Chapter 5 (pt.3)
CONTEXT AND IDENTITY
The Social Individual
In order to deal at all with conceptions of individuals and roles in sociology or anthropology today, a certain amount of conceptual rethinking is required. Sociology tends to have a very considerable individualist bias in its conceptualizations, research and theorizing, despite its roots in nineteenth and early twentieth century searching after lost community. More indeed than anthropology, sociology has taken individuals to be the basic building blocks of society, and more and more recently has come to deal with them as though they exist in abstraction, and with some essential logical priority. Cooley - among our all but forgotten founders - argued strongly against this notion, saying:
Self and society are twin-born, we know one as immediately as we know the other, and the notion of a separate and independent ego is an illusion. (1909, p. 5)
Recent exchange and transactional theorists have of course carried methodological individualism to an extreme, but it was generally implicit in earlier functionalist conceptions of roles, for example, as we would do well to look for a moment at just what sort of individual was implied. In the first place, he was not a psychological individual, not an essentially integrated unity. It was this which Homans rejected when called for sociology to 'bring men back in' (1964). The sociological individual was - despite the syntactical contradiction - divided. He was seen as made up of a number of discrete attributes, described in the language of status and role, age and ethnicity, education and mental health. The problem, the contradiction in this approach came about when it treated these attributes as reified units on their own - and measured them in terms of surveys or even observations of persons. It is no accident, in particular, that survey research became the great empirical spouse of functional theory. By sheer aggregation of numbers it presented the illusion of being intrinsically sociological, while remained intrinsically individualist in method. No meeting could be forced between grand theory and abstracted empiricism (Mills, 1959a) because the former provided only language ( a priori ) to the latter, while the latter gave only examples to the former. Pure form met pure content, as it were. The key to this non-confrontation was the conceptualization of society and individual as two given alternatives, the latter the unit of the former, with nothing else in between. (See Chapter 6 for further consideration of the reciprocal definition of individual and society).
That this individualism was not as strongly the case until recently in anthropology is shown by Evans-Pritchard's celebrated definition of social structure as "the relations between groups of persons within a systems of groups" (1940, p. 262). Role theory never had the same currency in anthropological theories of structure and function as it had in sociological functionalism. Lately, a number of anthropologists such as Barth (cf. esp. 1966) and to a lesser extent the network theorists (cf. Mitchell, 1969, Barnes, 1973) have introduced a greater methodological individualism into anthropology. Their arguments interestingly parallel the earlier exchange-perspective critiques of sociological theory, and Barth's transactionalism draws specifically on utilitarian economics, just as Blau's version exchange theory had done before it (Blau, 1964; see also Davis, 1973). Network theory came closest among all the perspectives noted to stressing social relationships as units of social structures or organizations. It was a like stress on relations which formed the better side of role theory.
As Goodenough has noted, Linton in his early formulation of the concepts of status and role defined the former as 'a collection of rights and duties', but then went on to discuss statuses as categories of persons (Linton, 1936, pp. 113-4; Goodenough, 1965, p. 2). Role was simply the dynamic aspect of status, and statuses were individual attributes. Goodenough, regarding this as a universal problem in writings on the subject, sets out to overcome it by distinguishing 'social identities' from statuses and 'identity relationships'. These are, respectively, positions, taken in a categorical sense, sets of rights and duties, and the ordered distributions of such of rights and duties. Goodenough is particularly concerned to deal with the semantics (both social and linguistic) of relationships. His method is to search for the (possible and actual) pairs of reciprocated rights and duties which obtain between social actors. The problem with his approach is that in concentrating on rights and duties he deals exclusively with the formal requirements of social relations on individuals, and produces a completely static analysis. Roles become merely the behavioural components of statuses, that is, the way in which the relationship is actualized. Although Goodenough suggests that his usage of role is "equivalent to a comprehensive 'role-set' in Merton's (1957, p. 369) terms" (op. cit., p. 16), and that Merton merely 'refines' Linton's work while incorporating the same basic problem (p. 2), Merton's is actually a much more complex and subtle analysis. Setting his discussion in the context of a consideration of reference groups, Merton never loses touch with the importance of social relationships as such , both socio-psychologically (Merton and Rossi (Kitt), 1950 and structurally, macro-sociologically, 1957, pp. 281-386).
Merton does not assume (as Goodenough does) that culture is simply and completely effective in ordering the social life of everyone. Rather he assumes that social life is made up of a complexity of groups and relations, each exherting their various pulls on individuals and each other, each a part of the whole, but not without internal conflict and/or contradiction. We cannot here either review role theory in detail or event take up most of the points raised by Merton's brilliant and lengthy 1957 paper. The latter, however, will provide us with a fruitful context and point of departure for the presentation of a point of view on roles in the continuous activity of individual lives, that is, as actions taken by social individuals whose lives are essentially linearly temporal in character, and not simply as the behavioural aspect of a reified concept of status. Merton, despite the many virtues of his article, does seem to fall - albeit complexly - into the latter problematic usage. Though we shall keep the dramatic analogy implicit in the term 'role', we shall emphasize that the extent to which roles are 'given' is greatly variable, and note that this variance limits the extent to which individual statuses imply distinct roles.
Merton's topic in the paper at hand is essentially the construction of problems in the theory of reference groups (which includes, as he points out, reference individuals). He is dealing primarily, thus, with the units in relationship to which individuals derive norms and formulate plans. His procedure is first to enumerate the "functional types of reference groups", the meaning of the concepts of membership and non-membership, and the dynamics of selection of both reference and membership groups. He then takes up a number of group properties (a provisional list of twenty-six) and looks finally at "the structural context of reference group behaviour". It is this last with which we are most directly concerned, since Merton defines this context as being made up of "role-sets, status-sets, and status-sequences" (1957, p. 368). The organization of his essay is both revealing and relevant, however. In particular, it is important to emphasize that he discusses the processes of reference group behaviour as properties of groups and social structures, not only in a socio-psychological frame of individual motives and/or decisions. Merton includes considerable - and cogent - discussion of the concepts of group, membership and the like, but inasfar as his ultimate consideration of statuses and roles is concerned, groups are relevant as units - structures - which involve persons in patterned relationships with each other. To be sure they are more - this is where membership comes in (cf. pp. 284-88) - and such other units as social categories and interpersonal contract relations may produce patterned relationships (see Chapter 6). Here it is his three concepts of role-sets, status-sets and status sequences which we need to look at in order to better understand the social nature and significance of individuality among the Tallensi. We shall add relatively little to Merton's discussion, but it may be emphasized that most using the concepts after Merton have had nowhere near the fullness of his vision, and have narrowed them greatly.
The most apparent distinguishing factor of Merton's essay within the context of structural-functional theory in which it was written is its emphasis on relationships. Where Parsonian functionalism was concerned primarily with inclusive and (definedly) integrated systems from largest to smallest, Merton focused in on the importance of sets of relationships among social actors. In this he overcame the trap of assuming the efficacy of culture, that is, of assuming that labels and systems of labels directly and simply create or reflect social reality. The intellectual context of his innovation is similar to that which produced network analysis in anthropology, although Merton's approach is more sophisticated and does not attempt to have networks stand on their own as inclusive of all social reality and/or as replacing structural-functional theory. Ápropos of this parallel, consider that Merton speaks generally in his paper of
those involved in interlocking networks of social status and social role . . . (p. 361)
The difference is primarily that Merton finds considerable continuity in these relations - sees the relations patterned by the more general statuses - where the network theorists focused most often on situations of considerable fluidity, where norms of larger scale consensus did not produce roles so directly attendant on statuses. Network analysis thus achieved its strongest focus (cf. Mitchell, ed. 1969) in studies of the rapidly growing towns of the Central African Copperbelt where relations were precisely not multiplex (or at least minimally so) and were most open to creation anew between individuals.
The second important differentiating aspect of Merton's paper is its discussion of temporal dimensions of social relationships, of statuses and roles. This is important in terms of considerations of changes in the social character of actors as revealed by changing patterns of selection of 'reference individuals'. As Merton suggests,
much of such selection is not idiosyncratic but is patterned by structurally determined and statistically frequent career sequences, actual, anticipated or desired. (p. 303)
This emphasis of Merton's draws attention to the fact that statuses and roles do not simply exist in abstraction in some equally abstract social system, but are concretely manifested in individual lives. They are thus inextricably bound up with social individuals, and their full meaning becomes apparent in their interconnections, both synchronically and diachronically. The terminology which Merton develops for dealing with this multidimensionality is as follows: each individual social personality (Radcliffe-Brown's not Merton's term) includes a number of statuses - positions in structures of related positions - these, grouped together by the single actor who occupies all of them, are termed the 'status-set'. Each of these statuses, according to Merton, implies a number of different roles, that is, behaviour in accord with the expectations of each of a given number of significant others (reference groups or individuals, or better, reference partners). Thus for each of his statuses, the social individual is involved in a number of different roles - a 'role-set'.
The concepts of role-set and of status-set are structural and refer to parts of the social structure at a particular time . Considered as changing in the course of time, the succession of statuses occurring with sufficient frequency as to be social patterned will be designated as a status-sequence. . . (p. 370).
Role sequences are then (but problematically) seen to follow on status sequences.
There are two weaknesses in Merton's treatment of the temporal dimension of social relations (and implicitly of social identity). One is that he treats changes over time only in terms of an idea of 'development', that is, only as 'progresses' on pre-ordained sequences (cf. p. 385). For the most part he sees these as connected to the life cycle, and there are no doubt many such. There is a further justification for this view in that it is a common aspect of everyday perception to identify patterned and recurrent sequences - and thus either progression or regression along them. Sequences are significant outside of their static definitions, however. In other words, specific temporal contiguities are of import, not only whole socially defined sequences. This is linked to the second weakness. Merton does not treat this paterning into sequences as a variable, but rather asks only for "sufficient frequency as to be socially patterned". The extent of social patterning, I argue (above, here, and in Chapter 6 below) is an important variable in the relative social stability of societies, directly influencing the extent of accurate anticipation possible among social actors. From other references it would appear that Merton is aware of the differences in time perspective attendant on such factors as different rates of social mobility, so the matter is perhaps more one of lack of explicitness or emphasis than weakness of conceptualization (cf. pp. 294, 385).
A conceptualization missing from Merton's set is one of the social individual as a whole We have introduced Radcliffe-Brown's (1922) term 'social personality' (Moore's "social identity" is similar in import, cf. 1969, p. 396) as preferable to a more complex usage involving status-set, but this term denotes the sum of an individual's statuses at a single time, not his succession or procession of identities over time. It does have the advantage of including attributes - such as age - which, although they affect social roles, are not precisely statuses. (Merton is actually a bit unclear on what the demarcation is between such an attribute and a status. Consider for example 'class', as an example of a problematic status/attribute. The unclearness comes from our immediate tendency to assume that class is a category, and a characteristic of an individual - for this is how the term is most frequently used in empirical research - and from the contrasting but implicit relativity of the term, which makes us look for the other element of any statement of class). The social individual as a unit becomes important in a number of ways. His past is relevant not only to his own psychology, but to his social existence, since it is remembered by others as well as by himself (in addition to whatever material effects it may already have had). His future is important, not only as he is subject to 'anticipatory socialization', but as he is the object of anticipation. It is thus, for example, that the potential of youth (a clearly time-laden concept) is weighed against the experience of age. In fact, the sheer quantitative rates of continuity vs. change in an individual's social identity are important, both for each individual as he is perceived by and planned for by others, and overall as one variable in the stability of a social order.
One last matter must be considered before we go on to social individuality among the Tallensi. This is Merton's exclusive consideration of role as 'the behavioural enacting of the patterned expectations attributed' to a given position in a given relationship (thus expanding Linton, cf. Merton, 1957, pp. 368-9). While he avoids the assumption that there is a single distinctive role for each position, he still assumes that the role is a matter of expectations which are fulfilled or not fulfilled. Goodenough (1965) modifies this by noting that cultural definitions are not totally specific and encompassing; that is, they leave different avenues open for meeting role-expectations, but he does not alter the fundamental assumption. I suggest that as actors, human beings are not totally devoted to meeting other people's expectations (including anticipated other people such as 'posterity'; see Merton's discussion of non-conformity, 1957, pp. 357-68). There seems to me no more empirical reason to choose to treat continuity as normal and change as exceptional in social theory than the reverse; perhaps less. In varying degrees, almost all mainstream sociological theory has done this in recent years (the streams I name as not in the main are evolutionary and Marxist), and Merton's somewhat revolutionist functionalism is no exception. Thus Merton sets out conflict and instability as 'disturbances' to the role-set, implicitly assuming that it exists somehow before them (p. 371; to be fair, he notes that perfect stability and compatibility in role-sets are 'historically rare').
In the ensuing discussion then, I shall treat roles as what actors do in their various social identities, perhaps explicitly on the basis of one or another status, but not as what someone thinks they should do, whether someone be another actor in a role-set or a sociologist. I shall refer to role-expectations explicitly as such. The degree to which actual roles played conform to expectations is too important a variable to lose in conceptual indistinction. In addition, the negotiated settlements of what should be expected (discussed in sociological literature from Thomas to Goffman (very clearly indicate the extent to which is is advantageous to see roles as temporally concrete in their performance. There is the Hamlet that Shakespeare wrote ( a set of formal rules, but not completely unchanging or inviolable), the Hamlet that each theatregoer expects (role-expectations) based on his or her varying knowledge, previous experience and even preferences, and there are the Hamlets given by any number of actors (roles, performed better, worse and differently). There are also, to complete the analogy, as many different perceptions of each Hamlet as there are members of each audience - but there is also a quantitative variance to their perceptions; at least ideally, they may be compared. And surely the theatre, like social life, knows eras of audience solidarity, ears of dissent, and eras of alienation.
Identification by Contraposition - The Social Situation - Hierarchical Inclusivity - The Social Individual - Static Aspects of Individual Identity - Temporal Dimensions of Individual Identity