Reading Guide to: Manis, J and Meltzer, B (1972) Symbolic Interaction: A Reader in Social Psychology, 2nd edition, Boston; Allyn and Bacon, Inc.

Meltzer 'Mead's Social Psychology'

For Mead, society was a matter of group life, and it was to be studied through grasping the intentions of others, and acting according to those intentions. What was needed was some social psychological mechanism to get at these intentions. This was possible, for example, by completing the acts of others through grasping symbols which stood for the intended act. This implies a consensus about the symbols, and an underlying set of common values. We are required to treat the self as if it were other, in order to understand this process. A key mechanism here is role taking, as the first step towards treating the self as a source of intentions which become symbolised: once we do this we can grasp the possibility of doing it for and with others, and thus consider social behaviour as incorporating the behaviour of others.

This self became an object for its own actor. It was formed through the definitions of others. Role taking was possible through symbols such as speech, and the definition of others, which required some social context. This capacity developed through several stages: (1) through imitation by infants; (2) through play, which directs action back towards the self and to others, but offers no unitary position; (3) through the game, where there is a consistent role for the actor against a number of others, as in team games, permitting some composite Generalised Other to emerge, which expresses the common standpoints of the group. The latter stage provides a context-free generalised set of expectations.

The self becomes split: the 'I' is the spontaneous initiator of action, while the 'me' is the incorporated other, the Generalised Other internalised, or even some particular other. Acts begin with the spontaneous 'I', then they become socialised or steered. It is the 'I' that provides novelty, creativity, and reflexivity. Thus the social and the individual are interwoven inside individuals, and this produces our inner mental life ('mind'), our agency, and the possibilities of social change through reflection. Both self and mind arise from social processes.

The mind is inextricably social, rather than some prior entity. It is a process, and it only appears through symbolic interaction. The mind selects stimuli rather than just responding to them: it adjusts, but it is not primarily cognitive (conscious or reflexive) and acts more from a sense of inhibition, or hesitation, pausing for thought and offering imaginative variation. It does not just release drives. It is activated from the perception of problems, and it attempts imaginative solutions to those problems, leading to reflexivity and abstract thinking. Problems themselves are socially selected, as are perspectives on them, so the mind is constantly adjusting and checking social consensus, which provides it with a social function too. Social life therefore provides the internal dialogue between 'I' and 'me'.

Selective perceptions are directed at objects, and symbols are attached to images. Objects are actually constructed by our intentions, and this is a shared activity, because shared symbols are necessary in this construction. Meanings are defined as an 'image of a pattern of action which defines an object'.

An act involves a process which is established actively. It is a synthesis, not a mere reaction to some stimulus. It begins with an impulse, and this leads to imaginative possibilities which are channelled, and eventually become whole projects which are interlinked: there are no simple and discrete chains of stimulus and response. A motive is the same as a rationalisation, involving the definition of a objective of an act at any one time. Acts include automatic, blocked, incomplete, and retrospective types.

In terms of criticism, Meltzer thinks that:

  • These concepts are vague, and this arises from the different aims of Mead's work (these include 'political' ones, but also negative ones designed to attack behaviourism)
  • Impulses are not clarified, and they seem to be both biological and social.
  • Definitions of meaning and mind are not well-developed, and seem to refer to self consciousness only. Role-playing is also obviously only a conscious mechanism for learning. There is no discussion of any unconscious level.
  • The 'I' is only a residual category, designed to avoid sociological determinism, but again it emphasises the  conscious level, and there are problems with habitual acts.
  • The concept of Generalised Other assumes that there is just one -- that is, that no conflict arises.
  • There are problems with the idea of images as mental representations of objects.
  • Mead omits consideration of content in his concern for the technical aspects of interaction, and affect especially is omitted (Cooley is better, thinks Meltzer).

The whole mechanism is too abstract to develop a proper methodology, and we are given no clues about how to proceed, or what actual evidence to gather [see Blumer and his rather vague prescriptions here].

However, lots of people were influenced by this work, which parallels Weber and the concept of verstehen.

Meltzer and Petras 'The Chicago and Iowa Schools of Symbolic Interactionism'

All the main figures saw self and society as interdependent, and all stressed the importance of the reflective self. The need for association became axiomatic in the development of reflexive selves, to replace the early emphases on biological and psychological mechanisms (although these remained in the work of Mead). Interaction became important as a research technique too, and two schools developed, around Blumer and Kuhn (M.H. Kuhn, that is).

Differences between the schools turned on choices of a methodology, and whether these should be humanistic or scientific. There are clear links with the German debates here. Blumer favours the former, which leads to a form of sympathetic introspection as a method. Kuhn opted for operationalism and science, involving testing, the development of attitude scales, or collections of statements about the conception of the self: these amounted to operational definitions of 'self'.

They differed over determinacy too. Blumer saw an interplay between the spontaneous and the socially determined, via the mechanism of the 'I' and the 'me' , and this provided a certain indeterminacy at the same time (whether from the 'I' itself, or from the I/me relation, or from both is unclear). Kuhn saw that social definitions define the actor, as in role theory, and thus the self became that which one's reference groups defined.

Blumer was more concerned with process than structure, with reflexive self interaction, and with tentative and exploratory research. Kuhn thought there would be stable traits in some 'core' self. The debate was played out in terms of the ambiguities of the concept 'role': this was something to be played [at] for Blumer, but something to be internalised for Kuhn.

There were different levels of interaction for Blumer, including nonsymbolic types (based on non-human responses). These were little discussed by Mead, and ignored altogether by Kuhn, who focused on the cognitive and rational level only.

There were some convergences too, so that both schools were interested in role theory and reference groups, and in achieving some kind of clarity, if not exactly operationalism.

Blumer, H 'Sociological Analysis and the "Variable"'

The use of variables in empirical social science is growing, but randomly, as a rather chaotic collection. Actual variables are often isolated far too simply and necessary dimensions of complexity are omitted. What is required is much more reflection on the problem, to isolate 'its genuine parts' (93). Looking across a number of studies, the selection of variables seems quite arbitrary. There is rarely a proper account of how these variables have been generated [and relying on empirical clusters has problems of its own].

Even generic variables such as 'social cohesion' seem to display considerable variations and are used in different and inconsistent ways [and many examples are given, page 94]. They are often localised, for example,  rather than used to generate some more general theory, yet rarely is the influence of context properly analysed. It is unlikely that much theoretical progress has been made, or will be made: 'The variable relation is a single relation, necessarily stripped bare of the complex of things that sustain it.' (95).

What we need is more investigation of the processes of interpretation and definition [as in symbolic interaction] in everyday life, especially how groups come to develop collective definitions and 'attach meanings' (96). Any social regularities arise from these stabilised patterns. Proper sociological variables would also have to be the results of such interpretation.

The classification of variables into 'dependent' and 'independent' represents an arbitrary split in the process of interpretation, however: causes/initiators and their outcomes are separated out, rather than looking at the process itself. But 'causes' must be interpreted as such first by the actors in a social context. Variables cannot produce their own meanings, nor can interpretation itself merely be treated as an intervening variable. Using the term 'variables' implies that there are discrete, clear, and independent factors at work in social action, but in real social contexts these often form an 'intricate and inner-moving complex (100 ) Any abstractions must therefore stand for complex meanings and not simple factors.

The best approach, surely, must be to try and recapture meanings, definitions and interpretations, and their connections using techniques such as participant observation --'[approaching] the study of group activity through the eyes and experience of the people who have developed the activity' (101). Variable analysis should be restricted to those areas of social life that are not produced by interpretation -- including 'patterns of interpretation which are not likely to be detected through the direct study of the experience of people' (102).

Blumer, H 'Society as Symbolic Interaction'

Although there is no particularly systematic statement of this position, Mead offers the best statement of fundamentals, although there is so little material, that it is not surprising that different versions exist of symbolic interactionist methodology. The main point is that human beings interpret or define each other's actions, and do not merely react to stimuli: there must be 'a process of interpretation between stimulus and response in the case of human behaviour' (145).

Human beings have a self, which is both subject and object, and which self-indicates, that is indicates things to oneself. This is conscious life. This is 'the mechanism that is involved in interpreting the action of others' (146). To indicate something in this way is to objectify it, to confer meaning on it. As a result, human consciousness is [an intentional] process, not a simple release [of drives], and not a simple reaction of ego to an objective world, or an event which is caused. Human action as a process extends into the future and is thus emergent.

Group life develops as a result of the alignments of human actions, through important processes such as role taking. This, and the preceding arguments, 'can be easily verified empirically.... The reader is challenged to find or think of a single instance which they do not fit' [!] (146).

This is a perspective that opposes sociological determinism and holism, including functionalism and systems thinking and positivist approaches in methodological terms, as well as psychological reductions of the self. It is not possible even to think of motives as some kind of cause which precede action [so that is J S Mill dealt with -- see file]. Terms like 'social structure' merely indicate common understandings or definitions, 'how to act in this or that situation' (151) The relatively common occurrence of misunderstanding or blocked collective action indicates undefined situations, and here 'it is necessary to trace and study the emerging process of definition which is brought into play' (151).

Human being 'catch' collective understandings through role taking, and this has a direct connection with the research technique of participant observation. Social structures set the conditions for such role taking but they do not determine it. On the contrary, there are lots of cases of nondetermined, transient interactions [of the kind studied by Goffman, of course]. Only the study of human interactions and meanings can properly explain individual and collective action.

Cooley, C 'Looking - Glass Self'

In lots of cases in social interaction individuals develop an imaginative perception of how they look to others. We imagine our appearance to the other, we imagine that their judgment about appearance, and we develop 'some sort of self feeling such as pride or mortification' (231). This is no mechanical reflection, and its effect depends on the significance and the characteristics of the other: we wish to seem brave if the other is brave, and so on, and we adjust our behaviour accordingly.

Children can be seen to develop this 'self feeling of the looking glass sort' (232), first by observing the effects of their actions on others [and a great deal of anecdotal evidence ensues, including a recollection from Darwin about how he deliberately hid some fruit and pretended to have discovered it, in order to make a favourable impression on his father]. Performances are soon adjusted to different others --'selective interest, admiration, prestige, is obvious before the end of the second year' (232). Strong emotions are attached to the results of such impression management [more anecdotal evidence follows -- you can see the thrust of Gouldner's argument here since it seems to be infantile manipulation and cynicism that is being described, and taken as human nature]. Initial successes construct 'a social "I"' (233 ) followed by greater precision in fullness and imagination.

'Even adults... make no separation between what other people think and the visible expression of that thought... [and] there is progress from the naive to the subtle in socially self assertive action... an endeavour to suppress the appearance of doing it [things for effect]... affection, indifference, contempt, etc are simulated to hide the real wish to affect the self-image' (233).

'Primary Group and Human Nature'

The other piece from Cooley in this collection is an extraordinary discussion of 'human nature' and its effects on primary groups. Human nature is identified with unique human qualities. especially 'sympathy and the innumerable sentiments into which sympathy enters, such as love, resentment, ambition, vanity, hero worship, and the feeling of social right and wrong' (158) .

Human nature is regarded as permanent and universal, although 'there are differences of race capacity, so great that a large part of mankind are possibly incapable of any high kind of social organisation' (159). However, there are equally large differences between members of the same race, and 'The more insight one gets into the life of savages, even those that are reckoned the lowest, the more human, the more like ourselves, they appear... [so that]... the difference [in actual societies?] is neither in human nature nor incapacity, but in organisation, in the range and complexity of relations, in the diverse expression of powers and passions' (159). Human nature is also a 'group nature or primary phase of society, a relatively simple and general condition of the social mind' (159). What this means is 'that society and individuals are inseparable phases of a common whole, so that wherever we find an individual fact we may look for a social fact to go with it... we must learn to see mankind in psychical wholes', as in our own common experience of the 'we - feeling' (160).

Denzin, N 'The Research Act'

Sociology needs theory, which offers some abstract explanation; methodology, how methods are used to do research and make it public; and the sociological imagination, as in Mills (the ability to develop a number of perspectives, to try and build up a picture of totality, looking at the social world in new ways). We must avoid doctrinaire applications of methodological principles to research, but instead try to develop a series of methodological rules based on symbolic interactionism .

First , we should be interested in symbols and in self-interactive processes rather than in trying to acquire a snapshot of social life, as in attitude surveys. We need to understand both the symbols themselves and the concrete settings in which action occurs --  both are required for a full investigation. It is, for example the interaction between dope smokers and dope non-smokers that leads to one becoming a marijuana user [a reference to Becker, here].

Second, we need to learn to take the role of the other as researchers, to learn the other's language rather than trying to develop some objectivist approach, which is really only the same as imposing our own meanings and preconceptions. Sociologists' meanings must connect to the meanings used by the actors.

Only then can this understanding be interpreted in the abstract terms of the sociologist. This is never going to be a smooth transition, and there will always be an 'irreducible conflict' (81) between the two sorts of conceptions.

Sociologists should always search for a social context of the interaction, to trying grasp the impact of 'broader social structures' (81), for example how individuals and groups consider and define the situation. Life histories and participant observation are best at covering these processes. However, sociologists may wish to ask specific questions about the typicality or range of possible activity.

Because we recognise that sociological methods are symbolising processes themselves, and that sociological observation itself is a form of interaction, there is an inevitable relativism about sociological research. This can partly be controlled by using multiple methods aimed at discovering different discrete instances of a concept. Reflection, the use of the sociological imagination, and imagined experiments can be just as fruitful as rigorous application of methods

We should choose theory as a guide to our understanding, but let actual findings modify theory. We can use concepts in stages, first using them to sensitise us, then we might try to operationalise them. This sort of movement is illustrated in work such as Goffman's on stigma. Eventually we might develop universal, 'formal' concepts, content-less abstractions which are then joined together in theory. The example here is the work of Simmel on forms of interaction. It is this ambition which are shared by Goffman [in Interaction Ritual -- the selection quoted reads very much like Durkheim on the need to limit and restrain individual variation in collective rituals: 'functional theory and certain portions of symbolic interaction', according to Denzin, page 89], leading to the development of universals in his work.

This is the proper task for sociologists, rather than opting for more specific theories of the middle-range, as in Merton. [Becker's study of marijuana use, cited extensively throughout, also claims to head towards important generalisations, ending in statements which 'may tentatively be considered as true of all marijuana users in this society'--  Denzin quoting Becker, page 90]. This is not the same as grand theory itself, however, because it is abstracted from empirical examples .