What's Wrong With Ethnography?

Martyn Hammersley

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781136115561
• Edition: 1
• Topic: Scienze sociali
• Subtopic: Sociologia

This stimulating and refreshing study, written by one of the leading commentators in the field, provides novel answers to these crucial questions.
" What's Wrong With Ethnography provides a fresh look at the rationale for and distinctiveness of ethnographic research in sociology, education and related fields, and succeeds in slaying a number of currently fashionable sacred cows. Relativism, critical theory, the uniqueness of the case study and the distinction between qualitative and quantitative research are all examined and found wanting as a basis for informed ethnography. The policy and political implications of ethnography are a particular focus of attention. The author compels the reader to reexamine some basic methodological assumptions in an exciting way", Martin Bulmer, London School of Economics.

Part I
Ethnography, Theory and Reality

1
What's Wrong with Ethnography?
The myth of theoretical description

In the past thirty years, ethnography and other forms of qualitative method have moved from a marginal position in many social science disciplines towards a much more central place. Of course, in the case of social and cultural anthropology ethnography has always been the predominant method; but it now has a strong presence in sociology and social psychology, as well as in applied areas like education and health. While this shift in methodological opinion among social researchers is to be welcomed in some respects, it is important to emphasise that ethnography offers no immediate solution to the problems that currently face social research. Indeed, I shall argue in this chapter that it suffers from a disabling defect itself: it is guided by an incoherent conception of its own goals.¹

The rationale for ethnography is based on a critique of quantitative, notably survey and experimental, research. The validity of this research is challenged on a number of grounds:

1. That the structured character of the data collection process involves the imposition of the researcher's assumptions about the social world and consequently reduces the chances of discovering evidence discrepant with those assumptions.
2. That making claims about what happens in 'natural' settings on the basis of data produced in settings that have been specially set up by the researcher - whether experiment or formal interview - is to engage in a largely implicit and highly questionable form of generalisation.
3. That to rely on what people say about what they believe and do, without also observing what they do, is to neglect the complex relationship between attitudes and behaviour; just as to rely on observation without also talking with people in order to understand their perspectives is to risk misinterpreting their actions.
4. That quantitative analysis reifies social phenomena by treating them as more clearly defined and distinct than they are, and by neglecting the processes by which they develop and change.
5. That quantitative analysis assumes that people's actions are the mechanical products of psychological and social factors, thereby neglecting the creative role of individual cognition and group interaction.

This is not an exhaustive list of the criticisms directed at quantitative research by ethnographers, but it reveals the main assumptions underlying advocacy of qualitative method: that the nature of the social world must be discovered; that this can only be achieved by first-hand observation and participation in 'natural' settings, guided by an exploratory orientation; that research reports must capture the social processes observed and the social meanings that generate them. On the basis of these assumptions, ethnography is directed towards producing what are referred to as 'theoretical', 'analytical' or 'thick' descriptions (whether of societies, small communities, organisations, spatial locations or 'social worlds'). These descriptions must remain close to the concrete reality of particular events, but at the same time reveal general features of human social life.

The claim to integrate description with theory is one of the most distinctive characteristics of ethnography. In this chapter I shall argue that this concept of theoretical description is problematic. It can be interpreted in several ways, all of which raise difficult questions. Furthermore, I believe that ethnographers' adoption of the goal of theoretical description has led to a fundamental misconception about the nature of the descriptions and explanations they produce. I will examine each aspect of this dual goal of ethnography in turn.

Ethnography as Theory

While ethnography places great emphasis on description, it claims to offer a distinctive kind of description: theoretical description. But the nature of this distinctiveness is not very clear. On the one hand, descriptions cannot be theories. Descriptions are about particulars (objects and events in specific time-place locations), whereas theories are about universals (relations between categories of phenomena that apply wherever those phenomena occur). On the other hand, all descriptions use concepts which refer to an infinite number of phenomena (past, present, future and possible). And all descriptions are structured by theoretical assumptions: what we include in descriptions is determined in part by what we think causes what. In short, descriptions cannot be theories, but all descriptions are theoretical in the sense that they rely on concepts and theories.

Given this, what sense can we give to ethnographers' claims that their descriptions are distinctive in being theoretical? There seem to be several interpretations of this idea to be found in the ethnographic literature. I shall examine the following possible rationales, treating theoretical descriptions as:

1. insightful descriptions;
2. descriptions of social microcosms;
3. applications of theories;
4. developments of theory through the study of crucial cases.

These rationales are often not clearly distinguished in the ethnographic literature, nor are they necessarily mutually exclusive.

i Insightful descriptions

nhe title of Everett Hughes's collection of papers The Sociological Eye nicely captures the dual emphasis of ethnography. The analogy with vision indicates the priority given to description, while the adjective 'sociological' signals (though does not specify unambiguously) the distinctive character of the view offered. For many, the value of Hughes's work lies in the provision of insights coming from the use of analogies and metaphors embodied in sociological concepts. In his introduction to the book, Hughes comments:

when people say of my work ... that it shows insight, I cannot think what they could mean other than whatever quality may have been produced by intensity of observation and a turning of the wheels to find a new combination of the old concepts, or even a new concept.

(Hughes 1971: vi)

From this point of view, the aim of ethnographic description is to present phenomena in new and revealing ways. Much of Hughes's work in the sociology of occupations, for example, involves analogies between high and low status occupations that are designed to display the similar concerns, problems and strategies of people who are conventionally thought to be very different: the doctor and the prostitute, the SS guard and the social worker etc. Analogies also play a central role in the writings of Erving Goffman: for example, social life as drama or as a compendium of games. In addition, he deluges his readers with less comprehensive concepts that cast new light on familiar features of our everyday social world. One of the most strikingly effective, of course, is the concept of 'total institution', by which Goffman draws parallels between life in such diverse institutions as mental hospitals, prisons, monasteries and ships (Goffman 1961).

There is no doubt about the suggestiveness of the work of Hughes and Goffman. Their writings lead us to see things differently, to see possible parallels and links that we had not noticed; and, perhaps most important of all, they enable us to free ourselves from those frameworks that we employ so routinely that we have come to take them for reality. But we must ask: on what basis are we to assess candidate insights? Are insights simply of value in themselves, or are they valuable because they are useful for other purposes? More precisely, are insights the end products of social research or does their value derive, for example, from their role as potential hypotheses that can be developed and tested by more systematic and rigorous research?

If the insights are intended as end products with their own intrinsic value, then this represents adherence to a conception of social research that is discrepant not only with that of quantitative researchers, but also with that expressed by many ethnographers.² At the very least it seems that such a view would abandon the idea that the representational validity of candidate descriptions, explanations and theories needs to be tested. Instead, the value of accounts is taken to depend on their novelty, aesthetic appeal or political appropriateness. Ethnographers sometimes seek to justify their work on the grounds that it challenges taken-for-granted assumptions. But are all novelties to be valued? Should all taken-for-granted assumptions be questioned? Surely not.³

Lofland (1988) has suggested that interactionist ethnography is wedded to a kind of political anarchism, a green philosophy and politics that is at odds with its blue and red competitors. And it is true that a strong theme in much ethnographic research is what we might call an urbane romanticism that celebrates the diverse forms of rationality, skill and morality to be found among ordinary people, including (indeed especially) those who are conventionally regarded as irrational and/or immoral. But if political advocacy is the function of ethnography, why is the politics so rarely made explicit? And on what basis are we to distinguish between ethnographic insight and political prejudice? Furthermore, what distinctive role can ethnography play in political advocacy? Is it anything more than an effective rhetorical device?

Of course, ethnographic insights need not be tied to any particular political view; they may simply be offered for others to use however they wish. The rationale here might take something like the following form:

The purpose of ethnographic analysis is to produce sensitising concepts and models that allow people to see events in new ways. The value of these models is to be judged by others in terms of how useful they find them: they are not intended to be theories in the conventional sense of the term, allowing prediction and control; or to represent privileged information of any kind. Rather, they are simply contributions to a public dialogue that should compete on equal terms with those from other sources. The task of the ethnographer is to add to our general store of sensitising concepts and models. Perhaps we might argue that additions or modifications to this stock are necessary as social conditions change, since such changes modify the appeal or usefulness of the models. There is no idea here of theories and concepts being progressively developed to approximate correspondence with reality. Models are used as and when they are found to be appropriate. And, indeed, aesthetic criteria might be as important if not more important than cognitive ones.

There are problems with this rationale. Does it assume that theories of a conventional kind are not viable in the domain of human social life, and that both factual and value claims must be judged solely in terms of their pragmatic utility or their market appeal? If so, on what grounds? And on what basis is it believed that consumers will, and should, evaluate the products of research favourably? We must also note that it is not at all clear that ethnographic research bears any necessary relationship to the production of insights. Might these not be produced on the basis of experience, or even armchair reflection, rather than research? And, indeed, is not this rationale very close to that of literature? If so, what is distinctive about ethnography, and why do we need it?

There are then, many questions surrounding the idea that theoretical descriptions are insightful descriptions. These need to be answered if this version of ethnography as theory is to be convincing.⁴

ii Theory as the description of social microcosms

A related argument for the theoreticity of ethnographic descriptions is the claim that the phenomena described, though apparendy particular and unique, exemplify universal social processes. The aim of ethnographic investigation, then, is to find the general in the particular; a world in a grain of sand. General features of human social life are to be illuminated through the description of particular events. Indeed, it is sometimes argued that it is only by means of this inductive, or discovery-based, approach that generic social processes can be understood. This strategy is contrasted with the hypothesis-testing character of much social research.

This is an idea that can be traced back to nineteenth-century historicism, and beyond to German romanticism. The historicist movement emphasised the diversity and particularity of human social life, and opposed those tendencies in philosophical and social scientific thinking that sought to understand particular social phenomena by applying abstract ideas about universals. Thus, they rejected both positivist attempts to develop natural sciences of human society, whose findings were universal laws parallel to the laws of nature, and Hegelian interpretations of history in terms of the progressive realisation of a universal rational spirit (Hammersley 1989a).

The historian Ranke was a particularly influential exponent of historicisrn. The overwhelming emphasis in his work is on the careful description of historical events, relying on the interpretation of written documents, 'to show only what actually happened' (Krieger 1977: 4—5). But at the same time Ranke distanced himself from history as mere antiquarianism. Ultimately, for him the goal of history was to detect 'the hand of God' behind or within particular historical events (Krieger 1977: 26-7, 137). Ranke believed that intuition, and perhaps even a certain genius, were required to discover the historically universal. Furthermore, the truths discovered might not be expressible in abstract form: at one point he comments that the universal is 'to be narrated not proved' (Ranke, quoted in Krieger 1977: 137). None the less, for Ranke the study of historical events only gained its value from its discoveries about the universal, and the spiritual and practical value that these had.

These ideas were developed further in Dilthey's attempt to discover the methodological foundations of the human studies. For example, he believed that through the study of 'representative individuals' we could discover general types of human individual. He regards this identification of the essential within the particular as characteristic of great art and literature:

The characters and situations presented to us in a poem, and the feeling-responses evoked by them in us, are typical of a segment of possible human experience, and of its value or significance to us. By contemplation of the type, our acquaintance with what it represents is widened, and our power to see its true significance is heightened. Art therefore, no less than science, but in a very different way, is a vehicle of truth.

(Hodges 1952: 113)

For Dilthey, the human studies partake of aspects of both science and art. They share art's goal of evoking the typical, but pursue it more systematically, seeking factual as well as symbolic truth. Furthermore, they are concerned not just with the study of individuals, but also with investigating cultural and social systems. Here too, though, the general is to be discovered through the study of the particular.

Something like this approach, the study of particular cases as social microcosms, is present in ethnographic thinking today. Ethnographic work often seems to amount to a celebration of the richness and diversity of human social life, but at the same time seeks to identify generic features. Ethnographers' commitment to the urbane romanticism that I mentioned above often serves an important function here. They trade on the significance that this perspective gives to descriptions of what otherwise might be regarded as fleeting and trivial events occurring in unimportant places. But even apart from this, the value of ethnographic work often depends on showing that the particular events described instantiate something of general significance about the social world.⁵ Furthermore, as with Ranke's conception of historical work, the aim is to do this through narration, rather than abstract demonstration. Thus, Krieger (1983: 179) comments approvingly on Street Corner Society:

Whyte's stories ... create his theory and are not just evidence for it. We understand his explanation of racketeering and, more basically, his theory of social organization, not because we are told its tenets in abstract or 'theoretical' terms (although we are told some of it in these terms in the end), but because we are led through a world in which we can develop an experiential sense of the way that individual actions relate to each other and to a larger whole.

Atkinson argues that this discursiveness is a characteristic feature of ethnographic writing. He suggests that the instances that ethnographers describe can be taken to stand for general cultural themes or social types, and that this process of representation is semiotic rather than statistical (Atkinson 1990:83-4). Like Krieger, he takes Street Corner Society as a key example, commenting that while some may believe that Whyte 'sidesteps a great many sociologically important questions', it is reasonable for a reader to conclude that the as...

Table of contents

• Cover
• Title
• Copyright
• Contents
• List of figures
• Acknowledgements
• Introduction
• Part I Ethnography, theory and reality
• Part II Ethnography, relevance and practice
• Part III Qualitative versus quantitative method
• Postscript
• Bibliography
• Name index
• Subject index