Interviewing,social work,and Chicago sociology in the 1920s

The AASW committees on the interview

Today, the composition, organisation, and operation of these groups have to be reconstructed mostly from published articles (Buell, 1925; Colcord, 1929; Myrick, 1926; Queen, 1928, 1929; Salsberry, 1927). Only the report of the Chicago group seems to have been published in full (Committee of the Chicago Chapter of the American Association of Social Workers, 1928). With the possible exception of the Chicago committee which had 26 members (Myrick, 1926: 122), the number of people involved in each group appears not to have been large, perhaps a dozen or so participants in each case. What members brought to the committees in terms of experience or expertise is now no longer clear, but presumably they were mostly experienced and active professionals. Two groups had explicit links to academic institutions. The work of the Twin Cities group was partly supplemented by the activities of students in Joanna Colcord’s advanced casework class at the University of Minnesota (Salsberry, 1927). The Kansas City group involved a partnership between the Sociology Department at the University of Kansas represented by the sociologist Stuart Queen and a committee of the local chapter of the AASW (Queen, 1928). There appears to have been relatively little direct collaboration between the four groups, although they were clearly aware of one another, and those that started later drew on material produced by earlier groups.

As part of their explorations, each group had to find ways of capturing what clients and social workers said and did in interviews. They had to conceptualise and analyse what was captured, and to translate those analyses into procedural recommendations for training and professional practice. It is clear that in their focus, concerns, and mode of operation the groups were strongly indebted to the work of Mary Richmond. Many of the topics they addressed dealt with issues Richmond had raised in her influential book Social Diagnosis (1917). Among these issues relating to how interviews were to be conducted and used in social casework one might note, among others, the importance of the initial contact, the need for social workers to pay attention to patterns that occurred with some frequency in casework situations, the emphasis on recording and documenting the processes at work in the interview and so on, all of which were subjected to careful scrutiny. ²

Working independently, mostly by trial and error, the groups seem to have operated in broadly similar ways. First, the group would discuss the topic in general before moving to a more practical phase of inquiry. Members would bring along a record of a recently conducted interview, which the group would examine and analyse. Subsequently, the group would try to develop some broader principles and conclusions based on the interviews examined. Most of the interviews that groups analysed appear to have been written out from memory soon after the interview with a client had been concluded. Occasionally, interviews had been stenographically recorded ‘without the client’s knowledge, of course’, as Buell’s (1925: 108) rather off-hand comment reveals, while the Minneapolis-St Paul group also made some use of interviews extracted from fictional sources (Salsberry, 1927).

Pragmatism, persuasion, and technique

The New York committee seems to have made least progress. According to Bradley Buell, who chaired the group, the final report was only produced with reluctance since some members disapproved of the ‘formulation of hypotheses lest they be taken as proved conclusions and translated into dogma’ (1925: 114). After looking at what they describe as ‘a meagre literature’ in fields such as personnel management, psychoanalysis, and salesmanship, and eliciting information from members about how they conducted interviews, the New York group seems to have concluded that it was not much the wiser, although it did identify common elements in the interview and recognised that interviews might proceed through a variety of stages.

Like the New York group, the other groups were somewhat tentative in their conclusions, but, drawing largely on various strands of pragmatist thinking, were rather more open to articulating general principles. In her introduction to the Chicago group’s report Helen Myrick, chair of the group, declared, ‘The underlying philosophy of the Committee throughout its existence is expressed by Dewey: ‘The purport of thinking, scientific and philosophic, is not to eliminate choice but to render it less arbitrary and more significant’’ (Dewey, 1925; Myrick, 1928: xiii). In his contribution to the Chicago group’s report, Harry Lurie acknowledged his indebtedness to the work of Cooley (with whom he had studied at Michigan) and John Dewey. Further cementing the relationship to pragmatism, Myrick had previously been associated with Ada Sheffield (Myrick and Sheffield, 1925), who maintained close contact with the group during its deliberations, and who was herself part of a circle around Dewey ³ (Siporin, 1986).

Sheffield took a ‘proto-social constructionist’ approach to social casework (Shaw, 2015a; 2015b). While not free of assumptions about the role of heredity in shaping people’s lives (1922: 10), Sheffield insisted that social workers should always see an individual’s situation as located within its wider social and relational context, and that people within a given situation might interpret it from a variety of standpoints, something that encouraged her to advocate careful, detailed case recording free from commonplace or superficial glosses (1922: 44–54). Lurie, like Sheffield, emphasised the importance of seeing the client as ‘a living, inter-acting personality’ located within a social setting that both influenced and was influenced by the client’s actions. ⁴ Thus, ‘the relationship between the social worker and the client is not one-sided. The client is not an inanimate object who is to be worked upon by the expert’ (Lurie, 1928: 8). Instead, the interview became a dialogue in which interviewer and client reacted to one another through a complex series of verbal and non-verbal expressions.

One also sees something of this conception in the work of the Kansas City group. According to its chair, Stuart Queen, the group began its work by reviewing ‘Park and Burgess’ chapters on contacts and interaction together with some of Mead’s articles on the conversation of gestures’ (1928: 546). Queen had attended and been much influenced by Mead’s lectures while at the University of Chicago (Huebner, 2014; Shaw, 2014), and, as Shaw notes, was particularly taken by Mead’s concept of the ‘conversation of gestures’ (2014: 766). This was a notion the Kansas City group applied to the interview, viewing its unfolding as a sequence in which the actions of each participant were interpreted by the other as a basis for subsequent action.

William Keith has drawn attention to the emergence in the 1920s of what he calls the ‘American discussion movement’ (2007). Members of this movement, many with an interest in adult education, included Alfred Sheffield (Ada Sheffield’s husband), Eduard Lindeman, Harry Overstreet, Mary Parker Follett, and others, all of whom were strongly influenced by John Dewey. Fundamental to the discussion movement, Keith suggests, was the assumption that American democracy was better served by the use of deliberative processes of discussion based around co-operation and problem-solving than by conflictual styles of decision-making. The discussion movement was important to the later development of group work in social work (Siporin, 1972). However, several of the committees also drew on the work of particular figures within the movement. Lurie (1928), for example, acknowledged the influence of Mary Parker Follett, for long a rather neglected figure now regarded as important to the historical development of research on organisational management (Follett, 1924; O’Connor, 2000). The Minneapolis-St Paul group was particularly influenced by the philosopher Harry Overstreet; an ‘immensely confident and glib’ figure, as Cotkin describes him, who produced ‘brilliant works of popularization in the pragmatic mode’ (Cotkin, 1994: 291).

What social workers took from people like Overstreet and Follett was not that it might make sense, for example, to treat the interview as a discussion. Rather, they looked to them as experts on techniques of persuasion. ⁵ Social workers could only enhance their technique, Lurie argued, by becoming more theoretically aware, singling out Follett as an example of relevant, new psychological thinking (1928: 11). Pearl Salsberry (1927: 132) concluded her report on the work of the Twin Cities group by quoting Overstreet’s book, Influencing Human Behavior:

… the salvaging of human life consists not simply in having high ideals. It consists as much in having the knowledge ‘how.’ We need, in short, to know how to interest our fellows; how to arouse their expectation; how to build up habits of favorable response; how to lead and adjust and control. All this is the groundwork of our human ethics. (Overstreet, 1925: 5)

‘Technique’ – understood in a general way – was a recurrent preoccupation of social workers in the 1920s. This concern is generally traced to Abraham Flexner’s speech to the National Conference of Charities and Correction in 1915, where he questioned social work’s status as a profession on the grounds that it did not possess ‘a technique capable of communication through an orderly and highly specialized educational discipline’ (1915: 580). At the time this was interpreted as encouraging the development of skills that could be taught to novices and capable of development by more experienced social workers (see, e.g. Austin, 1983). Nevertheless, there was a certain amount of disquiet about a preoccupation with technical matters. Specifically in relation to interviewing, the social worker G. Eleanor Kimble invoked the well-known parable of the centipede who having become self-conscious about which leg followed which became incapable of walking. She acknowledged that she might be seen as asserting that ‘social workers are born, not made’ but declared ‘Well I am not afraid of old-fashioned ideas just because they are old’ (1928: 566). As already noted, there is a hint that an emphasis on technical matters might have caused dissent in the New York group with some members resisting production of a final report for fear that their conclusions might be deemed dogmatic (Buell, 1925: 114). Each of the other groups examining the interview also acknowledged that an over-reliance on technique might be counterproductive (Myrick, 1926: 121; Queen, 1928: 558; Salsberry, 1927: 126). All went on, however, to identify relevant techniques for use in the interview and to explore their deployment and use.

One further consequence of the interest in persuasion can be seen in the emphasis on the notion of ‘rapport’ common to the various groups examining the interview. In the 1950s, survey researchers especially tended to see rapport as an aspect of the overall climate of the interview, often related to the respondent’s identification with the interviewer (see, e.g. Hyman, 1954). Social workers in the 1920s, however, regarded rapport as a somewhat transitory phase at the beginning of the interview that was crucial for ‘establishing the client’s confidence in the agency’ (Salsberry, 1927: 125). Walkowitz (1990) observes that unflattering depictions of 19th-century charity workers as meddlesome busybodies re-emerged in the 1920s. Establishing rapport at the beginning of the interview was seen to induce confidence in the social worker by neutralising suspicion about the motive behind the interview and, it was thought, encouraged receptiveness to the social worker’s point of view. The importance attached to rapport can also be seen, however, as compatible with wider cultural shifts in this period. Susman documents in the advice books and etiquette manuals of the 1920s the growing importance attached to ‘paying attention to others so that they will pay attention to you’ (Susman, 1984: 218). According to Susman, this was one aspect of an emerging emphasis on ‘personality’ rather than older Victorian notions of ‘character’ as a fundamental basis for notions of selfhood. In addition, Stearns and Stearns (1986) note the development of less confrontational approaches to interpersonal relations in fields such as selling and personnel management, both areas, perhaps not entirely coincidentally, social workers turned to for information about interviewing techniques (Buell, 1925). ⁶

Although how to represent data from unstructured interviews might seem like a modern issue, Tice (1998) points out there was considerable debate in the 1920s about the construction of written case records. While many social workers used relatively standardised approaches, others tried to write in a more literary style that captured something of the human interest in a particular case or, less commonly, in ways that tried to give voice to the client. ⁷ The AASW committees examining the interview each developed formats for making visible the techniques social workers used in the interview, although the New York group went scarcely beyond speculating about whether the interview could be represented by means of a ‘graphic curve’ (Buell, 1925: 112–113). Perhaps following the lead of Mary Richmond, who had used a similar system (1917, Appendix 1), the Twin Cities group indicated where a particular technique had been used in an interview by making an appropriate annotation in the margin of a written transcript of what had been said (a procedure for analysing interview data that would become ubiquitous in the late 20th century especially following the advent of computer packages for the analysis of qualitative data). The Chicago group, in particular, experimented with various forms of ‘data display’ (Miles and Huberman, 1994: 44). Dialogue, narrative, and skeletal forms of representation, singly or in combination, were used to condense and order the interview by reference to an outline. This was, in effect, a kind of template designed to reflect the temporal development of the interview, on the one hand, while, on the other, posing a series of questions about how the interaction between interviewer and interviewee had developed. The Kansas City group likewise experimented with a variety of more or less complicated ways of representing the verbal and non-verbal elements present in the interview.

Social work

The work of the various groups was not without its critics. Raymond Howes, a well-known rhetorician, argued that many of the purposes of the interview identified by the Minneapolis-St Paul committee were modes of persuasion that could be better understood through well-known principles of rhetorical analysis (Howes, 1930). By the end of the 1920s, psychoanalytic theories and concepts had become more prominent in social work. From this perspective, the interview could now be seen as involving emotional and unconscious elements that were not necessarily well captured by a focus on overt behaviour and manifest content (see, e.g. Rannells, 1927; Robinson, 1928; Taft, 1924). Summarising this view, Virginia Robinson explicitly targeted existing studies of interview technique,

Alluring as the behavioristic approach may be in its offer of objective measures, definite procedures, and more assured controls, it fails for these very reasons to satisfy the needs of the case worker confronted with the problem of understanding a deep-laid pattern of emotional response and guiding a process in which her own patterns of interacting play a constant part. (1930: 85)

Industrial relations research

Outside social work, the activities of the AASW committees on the interview attracted most interest in industrial relations research. Important here was the influence of Mary Van Kleeck, the Director of Industrial Studies at the Russell Sage Foundation. Van Kleeck had produced pioneering studies of working conditions in a number of industries, often based on extensive and detailed interviewing (see, e.g. Van Kleeck, 1913). As Ferguson (1961) relates, Van Kleeck’s involvement with the Personnel Research Federation, an organisation set up to foster and co-ordinate research on industrial relations, resulted in a number of publications on interviewing. One of these was a best-selling textbook entitled How to Interview by Walter Van Dyke Bingham and Bruce Moore (1931). (On the popularity and influence of this book, see Farr and Tesluk, 1997.) Bingham, an industrial psychologist who had been deeply involved during First World War in the development and application of aptitude and other tests for the US Army, lauded the work of the AASW groups investigating the interview for ‘their splendid pioneer explorations’ in ‘familiar and yet little known territory’ (1929: 530, 531) and continued to devote attention to social work interviewing in subsequent editions of How to Interview. ⁸

Sociological research methods

The response of sociologists to the work of the AASW groups on the interview was somewhat muted. Ruth Shonle Cavan (1929) used the contribution social workers had made to systematising interview practice as a justification for her own ‘topical summary’ of the literature. She also identified a number of publications on the interview by social workers, which compared to other sources, she regarded as giving relatively complete information about the topic. Vivien Palmer (1928) in the methods textbook written for sociology students at the University of Chicago makes reference to the work of local social workers who examined the interview and summarises some of their ideas, but in general she associated social work with the older social survey tradition that Chicago sociologists thought was lacking in scientific rigour. The most detailed treatment of the various social work groups examining the interview is to be found in Pauline Young’s book Interviewing in Social Work, published in 1935. Displaying what Ian Shaw describes as a ‘constant easy movement between what today would be seen as different even dissonant literatures’ (2015b: S20), Young’s widely cited book (Platt, 2001) comprehensively synthesised the literature on social work interviewing with the wider sociological and psychological literatures on the topic. In the post-war period, survey researchers began to pay close attention to the interactional dynamics of the interview as a basis for examining the role of ‘interviewer effects’ (Hyman, 1954; Kahn and Cannell, 1957), ⁹ but by this point the social work interview and the clinical interview had come to be seen as largely synonymous.

Burgess and social work

A number of prominent social workers responded directly to the proposal Burgess made (Bruno, 1928; Eliot, 1928; Swift, 1928). These writers were by no means entirely hostile to the position he had advanced. Thomas D. Eliot, for example, while doubting that the goals and interests of sociologists and social workers were as closely aligned as Burgess had assumed, thought that sociology would provide social workers with a richer understanding of the causes of social problems, while Linton B. Swift saw social work entering a period of growing rapprochement with sociology. In both cases, the potential gain for social workers was the possibility of a more sophisticated understanding of the situations that confronted them. Alongside this, however, was an emphasis, politely expressed, on the practical problems involved in selecting or recording information while engaged in interviewing social work clients. Was it actually possible to be objective, or was bias of some sort always inevitable (Eliot, 1928; Swift, 1928)? Was vividness in the record a sign of verisimilitude or superficial selectivity (Bruno, 1928; Swift, 1928)? A common thread was the inevitable disparity between the goals and interests of social workers and sociologists. Social workers had to attend to individual circumstances and to attempt to change people’s lives, while sociologists looked for broader patterns and how social relations and social organisation endured.

Tice argues that the interest Burgess showed in social work was part of what she describes as ‘a vociferous crusade’ by sociologists to denigrate social work’s scientific aspirations (1998: 50). Given the degree of engagement Burgess showed with social work throughout the 1920s (Shaw, 2009) and the respectful way in which he discussed both the theoretical contributions of social workers and their work on the interview, this seems overstated. Burgess was at pains to insist that verbatim interviewing would serve the needs of social workers themselves and would be in line with enlightened thinking in the field. Nor does he argue that social workers should desist from interpreting interview material, rather that data and interpretation should be separated, so that it is clear whose interpretation is being offered. Nevertheless, as Tice suggests, it can be argued that the relationship between sociology and social work Burgess proposed was not one of parity. Shaw points out that undermining a fruitful reciprocal and egalitarian engagement between social work and sociology is a rather impoverished conceptualisation of the relationship between theory and practice, on the one hand, and social science and social work, on the other. ‘In both cases’, he says, ‘the tendency is to fall back on a rationalist position whereby “science” has inherent bragging rights over “practice”’ (2014: 763). It is interesting in this context to consider Gieryn’s (2006) contention that, rhetorically speaking, Chicago sociologists oscillated unendingly between two constructions of the city. Chicago could be seen, on the one hand, as a ‘laboratory’ that provided sociologists with a controlled environment in which various aspects of urban life could, specimen-like, be measured, dissected, theorised, and generalised. On the other, the city was a field site requiring close, intensive, personal involvement designed to uncover particular facets of the urban experience in often unexpected ways. The oscillation Gieryn describes was underpinned by a division of labour. While academics theorised and conceptualised, it was students for the most part who ventured into the field in search of intimate knowledge of the how social worlds making up the urban fabric were organised. Less often remarked, it was also students who were left to work out the technical details of research procedures. (Krueger’s (1925) work on the life history and Erle Fiske Young’s work (1925) on the social base map are cases in point.) It is not difficult to see a situation where social workers might have become treated in a similar way, as lower level participants who have an understanding of technical matters that gives them something useful to contribute in a practical sense but in the process points only to their perceived lack of scientific credibility. Moreover, in a context where social workers were mostly female and sociologists mostly male, the likelihood was high that the social worker, viewed in the case of the interview as at best a technician and at worst simply as an amanuensis for the sociologist, would become allocated to a kind of ‘auxiliary’ sphere, to use Deegan’s term (1988: 144), largely defined by gender. Quite possibly, Burgess simply assumed rank-and-file social workers would fall into line with his proposal provided it was presented as a matter of authoritative opinion and expressed in an emollient manner.

In fact, there are few clues to how Burgess did actually envisage the relationship between sociologists and social workers. It might be relevant here that he had begun teaching a course on ‘clinical sociology’ beginning in 1928 and continued it for some years after ¹¹ (Fritz, 1991). According to Fritz relatively little material about this course now survives but Burgess seems to have seen it as a vehicle for the entry of sociologists into positions in child guidance clinics. One possibility arising from this is that in advocating the verbatim interview Burgess might have been thinking less in terms of a direct exchange between social workers and university academics and more that the interview could provide a common element in the activities of social workers and professionally trained sociologists embedded in social work agencies.

What is now impossible to say is whether Burgess might have found a way, had circumstances been different, to put the relationship between social workers and sociologists on a more formal footing. The accumulation and reuse of already collected research materials was a fundamental, if nowadays often unremarked, aspect of Chicago sociology (though see Bulmer, 1980). Soon after the Local Community Research Committee was established in 1923, it undertook an exercise to establish what potential research materials were held by local civic bodies, social agencies, and the like (Burgess, 1929). Subsequently, what Burgess refers to as ‘a clearing house or central depository of social investigations and research’ (1929: 49) was established with a wide range of material being accumulated through donation, salvage, or exchange. (For more on the role of repositories, see Star and Griesemer, 1989.) Since social workers already had a well-developed system of ‘confidential exchanges’ (see Tice, 1998: 31–33), it is not inconceivable that a system could have been established to pass on information from interviews (presumably on a confidential or anonymised basis) either independently or via the central depository. In fact, Vivien Palmer, reporting on the project directed by Burgess that located much of the material for the central depository, advocated the establishment of a ‘community research exchange’ (Palmer, 1930: 37) that might have served as a model, although she gives little detail of how it might operate. By then, in any case, the local communities project was in trouble for its slow progress and relative lack of substantive outputs (Bulmer, 1980; Topalov, 2014) making the advent of such an exchange unlikely.

One interesting aspect of the Burgess article on the verbatim interview is that it is not noticeably Chicago-centric. Good practice is seen to exist in many locations. This raises the question of why Burgess does not seem to have tried to engage with his social work colleagues at Chicago. Both departmental folklore and what seems like a degree of structured dis-attention to putatively similar work elsewhere point to the apparently strained relationship that existed between the two departments. A proposal apparently made by Burgess to Edith Abbott for joint research, if not exactly rebuffed, seems to have languished ¹² (Shaw, 2015b). Any room for manoeuvre Burgess might have had was also possibly constrained by an already running project between social work agencies and researchers at Chicago dealing with uniform statistical reporting (McMillen and Jeter, 1929). It might have been this project Burgess had in mind some years later when he acknowledged that ‘Social agencies have almost invariably, and often with considerable inconvenience, been most hospitable to use of their records for research purposes’ (1934: 447). In the event, however, he still felt the need to complain that verbatim interviewing was needed to overcome the ‘inadequate or even distorted presentation’ of attitudes in case records.

Burgess and interaction in the interview

The virtues of the verbatim interview, as Burgess set them out, implied a role description of the social worker as someone who was interested, empathetic, and non-judgemental. This left out of the equation the persuasive element so central to the work of AASW groups on the interview. It also begged the question of how the interviewer’s role manifested itself through the course of the interview. These were matters the groups exploring the interview viewed as fundamental, even if they did consider them largely in terms of technique. Burgess, however, is critical what he describes as the ‘legalistic’ conception of the interview which in his view

focused upon the art of cross examination, upon all the little tricks of technique designed to elicit information which the informant desired to conceal, and upon ingenious methods by which the social worker could impose his will upon an unwilling applicant for assistance. (1928: 527)

While this prefigures to some extent what later became known as ‘nondirective’ interviewing (Rogers, 1945), what Burgess is reflecting here is an understanding of the interview that derives from the work of W. I. Thomas. Thomas distrusted the interview as a result of negative experiences he had when attempting to use the method on a trip to Eastern Europe (1912: 771–772). (This was one reason why he preferred to use ‘undesigned records’ such as documents.) Interviewing, he thought, was best used to obtain information from officials or local worthies of various kinds, who were often ‘in a position to give reliable data, and are often interested in doing so’ (1912: 771). ¹³

Removing the role of the interviewer from consideration by emphasising the words of the interviewee over the process by which they had been acquired underlined one strength of the social workers’ approach. One can see this in the study by Clifford Shaw that Burgess held up as something of a model for ‘the method of first-person reporting’ (1928: 529). Shaw (1927) conducted a study in which a 14-year-old boy who had been referred to a child guidance clinic was interviewed about his problems in the presence of other members of his family. The interview, recorded verbatim by a stenographer, soon turned into a bitter family quarrel that caused the boy much distress. As one critic, the social worker Linton Swift (1928), pointed out, the procedure raised both methodological questions and ethical issues. How did the presence of third parties in the interview, including the stenographer, affect the proceedings, and what was the impact on the boy and his treatment? In this Swift was echoing Virginia Robinson’s earlier rather strident attack on sociologists for what she saw as a cavalier approach to ethics. Failing to recognise that research procedures could have psychological consequences for those studied was, in her view, ‘as serious a breach of responsibility as for a surgeon to abandon a case after inserting a knife’ (1926: 300).