Introduction

The canonization process enshrining a candidate in the religious, literary, political, or scholarly pantheon is an intricate affair reflecting not only the luminary’s inherent graces and exceptional merits but also the conflicting agendas of successors vying for the deceased’s mantle. Those who made the canon early do not always meet the test of time and those overlooked may yet get an honorable textbook mention. The controversies surrounding locus classicus are common, as scholars debate what counts as an appropriation, misappropriation, and crypto-appropriation where the evidence points to an influence the interpreter fails to acknowledge or flatly denies.

Talcott Parsons’ reading of Max Weber is a case in point. Many critics found troubling the structural-functional interpretation that Parsons offered of this German luminary, first tentatively in The Structure of Social Action (1937), then more insistently in 1951 in The Social System (Cohen et al. 1975; Walliman et al. 1980; Swatos and Kivisto 2007). Without mentioning him by name, Hans Gerth and C. Wright Mills took a shot at Parsons in a 1946 essay where they maintained that “Weber represents humanism and liberalism” in the mold of Schiller rather than the bureaucratic tradition valorizing the “technical expert.” Intriguingly, the authors linked Mead and Weber as two thinkers exemplifying this humanistic concern: “Just as for George H. Mead the ‘I’ is ordinarily in tension with the social roles derived from the expectations of others, so for Weber the potentially charismatic equality of man stands in tension with the external demand of institutional life. For Mead, the tension between the I and the role-demands is resolved in the creative response of the genius. For Weber, the response of the charismatic leader to distress unifies external demands and internal urges” (Gerth and Mills 1946:73; cf. Mills 1959).

Taking a clue from this observation, I review Mead’s induction into the sociological canon and the ways his ideas were appropriated by disciples. My focus is on American sociologists who made a concerted effort to reconstruct Mead as a systematic social theorist and apply his ideas to empirical research. Two interpretations in particular are singled out here – “symbolic interactionism” and “structural interactionism.” The former was developed by Herbert Blumer and his students at the University of Chicago, the latter by Manfred Kuhn and scholars with links to the University of Iowa. The thesis I wish to advance in this paper is that both schools draw on key precepts in Mead’s theory, each has engendered a viable line of inquiry, and all contribute to the ongoing effort to weave together disparate strands of Mead’s unfinished project.

Mead’s Corpus and Textual Criticism

Today’s commentators are in agreement about the enduring significance of Mead’s thought for the discipline of sociology. Hans Joas (1985: 2) calls Mead “the most important theorist of intersubjectivity between Feuerbach and Habermas.” According to Randal Collins (1989: 263), “George Herbert Mead is no doubt America’s greatest sociological theorist.” And for Mary Jo Deegan (2008: 329), Mead is “the most important sociological theorist in the world from 1892 to 1938.” The current consensus about Mead’s place in the history of sociological thought has taken time to evolve. While both John Dewey and Alfred North Whitehead vouched for Mead’s standing in American philosophy, sociologists were tardy recognizing Mead’s status as a sociological classic. Reasons for this delayed reaction may have something to do with the novelty of his ideas, which even his closest colleague Dewey took time to appreciate.

The struggle Mead faced putting his thoughts on paper was another factor, as was the lack of timely publications. During his lifetime Mead did not publish a single book. His archives at the University of Chicago Regenstein Library contain the galleys for a volume comprised of 18 articles, most previously published. Reprinted in 2001 under the title Essays in Social Psychology (the book’s original title was Essays in Psychology), the collection did not find its way into print at the time when it was assembled around 1909–1910. Mead’s “publication block” (Collins 1989: 264), as some commentators suggest, could stem from his perfectionism or his over-commitment as a public intellectual. One more explanation comes to mind. Steeped in functionalist psychology and valuable for understanding Mead’s formative years, the book is likely to disappoint readers familiar with his path-breaking ideas about the dialectics of self and society. By the time Mead finished this project, he was on the verge of completing several pioneering studies which would appear in various journals between 1909 and 1913 and help establish his reputation as an innovative thinker. Perhaps Mead realized that the planned anthology needed an update and put off its publication. His failure to anthologize his works in later years reflects the fact that he kept reworking his ideas, tackling the same problem over and over again in the hope of finding better verbiage to frame his point.

Today’s commentators have access to a wide range of materials printed after Mead’s death, some based on lecture notes compiled by students who took classes with Mead between 1912 and 1929. Mind, Self, and Society (MSS) is the best known and the most controversial addition to Mead’s corpus (Joas 1985: 2–3, 235; Cook 1993: 15–19; Silva 2008: 142–145; Huebner 2012). The problem with this opus and several other posthumous publications based on students’ notes is that they raise questions of whether the published texts adhere to Mead’s true wording and intent. Even though the notes were based on a stenographic account, significant additions and alterations were made during the editing process. Charles Morris, who was primarily responsible for editing MSS, relied on Mead’s notes for lectures that he gave at the University of Chicago between 1927 and 1930, with the 1928 stenographic account forming the backbone of the book. Stressing the “verbatim” character of class transcripts, Morris acknowledged that the original notes were “rearranged, pruned of superfluous repetitions, and stylistically corrected” (Morris 1934: vi). In some cases, the lecture notes were supplemented by excerpts from Mead’s published works, the excerpts identified as such in the text.

As it turned out, the MSS subtitle – “From the Standpoint of a Social Behaviorist” – was supplied by the editor. The term “social behaviorism” appears nowhere in the original notes (Cook 1993: 70–71). “I never heard Mead use the term in the many lectures of his which I attended,” observed Herbert Blumer (2004: 164), “nor in the many discussions that I had with him as his research assistant. I have no recollection of his use of the term in his writings, [and] if he used it, it was on exceedingly rare occasions. It is safe and accurate to say that the term social behaviorism was introduced into Mead’s thought not by Mead but by others” (Blumer 2004: 164). Suspiciously close to the behaviorist stance Morris took in his signature work Signs, Language, and Behavior, this wording suggests that the politics of interpretation might have been at work in this framing (see Morris 1938, 1955). Given the heated polemics about Mead’s behaviorism that divided Mead scholars in later years, choosing the right moniker is hardly an idle matter. Such liberties made some commentators charge that MSS is too compromised by its “unreliability, insufficient precision in the recording of Mead’s words, and the obfuscation of the coherence of Mead’s thought” (Joas 1985: 2). Similar concerns were raised about other posthumously published volumes based on student notes, notably David Miller’s 1982 publication stemming from the course on social psychology that Mead gave in the winter quarter of 1912 (Miller misidentified the manuscript as based on the course taught in 1914). Still, no attempt to frame Mead is exempt from the politics of interpretation, and those who claim otherwise invite an especially close scrutiny.

The importance of textual analysis is beyond doubt, particularly when the researcher tracks the evolution of Mead’s thought. Dan Huebner’s painstaking attempt to reconstruct how MSS was put together makes important strides in this direction. Huebner’s archival research reveals the limitations of the published work – the problematic choice among alternative sets of student notes, questionable judgment on what constitutes a dispensable repetition or irrelevant aside, errors in rendering certain technical terms (Huebner 2012). The very decision to publish lecture notes that were not intended for publication raises ethical questions. However, such concerns apply not only to other posthumously published volumes based on student notes like Movements of Thoughts in the Nineteenth Century, but also to compilations based on the fragments and unfinished manuscripts that Mead did not necessarily intend for publication. Robert Merton (1994) made a convincing case for the role of what he called “oral editions” as the legitimate form of circulating scientific knowledge. We should bear in mind, also, that Mead was a great conversationalist who did much nimble thinking on foot, and that Mind, Self, and Society offers a glimpse into his mature thought where he ties together the diverse strands of his sprawling theoretical corpus. Finally, it would be churlish to deny credit to Mead’s early editors whose efforts brought Mead to the attention of wider audiences and helped propel him to the center of sociological analysis. Until a critical edition of Mead’s papers is published, we should make the most of the materials in print, with due regard for the textological criticism available today, and various collections of previously published works by Mead (1964, 1982, 2001, Petras 1968; see also Mead Project, https://www.brocku.ca/MeadProject/inventory5.html; Pragmacyberlibrary, http://cdclv.unlv.edu//archives/pragmat_lit.html).

The Scholarly Canon and the Reception of Mead’s Ideas

Four volumes of Mead’s work that appeared after his death produced a steady stream of reviews and occasioned mostly good press, which seemed to bode well for the future of Mead scholarship (Merton 1935 as cited in Hamilton 1992; Murphy 1935, 1939; Morrow 1935; Faris 1936; Hook 1936; Bierstedt 1938; Burke 1939). Some reviewers confidently predicted that “Mead will enjoy comparable stature with James and Dewey in the extent of his influence” (Bierstedt 1938: 28). Given the powerful endorsements from the likes of Dewey and Whitehead (Cook 1993; Shalin 2000), one had reasons to expect Mead’s reputation to spread quickly, yet when Maurice Natanson published his study of Mead in 1956, he observed with dismay that this “truly major American philosopher is today largely unknown, frequently misunderstood where he is known, and, more often than not, simply ignored” (Natanson 1956:1). The failure to come to grips with Mead’s legacy paralleled a decline in the fortunes of American pragmatism which mid-way through the 20th century was seen by many in this country and abroad as an intellectual current whose progenitors had little to offer to contemporary debates dominated by analytic philosophy (Westbrook 1991; Cook 1993; Joas 1993, 1997; Shalin 2011). As Paul Rock (1979: 166–167) put it, interactionists claiming Mead’s legacy “relied on the authority provided by the virtually overthrown pragmatism.”

Mead’s reputation among sociologists took an equally long time to gel. The famous Introduction to the Science of Sociology that saw its first printing in 1921 did not mention Mead, nor did the second edition that came out in 1924 (Park and Burgess 1969; cf. Park 1955). Among the readings collected in this volume one finds the articles by Charles H. Cooley, Dewey, Georg Simmel, Emile Durkheim, Leonard T. Hobhouse, Ludwig Gumplowicz, and scores of other luminaries, yet the editors did not see fit to sample Mead’s writings. This is puzzling, given that Robert E. Park and Ernest Burgess knew Mead personally, were well aware of his work, and encouraged their students to take his classes. The absence of Mead in this sociology primer suggests that he was not considered by sociologists a principal figure at the time. We find an occasional reference to Mead in the works of Park and his colleagues following publication of Mind, Self, and Society, but the references are brief, generic, and not nearly as laudatory as those accorded to James, Dewey, and Cooley.

Until the 1950s, several milestone publications failed to recognize the centrality of Mead to the sociological enterprise, with such classics as The Structure of Social Action (1937) and The Social System (1951) omitting his name altogether. Later on, Parsons et al. (1975: 126) conceded that “there is a very substantial area of overlap in the structure of conceptualization between social interaction theory, notably in the version presented by George Herbert Mead himself, and the theory of action,” that “an intensive reading of Mead’s work was not undertaken by me until fairly well after the completion of The Structure of Social Action,” and that “this was doubtless an oversight.” Elsewhere he wrote, “I came to attribute great importance to the convergence of Durkheim and Freud – a convergence which extended in an attenuated sense to Weber, but very importantly to the American social psychologists, especially to G. H. Mead” (Parsons 1970: 2). Had he chosen to revise The Structure of Social Action, intimated Parsons (1949: xv) in the preface to the second edition, Mead would have figured prominently in the update; the “decision not to embark on a thorough revision of the book represents the judgment that in the present situation of social science [empirical research constitutes] the more fruitful channel for a major investment of time and energy” (Parsons 1949: vx; cf. Parsons 1954: 14).

Robert K. Merton was among the first to review Mind, Self, and Society and give this seminal volume a nod of approval, in part because of Mead’s emphasis on the social nature of truth (Merton 1935). Yet Merton had little use for Mead’s ideas in his own work, relegating the relevant comments to footnotes. When he did mention Mead, it was typically in the context of “Cooley’s still enduring formulation” presaging the reference group theory and “George Mead’s extension in turn” (Merton 1968: 19n39).

C. Wright Mill’s voluminous dissertation on the influence of pragmatism in American sociology that he had defended in 1948 recognized Mead in passing. In a later note, Mills signaled a change in his attitude toward Mead: “An account of George H. Mead must be included. It is true that many features of Mead’s thought are treated by consideration given to the work of John Dewey. However, in view of the course of the pragmatic movement and of Dewey’s differential evaluation of Mead and [William] James, the inclusion of James and the omission of Mead is an unrepresentative act that is intellectually unwarranted” (Mills 1966: 464).

Blumer offered perhaps the earliest extended assessment of Mead as a social psychologist. It appeared in a chapter titled “Social Psychology” published in 1937 in a volume edited by Peter Schmidt under the heading Man and Society: A Substantive Introduction to the Social Sciences. In this essay Blumer introduced the term “symbolic interaction” that would mark a particularly influential branch of interactionist sociology. Mead is given more prominence in the text than other thinkers, yet the treatment was brief, a few pages in a 50 page article, and it focused on Mead’s concept of role-taking and self-formation (Blumer 1937: 180–185). The paper provides an overview of the field of social psychology based on William McDougall, Charles Cooley, Gordon W. Allport and Luther Bernard, with a bit of Dewey, W. I. Thomas, and Florian Znaniecki. The chapter was not meant to elucidate Mead’s theoretical corpus and the sociological implications of his ideas. In spite of its terminological innovation, it seems dated, and this might be the reason why Blumer chose not to include this text in a pivotal collection of his essays (Blumer 1969). After this early engagement, Blumer had little to say about Mead until the 1950s, his early scholarly output containing but few references to Mead. In his 1940 article “The Problems of the Concept in Social Psychology” Blumer (1969: 178) evokes Mead’s name once, while discussing corporeal actions that “can be translated into a space-time network or brought inside of what George Mead has called the touch-sight field.” Starting in the mid-50s, Mead’s name began to crop up in Blumer’s publications with increasing frequency. In 1962 Mead reemerged as a central figure in Blumer’s writing with the publication of an essay titled “Society as Symbolic Interaction” (Blumer 1969: 78). A major assessment of Mead appeared four years later in a paper “Sociological Implications of the Thought of George Herbert Mead” (Blumer 1969; cf. pp. 61, 65, 94, 101).

About the same time Mead began to attract national and international attention, with three monographs appearing in quick succession (Victoroff 1953; Pfuetze 1954; Natanson 1956). Conferences devoted to Mead were staged from the 1960s on, supplemented by the anthologies showcasing Mead scholarship (Corti 1973; Gunter 1990; Aboulafia 1991). In Russia, Igor Kon and Dmitri Shalin (1969) examined Mead’s concept of self (this paper evolved from the thesis on Mead and Lev Vygotsky that Shalin defended in 1970 at the University of Leningrad). Miller (1973) published a study of Mead’s pragmatist philosophy which included an extensive biographical survey. In Germany, Joas wrote a comprehensive analysis of Mead’s ideas in their historical development (1979, 2001), the English translation of this influential investigation appearing in 1985. The heightened attention to Mead mirrored the growing interest in pragmatism around the world (Bernstein 1966, 1972, 1992; Rorty 1979, 1982; Habermas 1984, 1987; Campbell 1983; Tugendhat 1986; Shalin 1986a, b, 1988b, 1989, 2000, 2007, 1992; Halton 1986; Joas 1993, 1997; Shusterman 1991; 1997; Kilpinen 2000, 2013). Meanwhile, monographs have been added steadily to the body of Mead scholarship (Goff 1980; Baldwin 1986; Aboulafia 1986, 2001; Cronk 1987; Rosenthal 1991; Cook 1993; Feffer 1993; Kravchenko 2006; da Silva 2007, 2008). Several dozen Ph.D. theses on Mead were defended between 1950 and 1990 (see Mead Project, https://www.brocku.ca/MeadProject/inventory5.html for details). A recent collection of essay edited by Peter Burke and Krzytof Skowronski (2013) explores a range of ideas that are likely to inform Mead scholarship in the coming decade.

Mead and Sociological Cannon

In the 1960s, textbooks in sociological theory began to catch up with the trend by offering separate chapters on Mead. The 1966 edition of Robert Nisbet’s The Sociological Tradition still had no reference to Mead, but Don Martindale’s The Nature and Types of Sociological Theory (1960) made room for Mead, which would increase with every new edition. By 1970, all textbooks in sociological theory and history of sociological analysis allotted ample space to Mead (e.g., Coser 1971; Warshay and Warshay 1986; Bottomore and Nisbet 1971; Wallace and Wolf (1980). And when the primers in symbolic interactionism flooded the market, they all gave Mead’s ideas the pride of place (Rose 1962; Manis and Meltzer 1967; Meltzer et al. 1975).

Several rich veins in Mead’s corpus drew close attention: the temporal structure of reality (Flaherty and Fine 2001; Bergmann 1981; Maines et al. 1983), the genesis of language (Ionin 1975; Glock 1986; Koczanowicz 1994), the politics and ethics of reform (Lichtman 1970; Ropers 1973; Deegan and Burger 1978; Deegan 2008; Shalin 1988a; Schwalbe 1988), the relationship between pragmatism and sociology (Kuklik 1973; Fisher and Strauss 1979; Joas 1980, 1993; 2008; Harvey 1986; Shalin 1986a; see also Intercyberlibrary, http://cdclv.unlv.edu//archives/interactionism/index.html).

Why this sudden upsurge of interest in Mead as a sociological thinker? The decline of Marxism sent its adherents searching for a viable alternative, with some embracing postmodernism as a creed antithetical to authority and others attracted to the democratic ethos of Dewey and his pragmatist colleagues. The growing dialogue between European and American sociologists contributed to the enshrinement of Mead in the sociological cannon. Alvin Gouldner stressed Mead’s role as a vital link between the Continental and American sociological traditions. “The most important influence of Romanticism on Academic Sociology in Europe, is to be found in Weber, while its most important influence in American sociology is through George Herbert Mead. . . . The purest vein of Romanticism in American sociology is, I believe, to be found in the ‘Chicago School’, which had the most concentrated exposure to the German tradition and was, in fact, established by many (A. W. Small, W. I. Thomas and R. E. Park) who were directly trained in it. Currently, its leading exponents are Anselm Strauss, Erving Goffman and Howard Becker” (Gouldner 1970: 116 and 1973: 345; cf. Shalin 1984). Commenting on the reasons behind the growing popularity of interactionist ideas and qualitative research, Gouldner singled out the growth of the middle class in postwar America, the insecurities endemic to the periods of heightened social mobility, and the tendency to downplay class divisions and the structural sources of inequality in the United States (Gouldner 1970: 378–390).

Jeffrey Alexander (1987: 215) had a different explanation for the upsurge of the interactionist paradigm. He linked it to the confrontation between the once dominant Midwestern tradition represented by Chicago sociologists and the emerging functionalist paradigm articulated by scholars trained at Harvard and Columbia University:

Parsons was part of a younger group of American sociologists who were breaking away from the institutional and intellectual hegemony of the Chicago school, from the Pragmatic sociological tradition which – in their eyes at least – has been relatively untheoretical and individualistic. It was in 1936 that this rump group of mostly Harvard and Columbia trained sociologists had set up the American Sociological Review in opposition to the Chicago journal, the American Journal of Sociology (Alexander 1987: 215).

Symbolic interactionism, in this reckoning, was an attempt to shore up the sagging fortunes of Chicago sociology.

Sociologists trained at the University of Chicago agreed that interactionism was the “reaction to the eclipse of the Chicago school by the ascendant functionalist project” (Colomy and Brown 1995: 23), that the “general theory at Harvard under Talcott Parsons and the development of survey research and functionalism with Robert Merton and Paul Lazarsfeld at Columbia were where the new and exciting mainstreams were flowing” (Gusfield 1995: ix). However, the interactionist exegesis praised the maverick status of the Meadian tradition in its post World War II incarnation – “the Second Chicago School” (Carey 1975; Fine 1995; Abbott and Gaziano 1995; Colomy and Brown 1995). Looked at from this vantage point, “Blumer went on the offensive [in the words of Shibutani, as] ‘the spokesman for a small minority of dissidents’” (Colomy and Brown 1995, p. 23). Meanwhile, the battle for Mead’s legacy and the right to wear the successor’s mantle intensified.

Critics outside the interactionist circles were unimpressed with Blumer’s reading of Mead. “Mead’s most important successors have emphasized his indeterminacy, the strain in his work that emphasized individualism,” contended Alexander (1987: 214), “contemporary interactionism has moved away from Mead’s own institutional and collectivist thrust.” Robert Bales accused the founder of symbolic interactionism of neglecting scholarly cannons: “Blumer seems to have a way of leading us to ultimate antinomies, created by the process of abstraction, rather than to operational solutions based on a process of empirical research” (Bales 1966: 42). Echoing some of these charges, Joan Huber asserted that pragmatism-inspired research was plagued by “the ambiguity toward the logico-theoretic component of their work,” the failure “to spell in advance and in detail what is expected,” and the inability to “integrate the findings into a larger body of work” (Huber 1973a: 62–63; Blumer 1973; Huber 1973b). Jenice Reynolds and Larry Reynolds (1973) blamed symbolic interactionism for its “astructural bias” and the failure to come to grips with the realities of social class and power (see also Shaskolsky 1970; Ropers 1973). Warshay and Warshay (1986: 265, 260) contended that Blumer was guilty of “the individualizing and subjectivizing of Mead,” that his codification “went against Mead’s [insistence on] working from the outside to the inside rather than the other way around,” and that the attempt to latch on to Mead’s legacy could be seen as a stab at the “legitimation of Blumer.”

The polemics within the interactionist community were every bit as heated, with two competing interpretations of Mead vying for supremacy. The first current, symbolic interactionism, stressed the emergent character of social interaction and traced its roots to Chicago sociologists who did their pioneering work between 1920 and 1940. Sociologists working in this tradition today generally accept Blumer’s reading of Mead. The second current, structural interactionism, was associated with the Iowa school and concentrated on building and testing theory of the self. Manfred Kuhn spearheaded the work of this group but his untimely death prevented him from establishing himself as its leader. His followers looked to remedy what they perceived to be the weakness of Blumer’s symbolic interactionism – disregard for the structural properties of everyday interactions. Thus, the battle for Mead’s legacy was joined.

Symbolic Interactionism and Chicago Sociology

Mead knew personally the principal figures in early Chicago sociology, with one of whom, W. I. Thomas, he had a close relationship. When in later years he surveyed the evolving domain of sociology, he referred to his colleagues as standard bearers in the new science of society. Here is Mead’s take on sociology as gleaned from a fragment dating back to the 1920s: “In recent years another discipline emerged, that of sociology. Both its definitions and its procedure are uncertain. It approaches nearer to a physiology of the social process than to the anatomy of social institutions. It is interested in the sociality that finds its expression in institutions and their functions, and it is particularly interested in this sociality as an expression of the experience of the individual; it has a special affinity, therefore, with social psychology. [The sociological] science is in this sense a more general one than that of the other social sciences” (Mead 1938: 498–499). As this fragment suggests, Mead saw Chicago sociology as akin to, if not identical with, social psychology, a field focused on the dynamics of self and society: “[I]t is evident that the acceptance by the sociologists of a society of selves in advance of inner experiences opens the door to an analysis which is behavioristic. I refer to such analyses as those of W. I. Thomas, Park and Burgess, and Faris” (Mead 1956: 305–306). Mead goes on to observe that “in many respects Cooley’s analyses are of this type, but they always presuppose a certain normal social order and process as given. . . . To this type of analysis Cooley’s assumption of the psychical nature of society closes the door. And it commits him to a conception of society which is mental rather than scientific.” By juxtaposing Chicago behaviorism to Cooley’s introspectivism, Mead signals his preference for empirical methods of research. From the 1920s on, W. I. Thomas, Robert Park, Ernest Burgess, Louis Wirth, Ellsworth Faris, and their students seized on the process-oriented pragmatism and boldly applied its tenets to empirical problems that the urban environment in Chicago and elsewhere posed to the budding sociologists.

Among the young scholars exposed to the Chicago tradition was Herbert Blumer. He wrote his thesis under Ellsworth Faris and defended it in 1928. As a graduate student, Blumer took classes with Mead, for whom he served as a teaching assistant. A term paper Blumer wrote for one of Mead’s classes bears his teacher’s comment: “A most satisfactory statement opening up the field of the world of objects within which the self arises” (Blumer 2004: 1). When Mead fell ill, he asked Blumer, then an assistant professor at the sociology department, to teach his class in advanced social psychology. Blumer took over the class and kept teaching it for several decades until he moved to Berkeley in 1952. He was a key figure in training the younger generation of scholars who formed the Second Chicago School (Fine 1995). Strauss recalls that he heard about Mead not from Everett C. Hughes or other faculty members but from Blumer who directed him to Mead’s work, which would become locus classicus for the postwar generation of Chicago sociologists. “George Herbert Mead was, as we referred to it, the Bible,” remembers another student of Blumer (Gusfield, cited in Fine 1995: 6). “We read Mead with great intensity, discussed it with great intensity, and the symbolic interactionist orientation seemed to us at the center of what sociology was. A few years ago, Anselm Strauss remarked to me that ‘We didn’t think symbolic interaction was a perspective in sociology; we thought it was sociology.’”

While Blumer brought up Mead with increasing frequency in the 1950s, he did not attempt to codify his sociological position until the following decade when, scorned by critics and urged by supporters, he published a collection of his papers Symbolic Interactionism: Perspective and Method. Except for the sixty page introductory essay written for this occasion, the book was comprised of previously published material. In his lead article titled “The Methodological Position of Symbolic Interactionism,” Blumer articulated the theoretical premises and provided methodological blueprints for sociologists who sought in William James, Dewey, and Mead their philosophical ammunition. Blumer’s own empirical work was in the area of collective action, labor negotiations, and public opinion (Blumer 1969, chs. 11–12). In the last three decades of his intellectual career he was preoccupied with codifying the theoretical and methodological principles embedded in Mead’s corpus.

“A cardinal principle of symbolic interactionism,” according Blumer (1969: 7), is that “any empirically oriented scheme of human society, however derived, must respect the fact that in the first and last instance human society consists of people engaging in action.” Three tenets undergirding symbolic interactionism (SI) stipulated (a) “that human beings act toward things on the basis of the meanings that the things have for them,” (b) “that the meaning of such things is derived from, or arises out of, the social interaction that one has with one’s fellows,” and (c) “that these meanings are handled in, and modified through, an interpretation process used by the person in dealing with the things he encounters” (Blumer 1969: 2). These postulates were supplemented by six “basic ideas, or ‘root images’” framing the SI perspective on social groups, social interaction, social objects, human beings, individual action, and the joint social act (Blumer 1969: 6). These root images (1) defined group life as a network of acts fitted together by individual and collective actors; (2) cast human agency as a protean force shaped by and continuously reshaping society; (3) drew attention to the symbolic media informing the individual’s attitude toward the objects; (4) emphasized the centrality of self-consciousness in social life; (5) singled out the agent’s ability to interact with itself and consciously construct one’s own action; and (6) put forward the notion of a “joint act” as a central unit of sociological analysis (Blumer 1969: 6–20). These precepts were meant to demarcate the SI approach from the Parsonian perspective that “ascribes behavior to such factors as status position, cultural prescriptions, norms, values, sanctions, role demands, and social system requirements” (Blumer 1969: 7).

The SI creed drew fire inside and outside the interactionist community. As we size up this controversy, we should remind ourselves that Blumer wasn’t just a theoretician, that he inspired a generation of researchers who went into the field and came back with valuable ethnographies. On that score, SI was a success. The second generation of Chicago sociologists have studied deviance and drug use (Lindesmith 1947; Becker 1953, 1955, 1963; Matza 1969); the medical profession and its training (Becker and Geer 1958; Becker et al. 1961); total institutions and stigma (Goffman 1961, 1963); alcoholism and the temperance movement (Gusfield 1963; Cavan 1966; Wiseman 1979), and much more (Rose 1962; Gans 1962; Meltzer 1959; Manis and Meltzer 1967; Meltzer et al. 1975; Glassner 1980; McCarthy 1984; Perinbanayagam 1975, 2005). Methodologically, the work done in this tradition required participant observation and field work, sensitivity to local scene and changing realities, theorizing in situ and fitting concepts to emergent transformations. Here is how Joseph Gusfield captured the ethos of Chicago sociology in the late 1940s and 1950s:

The preference for descriptive material and observation made us suspicious of records and questionnaires – of data torn from the context of their creation. Action was too situated, too contextual to be understood at the high levels of much macroanalysis. Meanings were not often assuredly understandable without an experience with those we were describing. The simple survey, probably the major contribution of sociology to American life, is still viewed with a degree of doubt and humility. . . . [T]he Chicago school was open to a more artistic, improvised, and situated mode of sociology than implied in the tenets of research design. . . . The second school shared with the first an identification with the less respected, less established elements in the society and a notable dose of skepticism and disrespect for the well-off, the authoritative, and the official (Gusfield 1995: xii-xiii).

Not all sociologists claiming Mead’s legacy were impressed with this framing of Chicago sociology. Joas (1981, CA III, p. 85, 94) found Blumer’s interpretation wrongheaded, contending that his writings “cannot be regarded as the authoritative interpretation of the works of Mead or other scholars of the Chicago tradition” and proposing an alternative interpretation based on “theory of practical intersubjectivity.” A different line of attack on Blumer came from scholars embracing Mead’s affinity with Charles Peirce. According to David Lewis, “there were two opposing branches of American pragmatism: (1) the social realism of Peirce’s pragmatism, and (2) the social nominalism of James’s pragmatism. The writings of Dewy and Mead reflect the influences of both of these branches. Dewey, however, was more closely aligned with James while Mead’s work is more consistent with Peirce’s principles” (Lewis 1976: 137; cf. Alexander 1987: 203). Blumer’s symbolic interactionism, Lewis charged, was subjectivist and individualist and ultimately antithetical to Mead’s behaviorism and collectivism, insofar as Blumer postulated “the self definition process as occurring in an indeterministic system in which a freewheeling, unique individual ‘defines’ things as he sees fit, and the parameters of that definition are not constrained by the social structure of that individual’s society” (Lewis 1976: 148; Lewis and Smith 1980).

Not one to shrink from a challenge, Blumer vigorously defended his position against critics (1973, 1975, 1977, 1979, 1980). To structural functionalists Blumer answered that they have “an erroneous understanding of the symbolic interactionist point of view that stems from George Herbert Mead, a misunderstanding traceable in large measure to Parsons’ failure to grasp the nature of the ‘self’ as seen by Mead. For Mead, the self is far more than an ‘internalization of components of social structure and culture.’ It is more centrally a social process, a process of self-interaction in which the human actor indicates to himself matters that confront him in the situation in which he acts, and organizes his action through his interpretation of such matters” (Blumer 1975: 136). Mocking Bales, Blumer wrote that “Mead must shudder in his grave at such butchering of his thought. For Mead an object is something that is designated and does not exist apart from designation. . . . If there is any single matter that is abundantly shown in Mead’s lengthy analysis of ‘object,’ it is that objects do not have an inherent or self-conscious character, this is a cornerstone of his pragmatic position” (Blumer and Bales 1966: 45–46). To Lewis’s criticism, Blumer responded that “this is a wild and foolish caricature of the view that I have presented of the human being as an actor. It ignores (1) that the actor has to fit his developing action into the on-going actions of others and thus is necessarily constrained by their actions; (2) that the actor in defining his situation is guided by the scheme of definition that he has got from the group; (3) and that in forming his action the actor is guided by the generalized roles from which he addressed himself” (Blumer 1977: 154). Blumer was equally unsparing to Clark McPhail and Cynthia Rexroat, charging that their social behaviorism fits John B. Watson’s scheme more than Mead’s, that “Mead did not have such a debased view of human social behavior as to equate it with reflex movements,” and that Mead would have rejected “such a constricted view of human social behavior as to confine it to small portions of simple, overt, quantifiable actions” (Blumer 1980: 177; cf. Blumer 1973).

While vigorously defending the symbolic interactionist paradigm, Blumer continued to refine his formulations. He went some distance to join issues with his critics, softening his opposition to the insights championed by the paradigms he once rejected outright. We can see this in the nuanced treatment Blumer (1981) gave Mead and interactionism in a little known paper he wrote a few years before his death. Published in the volume The Future of the Sociological Classics, Blumer acknowledged in this piece that much of human interactions is guided by precepts “which are already patterned by previous understanding,” that “fixed and repetitive social acts are abundantly familiar to us, and the social life of any human society, certainly of one in a settled state, is made up predominantly of such routines and rituals” (Blumer 1981: 150). His formulations now gave more prominence to the generalized other as a collective agency guiding individual conduct: “To take the role of the ‘generalized other’ means that the participant is indicating to himself the way in which his ‘community’ views the situation in which he has to act. Such interpretations using the community definitions or symbols are made at the beginning of a social act in order to identify the social act and thus get one’s bearings. . . . The interpretations which lay the basis for his act arise out of the process of taking community roles, taking the roles of other participants and judging his own intentions and prospective action” (Blumer 1981: 151). The paper acknowledged the reality of power, the fact that humans “make adjustments of their actions through coercion,” and it widened room for cultural imperatives: “Without doubt, there are many instances of joint or collective social action which are culturally prescribed (as in the case of routinized and ritualized behavior) or which, in Mead’s language, are under the sole control of the generalized other” (Blumer 1981: 153). Power and conflict, Blumer (1981: 156–157) contended, are endemic to the social scene: “[T]he struggle for power or the exercise of power takes place . . . in those acts in which participants have to vie with one another with the deliberate control of a developing social act . . . in one situation a given participant may assess available resources as giving him control over the developing social act, whereas in another situation involving the same participants he may interpret his position as weak.”

It is unlikely that Blumer saw his mature statement as anything other than clarifications of what he had said before. He did not mean to cede theoretical ground as much as to assert that his formulations were flexible enough to accommodate issues central to other sociological perspectives without abandoning the research focus on flesh and blood humans engaged in everyday life. Yet, the new emphasis on robust community definitions and structural constraints refocused the interactionist inquiry on routinized behavior that did not figure prominently in his earlier formulations. Even in its revised form, Blumer’s interactionism remained largely ahistorical and less than amenable to structural analysis, as he shunned an inquiry into the origin and evolution of cultural patterns and had little to say about the unequal distribution of definitional resources. It is true that the imperatives embedded in a particular social structure do not automatically force interactants to assume the combat position (why else would Marxists keep chasing laborers in the hope of raising their class consciousness). Still, the power affecting the outcome of a definitional conflict is not just in the eyes of the beholder. Where the individual’s agenda is implicated, the same thing may be put to different uses and acquire situationally specific meaning, yet in most cases the hammer will be used to hammer a nail, the Taliban male will force his wife to wear a burqa, and the boss’s interpretation will carry more weight than that of a subordinate. The balance of power as it transpires in a situation does not originate within its fold, and it requires the kind of analysis that Blumer’s interactionism is ill-suited for.

Certain precepts found in Mead do not readily fit into the SI schema, notably the notion of the generalized other. Blumer grappled with this concept mightily, especially in his later years (see Blumer 1981: 167). We can see how much he was bothered by its implications in his posthumous publications, notably in his correspondence with David Miller reprinted in the volume edited by Blumer’s student, Thomas Morrione (Blumer 2004). In a letter of May 7, 1979, Blumer asks his friend, “How do you regard and treat the ‘generalized other’ when one is required to speak of the voices of the community? It is evident that in many societies, particularly modern societies, the community or the society does not speak with a single voice. . . . I do not find that Mead dealt with this matter of divided, contradictory, or uncertain ‘voices’ of the community” (Blumer 2004: 110). In another letter Blumer says, “The more I think about Mead’s concept of the generalized other the more I am led to believe that the concept stands for a ‘mechanism’ of adjustment rather than for a given organization of one’s society . . . one can ‘cut out’ a generalized other from the community life that is different from generalized other cut out by other members of the community” (Blumer 2004: 117, 112). What Blumer was getting at was the conflation of two perspectives – one taken by outside observers expostulating the dominant interpretations, the other assumed by flesh and blood agents liable to misread the norms and eager to sell their interpretations to others. “The voice of the community as delineated by a skillful outside observer might be very different from the way in which the voice appeared in the makeup of the individual. . . . I do not think a skillful outsider can, by observing the behavior of people, know for sure what the G.O. is. He may make a good guess, but actual behavior often falls short of our ideals. And the G.O. stipulates how one ought to behave, not how he necessarily actually behaves. I believe Mead holds that by reflective thinking . . . we continuously change the G.O., but only piece-meal, so that there is a continuity between individualism and social change that builds on the past and uses the G.O as a servant to an open, progressive society of open selves” (Blumer 2004: 124–115).

In assessing Blumer’s legacy, we need to bear in mind that his perspective was primarily, if not exclusively, sitogenetic (Shalin 2000: 315). Blumer does not dwell on ontogeny – the socialization process though which the actor learns the ways of the generalized other and acquires a standard repertoire of roles. Nor does he pursue the phylogenetic inquiry, the historical explorations of how selfhood and cultural patterns emerge, norms are structured, and definitions are disseminated and enforced (Mead 1932, 1938). His focus is on mature, autonomous agents pursuing their personal and collective agendas in situ, especially in relatively unstructured settings. Much that happens in such situations is emergent and unpredictable, although the extent to which actors are free to impose their vision on the situation remains underconceptualized in Blumer.

It should be noted, also, that theoretical framing aside, empirical studies of Chicago sociologists trained and influenced by Blumer showed sensitivity to the power dynamics and structural constrains in society. Joseph Gusfield’s inquiry into the temperance movement starts with the cultural tradition and the network of institutions within which the alcohol abusers have to operate, and then demonstrates how and why these institutions fail to achieve their stated goals (Gusfield 1963). Jacqueline Wiseman (1979) mingled with people on skid row, the outcasts buffeted by the organizational networks and channeled through the assorted rehab facilities, yet she never lost sight of their agency and skillfully reconstructed the ingenious ways in which they worked the system to their advantage (Wiseman 1979). In her study of a Chicano community and street gangs, Ruth Horowitz (1983) discovered that the informal, freewheeling interactions are powerfully constricted by standards of honor and status hierarchy permeating this community. Gary Alan Fine researched the social context within which members of society acquire their reputations (2001), examined how interpersonal networks coalesce into enduring publics (2012), and explored with Patricia Turner (2004) how racial stereotypes formed in local communities perpetuate discriminative practice. Steeped in the symbolic interactionist tradition, such studies “focus on the total control by institutions and communities, as revealed in capitalism, culture, and microsystems” (Fine and Ducharme 1995: 128). They show that the interactionist concern for structure, power, and inequality is not ephemeral. We can see this most clearly in the work of Anselm Strauss and his associates (Strauss 1978, 1994) who have studied social orders as incessantly accomplished and transformed through negotiations by skilled groups of participants locked in adversarial/cooperative relations in hospitals and other large-scale organizations.

To the critics concerned about the lax methodology employed by symbolic interactionists, we should point out that the reluctance of Chicago interactionists to spell out their theories and hypotheses in advance was not an oversight. Rather, we should see it as a principled stance, as a “methodological expressivism” according to which researchers must act first, engage with reality, form abductive hypotheses, and build the theory from the ground up while clarifying their assumptions and perfecting their conceptual tools. Interactionist methodology avoids premature closure of the theoretical process, bidding its practitioners to remain on the lookout for the unforeseen and serendipitous in order to “discover[.] new categories as emergencies of the group’s changing activities” (Lindeman 1924: 192). Chicago sociologists are committed to what John Dewey and Arthur Bentley (1949: 104) called “unfractured observation” which binds the knower to the known, allowing the researcher to ascertain the questions to be asked and gauge reality by the indexical reactions that obdurate properties of the situation elicit from the knower’s bodymind. Looked at in the perspective, social research is a search for right questions to ask. “As I sat and listened,” recalls William F. Whyte (1943/1981: 303), “I learned the answers to questions that I would not even have had the sense to ask if I had been getting my information solely on an interviewing basis.” This is the rationale behind the “grounded theory” articulated by Barney Glaser and Anselm Strauss (1964) who demonstrated that participant observation is not inimical to systematic theory, provided we leave room for creative engagement with the emergent reality under investigation.

More work needs to be done, both theoretical and methodological, to shore up the Chicago tradition in sociology. Among the key challenges symbolic interactionists are facing is how to incorporate into their framework a full-blown theory of pragmatically understood social structure. This is the route worth taking, even if that means crossing the conceptual divides and reaching out to colleagues with alternative readings of Mead.

Structural Interactionism and the Iowa School

Another brand of interactionism found inspiration in Mead’s behaviorism. In 1979, McPhail and Rexroat published a strongly worded paper, “Mead vs. Blumer: The Divergent Methodological Perspectives of Social Behaviorism and Symbolic Interactionism” where they claimed that Blumer’s construction “bears no resemblance to Mead’s position,” that “Blumer’s naturalistic inquiry neither complements Mead’s methodological perspective nor facilitates the examination of Mead’s theoretical ideas,” that “Mead’s treatment of hypotheses is theoretically grounded, though not formally deductive, [whereas] Blumer’s treatment is virtually atheoretical, inductive empiricism” (McPhail and Rexroat 1979: 453–459). This line of attack recycled the charges that Manfred Kuhn aired in the early sixties when he announced the end of the “long era of the ‘oral tradition’” in symbolic interactionism, the era marked by the unsystematic attempts to “get it right,” “debate over orthodoxy,” “casuistry and criticism.” Now that this era is coming to a close, we enter “the age of inquiry in symbolic interactionism” marked by empirical investigation with all its “drudgery of testing and justification” (Kuhn 1964: 11–13). Kuhn singled out “Blumer, the young and promising heir apparent, [who] has published relatively little and has nowhere gathered together a rounded version of his point of view,” and who failed to transcend “the essential ambiguities and contradictions in the Mead statement – ambiguities and contradictions which were generally interpreted to be dark and inscrutable complexities too difficult to understand as long as the orientation remained largely in the oral tradition. Much of this confusion [reflects] a contradiction between determinacy and indeterminacy in Mead’s over-all point of view” (Kuhn 1964: 11–13). After reviewing the empirically grounded studies of self-identity, Kuhn outlined a program of quantitative research designed to move past “conjectural and deductive orientation – as represented by Cooley, Dewey, and Mead [toward] a derivative but developing set of generalizations, tested by empirical research” (Kuhn 1964: 18).

The name “Iowa school” as this strand of interactionism is called reflects Kuhn’s affiliation with the University of Iowa, although the Self Theory (ST) tradition is not confined to this locale. In a programmatic piece published a year after his untimely death in 1963, Kuhn cast his conceptual net broadly: “Perhaps the most significant contribution of the Iowa school research is simply that in which it joins the research of Miyamoto and Dornbusch, Deutsch and Solomon, Dick, Dintz and Mangus, McKee and Sheriffs, Stryker, Videbeck and Bates, and many others in demonstrating to some degree at least that the key ideas if symbolic interactionism could be operationalized and utilized successfully in empirical research” (Kuhn and McPartland 1954: 18). Remaining faithful to Mead’s tenets, most notably to his emphasis on the homology between the structure of the group and the structure of the self, these scholars set out to test Mead’s precepts.

Kuhn’s main contribution to interactionist research was his “Twenty Question Test” (TST). He began to administer it in the early 1950s to subjects who were presented with a single sheet of paper and asked to answer a simple question, “Who am I?” The instructions were as follows: “There are twenty numbered blanks on the page below. Please write twenty answers to the simple question ‘Who am I?’ in the blanks. Just give twenty different answers to this question. Answer as if you were giving the answers to yourself, not to somebody else. Write the answers in the order that they occur to you. Don’t worry about logic or ‘importance.’ Go along fairly fast, for time is limited” (Kuhn and McPartland 1954: 114). This procedure was premised on the notion that the self is best understood and operationalized as an attitude or a set of self-selected attributes. Such a “conceptualization is most consistent with Mead’s view of the self as an object which is in most respects like other objects, and with his further views that an object is a plan of action (an attitude)” (Kuhn and McPartland 1954: 112). Self Theory postulated that “the self is an interiorization of one’s positions in social systems. One may assume from this orientation that variations in such self-identifications are equivalents of variations in the ways in which individuals in a society such as ours have cast their lot within the range of possible reference groups” (Kuhn and McPartland 1954: 118).

Kuhn and his coauthors reported several findings on how humans frame their selves, including the general preference for the positional self-characterizations such as “girl,” “husband,” “Baptist” (Kuhn calls them “consensual”) over the attributional self-framings like “happy,” “pretty,” “bored” (those are labeled “subconsensual”). One finding the authors singled out concerned religious affiliation: respondents with non-mainstream religious preferences (e.g., Roman Catholics, members of small sects) tended to bring up their religious credentials earlier in their answers than those in mainstream religious and nonreligious groupings. Self-attitudes reflect our standing in society, the authors concluded, with the more prominent positions marked by a greater salience in our self-conception as revealed by TST. In assessing their technique, the researchers expressed confidence that “self-attitudes might be studied in a fairly direct manner by collecting statements of role preference and role avoidance, role expectations, models for the self, and the like” (Kuhn and McPartland 1954: 113).

Shortly after this ground-breaking article, Frank Miyamoto and Sanford Dornbush (1956) published an influential study, “A Test of Symbolic Interactionist Hypothesis of Self-Conception.” Taking a clue from Mead’s concept of the generalized other, the researchers hypothesized that our self-assessments mirror the significant others’ opinions about ourselves. Working with 195 students drawn from college classes, sororities, and fraternities, the authors invited each respondent to (1) rate oneself on four personality traits, (2) rate every group member on the same characteristics, (3) predict how each group member would rate the respondent, and (4) assess how the group as a whole would rate the respondent on the designated personality traits. The researchers found a significant positive relationship between self-ratings on the assorted qualities (intelligence, self-confidence, physical attractiveness, and likability) and the ratings they received from other group members. The self-concept was more closely associated with the respondents’ perception of how others rated them than with the actual rating by significant others. The correlation between the self-rating and the respondents’ estimate of how the group as a whole rated them was stronger than the association between the self-rating and the respondent’s prediction of how separate group members assessed the respondent. The generalized other – objectively measured and subjectively perceived – turned out to be linked to the individual’s self-concept.

This design was replicated by Reader et al. (1960) and E. L. Quarantelli and Joseph Cooper (1966) who confirmed that “(1) the responses of others have an influence in shaping one’s self-definition and (2) that this self-conception is derived chiefly from the perception of the ‘generalized other’” (Reader et al. 1960: 76). In an interesting twist, the researchers discovered that “the self-conception of persons who do not think highly of themselves . . . appears to be determined largely by the perceived and actual responses of others. For those with a healthy ego, additional variables are necessary to explain self- conception” (Reader et al. 1960: 76). The authors hypothesized “that the subjects who have high self-ranking are operating with values derived from groups other than the ones with which we are concerned, while those with low self-ranking may judge their performance primarily on the basis of the values of their particular groups” (Reader et al. 1960: 76). Quarantelli and Cooper refined this experimental design by adding a time dimension. But their hypothesis that “anticipated self-rating is closer to the mean perceived future response of others to the actor than to the mean actual future responses of others” received only partial support (Quarantelli and Cooper 1966).

In theoretical terms, it should be noted, the ST main finding that “it is the perceived rather than the actual response of others that is the more important in the formation of self-concept” (Quarantelli and Cooper 1966, p. 283) echoes Blumer’s position – external variables affect behavior insofar as they are mediated by the agent’s interpretation. The question to be posed here is whether the aggregated response of significant others is a proper way of operationalizing the generalized other. Assessing the role that structural properties play in social interaction would become a priority for the thinkers like Ralf Turner, Peter Rose, Sheldon Stryker, George McCall and J. L. Simmons, structurally-minded interactionists looking for ways to apprehend the impact on behavior of the institutionally prescribed and publicly enforced expectations.

Already in his early writing, Stryker sought to balance Blumer’s interactionism with a version that makes room for structural constraints and the guiding role of language in shaping our identities and interactions. The world we inhabit is preinterpreted, insisted Stryker; the objects we encounter in everyday life, most certainly other human beings, have already been named, and we are not at liberty to ignore these collectively inaugurated objects. The general proposition he came up with in 1959 hewed closely to the SI premises: “Humans respond to a classified world, one whose salient features are named and placed into categories indicating their significance for behavior. In short, humans do not respond to the environment as physically given, but an environment as it is mediated through symbols – to a symbolic environment” (Stryker 1959: 438). Stryker was quick to elevate “a particularly important kind of category [which] is that called ‘position.’ Positions are socially recognized categories of actors, any general category serving to classify persons: father, sergeant, teacher are positions by this usage, as are playboy, intellectual, blacksheep” (Stryker 1959: 439). Our group affiliations and designated positions channel our behavior, constraining our capacity to impose a definition on the situation. The agent’s power to define the situated objects, particularly in the early stages of socialization, is no greater than our ability to rename things. Those unwilling to accept the names society gives them will be painfully reminded who they are and compelled to play the part. To make room for human agency and choice, Stryker (1968) introduced the notion of “identity salience” that signifies the place of a given identity in the hierarchy of selves peculiar to the individual or a group. This notion was paired with the concept of “commitment” that referred to behavioral alternatives available to and disparately favored by the actor. In his 1980 book, Symbolic Interaction: A Social Structural Version, Stryker laid down several postulates and affiliated hypotheses in a bid to advance a research program that he and his colleagues have pursued for several decades (McCall and Simmons 1978; Stryker 1968; Stryker and Serpe 1982; Serpe and Stryker 1987; Stryker and Vryan 2003; Burke 1980, 1991; Stryker and Burke 2000; Burke 2004; Burke and Cast 1997; cf. Maines 1977). The first two propositions on which this research program was based read as follows (Stryker 1980: 53–54):

Behavior is premised on a named or classified world. The names or class terms attached to aspects of the environment, both physical and social, carry meaning in the form of shared behavioral expectations that grow out of social interaction. From interaction with others, one learns how to classify objects one comes into contact with, and in that process also learns how one is expected to behave with reference to those objects. . . . Among the class terms learned in interaction are the symbols that are used to designate ‘positions,’ which are the relatively stable, morphological components of social structure. These positions carry the shared behavioral expectations that are conventionally labeled ‘roles’.

The world of social objects forms an expectation matrix from which agents derive their identities and into which they weave their actions as they strive to legitimize their personas in the eyes of other community members. The named worlds we inhabit furnish us with the resources facilitating or impeding our opportunity to claim and implement a given self. The unequal access to resources is a fact of life that situational opportunities available in routine encounters cannot fully neutralize. Interactants are also bound by the commitments they developed in their long-term relationships, which further constrain their identity choices. There is room for negotiations about the roles to be enacted, but situations will vary greatly in the amount of definitional freedom they accord to participants in following their preferred scripts. Role-making and role-taking (Turner 1962) are two sides of the social process balanced in a predictable way, with the role-making opportunities tethered to the power implicit in the authority and bound to asymmetrical positional capital available to interactants. While rebellion or withdrawal is an option, the cost of straying too far from the script can be prohibitive, compelling actors to trim their improvisational forays and bring their dissent in line with what the social structure withstands without crashing the willful agent. Taking into account the identity hierarchy, Stryker predicted, the individual will implement the self that is most salient, reflecting the person’s strongest commitments, promising the best rewards, and meeting with the approval of the most significant others. With this theoretical framework in place, Stryker and his associates proceeded to inquire how individuals affiliated with different groups structure their selves, which identities are likely to be activated when the situations call for contradictory responses, who is more likely to dissent in conflict prone situations, and how activists involved in protest movements rationalize their self-preferences and square off their deeds with their stated beliefs and identity commitments (Stryker and Burke 2000; Stryker and Vryan 2003; Burke 2004). As Peter Burke (2004: 16) pointed out, “Stryker’s approach maintains the importance of the individual as an agent, especially in the concept of role-making, but recognizes it to be limited to the extent that the identity is embedded in a more closed structure, and when greater commitment is due to the many others with whom one interacts in relation to the identity.”

As this exposition indicates, the Iowa School interactionists moved structural constraints to the center stage of theory and research grounded in Mead. Structurally-oriented interactionism recognizes that unequally distributed resources limit the agent’s definitional prowess. They are well positioned to explain the dual role that symbolic media plays in constraining conduct and enabling social creativity. Self Theory research strategies are also better aligned with the scientific canons of theory building and hypothesis testing. This branch of interactionist sociology has blind spots of its own, its methodological advance achieved at a price.

While ST researchers acknowledge the creative powers of human agency, they do not make much conceptual room for it, leaving its somatic-affective underpinnings unexplored. Refocusing ST on the agency’s noncognitive facets would strengthen structuralist interactionism. The ST conceptual framework is also weak on describing society-in-the-making, society as a structured process that simultaneously reproduces and updates its systemic properties. “Structure,” as Dewey ([1929]1958:72) noted, is “an evident order of changes. The isolation of structure from the changes whose stable ordering it is, renders it mysterious. . . .” “You cannot have a process without some sort of a structure,” agreed Mead, “and yet the structure is simply something that expresses this process as it takes place” (1936:164); cf. Conner and Dickens 2015. Interactionists steeped in participant observation have reasons to be uneasy about substituting paper-and-pencil outputs with overt behavior unfolding over time. Efforts to employ qualitative methods in identity studies have been made in the past, which included videotaping and systematic monitoring role-taking behavior, but questions about the ecological validity of the relevant generalizations remain (see the methodology section in McCall and Simmons 1978: 252–273).

One last point about the future direction of ST research should be made here. Mead’s writing on language has very little to say about grammar – the structural dimension of linguistic communication, and the ST framework can rectify the situation. Its proponents need to check neighboring disciplines like sociolinguistics and anthropology where researchers grapple with the notions of “emergent grammar” and “grammaticalization” (Hopper 1987, 1988, 1998; Hopper and Traugott 2003; Helasvuo 2009; Edwardes 2010). The work of Paul Hopper and his colleagues deserves special attention in this respect as it echoes pragmatist ideas and deploys strategies familiar to interactionist sociologists. Just as interactionists struggle to do justice to social norms, the linguists steeped in everyday discourse and attuned to pragmatics strain to accommodate paradigmatic qualities of grammatical forms.

“[G]rammar is always emergent but never present,” “it never exists as such, but is always coming into being,” [t]here is, in other words, no ‘grammar’ but only ‘grammaticization’ – movements toward structure which are often characterizable in typical ways” (Hopper 1987: 148). The “emergence of grammar as byproduct of usage and frequency is complicated by the role of institutional norms of various kinds that may artificially extend or restrict the range of emergent regularity and fix standardized forms as dictionaries, grammar books, style manuals, and many institutional artifacts,” explains Hopper (1998: 160). “Emergent Grammar is meant to suggest that structure, or regularity, comes out of discourse and is shaped by discourse as much as it shapes discourse in an on-going process. Grammar is hence not to be understood as a pre-requisite for discourse, a prior possession attributable in identical form to both speaker and hearer. Its forms are not fixed templates but are negotiable in face-to-face interaction in ways that reflect the individual speakers’ past experience of these forms, and their assessment of the present context, including especially their interlocutors” (1987: 142). Martin Edwards, a linguistic anthropologist, took one step further in foregrounding the isomorphism between societal and linguistic structures thus: “Human societies have all the features we find in grammar. They are segmented, consisting of individuals who are nonetheless capable of working together to produce solutions where individuals cannot. They are differentiated, with individual taking particular complementary roles to solve problems. . . . Finally, societies are hierarchical, with some individuals deferring to others in ritualized ways which sometimes appear not to promote the deferring individuals’ reproductive fitness” (Edwards 2007: 24).

There are key differences in the way linguistic and societal structures come into being, and these differences must be spelled out. Yet interactionists would be remiss if they ignore the recent developments in neighboring social sciences where scholars have been experimenting with kindred conceptual strategies that allow framing structural processes while leaving ample room for human agency and emergent transformations.

Conclusion

Coming back to the Weber-Mead comparison drawn by Gerth and Mills, we can make several observations about the canonization process and framing strategies that disciples and followers deploy in rendering a classic thinker. Writings, encapsulations, conceptual strategies deployed by a canonical author are usually too diverse and paradoxical to lend themselves to an immaculate rendering. What makes a corpus classicus exemplary is that it often straddles more than one discourse and strives to bring into one continuum ideas which contemporaries find contradictory. This sensitivity to seemingly incompatible ideas is often lost on the followers who maximize one insight at the expense of the other in their quest for an orthodox reading and control over the cannon. Weber’s inquiry into modernity’s bureaucratic proclivities and the iron cage of rationalization give ample fodder to theorists like Parsons who championed society’s structural and integrative dimensions at the expense of its voluntaristic and emergent properties. For Gerth and Mills, on the other hand, Weber was first and foremost a liberal thinker preoccupied with charismatic transcendence and historical transformation of the status quo. Interpreters tend to fall on different sides of this divide, but the challenge is to grasp the original mind in all its contradictions and possibilities.

A similar penchant for bridging contradictions we find in Mead. Some of his followers build on this thinker’s abiding concern with the collective nature and institutional moorings of mind, self-identity, and human interactions – hence, their preoccupation with the generalized other, institutional constraints, and the normative “Me.” Others latch onto the creative underpinnings of agency, emergent properties of social interaction, critical abilities of the fully developed mind, and the “I” phase of personality dynamics. Drawing on a Weberian vocabulary, we can say that Blumer zeroed in on charisma-writ-small or situational charisma – the agent’s ability to reimagine the established world and bring into existence an alternative reality, however fleeting and marginal the emergent transformation might be. For structural interactionists like Stryker and Burke, agentic creativity has a good deal to do with the institutionally determined and unequally distributed resources. Each group of interpreters appeals to Mead as a way to support their reading and finds ample verbiage to back up their views. Yet each tends to miss the dialectical sensibilities feeding the founder’s pragmatist imagination.

Mead shuns the dichotomies commonly found in social theory – nature and culture, behavior and institution, self and society and looks for a conceptual strategy that places polar terms on an equal footing, rendering each contingent on its other. “Human society as we know it could not exist without minds and selves, since all its most characteristic features presuppose the possession of minds and selves by its individual members; but its individual members would not possess minds and selves if these had not arisen within or emerged out of the human social process. . . . The organization and unification of a social group is identical with the organization and unification of any one of the selves arising within the social process in which that group is engaged” (Mead 1934: 144). Given this dialectical approach, it is understandable if his interpreters choose a particular strand in his sprawling theoretical corpus and develop it without special regard to the overall design. If the canonization process under review has been overheated at times, it is not because interactionists are an especially cantankerous bunch, spirited though they sometimes are, but because Mead left no systematic statement of his sociological theory and, more importantly, because his formulations lend themselves to alternative interpretations. “[T]he commentators claim to discover a new Mead, altogether different from the older Mead we thought we knew,” observed Fine and Kleinman (1986: 215), but “the new Mead is, of course, the old Mead somewhat twisted out of shape” (Johnson and Shifflet 1981: 109). It matters, too, that the successful claim to the framer’s mantle carries with it the promise of a power base within the discipline. Continuing the analogy, we can say that the open canon politics of interpretation is aligned with legal pragmatism that envisions the Living Constitution as a constantly evolving document whose meaning is historically emergent and inseparable from practice (Shalin 2011: 239–295).

In the light of prior discussion, I propose to distingish two interpetive strategies for framing the classic and constructing the canon—the politics of orthodoxy and politics of the open canon. The politics of orthodoxy aims to ascertain the original meaning and close the textual canon, creating a template relative to which alternative interpretations are judged to be incomplete, unorthodox, or wrongheaded. The strict constructionist reading of the U.S. Constitution falls into this category, with textual purists claiming to have located the true meaning or original intent in the text.  The open canon strategy proceeds on the assumption that the classic text lends itself to conflicting interpretations, that room for the honest difference of opinion must be safeguarded, and that the capacity to further productive inquiry rather than the adherence to an orthodox view recommends a particular construal as viable and canon-worthy.

While control over the sociological canon and school-building imperatives are implicated in framing a classic, we should not overlook the upside of the drive to render him or her in a logically consistent fashion. By committing themselves to a particular framework and fleshing it out in empirical investigations, researchers push the seminal insights beyond their original confines, discover their limitations, and open up new theoretical and methodological horizons. We should bear in mind that the range of attempts to frame the interactionist agenda is broader than any summary can capture. As Mary Jo Deegan and other feminist scholars point out (Deegan 1988), the very canon of early Chicago Sociology needs to be broadened to include the likes of Jane Addams and her colleagues who pioneered some of the observational and mapping techniques for which the first Chicago School would become famous.

A broader review of contemporary social interactionism would include the dramaturgical sociology championed by Erving Goffman and his students (Goffman 1959; Cavan 1966; Lofland 1966; Lofland 1998; Heilman 1976; Zerubavel 1985). An influential group of sociologists has formed around Norman Denzin and his colleagues who pursue “interpretive ethnography” that explores the role of narrative structures in social life and promotes a dialogue with postmodernist theories (Denzin 1996; Clough 1998; Richardson 1990; Ellis 1996). A nascent program of biocritical hermeneutics investigating interfaces between biography, theory, and history is another example of the emerging trends in interactionist sociology (Cavan 2014; Deegan 2014; Delaney 2014; Scheff 2014; Shalin 2014a, 2014b). Not all of these research programs find their inspiration in Mead, but each branch upholds the spirit of interactionism and pragmatism. Preoccupation with everyday life and the intense focus on the self-in-interaction remain central to this sociology. Interactionists continue to favor observation in natural settings and privilege participant observation data – a methodological stance reflecting the uncertainty relationship between validity and reliability and the frequent need to choose one over the other. In the domain of theory, interactionists keep wrestling with the task of incorporating structural, grammar-bound, power-inflected phenomena into their schema without downplaying human agency.

As Anthony Giddens (1979: 50) observed, “Mead did not elaborate a conception of a differentiated society,” which is why there has been “a partial accommodation between symbolic interactionism and functionalism in American sociology: the former is held to be a ‘micro-sociology’, dealing with small scale interpretational ‘relations,’ while more embracing ‘macro-sociological’ tasks are left to the latter.” But then, his account glosses over diverse strands in interactionist sociology. It is questionable, as Giddens asserts, that the Meadian analysis is incapable of offering “any interpretation of social transformation” (Ibid.). Mead was not a sociologist by training; he was first and foremost a philosopher who opened up new theoretical vistas and worked hard to effect real social change. Structural transformation engaging every point in the bio-psycho-social continuum was on his agenda, both theoretically and practically. Sociologists claiming Mead’s legacy must remain committed to this agenda.