Copyright Sociological Research Online, 1999


Max Travers (1999) 'Qualitative Sociology and Social Class'
Sociological Research Online, vol. 4, no. 1, <>

To cite articles published in Sociological Research Online, please reference the above information and include paragraph numbers if necessary

Received: 05/11/98      Accepted: 23/3/99      Published: 31/3/99


This paper contrasts two approaches that qualitative researchers can adopt towards studying class and status divisions, drawing upon issues raised by Gordon Marshall (1988) in his paper about working class consciousness. It is suggested that researchers influenced by Marshall, and recent feminist ethnographers, whose central concept is class, ultimately adopt a competitive stance towards common-sense understanding and experience. Sociologists who seek to describe how members of society understand their own activities, such as the community studies tradition in anthropology, Pierre Bourdieu, and ethnomethodology, often conceptualise class in terms of status. These different ways of understanding qualitative data need to be understood in the context of foundational debates in nineteenth century sociology about action and structure, and indicate the continuing relevance of the Marx/Weber debate in discussions about social class.

Class; Discourse; Ethnography; Ethnomethodology; Life-History; Qualitative; Stratification


Perhaps the greatest achievement of British postwar sociology has been the cumulative programme of research, methodological debate and investigation, using mainly quantitative methods, which has been pursued by three generations of neo-Weberian researchers on social class. This has recently been attacked on two fronts, firstly by postmodernists who dispute the existence and relevance of social class (Joyce 1995; Pakulski and Waters 1996),[1] and secondly by feminist researchers, who have argued that quantitative methods cannot address how women understand and experience class in everyday life (Skeggs 1997; Reay 1998a, 1998b; Mahony and Zmroczek 1997). According to Diane Reay, 'what is required are British-based ethnographic examinations of how class is "lived" in gendered and raced ways to complement the macro versions that have monopolised our ways of envisaging social class for far too long' (Reay 1998a, p.272).

In this paper, I want to draw attention to some methodological issues that are relevant for researchers doing qualitative research in this field. I will begin by examining the (1988) arguments made by Gordon Marshall in favour of using 'sociological ethnography' to investigate social class. I will suggest that this is similar, in some respects, to the methodological argument made by Herbert Blumer in his critique of quantitative sociology in the 1950s, although Marshall does not think through the more radical implications of that critique for class analysis. The central principle informing the programme of research advocated by Blumer is that the analyst should respect and address the actor's point of view, which requires not simply employing qualitative rather than quantitative methods, but giving up the whole set of methodological assumptions that quantitative analysts use to describe and explain the causes of human actions. Marshall, on the other hand, wants to use qualitative methods to address questions in the same way as the quantitative traditions he is critiquing: the kind of investigation he proposes is based on the premise that the analyst knows more about the structural forces shaping our actions than ordinary members of society. The distinction between these two approaches, which are rooted in foundational debates about epistemology and method in sociology, will be a central theme in this paper, since they lead to different ways of conceptualising the study of social class, and conducting qualitative research.

In the second section in the paper, I will review some British studies which have taken up Marshall's agenda. I will focus on Fiona Devine's (1992) qualitative research on affluent workers in Luton, and on two feminist ethnographies by Beverley Skeggs (1997) and Diane Reay (1998b). I argue that, despite the best intentions of their authors, these studies only tell us a limited amount about the way class is 'lived'. This is because, although they are sensitive towards the problem of how to relate lay and analytic perspectives, they ultimately adopt an ironic or competitive stance as analysts (Watson 1998) towards how ordinary members of society understand the issue of class.

In the third section of the paper, I will discuss two approaches to class that adopt a non-competitive or explicative stance towards common-sense understandings and knowledge. These are the community studies tradition in anthropology, and Pierre Bourdieu's (1996; first published 1979) Distinction which looks at the cultural practices constituting what he terms the habitus of different social groups. Status, rather than class, becomes the central analytic concept for these traditions.

The fourth and final section of the paper sets out a proposal for pursuing qualitative research on class and status divisions from an ethnomethodological perspective. I suggest that what distinguishes ethnomethodological research is not so much method, since any tradition can employ qualitative methods, but a commitment to the principle that any feature of social life must be demonstrably relevant to social actors in particular courses of action. My own preference, however, would be for ethnographies in a number of sites, modelled on the community studies tradition. This could form part of a wider programme of research investigating different dimensions of stratification in contemporary Britain.

Gordon Marshall's Critique of Quantitative Class Analysis

Gordon Marshall is an unusual figure in British class analysis in that he has written at least one paper which is highly critical of quantitative research, and argues that 'sociological ethnography' offers a 'more appropriate and productive research strategy for the study of class consciousness' (Marshall 1988, p.124). This paper is worth considering since it raises a number of important methodological issues, which are relevant to anyone planning to undertake a qualitative study of social class.

Marshall retains the same problematic as other researchers ('why, in advanced capitalist societies, have working classes not become revolutionary classes'), but argues that theories of working class 'instrumentalism' or 'ambivalence' were not 'substantiated by the survey and attitudinal data from which each has been derived' (Marshall 1988, p.98). This is because the variables used in these studies - such as class consciousness or class action - are vague and ill-defined, and because survey methods fail to address the complexity of class consciousness. Instead, he proposes that

... research energies and resources should be channelled in the direction of intensive, longitudinal ethnography, in which different aspects of consciousness are located firmly in the context of class practices - everyday work in the factory, leisure time at home and in the club, the recent protest, the strike - and conceptualised at the outset as inherently dynamic phenomena. (Marshall 1988, p.121).

Although it is not cited by Marshall, a similar argument against quantitative methods can be found in a 1969 (originally published 1956) paper by the American symbolic interactionist Herbert Blumer (1969) called 'Sociological Analysis and the "Variable"'. Blumer suggests that quantitative analysis is an inappropriate method for studying human beings, and can never produce meaningful findings other than bare statistical correlations. This is firstly because variables can never be precisely defined, as they can in the natural sciences, but also because human life involves an 'interpretative process', rather than a behavioural response to variables. He asks, for example,

... how can variable analysis include the process of interpretation? Presumably the answer would be to treat the act of interpretation as an 'intervening variable'. But what does this mean? If it means that interpretation is merely an intervening neutral medium through which the independent variable exercises its influence, then, of course, this would be no answer. Interpretation is a formative or creative process in its own right. It constructs meanings which ... are not predetermined or determined by the independent variable. (Blumer 1969, p. 135).

Marshall also notes that 'consciousness is ... an integral component of social action, rather than a distant something that somehow causes or is caused by it'. This leads him to a radical conclusion about the methods that should be used in class analysis. Because Marshall is normally associated with research based on social surveys, and one might, therefore, expect this paper to advocate some kind of mix between qualitative and quantitative methods, they are worth quoting in full:

Is this simply to reiterate the familiar plea for a proper balance between the validity sought through intensive fieldwork techniques and the reliability guaranteed by statistically sophisticated social surveys? I think not. It is difficult to see how the shortcomings of social surveys, as a means of studying class consciousness, can be overcome simply by a more thorough piloting of survey questions or a more careful aggregation of survey data. It is not that qualitative research of the kind here advocated complements surveys in this field. Rather such research provides a more suitable alternative to the large scale survey, and it generates data that the latter are incapable of uncovering no matter how much care is taken in piloting the survey and in compiling the questions. Survey research cannot report on action directly observed; consciousness can only be studied contextually as a component of class practices; therefore surveys are inappropriate to the study of class consciousness. (Marshall 1988, pp. 122-123).

Although there are some similarities between the methodological arguments of Marshall and Blumer in their advocacy of ethnography and critique of quantitative methods, there are also important differences in the way they theorise action and consciousness. Blumer did not specifically write about the methodological issues involved in class analysis, but it can be regarded as part of his more general interest in how individuals interpret the meaning of objects in the social world. Objects can include 'physical objects, such as trees or chairs; other human beings, such as a mother or a store clerk; categories of human beings, such as friends or enemies' and also 'guiding ideals, such as individual independence or honesty' (Blumer 1969, p. 2). A symbolic interactionist approach to social class would, therefore, involve addressing how people use the concept of class in understanding the world around them. Blumer's writings on racial prejudice (Blumer 1959; Blumer and Duster 1980) suggest how he might have conceptualised the study of working class consciousness as a collective perception of shared economic interest. This perception is not static, but can change in unpredictable ways depending on how people interpret fresh events.

One important methodological principle, is that the analyst should avoid imposing theoretical categories or problematics on qualitative data (Blumer 1969, p.48). Instead, analysis should proceed inductively, a method that was later formalised by Barney Glaser and Anselm Strauss (1967) in their text on grounded theory. If, therefore, those being studied do not understand their actions in collective terms, or even think of themselves as members of a class, then sociologists should not use the term in describing their activities.

Marshall is also critical of analysts who bring prior theoretical and political commitments to data, and advances, what might seems on a first reading, to be a purely methodological case in favour of using qualitative methods to research working-class consciousness. The paper also suggests, however, that sociologists in the 1970s (including those in the Marxist tradition) underestimated the extent to which the working class was still capable of becoming a revolutionary force, given the right economic conditions.[2] In the context of his other work, it seems likely that Marshall hoped that qualitative research would reveal a reality, more conducive to his own political viewpoint, by improving upon the findings of surveys.[3]

I hope to have shown that Marshall and Blumer do differ considerably even though each advocates the use of qualitative methods. Blumer views these as a means of explicating the actor's point of view, without judging this from the superior vantage point of the analyst. Marshall, however, either does not see a problem in reconciling the potential conflict between sociological theories about class, and 'consciousness ... as an integral component of social action', or views qualitative research as a means of confirming his own theoretical and political views about the working class.

Whichever interpretation of Marshall is correct, the distinction between 'competitive' and 'explicative' approaches towards common-sense knowledge is a useful device that distinguishes traditions with different epistemological and methodological assumptions in this field. It is also a dualism that successive theorists claim to have bridged or transcended, but which remains stubbornly present in sociological analysis. In the rest of this paper, I will be contrasting qualitative approaches towards class that represent the two sides of the dualism. I will begin by looking at some recent British qualitative studies about social class. I will then contrast these 'competitive' approaches with 'explicative' ethnographic and qualitative studies, which adopt a similar methodological stance to Blumer, and conceptualise class as status.

'Competitive' Approaches Towards Social Class: Some Recent British Studies

Marshall never went on to conduct the kind of studies recommended in his 1988 paper, and most research on stratification in Britain has continued to employ quantitative methods in collecting and analysing data. There has, however, also been a growing interest in the use of qualitative research methods, owing to a dissatisfaction with the increasingly abstract and technical character of research in this field. Rather than giving a comprehensive review of developments, I will focus on three recent studies. My general argument will be that they represent an important development in addressing 'how class is lived', but ultimately become a vehicle for the authors' political views, rather than a thorough-going investigation of working class life or consciousness.

Devine's Research on Car Workers in Luton

One study which has engaged with the theoretical and methodological issues raised by Marshall's paper is Fiona Devine's (1992) Affluent Workers Revisited, which is based on semi-structured interviews with thirty workers employed at the Vauxhall car plant in Luton, and their wives. John Goldthorpe, David Lockwood, Frank Bechhofer and Jennifer Platt conducted a similar study, using structured interviews, in the 1960s (Goldthorpe et al 1969), and discovered that workers were content with their economic and social position. This suggested that a new affluent working class voted for Labour instrumentally, in the hope that it would improve living standards, rather than through the "almost unconditional allegiance" it received in the past (Goldthorpe et al 1969, p. 31).

Devine, who conducted research during the economic recession of the 1980s, found that working class families did have some degree of class-consciousness. A majority of interviewees, for example, revealed that they would prefer a more equal society, although they were disillusioned with the record of Labour in office. A key claim is that these findings were only obtained using semi-structured interviews: the suggestion being that the closed questions used in the Affluent Worker study may have resulted in a distorted view of the working class.

The qualitative approach adopted in Devine's study reveals a wealth of information about the leisure pursuits and family life of these workers, their political views, and their reasons for moving to Luton, which creates substantial problems for Goldthorpe et al's original thesis. Her interview data does not, however, suggest that these workers were any more a potential revolutionary force in the 1980s, than the Affluent Worker study had found to be the case in the 1960s, despite the fact that many were pessimistic about the future of the car industry. Nevertheless, Devine concludes that the modest support expressed for higher taxes on the idle rich indicates that "members of the working class do not normatively accept or endorse the prevailing system", and that they only accept capitalism because of 'everyday economic realities' (Devine 1992, pp. 209-10).

This is a good example of how common-sense knowledge and perspectives are handled by analysts who view consciousness as being shaped or determined by the actor's structural location in society. Devine obtained more complete data about the political views of her informants through using qualitative methods, but she ultimately interprets and explains these from her own theoretical perspective as a sociologist. There is, in my view, no evidence from this study that the working class would prefer, or even can imagine, a different social and economic system to capitalism, despite their experience of redundancies, insecurity and wage-cuts during a period of economic recession.

Feminist Qualitative Research on Social Class

Feminist sociologists have, for some time, felt a sense of grievance towards class analysis, mainly because of John Goldthorpe's continuing insistence that women derive their class identity through the occupation of their partner (which he defended in 1983 on the basis that most women were economically dependant on men). Goldthorpe is one target for three recent feminist books on social class (Skeggs 1997; Reay 1998b; and Mahony and Zmroczek 1998), since they argue that the tables, produced by quantitative researchers from their surveys, do not address how it feels to be a working class housewife (a non-category for the purposes of neo-Weberian analysis). However, they are equally concerned to challenge the view, promoted by postmodernists, and some politicians, that class no longer has any meaning or political significance in contemporary Britain.

Diane Reay interviewed thirty three mothers of pupils in two primary schools, catering to working class and middle-class children, while Beverley Skeggs conducted a longitudinal ethnography maintaining contact with eighty three women over a twelve year period. Both draw upon the work of Pierre Bourdieu, and I will be discussing how his ideas and concepts are employed by these sociologists in the next section of the paper. Here, I again want to focus on how they understand the relationship between lay and analytic perspectives.

Reay notes that a majority of the mothers she interviewed 'did not mention social class until I asked them to identify in class terms' (Reay 1998a, p.269). A significant proportion of her sample from 'objective' working class backgrounds either saw themselves as 'middle-class', or felt that the term had no relevance to their lives. However, they often made a regular contrast between 'people like us' and more privileged groups in society. Reay argues that these responses demonstrate that the attitudes of these women were influenced by their economic position, even though they denied the existence of class.

Skeggs also found that some of the women she interviewed expressed resentment towards wealthy members of society (Skeggs 1997, p. 92). However, most did not describe themselves as 'working class', and seemed to regard this as a way of describing the homeless, certain categories of single-parents, or the unemployed (what some sociologists call the 'underclass'). However, she views this as an attempt to 'pass' as middle-class, and to distance themselves from what Reay, drawing on Goffman (1964) terms a 'spoiled' identity.

Each of these studies listens to what working class women say about their own lives, but ultimately rejects how they understand their own activities. Skeggs expresses gratitude to her informants for making her question the categories of class used in the academy, and is unusually sensitive towards the problem, faced by all ethnographers, of how to reconcile lay and academic perspectives. However, she still insists on 'the centrality of class', even though this is something 'which they consciously try to disclaim' (Skeggs 1997, p. 94). In this sense, feminist sociologists do not differ substantially from the quantitative traditions they criticise, which begin from the premise that class location can be defined without reference to the subjective understandings of actual members of society.

'Explicative' Approaches: Viewing Class as Status

It is perhaps understandable that sociologists whose central problematic is class usually adopt a competitive stance towards common-sense knowledge, since it is hard to find much evidence that members of the working class do understand their lives in class terms. There has, after all, been little class-conflict in Britain, involving widespread violence, or a real threat to the legitimacy of ruling elites, since the first half of the nineteenth century, and even the class basis of Chartism (which some would regard as a revolutionary working class movement) is disputed by many historians.[4]

There is, however, an alternative tradition in sociology which has adopted a different approach towards studying social divisions. Instead of looking for class-consciousness, it has focused on the status divisions that people recognise in their everyday lives. I now wish to discuss two examples of work in this tradition: the community studies conducted by American and British anthropologists earlier this century, and the more recent work of Bourdieu on status divisions in France. I will suggest that each attempts to address and respect how social actors understand their own activities, which can be contrasted with the competitive stance of the class analyst.

The Community Studies Tradition in Anthropology

Marshall discusses a number of ethnographies in his (1988) paper, including the studies of factory life and industrial conflict by Lane and Roberts (1972), Beynon (1973), Kornblum (1972) and Burawoy (1985), and Willis' (1977) study of the reproduction of working class culture in schools. However, he makes no reference to an older body of work which preceded neo-Weberian quantitative research on social class: the community studies tradition in anthropology.

The fact that these studies are now largely forgotten, and not taught on sociology degrees (although see Crow and Allan 1994), illustrates the extent to which the disciplines have grown apart during the twentieth century. The best-known American research is the study of Middletown conducted by Robert and Helen Lynd (1929), and the multi-volume Yankee City series by Lloyd Warner and his associates (Warner et al, 1959a-e).[5] British work in the same tradition includes Margaret Stacey's (1960) study of Banbury, and Dennis et al's (1969) Coal is our Life.

In contrast to the approach adopted by class analysts, a central methodological assumption is that one builds up a picture of stratification inductively by examining the lives of ordinary members of society in the round. The foreword to volume 5 of the Yankee City series faults attempts to find underlying factors which explain social life, in the manner of biologists and psychologists, which are 'not unlike groping behind the scenes and digging under the stage, disregarding the comedies, tragedies and dramas in plain sight' (Warner 1959e, p.v). Instead, it suggests that 'experience with social phenomena is bringing us nearer and nearer to a realisation that we must deal directly with life itself, and that the realities of social science are what people do'. In this literature, the life of different groups in these cities, and their feelings towards other groups, are conveyed through a series of ethnographic vignettes, although Warner also developed a technique to map out local stratification systems through collecting responses to a questionnaire.

The central concept used to organise the materials (which is grounded in how members of these communities understand their own actitivies) is status rather than class. Chapter 8 of the Banbury study, for example, identifies a number of status-groups within the working and middle-classes. We learn about the areas where they live in the town, the life-style of the different groups and what Lockwood (1966) would term their images of society. Although often criticised for their conservative bias, these British and American studies reveal a dimension of stratification which is concealed, or neglected by class analysis.[6]

Bourdieu's Distinction

Arguably the most ambitious, and theoretically sophisticated sociological study about class in recent years is Pierre Bourdieu's (1996; first published 1979) Distinction. This looks in some detail at the cultural practices constituting what he terms the habitus of different social groups in France. It draws upon surveys of taste and aesthetic judgment, government statistics about consumption patterns, and a number of ethnographic vignettes documenting class as a way of life, and set of cultural attitudes.

Bourdieu is usually associated with the structuralist tradition in sociology, and for advancing a critique of western societies, influenced partly by Marx, which would place him in the previous section of this paper as a theorist with a 'competitive' stance towards common-sense knowledge.[7]

This is certainly how his ideas are used by Diane Reay and Beverley Skeggs in their research on social class. A central theme is that people are born into an unjust social system in which stratified social groups have access to different types of economic, cultural and social capital. Reay describes how the middle-class women in her sample drew upon economic capital to improve the prospects of their children by hiring tutors, and social and cultural capital to make greater demands on schools. Skeggs employs the same framework when she suggests that working class women attempt to compensate for their low economic capital by investing in respectable middle-class clothing.

Although this is a legitimate reading of Bourdieu, I would argue that there is another side to his work, particularly evident in Distinction, which is closer to the explicative tradition. In common with theorists like Anthony Giddens and Dorothy Smith, Bourdieu employs ideas from structural traditions in sociology to explain stratification patterns, but he also draws upon interpretive approaches such as ethnomethodology in addressing how these are produced and recognised by social actors. He explains the relationship between the two elements of structure and action in this extract from Distinction:

The cognitive structures which social agents implement in their practical knowledge of the social world are internalized, "embodied" social structures. The practical knowledge of the social world that is presupposed by 'reasonable' behaviour within it implements classificatory schemes...historical schemes of perception and appreciation which are the product of the objective division into classes (age groups, genders, social classes) and which function below the level of consciousness and discourse. Being the product of the incorporation of the fundamental structures of a society, these principles of division are common to all the agents of the society and make possible the production of a common, meaningful world, a common-sense world. (Bourdieu 1996, p. 468).

This is written by a structuralist, who sees cultural distinctions as part of 'the fundamental structures of society', but Bourdieu also wishes to address the 'practical knowledge' that social actors use in producing and recognising social divisions.

Although it is difficult to separate the objectivist and subjectivist tendencies in Bourdieu (what I have been calling 'competitive' and 'explicative' approaches in sociology), I would argue that Distinction can be read as a non-judgmental study of French society, which seeks to address how people understand and produce social divisions and identities. In common with other researchers in the explicative tradition, Bourdieu views status as a more important category than class. There is also no suggestion that people in lower social groups resent those with higher status, or oppose the legitimacy of the entire system. This can be contrasted with the approach of many British sociologists, influenced by Marx, who have little interest in status divisions, and adopt a competitive stance towards common-sense knowledge.

Ethnomethodology and Social Class

In the remaining part of my paper, I want to discuss how the research tradition of ethnomethodology can be used to study social class. This is worth considering, because ethnomethodology has a more thorough-going commitment towards addressing the actor's point of view than most 'explicative' approaches in sociology, including symbolic interactionism. I will begin by considering how ethnomethodologists approach the study of stratification, drawing on a discussion by Mathew Speier (1973). I will then discuss some ways to investigate class and status from this theoretical perspective.

Ethnomethodology and stratification

Ethnomethodology is not usually associated with the study of social class, or stratification generally, although I have suggested that Bourdieu does address the production and recognition of status divisions along ethnomethodological lines in Distinction. One of the few discussions by an ethnomethodologist can be found in Mathew Speier's (1973) How to Observe Face-to-Face Interaction.[8] He suggests that the normal strategy of the sociologist is to use some category (such as income or occupation) to identify a group of people who form a social class:

However, the sociologist's lumping of people into various socio-economic categories of membership tells us very little about the structure and process of everyday life. Rather it is the lumping that everyday participants do that tells us that. Using common-sense we can entertain the idea that this or that category is applicable to this or that person in society....But what we want to discover, so as to come out with news about society's workings, is precisely how members decide for themselves what social categories of membership they are in at any interactional moment in their daily intercourse. (Speier 1973, p. 185).

This way of conceptualising class is sometimes confused with the view expressed by some postmodernists, and right-wing commentators, that class does not exist, because so many other categories may be important for social actors in particular situations (Pakulski and Waters 1996, p. 196). Saunders asks, for example,:

Are we seriously to believe that in their everyday lives people think of themselves as members of a class, rather than, say, as British, as parents, or as white or black...? On holiday in Spain we feel British, waiting outside the school gates we know we are a parent....and answering questions, framed by sociologists with class on their brain, we are working class. (Saunders 1989, pp. 4-5).

Speier also suggests that

... it may actually turn out that the concept of class has to be either discarded or reserved for political rhetoric, or at the very most, reserved for a common-sense short-hand method of referring to vast collections of persons who are categorised for the convenience of a theoretical argument, and categorised from without their own interactional patterns of life. (Speier 1973, p. 186).

This is, not, however, a wholesale dismissal of the concept of class, but draws attention to 'a whole world of identification techniques and behaviours that must be consulted in the natural activities of members themselves'. It would be more correct to view ethnomethodology as having an agnostic attitude to class, or for that matter status divisions, in that their relevance has to be demonstrated for members of society in actual situations. The objective should be to address the interpretive and communicative methods used by members of society to display and recognise class and status in everyday life.

Addressing Class as an Ethnomethodologist

How then might one pursue an ethnomethodological study of class? One of the interesting features of this tradition (which makes it distinctive, although by no means unique, in the human sciences) is that there is little, if any, emphasis on research methods. Instead, Garfinkel and Sacks provided a set of maxims, or methodological principles that can assist the researcher in viewing aspects of the world that we normally take for granted as research topics. These include the principle of 'ethnomethodological indifference' towards moral and political questions (which are treated as matters for members rather than the analyst), and the study policy which invites researchers to view social facts as accomplishments.[9]

For this reason, there might be any number of ways in which one could pursue a study of class from this theoretical perspective. My own preference would be for four linked projects which attempt to address different dimensions of class as a lived phenomenon.

(1) An auto-ethnographic study

One technique recommended by Harvey Sacks is to begin an ethnomethodological investigation by describing something mundane or obvious in the world around us (Silverman 1998). This can be done without collecting interview data, or conducting ethnographic fieldwork, and is a good way to introduce undergraduates to this sociological perspective. We can see class around us in innumerable ways in the daily round of our everyday lives: in Britain it is still relatively easy to identify someone precisely in class terms by their dress, the newspaper they read, their diet, their accent, or whether they have private health insurance.[10] This offers a similar way of researching class to Bourdieu (and he discusses this at length in Distinction), but an ethnomethodologist would go further in describing how and when this classification gets done, and how it is relevant to our interactions with other people. This kind of investigation would reveal a great deal about the shared cultural knowledge which is used as a taken-for-granted resource by other sociological traditions when they research social class.

(2) Ethnographic research

Another project could involve spending time in a number of settings in which one would expect to find members of a particular social class. These could include: a public school, an Oxbridge College, a polo club; a stately home; a gentleman's club; a working man's club; a 'working class' home; a 'middle-class' home; a Rotary club; an upmarket health club; a factory shop-floor; and a picket line during a protracted industrial dispute.[11]

Here one might begin by interviewing informants, in an open-ended way, about their biographies, and their practical concerns in that setting. The interviewer could invite them to expand on anything that seems relevant to class or status divisions. Ethnomethodology adopts a distinctive perspective towards the research interview, in that the data is seen as a collaborative product of the interaction between interviewer and interviewee, rather than a means of obtaining reliable or valid information about the social world.[12] For this reason, it would also be important to conduct naturalistic observation over an extended period: the assumption is that, if class and status divisions are relevant, then people will talk about them in everyday situations.

Without prejudging the findings of such a study, one might expect 'respectable' citizens in a working class community to display their respectability by making derogatory remarks about the 'rough' working class; for such attributes as taste, educational level, occupation, income, accent and values to matter in middle-class families and institutions; and for the upper class to know the difference between 'old' and 'new' money. The objective would be to collect a corpus of data about how people in different social groups use categories and distinctions, which would update the findings made by community studies in the 1950s.

(3) The Experience of Mobility

One of Garfinkel's best known papers is his (Garfinkel 1984; first published 1967) study of Agnes, a transsexual, who had been brought up as a boy, but had then tried to live as a woman in his/her early twenties. What is methodologically interesting about the paper is how taken-for granted aspects of having a female gender are revealed through Agnes' efforts to pass. These included, for example, knowing how to cook like a woman, walk like a woman, talk like a woman, and share female experiences, a list which Garfinkel suggests could be extended indefinitely.

A similar technique which could be employed in researching class might be to focus on the experiences of those who have joined a different class. Many of the contributors in the Mahony and Zmroczek (1997) collection complain that they were required to change their dress, diet and accent in being accepted as middle-class. They were also made to feel ashamed of their working class origins.[13] This only tells us how a small group of people have experienced social mobility, rather than about working class life in general. However, those who do move between classes obtain an insight into the practical, interactional work involved in being a member of a class that is taken for granted by ordinary members of society. It would be interesting to pursue a project that attempted to identify different responses to social mobility.

(4) Professionals and Class

Perhaps the easiest and most productive project one might pursue as an ethnomethodologist, would be to look at groups with a technical or professional interest in class. Here there have already been a number of interactionist and ethnomethodological studies which show how teachers differentiate between children in class terms (for example, Rist 1970). It would also, however, be interesting to look at other groups whose day-to-day work involves the measurement and identification of class divisions, such as market researchers and their clients. Bourdieu uses the data produced by these occupations to support a sociological analysis of status and class divisions. An ethnomethodologist, on the other hand, would want to know how the work of classifying gets done, and how class features as a practical matter in the worlds of commerce and politics.

A Qualitative Future for Class Analysis?

As Marshall (1997) has recently suggested, there is no reason to suppose that the quantitative tradition in class analysis has run out of steam. There continues to be a healthy level of methodological debate among specialists about different techniques for analysing statistical data, and new problems to be overcome in comparative studies of social mobility or patterns of social inequality. Governments also continue to require some means of measuring the distribution of property and wealth, and research councils will continue to fund teams of researchers to conduct large-scale social surveys, and analyse the data using quantitative methods.

Many will, however, continue to feel that there is also a need for a programme of research that employs qualitative methods in researching social class. Although this can be presented as a debate about methods, I have tried to show that more fundamental epistemological issues divide researchers working in this field.[14] These can all be related back to methodological debates in nineteenth century sociology about action and structure, and indicate that the Marx/Weber debate (although it might be regarded as old hat by some sociologists) is still highly relevant in discussions about social class.

Although Weberians continue to use status as a central category (for example, Lockwood 1992; Scott 1996; Crompton 1998; see also Savage et al 1992), most debate and research in Britain, at least in recent years, has been concerned with arguments about class. This has been, and continues to be, premised on the assumption that structures exist independently of how people understand their own position in society, and the analyst ultimately adopts a competitive stance towards common-sense knowledge and experience. Marshall arguably saw qualitative research methods as a means of finding evidence of working class consciousness that was concealed by survey research. Devine's study of car-workers in Luton, and the two feminist ethnographies I have discussed, are based on similar assumptions.

I have tried to show that the community studies tradition, and also the work of Bourdieu have much to offer in conceptualising class as status, and that ethnomethodology offers a useful way of investigating how members of society understand class and status divisions. These approaches differ from most British studies in adopting an explicative rather than competitive stance towards common-sense knowledge. Although they offer fewer resources for advancing a political critique about class or status inequalities, their methodological assumptions allow them to address dimensions of human life and everyday practices which are neglected or idealised by competitive approaches.

One central issue in the Marx/Weber debate concerns whether sociologists should aspire to develop theoretical explanations which are superior to those of social actors, and enable sociology to have a political role in addressing inequality and injustice. Layder (1998) contrasts the Marxist/Durkheimian view that structures have an independent existence shaping and constraining our actions, with what he views as the limited approach of interpretive traditions such as ethnomethodology and poststructuralism which never advance theoretically beyond members' understandings.

Harvey Sacks might respond that sociology cannot progress as a scientific discipline until it becomes more serious about describing phenomena such as class and status divisions (Silverman 1998). This does, not, however, mean that interpretive sociologists cannot contribute to debate about moral and political questions. Weber suggested that most commentators avoid confronting 'inconvenient' facts, or thinking critically about their own political assumptions (Weber 1991, p. 147). It could be argued that a great deal of contemporary sociology is vulnerable to this criticism, and has very little real political content, other than a romantic identification with the underdog. My own (Travers forthcoming, July 1999) study of immigration control is intended to cause politicians and others to think more critically about this area of policy, by describing a range of institutional and practical perspectives which are idealised by structural approaches (see also Travers, 1997). In the case of stratification, it might be argued that sociologists need to do more than simply identify with subordinate groups like the 'working class'. To be taken seriously outside universities, we also need to develop a convincing analysis of the causes of economic inequality, and a political programme which engages with the electorate.

It remains an open question whether a dissatisfaction with quantitative analysis will result in more researchers conducting qualitative research on social class. I would argue that the feminist ethnographies discussed in this paper have made a promising start, but that more could be done by drawing upon a wider range of traditions. It goes without saying that such investigations should not be confined to the study of the working class, and may result in new ways of conceptualising other dimensions of stratification, such as ethnicity, gender, age and locality.

Although I have proposed some ways of researching class from an ethnomethodological perspective, I would like to see sociologists from a range of traditions using qualitative methods to explore class and status divisions. This agenda might interest those who feel that quantitative analysis has become a dry technical exercise and is too removed from how people understand their own lives. As Prus (1996) suggests, it also offers a means to combat the philosophically-driven excesses of postmodernists, whose thesis about the death of class is also at odds with our everyday experience.


1 For a discussion of the postmodernist literature, see Bradley 1996.

2 Much of Marshall's work is informed by a commitment to Keynsian economics and social justice (see, for example, Marshall et al 1988, and Marshall et al 1997).

3 Marshall was responding to numerous studies, conducted during the 1960s and 1970s, which had investigated working-class consciousness, mainly using survey methods (see, for example, Bulmer 1975 and Newby 1977). The inspiration for these studies was Lockwood's (1966) paper on images of society, in which he suggested that different groups of workers developed particular values as a result of their structural location in the economy.

4 History, like sociology, is more about interpretation, and argument, rather than discovering the facts, and a Marxist historian would have no difficulty in demonstrating the existence of class conflict. The biggest critics of traditional Labour history in recent years have been poststructuralists (see, for example, section F of Joyce 1995), although they owe rather more to Weber than Derrida in their approach to the past. Joyce (1990) argues, along Weberian lines, that nineteenth century British society is best understood using the concept of status rather than class.

5 What is extraordinary about these studies is their scale and ambition when compared to the projects that get funded today. Each involved a team of researchers living in a city for several years, and producing several volumes of findings about all aspects of urban life.

6 The conservatism of these studies derives from the fact that they were writing about small towns in periods of economic prosperity, although Warner believed that America benefited from economic inequality, provided there were no barriers to social mobility (see also Davis and Moore 1945).

7 See, for example, Jenkins (1992).

8 Other ethnomethodological papers about social class include Gobo (1995) and Coleman (1996); see also Hiller (1973).

9 For an introduction to these methodological principles, see Heritage (1984).

10 Remarkably, there are some commentators and politicians who believe that this is no longer the case, including the current prime-minister Tony Blair who has recently suggested that most people in Britain are middle-class (The Guardian, 15th January 1999, p.3). However, the fact that class is constantly discussed by the media, and contested in public life, is one of the ways in which it is part of everyday life. For some compelling evidence about how class distinctions still matter, see Adonis and Pollard (1997).

11 Perhaps the most promising avenue of investigation, if one wanted to find class consciousness, might be to study an industrial dispute. An ethnomethodologist would, however, adopt a different approach to the class-analyst in that the objective would be to understand the perspective of workers in these conflicts, rather than using them as a vehicle to make a political point about working class consciousness. For an example of a study by a class-analyst, see Fantasia (1988) which makes a general case about the potential radicalism of American workers, through describing a wild-cat strike. See also Fantasia (1995), reviewing research on class-consciousness in America, which draws partly upon Marshall.

12 For a discussion of the issues involved in analysing interview data from an ethnomethodological perspective, see Silverman (1993), chapter 5.

13 The psychological predicament of the parvenu - the person who has moved into a class above his or her original station - has inspired many films and novels. This is one case of what the Chicago School sociologist Everett Stonequist (1936) called 'the marginal man': a general term for people who find themselves 'caught between two cultures'. One example Stonequist uses are Polish and Jewish immigrants to America in the 1920s and 1930s who were under pressure to assimilate. They too were made to feel ashamed about their ethnic origins (for example, their accent, or religion), but were also ashamed to feel ashamed. For studies which address the shame felt by working class people, see Goffman (1964) and Sennett and Cobb (1972).

14 For an interesting discussion of how different sociological traditions understand the 'qualitative'/'quantitative' issue, see Halfpenny (1979). One point he makes is that sociologists handle qualitative data in different ways depending on their epistemological assumptions. Marshall, like many British sociologists, has a broadly positivistic orientation to data-collection. He recommends ethnographic research as a means of measuring the extent of working class consciousness, provided steps are taken to obtain a reliable sample, and ensure a valid account uncontaminated by observer bias (Marshall 1988, pp. 122-23). On the other hand, an interpretivist would accept that there may be many different ways of understanding the same event or social setting. This has led many qualitative sociologists in the direction of either poststructuralism (with its celebration of multiple versions), or ethnomethodology (in which the process of creating versions becomes the topic).


ADONIS, A. and POLLARD, S. (1997) A Class Act. Penguin: Harmondsworth.

BEYNON, H. (1973) Working for Ford. Penguin: Harmondsworth.

BLUMER, H. (1959) 'Race Prejudice as a Sense of Group Position', Pacific Sociological Review, vol. 1, pp. 3-7.

BLUMER, H. (1969) 'Sociological Analysis and the "Variable"' in Symbolic Interactionism: Perspective and Method. University of California Press: Berkeley.

BLUMER, H. and DUSTER, T. (1980) 'Theories of Race and Social Action' in Sociological Theories: Race and Colonialism. UNESCO: Paris.

BOURDIEU, P. (1996) Distinction. Routledge: London.

BRADLEY, H. (1996) Fractured Identities: Changing Patterns of Inequality. Polity: Cambridge.

BULMER, M. (editor) (1975) Working Class Images of Society. Routledge: London.

BURAWOY, M. (1985) The Politics of Production: Factory Regimes under Capitalism and Socialism. Verso: London.

COLEMAN, W. (1996) 'Class Analysis: Some Sceptical Notes', unpublished paper.

CROMPTON, R. (1998) Class and Stratification. Polity Press: Cambridge.

CROW, G. and ALLAN, G. (1994) Community Life. Harvester Wheatsheaf: Hemel Hempstead.

DAVIS, K. and MOORE, W. (1945) 'Some Principles of Stratification', American Sociological Review, vol. 10, pp. 242-9.

DENNIS, N., HENRIQUES, C. and SLAUGHTER, C. (1956) Coal is Our Life. Tavistock: London.

DEVINE, F. (1992) Affluent Workers Re-Visited: Privatism and the Working Class. University of Edinburgh Press: Edinburgh.

FANTASIA, R. (1988) Cultures of Solidarity: Consciousness, Action and Contemporary American Workers. University of California Press: Berkeley.

FANTASIA, R. (1995) 'From Class Consciousness to Culture, Action and Social Organization', Annual Review of Sociology, vol. 21, pp. 269-87.

GARFINKEL, H. (1984) 'Passing and the Managed Achievement of Sex Status in an Inter-Sexed Person' in Studies in Ethnomethodology. Polity: Cambridge.

GLASER, B. and STRAUSS, A. (1967) The Discovery of Grounded Theory. Aldine: Chicago.

GOBO, G. (1995) '"Class" as Metaphor: On the Unreflexive Transformation of a Concept into an Object', Philosophy of the Human Sciences, vol. 25, no. 4, pp. 442-467.

GOFFMAN, E. (1964) Stigma. Prentice-Hall, Englewood-Cliffs: New Jersey.

GOLDTHORPE, J.H. (1983) 'Women and Class Analysis', Sociology, vol. 17, no. 4, pp. 465-488.

GOLDTHORPE, J.H., LOCKWOOD, D., BECHHOFER, F. and PLATT, J. (1969) The Affluent Worker in the Class Structure. Cambridge University Press: Cambridge.

HALFPENNY, P. (1979) 'The Analysis of Qualitative Data', Sociological Review, vol. 27, no. 4, pp. 799-825.

HERITAGE, J. (1984) Garfinkel and Ethnomethodology. Polity: Cambridge.

HILLER, P. (1973) 'Social Reality and Social Stratification', Sociological Review, vol. 21, no. 1, pp. 77-100.

JENKINS, R. (1992) Pierre Bourdieu. Routledge: London.

JOYCE, P. (1990) Visions of the People: Industrial England and the Question of Class, 1840-1914. Cambridge University Press: Cambridge.

JOYCE, P. (editor) (1995) Class. Oxford University Press: Oxford.

KORNBLUM, W. (1972) Blue Collar Community. Chicago University Press: Chicago.

LANE, T. and ROBERTS, K. (1972) Strike at Pilkingtons. Fontana: London.

LAYDER, D. (1998) Sociological Practice: Linking Theory and Social Research. Sage: London.

LOCKWOOD, D. (1966) 'Sources of Variation in Working Class Images of Society', Sociological Review, vol. 14, no. 3, pp. 244-67.

LOCKWOOD, D. (1992) Solidarity and Schism: The Problem of Disorder in Durkheimian and Marxist Sociology. Clarendon Press: Oxford.

LYND, R. and LYND,H. (1929) Middletown. New York.

MAHONY, P. and ZMROCZEK, C. (editors) (1997) Class Matters: 'Working Class' Womens' Perspectives on Social Class. Taylor and Francis: London.

MARSHALL, G. (1988) 'Some Remarks on the Study of Working Class Consciousness' in D. Rose (editor) Social Stratification and Economic Change. Hutchinson: London.

MARSHALL, G. (1997) 'Class and Class Analysis in the 1990s' in Repositioning Class: Social Inequality in Industrial Societies. Sage: London.

MARSHALL, G., ROSE, D., NEWBY, H. and VOGLER, C. (1988) Social Class in Modern Britain. Unwin Hyman: London.

MARSHALL, G., SWIFT, A. and ROBERTS, S. (1997) Against the Odds? Social Class and Social Justice in Industrial Societies. Clarendon Press: Oxford.

NEWBY, H. (1977) The Deferential Worker. Allen Lane: London.

PAKULSKI, J. and WATERS, M. (1996) The Death of Class. Sage: London.

PRUS, R. (1996) Symbolic Interaction and Ethnographic Research. State University of New York Press: New York.

REAY, D. (1998a) 'Rethinking Social Class: Qualitative Perspectives on Class and Gender', Sociology, vol. 32, no. 2, pp. 259-275.

REAY, D. (1998b) Class Work: Mother's Involvement in their Children's Primary Schooling. UCL Press: London.

RIST, R. (1970) 'Student Social Class and Teacher Expectations: The Self-Fulfilling Prophecy in Ghetto Education', Harvard Educational Review, vol. 40, no. 3, pp. 411-451.

SAVAGE, M., BARLOW, J. DICKENS, A. and FIELDING, T. (1992) Property, Bureaucracy and Culture: Middle Class Formation in Contemporary Britain. Routledge: London.

SAUNDERS, P. (1989) 'Left-Write in Sociology', Network (May), pp. 4-5.

SCOTT, J. (1996) Stratification and Power: Structures of Class, Status and Command. Polity, Cambridge.

SENNETT, R. and COBB, J. (1972) The Hidden Injuries of Class. Faber and Faber: London.

SKEGGS, B. (1997) Formations of Class and Gender: Becoming Respectable. Sage: London.

SILVERMAN, D. (1993) Interpreting Qualitative Data. Sage: London.

SILVERMAN, D. (1998) Harvey Sacks. Polity Press: Cambridge.

SPEIER, M. (1973) How to Observe Face-to-Face Communication. Goodyear Pub. Co.: Pacific Palisades, CA.

STACEY, M. 1960 Tradition and Change. Oxford University Press: Oxford.

STONEQUIST, E. (1936) The Marginal Man. Charles Scribener: Boston.

TRAVERS, M. (1997) 'Preaching to the Converted? Improving the Persuasiveness of Criminal Justice Research', British Journal of Criminology, vol. 37, no. 3, pp. 359-377.

TRAVERS, M. (July 1999, forthcoming) The British Immigration Courts: A Study of Law and Politics. The Policy Press: Bristol.

WARNER, W.L. and LUNT, P. (1959a) The Social Life of a Modern Community. Yale University Press: New Haven.

WARNER, W.L. and LUNT, P. (1959b) The Status System of a Modern Community. Yale University Press: New Haven.

WARNER, W.L. and SROLE, L. (1959c) The Social Systems of American Ethnic Groups. Yale University Press: New Haven.

WARNER, W.L. and LOW, J.O. (1959d) The Social System of the Modern Factory. The Strike: A Social Analysis. Yale University Press: New Haven.

WARNER, W.L. (1959e) The Living and the Dead: A Study of the Symbolic Life of Americans. Yale University Press: New Haven.

WATSON, D.R. (1998) 'Ethnomethodology, Consciousness and Self', Journal of Consciousness Studies, vol. 5, no. 2, pp. 202-223.

WEBER, M. (1991) 'Science as a Vocation' in H. Gerth and C. Wright Mills (editors) From Max Weber: Essays in Sociology. Routledge: London.

WILLIS, P. (1977) Learning to Labour. Gower: Aldershot.

Copyright Sociological Research Online, 1999