep immersion into the cultural life of others happened almost accidentally, when Boas spent the long arctic winter on Baffin Island. This extended period of exposure and participation in the life of the Inuit led, in the American school of anthropology of which Boas was the creator, to a similar set of prescripts: anthropologists should take time to observe and participate in the cultural life of the communities they studies; they should observe everything, from material culture over rituals and economic activities to political relationships and structures of power and authority.

And - Boas insisted - they should learn the language of the community they observed, as an entrance into the deeper layers of social and cultural structure. Boas' 1911 "Introduction" to his monumental Handbook of American Indian Languages became (and remains) the foundational text for a new and highly sophisticated type of field linguistics, as well as for grasping the deep connections between language and culture.

Boas' students rigorously applied these methods to the study of the rapidly disappearing Native American communities in the early 20th century. Boasian anthropology was comprehensive in its descriptive aims; but these aims were animated by a number of far more important theoretical propositions.

Two fundamental theoretical propositions

One foundational theoretical proposition was, and remains, that aspects of sociocultural phenomena and processes must be seen as somehow connected. Every feature of sociocultural life is in itself a mere detail, evidently, but details start making sense as indices of something larger and more fundamental when they are connected in patterns. These patterns connecting a multitude of - in themselves hardly significant - small features point towards the system governing the occurrence, distribution and function of such features. This first theoretical proposition marks, in effect, an ecological approach to human social and cultural life. 

The second major theoretical proposition was, and remains, cultural relativism, understood in two ways.

• One: Boas and his followers strongly emphasized that the study of sociocultural systems should be free of predefined benchmarks and criteria. The system had to be studied in its own right, not in contradistinction with other systems (such comparisons would only be possible as juxtaposition, after the study of the different systems had been concluded). And certainly not in terms of "superiority-inferiority" terms, in which particular systems would be measured and evaluated according to the standards of other ones (usually: those of the West). 
• Two: the study of sociocultural systems had to be free of such benchmarks and criteria, because such systems have a logic of their own. They are organized according to socioculturally specific benchmarks and criteria, and these can only be found when other such yardsticks (e.g. those of the West) do not influence one's analytical gaze. The task of ethnographic analysis is precisely to identify the endogenous (often called "emic") logic of sociocultural systems - the glue that connects the multitude of details and turns them into a system - and to abstain from imposing exogenous benchmarks and criteria (often called "etic") onto the phenomena and processes studied.

These theoretical propositions - especially the emphasis on the internal logic of sociocultural systems - were highly influential in the development of structuralism. In fact, one of the key figures of 20th-century structuralism was an anthropologist, Claude Lévi-Strauss. The propositions also set ethnographic work apart from more prescriptive and positivist branches of science, and they exposed ethnocentrism and colonialism as epistemological filters on knowledge about "the Other". They have been highly influential in the work of scholars across the social sciences and humanities - one can think of Pierre Bourdieu, Gregory Bateson, Bruno Latour, Erving Goffman and Carlo Ginzburg as shining examples. The Chicago School of Sociology brought ethnography "home", so to speak, by introducing it as a mode of studying modern urban life, and later, Grounded Theory pushed ethnographic principles into mainstream quantitative-qualitative social science (an elementary aspect of Grounded Theory often overlooked these days).

Methodological principles

Ethnographic studies can be characterized by the following methodological principles.

1. The big in the small: In line with the first theoretical proposition above, ethnography starts from the infinite richness of small details, and works its way up to insights of a systemic nature. The leading question is therefore: what is the larger story documented by these details?

2. The situated nature of human sociocultural behavior. Human behavior is always behavior in a highly specific context, and derives its meaning and effect from how it fits in such contexts. Any specific feature to be studies needs to be "situated" in its context, and decontextualizing it would curtail whatever can be learned from it.

3. Intersubjectivity in knowledge construction. Ethnographic research does not claim to be "objective" - and exposes such claims as unwarranted and unsubstantiated. People inhabit a world which they interpret, and they approach that world on the basis of what it means for them. This princple also holds for researchers themselves: what ethnographers construct in the way of knowledge is mediated by the subjectivities of all the parties involved. It is intersubjecive. Observe that perhaps some of the most powerful statements on this topic were made by a sociologist, Pierre Bourdieu.

4. Concerned with ecological validity, not representativeness. Ethnographic research yields hypotheses (like any other science) that have to be ecologically valid in the sense that they need to reflect the lived experienced of the participants, not those of an a priori theoretical framework or those of a (exogenously defined) broader population. Whether ethnographically gathered knowledge is "representative" of the experiences of everyone is irrelevant; it has to be knowledge that maps iconically the knowledge of the participants in research. From within ethnography, and guided by this principle of ecological validity, devastating critiques were (and are) addressed at mainstream survey-and-statistical research, notably by Aaron Cicourel.

5. Empirical means "naturally occurring". There are branches of science in which the term "empirical" is reseved for research in which carafully doctored (repeatable) experiments are central. In ethnographic research, the term "empirical" means precisely the opposite: work is empirical when it gathers and addresses "naturally occurring" data produced in real-life contexts. Such data are by their very nature unique, and the phenomena and processes they document are by their very nature not repeatable. "Reality is not an experiment" is an ethnographic motto. This is what informs the term "clinical" used at the outset in defining the observational stance in ethnography. In medical training, the "clinic" is where one meets real patients and examines real bodies, not objects manufactured in laboratories. Ethnographers, consequently, are usually very much "out there" in the field.

Beginners' mistakes

Like any other branch of science, ethnography has an image in the eyes of non-practicioners and is quite often badly misrepresented in non-ethnographic literature. What follows is a review of some very widespread mistakes made in discussing ethnography.

1. Ethnography is a method. This is how ethnography is very often sketched: as a particular set of methods of data-gathering. The iconic type of data here is usually "interviews". In research proposals, one often reads phrases such as "I will perform ethnographic interviews", in ways that suggest that any form of interviewing and usage of interview data would in itself be ethnographic. It isn't. If interviews are gathered just to turn the data into standardized corpora (usually deleting the interviewer's questions) there is nothing ethnographic about them. They have been turned into a decontextualized sample of text, the exact opposite of what ethnography proposes. The thing is: ethnography is a theoretical complex, a paradigm if you wish, in which ontological, epistemological and methodological statements define what is ethnographic and what is not. In actual fact, few branches of social science and humanities have developed such a robust set of theoretical constructs, and such an extreme degree of self-critical reflexivity in practice, as ethnography.

2. Ethnography is unscientific because it is subjective. One hears this remark often from more positivist or "objectivist" corners of science, and very often Karl Popper's discussion of objectivity is invoked as authoritative backing. It should suffice to quote Popper himself (in his celebrated text The Logic of the Social Sciences) as rebuttal to such uninformed claims: 

“It is a mistake to assume that the objectivity of a science depends upon the objectivity of the scientist. And it is a mistake to believe that the attitude of the natural scientist is more objective than that of the social scientist. The natural scientist is just as partisan as other people, and unless he belongs to the few who are constantly producing new ideas, he is, unfortunately, often very biased, favouring his pet ideas in a onesided and partisan manner” (p.95). Popper continues: “In formulating this thesis I have said that it is practically impossible to achieve the elimination of extra-scientific values from scientific activity. The situation is similar with respect to objectivity: we cannot rob the scientist of his partisanship without also robbing him of his humanity, and we cannot suppress or destroy his value judgments without destroying him as a human being and as a scientist. Our motives and even our purely scientific ideals, including the ideal of a disinterested search for truth, are deeply anchored in extra-scientific and, in part, in religious evaluations. Thus the ‘objective’ or the ‘value-free’ scientist is hardly the ideal scientist.” (p.97)

It is precisely in ethnography that the researchers' potential and actual biases and cultural accents - his/her subjectivity - are fully acknowledged and addressed, and used as a keystone of the program in the form of intersubjectivity. No attempt is made to hide the researcher's agency as a co-constructor of ethnographic knowledge behind the safe walls of "method" or any other "objective" construct. There is more Popper in ethnography than in several other overtly "Popperian" approaches to science.

3. Ethnography is anecdotal and only addresses "small" things. Wrong: see what was said above about details, connections between details, and systemic features of social and cultural life. There is a full theoretical foundation for this option and strategy, vastly more solid than the ones often offered in, for instance, survey-and-statistics research. If applied across the board, this allegation of anecdotism applies to any form of clinical investigation - including that performed by your GP and regional police department detectives and the judges in your courts of law.

4. Ethnography does not yield "facts", usually because things can only be "facts" when they are representative, i.e. quantitatively established, with "hard" figures backing them up. As we have just seen, there are very different modes of establishing "facts", and ethnography (along with any other clinical approach) is one very well established such mode. Ethnographic facts have the advantage that they have validity, as described above - a thing usually very shabbily addressed in approaches that aim for representativeness. It will not be of great help to challenge the verdict of a court on the crimes of an individual by arguing that the judges' conclusion is not representative and can, therefore, not be factually true.

See also the wiki entry on Digital Ethnography