Myths of the Chicago School

3. Chicagoans as ethnographers

3.7 Participant observation and the 'Chicago School' approach

Participant observation was not, then, a recognisably Chicago research practice until at least into the 1940s. Thus, at the 1939 conference on the Polish Peasant study, which concentrated on the human document as a research tool, a forum at which methodological issues were extensively discussed, there was but a single mention of participant observation: and this in reference to documents collected in an extensive survey in Sweden.

The Swedish material has an all-round superiority in the fact that it includes the medical examination, the life history, the controlled interview, the letters to relatives, friends, sweethearts, lawyers, etc., and the testimony of 'participant observers'. (Social Science Research Council, 1939, p. 133)

Furthermore, there is in the transcript of the conference virtually no reference to a specifically Chicago approach with the exception of two references, by Thomas, both to the collection of human documents in the early part of the century

This movement toward the collection of human document material was going on inevitably, anyway, that is, in Chicago. So this work was merely another influence on the concrete trend in sociology…. [besides the Swedish study and a collection of material from the Jewish Daily Forward] it is not necessary to mention the important collection of human documents by sociologists of the University of Chicago where the practice has been extensive and refined. (Social Science Research Council, 1939, p. 130–2)

Participant observation was still unusual at Chicago in the 1940s. For example, in his doctoral thesis submitted to the Chicago sociology department in 1940, Daniel adopted a mixture of methods including participant observation. Far from taking the method for granted, Daniel tentatively referred to the 'participant observation type method', as though it were not common-place. Indeed, it was Whyte (1943) who probably provided the first example of a 'pure' participant observation study. Even so, this was not at the instigation of the University of Chicago, as correspondence between Burgess and Wilson (Wilson 1940) indicates. Indeed, Wilson talked of Whyte's study not as emulating the Chicago approach but rather 'you see his technique is the interview technique like Bakke's of Yale'. The only 'concession' Wilson made to a Chicago orientation is to say that Whyte is 'from the sociological or socio-economic aspects of current phenomena, on what I might call the clinical side, which has been favored by W. I. Thomas and Park and you.'

Whyte, himself, noted that 'I owe a great personal debt to Conrad M. Arensberg, now at Columbia University, from whom I learned my field-work techniques' (Whyte, 1955, p. vii). And it was Whyte, in his methodological appendix to the second edition of Street Corner Society (Whyte, 1955), who was one of the first researchers to discuss the problems of participant observation directly and in some depth. This belies the view that 'in the period from 1920 to 1940 people who called themselves students of society were familiar with personal documents and participant observation' (Bogdan and Taylor, 1975, p. 4).
 
Although he completed his research as a Chicago graduate student, Whyte had begun his study of an area of Boston as a Harvard graduate. The only contribution that Chicago seemed to have made to Whyte's discussion of participant observation methodology was indirectly through Arensberg having worked on Warner's 'Yankee City' research. This further supports my suggestion that participant observation at Chicago resided in its community studies.
 
It would seem, then, that the first classic 'pure' participant observation study was neither conceived at Chicago, nor, did its fieldwork orientation owe much, if anything, directly to a Chicago tradition. [7]

While Blumer suggested the efficacy of observational studies in the 1940s, and a 'tradition' of symbolic interactionist observation research ostensibly deriving from Chicago emerged in the 1950s, there is little to suggest that the 'Chicago School', prior to 1940 was single-mindedly pursuing participant observation research [8]. When Becker undertook ethnographic research in the late 1940s and early 1950s he studied the 'classics' of social research

such books as Street Corner Society, the Polish Peasant in Europe and America, and back I should say to Charles Booth and Henry Mayhew was something that a lot of us read. You know, taking that to be a kind of model of how ethnography might proceed, of the kind of detail that one would want to know about people living in a city, for instance, or about occupation. (Becker, 1979, p. 7)

The 'Chicago tradition' prior to 1940, then, was broadly ethnographic rather than specifically concerned with participant observation. It aimed at finding out things 'you hadn't thought of' rather than adopting the insider perspective of the subjects. That was to occur later when 'ethnographic research began to be informed by multiple theoretical viewpoints' (Becker, 1979, p. 9).

Indeed, as late as the 1960s there was considerable advocacy of participant observation as an appropriate and novel method of research. Notable is Polsky's (1971) advocacy of participant observation for the study of deviance. He claimed that the naturalistic perspective was little used and the main target of his critique was the well known and highly regarded criminological work of Sutherland. Similarly, as late as 1969, Coleman pointed to the lack of 'in situ' observation, which is commonly regarded as the hallmark of Chicago sociology. Coleman (1969) noted that sociologists have inadequately used observation; rather they have tended to depend on individual's reports of their own behaviour, (through questionnaires, life histories, interviews). Not enough, he suggested has been done on observation in situ. He pointed to the doctoral work of both Stinchcombe and Barker, on high school students and children respectively, which used direct observation and to

the work of Garfinkel and his students, in which the investigation presents verbal stimuli, not as an interviewer, but as a member of the same social system.... More useful, however, than at least the existing disappointing results of Garfinkel's work is the work in participant observation carried out by such sociologists as Howard Becker, and work described by Webb, Campbell, Schwartz and Sechrest in Unobtrusive Measures. (Coleman, 1969, p. 112)

Coleman may have underplayed the extent to which, by the late 1960s, participant research had been used: but even so his comments on the relative novelty of direct, in situ, observation belies the view that such an approach is intrinsic to the Chicago heritage. The problem in the past, he argued, was that observation was difficult because there were few aids. This, he argued, had changed thanks to electronic aids, awareness of time sampling, space sampling and sampling of roles and so on, all of which increase reliability and validity.
 
Without doubt some of the work done at Chicago involved methods more usually located within ethnographic orientations. However there is very little to suggest that Chicago adopted other than a nomothetic perspective on the analysis of the social world until the 1930s, when, in embryonic form, a proto-phenomenological scepticism began to emerge in the work of Blumer and Wirth. (This is examined further in chapter five). Prior to this, the interactionist perspective, reliant as it was upon Thomas' general sociological orientation, clearly adopted nomothetic principles and, with it, a degree of methodological eclecticism, as will be illustrated below, with few pieces of work that could be described as adopting a single method. However, that does not mean that there were not clear preferences expressed by members of the Department for particular methodic devices. On the one hand, Park used to adopt an attitude of extreme scepticism towards statistics, which became more acute as he grew older (Matthews, 1977). Ogburn, on the other, was keen to develop the measurement of social phenomena. Burgess, tended to mediate between the two and make use of qualitative and quantitative techniques.

Ogburn knew very little of the Park sociology and Park knew nothing of statistical methods. I think Park would sometimes make some grudging remark about it, disapproving of the fad for statistics. Ogburn was a Southern gentleman and he generally didn't make personal remarks about any-body, but his wife did. And the bitter personal feelings between the wives suggested that underneath both men had some feeling about the matter but on the surface they both co-operated well and I once overheard a good part of a department meeting and everything went well. (Faris, 1972)

Although the 'Chicago School' myth would have it otherwise, the Chicagoans were as involved in developing quantitative techniques as they were in qualitative procedures. There were debates at Chicago, and there was, it seems, 'a tension in the Department about the Ogburnian's and Parkian's' (Cottrell, 1972). However, there is little evidence that any long standing alliances or factional divisions characterised the department. 'I'm not sure that they grouped into factions very much, at least I wasn't aware of it' (Faris, 1972). Importantly, there was no division into competing nomothetic and ideographic camps at Chicago. 'The two sides, the statistical and the social interactionists both wanted to build a scientific sociology' (Dollard, 1972).
 
In effect there were three debates at Chicago: those relating to quantification and the associated discussions about the efficacy of case studies and statistics; those relating to instinct theory; and those, somewhat more bitter and less resolvable, revolving around Freud's theories. Of these, the instinct theory debate was the least active. Instincts, for many, were a dead issue as Faris had laid to rest the residual instinct notion embodied in Thomas's wishes. The other two debates were active.  

Park would preside and he would rumble on about stupid Freudians and statisticians, but Burgess would nod as if he was approving of what Park said and he would look across to where some of us were sneaking out and getting courses on statistics and courses on Freud and he'd twinkle because he was doing the same thing.... He got very interested in the Freudian contribution to sociology, to social theory. (Cottrell, 1972)

The Chicagoans role in developing quantitative techniques is explored in Chapter four. Chapter five takes up the theoretical debates, which involved Freudianism.

Notes

[ 7] When Whyte was awarded a three-year junior fellowship at Harvard, following his graduation from Swathmore in 1936, he decided he wanted to research a slum district with a view to investigating racketeering. Whyte was an economist not a social researcher and when he located 'Cornerville', which he thought looked just like he imagined a slum area should, he had no real idea how to go about his research. In 1936, there was little existing literature to help him. Apparently only two published studies were available, Middletown by the Lynds (1929) and Greenwich Village by Carolyn Ware (1935). Neither of these were what Whyte wanted as they focused on social problems and not social systems. He wanted to see how a local community worked rather than investigate its particular social difficulties (Madge, 1963, p. 213). As Whyte acknowledged, it was Conrad Arensberg, another junior fellow at Harvard, who was to be a big influence on his approach. Arensberg, along with Solon Kimball had adopted an anthropological approach to the study of a community in Ireland (Arensberg and Kimball 1940). So Whyte did not derive his method from the Chicagoans, he did not seem to consider their work to be of the kind of direct involvement research that he intended. This is a little surprising. Perhaps his lack of sociological background left him without the knowledge of the Chicago studies. Perhaps, in 1936, the Chicago studies were not seen as participant observer studies. Or perhaps Whyte dismissed them, too, as social problem research. It is, nonetheless, surprising that he was apparently unaware of Blumenthal's Small Town Stuff (1932a), which was concerned with the social system, rather than the problems, of a small community. In the event, due to force of circumstance, Whyte adopted a different approach to participant observation than the Lynds, Ware or Arensberg. His approach was to move into the close community of a particular group of people and rely heavily on a key informant and sponsor. He had not planned to operate in this manner but took full advantage of the opportunity when it arose. The result is the classic study (Whyte 1943a) to which a lengthy methodological appendix was added when it was published in its second edition in 1955. For the first time, we have an account of what it is like to research a group of people from ones own society with which the researcher has intimate contact. Many such studies have been done subsequently; but there were no such close and complete participant observation studies done at Chicago, or elsewhere, prior to Whyte's research in the late 1930s. Return

[ 8] As late as 1947, Wirth (1947) in reviewing the development of sociology over the previous thirty years suggested that not only had there been a move towards specialising in specific areas but also that sociologists needed specialised skills for empirical work. Wirth listed such skills, which did not include participant observation as such and was dominated by quantitative techniques: social statistics, sampling, population analysis, personal documents, prediction methods, attitude testing, public opinion polling, questionnaire construction, field interviewing and the mapping of social relations. Return

Next 4.1 Introduction