Social Attitudes and Nonsymbolic Interaction

Blumer, Herbert 1936. Social Attitudes and Nonsymbolic Interaction. The Journal of Educational Sociology 9(9): 515-523.

My chief interest in this paper is to treat in a more conspicuous fashion one phase of social attitudes and of their development that is usually ignored or given but minor consideration. I refer to their affective nature as set apart from their ideational content or symbolic character. In the usual discussions where some attempt is made to analyze the nature and, so to speak, to describe the structure of social attitudes, attention is given primarily to the symbolic character. (Blumer 1936: 515)

The affective structure of social attitudes.

In regarding the attitude as an orientation on the part of the individual, as a "set" of his musculature, as a tendency to act in a given way, or as an incipient preparation to a scheme of conduct, there is usually an implied emphasis on the meaning of the object or situation to which the orientation is had. It does not matter where the "meaning" is lodged in the structure of nerve and muscle, as the physiologically minded incline to believe, or in a set of images or mental constructions, or in the object. The point is that the attitude as usually depicted represents a plan of action dependent upon the meaningful character of the object or situation toward which it is directed. (Blumer 1936: 515)

As in older rationalist psychology, the semiotic "disposition to respond" (with a sign, to a sign) is dependent on the nebulous "object", a part of the three-pronged Peircean sign.

As such the "symbolical" character of the object incorporated in the attitude as a plan of action receives the stress; the affective nature of the attitude is ignored or given minimal attention. (Blumer 1936: 515)

Replace "symbolical" with "semiotic" and you have a rather common statement about sign theory. "The affective nature of the attitude" on the other hand, seems as paradoxical as "the passionate nature of the desire", that is, two terms of Firstness are counterposed. From Day's (1876) definition of affection, it would seem that the affective nature of the attitude concerns the love/hate relationship towards the object of an attitude.

This point stands out more clearly in the treatment given to the way in which the social milieu enters into the formation of an attitude to give it its social character. This treatment usually is expressed in the declaration that the social milieu "defines" the relatively unformed activity of the individual. The responses of others to one's own activity are regarded as signifying the line along which that activity may go. Here the thought is that these responses of others give the individual primarily a "realization," "interpretation," or "meaning" which represents the way in which the object of his act is socially interpreted and the way in which that object is likely to be construed on subsequent occasions. Hence the individual's attitude or approach to that object becomes organized on the basis of the symbolic character of the object as that has been outlined by the acts of others. To view the formation of attitudes in this way is not, in my judgment, intrinsically wrong, but it does tend, as remarked above, to emphasize the symbolic feature and to minimize the element of feeling. (Blumer 1936: 515-516)

The social construction of attitudes. Basically this touches the way signs grow in the community - Peirce described it quite poetically. An example of "the way in which the object of [someone's] act is socially interpreted" is the way Aaron Swart's downloading of JSTOR's database content could be interpreted either as an act of felony (intellectual content theft), liberation (data belonging to the public domain should be publicly accessible), or research (the scientific data hosted by JSTOR could potentially illuminate the relations between corporations, scientific work, and global climate change). The critique here essentially pertains to viewing attitudinal semiosis either as a public or personal matter - when it is viewed in terms of how "the acts of others" organize the individual's attitude towards objects, or in terms of how intimate feelings, i.e. affect, contribute to the formation of attitudes. This distinction is quite relevant in discourse on racism, and whether it's a matter of sociocultural conditioning (mediated, discursive influence) or personal experience (unmediated, intuitive influence). (The distinction between intuitive and discursive is Peirce's.)

It is this feeling side of the attitude that I wish to single out for consideration. I regard feeling as being intrinsic to every social attitude, and, as such, as differentiating attitudes from other types of orientation which in terms of definition would be regarded as attitudes by many writers. Common usage seems to me to carry an implicit recognition of the affective element. Thus we speak of attitudes toward such objects as parents, country, races, group, and professions. Sentiments and feelings are involved in the relations to such objects. Contrariwise, we do not speak ordinarily of an attitude to such things as, let us say, pencils, chairs, or doorknobs. Certainly, to such objects people in our culture have defined ways of acting represented by tendencies, muscular sets, or orientations. But in common parlance such sets or tendencies are spoken of as attitudes only when they are marked by some feeling. Thus a person may dislike to use pencils, or an Oriental may have an aversion to chairs which he finds it torturesome to sit in. In these instances, one would, I think, immediately speak of attitudes. An affective element has entered in. It is the presence of this element which seems to justify one in speaking of a given orientation or activity tendency as an attitude. (Blumer 1936: 516)

In this, Blumer is extending the Humean (was it Hume?) connection between cognition of emotion (namely, that cognition is not free from emotion but always even in some minimal influencing it). The way he defines attitudes likens it to what we commonly qualify as "social attitudes". Gustav Jahoda's (2007) paper about the attitudes towards the New Guinea natives in the diaries of Malinowski and Thurnwald immediate comes to mind. A similarly minded study of social attitudes in scholars has been conducted by Wolfgang Drechsler (2009) on Cassirer, Jung, and Uexküll from the perspective of political semiotics. Much like Peirce's definition of the sign, Clay's definition of consciousness, and Day's definition of the affections, attitude is (here) defined as a type of relation to an object, characterized by a type of orientation towards, or element of, feeling. At the end of the day they're all constructing rational - in some loose sense, phenomenological - classifications of experience, and doing so in philosophical terms that have an extensive history (the "object" reaches back to Ancient Mediterranean philosophy, even - if I'm not mistaken - to very Ancient South- and East-Asian philosophy and theories of mind). "Dislike" and "aversion" are emphasized because they belong to the category of affect, that is, love and hate. In this particular example, when a peoples is forced to acculturate with unaccustomed cultural practices, they may come to resent these practices as a mediated feeling towards the "overlord". The resting practices are a limited case, but the illustration of sitting in chairs for early 20th Century Chinese, one could add the anthropological study of squatting, which is "un-learned" during growing up by nearly every cultural group except the Slavs and Sudanese. Due to sitting in chairs, the natural resting position of squatting is lost by way of muscle atrophy. The original author of that study (Hewes 1957) noted that it had a component of social attitudes: the squat was (and perhaps still is) viewed as "barbaric", common to "thugs", for example, at least in groups in cultural contact with Russians.

In the theoretical discussions of the nature of attitudes there is, of course, plenty of declaration that attitudes may be marked by strong feelings, and most of the testing devices, as I am familiar with them, proceed on the assumption of the presence of this character. Yet the general tendency is to think of feeling as an ex parte element which may be added to certain attitudes but which is absent from others; the essential part of the attitude is held to consist in its orientation, in the implied symbolic content determining its direction. Such a view I believe to be wrong. Feeling is intrinsic to every social attitude - it is not to be treated as an additional element fused into some symbolic structure which is to be regarded as central to, or as the corpus of, the attitude. (Blumer 1936: 517)

This is an alternative route to argument that feeling is First, along with sensations, emotions, affections, desires, sentiments, and passions (Day 1876: 55).

I am not concerned here with any serious effort to consider the peculiar role or function of the feeling or affective side of the attitude. I believe, however, that this role is quite important. It seems that it is the affective element which ensures the attitude of its vigor, sustains it in the face of attack, and preserves it from change. Common usage seems to have caught this recognition and given it expression in the popular realization that no change a person's attitudes one must change his feelings. (Blumer 1936: 517)

While Day finds the function of feeling important in general, or in relation to social attitudes and altering them, others found this function not only "peculiar" but particular, as in the sign-process (Peirce), or even more particularly in the functioning of the linguistic sign (Ogden & Richards, Bühler, Jakobson). The general idea expressed in this paragraph pertains to something germane to cognitive dissonance research, and its finding that rationalization impedes resolution of dissonance, as beliefs and conceptions have an emotional component, outside of the immediate control of the conscious mind.

My purpose, then, is to call attention to two phases of attitudes: (1) a symbolic aspect represented in the specific direction of the tendency, and (2) an affective aspect assuring the attitude its liveliness, its movement, its vigor, and its tenacity. (Blumer 1936: 517)

At this time it was common in America to consider the referential and emotive functions in tandem and with synonyms like symbolic representation and affective assurance. In some quarters (i.e. Susanne Langer, Charles Morris, etc.) this is embodied in the distinction between "signs" and "symbols".

This affective aspect of the attitude is not only slighted in definition - it has not been given due consideration in the discussions of the process of interaction out of which attitudes arise. Here again the treatment has been weighted heavily on the side of the symbolic content, stressing the formation of the attitude on the level of communication; i.e., in terms of definition or of the conveying of a meaning. Such treatment has not given proper recognition to the fullness and diversity of what takes place in interaction, and so has yielded, in my judgment, only a partial statement of what is involved in the formation of attitudes. (Blumer 1936: 517-518)

The interactionist of course stresses the interaction process. By the "nonsymbolic interaction" in the title of this paper, Blumer must mean the "insignificant" communication gestures that constitute the Conversation of Attitudes in Mead's parlance, and Communization in Morris's. In effect, the argument is one familiar from later contributions of the Chicago school in sociology, which culminated in the study of nonverbal communication.

While we have only limited knowledge of what occurs in the interaction between human beings, I think one can recognize that the process has at least two levels, levels which perhaps represent extremes, with different admixtures of the two in between. I prefer to call the two levels the symbolic and the nonsymbolic. Little need be said here of symbolic interaction, since this is the one phase of interaction which has been given a great deal of treatment in the literature, although with results that are none too convincing. (Blumer 1936: 518)

This limited knowledge pertains to the "subliminal" in Wake's diagram. While Blumer distinguishes two levels, Bühler distnguishes three, Malinowski four, and Jakobson six. The symbolic and the non-symbolic roughly correspond to Langer's discursive and non-discursive, and Peirce's discursive and intuitive, though "intuition" he had to restore to an earlier meaning.

It is usually what is considered under the rubric of communication where that term is used carefully and with circumspection. Suffice it to say that on this level individuals respond to the meaning or significance of one another's actions. The gesture of the other is subject to interpretation which provides the basis for one's own response. We may say, roughly, that at this level of interaction the stimulus-response couplet has inserted a middle term in the form of interpretation which implies some checking of immediate reaction, and leads, as suggested, to directed response upon the basis of the meaning assigned to the gesture. (Blumer 1936: 518)

This is, indeed, early communication theory, though much from the 1930s has been largely forgotten. In essence, Blumer attempts here to slip "interpretation" into the stimulus-response sequence, much like Jakob von Uexküll, and "central processing" in cybernetic communication theory (Jurgen Ruesch). This is broadlly Peircean, as the interpretant, interpreter, or interpretation is placed between the subject and object.

Interaction on its nonsymbolic level operates, in my judgment, in an intrinsically different way. It is marked by spontaneous and direct response to the gestures and actions of the other individual, without the intermediation of any interpretation. That there is involved a lively process of interaction of this sort when people meet is, I think, undeniable, although it is difficult to detect. People are unaware of this kind of response just because it occurs spontaneously, without a conscious or reflective fixing of attention upon those gestures of the other to which one is responding. (Blumer 1936: 518)

By "spontaneous and direct" he comes close to intuitive. In essence, he is saying that nonverbal communication is unthinking action, where the conscious processing of signification is a secondary aspect, a parallel process on another level. As later researchers verified, not only is it difficult to detect, it is difficult to formalize, structure, and treat in a precise manner. The lack of conscious or reflective fixing of attention upon one's own responses to another's actions is treated by Clay in terms of "unconscious equivalents of interpretations" (1882: 311), and - especially with regard to conscious attention - by vice-judgments (1882: 47).

It is this nonsymbolic phase of interaction that should be considered with reference to the formation of the affective element of social attitudes. It is from this type of interaction chiefly that come the feelings that enter into social and collective attitudes. (Blumer 1936: 518-519)

Notice that now nonsymbolic interaction is a "phase", meaning a passing period of time. Before it was a level. It seems multidimensional. The "formation" here leads us to social constructivism, particularly the formation of social attitudes. In other words, communication reifies beliefs and gives them a shared emotional ground. We're dealing here with what Malinowski meant by "social sentiments". Henry Day is an invaluable resource in this regard, as he defines sentiments as "feelings which are characterized either by intelligence or endeavor" (1876: 94).

They arise from the unwitting, unconscious responses that one makes to the gestures of others. To state this point is one thing; to prove it, another. However, I believe a good case can be made for the assertion, and an appreciation of its validity can be given, by considering the phenomenon of impression, especially the formation of first impressions. It is a familiar experience in meeting people for the first time to discover in oneself immediate likes or dislikes, without any clear understanding of the basis of these feelings. Something in the form of a spontaneous and undirected response has taken place, establishing a feeling and providing a basis for one's judgment. (Blumer 1936: 519)

E. R. Clay treats this subject extensively, as it regards the vicarious nature of consciousness: that there are subliminal processes, especially in the species of experience he calls Emotive Perception and Latent Experience (Clay 1882: 188). His account is accusatory: there are mental representations that appear to be conscious, rational, argued, known, but in fact are counterfeits of attention (Clay 1882: 3) such as advertising and propaganda which stick to mind despite will to put it out of mind, counterfeits of remembrance (Clay 1882: 75) which are not true memories but inferences from retrospection, counterfeits of observation (Clay 1882: 135) borne from introspection, counterfeits of selective deliberation (Clay 1882: 207) pertaining to ideology, and counterfeits of purpose (Clay 1882: 214) that pretend to be true and deliberated intentions. counterfeits of charity (Clay 1882: 352) known in Christian "superior" morality despite mediocre altruistic behaviour, and finally counterfeits of argument (Clay 1882: 384) by which he means a logical fallacies and false equivalences in particular. While Blumer's point is that judgments have an unconscious component, a beginning in Firstness, he's merely emphasizing the role of feelings. Clay seems to go much further than that, actually devling on different aspects of how our judgments form, and whether they have an internal origin in feeling or if they're substitutes for feelings that have pervaded our becoming, infecting us from without and giving us the false impression that they're our own making.

Even when one can give some explanation of his feelings in terms of traits of the others, most frequently the designation of the traits follows the having of the feeling. Seldom, I think, in the give and take of social intercourse, is the having of impressions dependent upon a prior analysis of the symbolic value of the other's traits. An individual who approached all his social relations solely on the premise of such a preliminary analysis would, I think, be exceedingly awkward in making adjustments, assuming that he could get along at all. The very nature of the impressions seems to me to point to their immediacy. (Blumer 1936: 519)

First impressions are fast unconscious evaluations of the others' character, and they cannot be explained with great precision because this type of understanding requires prior intuitive experience. Some emotions (especially blended ones) cannot be described well, and if one tried to approach others with a descriptivist method it might - as many, from Sapir to the ethnomethodologists, have pointed out - exhaust conscious energy (Henry Spencer) and lead one to stumble. (Frequent and/or continuous object- and meta-level concurrence or intervention is consuming and fatigueing.) In terms of "adjustment", Blumer precedes the integrationists.

There is presupposed here a direct and spontaneous response to others which analysis can show more easily to be unwitting than to be conscious. Such impressions, it should be remarked, are not trivial. That they provide the immediate bases for the direction of conduct is clear; that they are less readily changed than formed I think will also be found to be true. Their consideration suggests that it is probably the organization set up by unwitting response which is the foundation of social attitudes; it is such organization that has to be changed if any significant alteration is to be made in these attitudes. (Blumer 1936: 519)

Impression formation is indeed a playground of immediate interpretants. Though the people who reviewed first impression literature in the 1970s found that first impressions are indeed readily changed with more communicative contact and more dynamic consequent impressions. The organization of immediate impressions is the foundation of social attitudes.

This suggested relation of the affective aspect of social attitudes to nonsymbolic interaction invites further analysis. On its stimulus side nonsymbolic interaction is constituted, I believe, by expressive behavior; i.e., a release of feeling and tension, to be distinguished as different from indication of intellectual intention, which properly comes on the symbolic level. Expressive behavior is presented through such features as quality of the voice - tone, pitch, volume - in facial set and movement, in the look of the eyes, in the rhythm, vigor, agitation of muscular movements, and in posture. These form the channels for the disclosure of feeling. It is through these that the individual, as we say, reveals himself as apart from what he says or does. Expressive behavior is primarily a form of release, implying a background of tension. It tends to be spontaneous and unwitting; as such, it usually appears as an accompaniment of intentional and consciously directed conduct. (Blumer 1936: 520)

This parallel occurred to me when thinking about placing "meaning" or "interpretation" between the stimulus and response in the functional cycle (Uexküll). Namely, that stimulus palallels expression because they're both on the level of Firstness, but in "indication of intellectual intention" Blumer seems to be writing lyric poetry and in the process conflating Secondness and Thirdness. This probably makes it one of Peirce's degenerate categories? Blumer uses "channel" in the very technical sense, in what today would probably amount to "modality", i.e. sensory channel distinguished via perception/input and action/output "organs" or "devices". With the "background of tension" he seems to be perpetuating the psychohydraulic model of emotions, mixing the energetic aspect of behavior with the internal tension (or intensity) of emotions.

There is, I think, common recognition that expressive gestures are especially effective in catching attention and creating impression. Stripped of expressive features, the act of the other person is not likely to incite or inspire, is missing in dramatic qualities, and requires some coercion of attention in order to be held before one. All of us have had experience with discourse whose symbolic content may have been of intrinsic merit but which failed to gain attention and failed to make an impression. Likewise, to take a contrary example, we are all familiar with the speaker, orator, or lecturer whose display of interest and enthusiasm, whose use of dramatic utterance, and whose lively play of expressive gesture all combine to overshadow a meager symbolic statement. It is the overtone of expressive gesture which makes the situation fascinating and effective. (Blumer 1936: 520)

An aspect of why the emotive and phatic functions are conflated as easily as the conative and referential functions. This is Wake's vortex (a watex?) in operation: consciousness is intentional, and emotions spring from and tend towards latent experiences. Blumer's "overtone" is comparable, compatible, and combinable with Richalds' "tone", resulting in the notion that what makes someone's expressions stimulating, fascinating ad effective is the relation we have with them. This could be further developed in the direction of generalized social attitude in regard to phaticity: there are those for whom greetings, small talk and gregariousness are perfunctory obligations, and those who delight in the ritual and enjoy the pullulation of behaviour patterns. The plus and minus poles in this instance varies with the nature of the interpersonal relation.

Expressive gestures seem to enjoy a special uniqueness in gaining ready and immediate responsiveness. Speaking metaphorically, one might declare that human beings are delicately attuned to one another on the level of expressive behavior. They seem to be especially sensitive to such display on the part of others. Expressive behavior exerts a claim on one's attention; to ignore it usually requires some act of decision, some justification to oneself as to why one does not attend to it. (Blumer 1936: 520-521)

What is emotional communication or emotional contagion? What is empathy or sembling? Or communization, or a conversation of attitudes. It's where intuition connects Jakobson's emotive and phatic functions in non- and paralinguistic mediums.

The peculiarity of nonsymbolic interaction, then, is that on the side of both stimulus and response it is spontaneous, direct, and unwitting, and that it operates between the parties as a rapid and especially facile channel peculiarly congenial to human beings. Because it is expressive on one side, it is likely to impressive on the other. The disclosure of affective states on the one side seems to arouse and influence feelings on the other side. (Blumer 1936: 521)

We're dealing with hypersemiotic communication (Fiordo 1989). That it "arouses and influences" is congenial with Ruesch & Bateson's terminology, i.e. mutual awareness and influence.

It is my belief that it is just this nonsymbolic phase of interaction which has been ignored in the usual theoretical discussions of how attitudes are formed inside of a social milieu. The treatment, as suggested above, in so far as it has risen above the mere statement that there are action and reaction, has tended to treat this formation on the symbolic level in terms of the defining activities of others, or the conveying of a meaning to the individual, which gives direction to his act. And most sophisticated attempts to change or transform attitudes have followed this theoretical lead by placing reliance on a symbolic content which conceivably might yield the individual a new picture of the object in question. Yet it is my feeling that both this theoretical interpretation and the practical efforts based on it seriously ignore the affective aspects of attitudes. The feeling element is a basic part of the attitude and has to be changed in order to have guarantees of a genuine transformation. (Blumer 1936: 521)

The implication is that affect is a transformative, rather than stabilizing, force. Blumer feels that the affective aspect has been ignored, and proceeds with an attempt to direct or inspire the transformation of the situation. What are feelings without emotions? asks La Roux.

I think this change is likely to be made effectively on the nonsymbolic level and not merely seeking to convey a new interpretation of the object. We are familiar with the frequent futility of trying to change a person's attitude through some form of intellectual conversation. One may convince him in argument, yet his feelings remain untouched. He retains, even though in a pertubed form, his previous attitude, with the original orientation to action which it stood for. However, the disclosure of feeling through some form of expressive behavior readily touches affective states - awakening, setting, disturbing, or modifying them. (Blumer 1936: 521-522)

In cognitive dissonance theory terms, he is describing the phenomenon of trivialization. The term "modifying" should be doubly emphasized, because "modification" comes up in the discussion of feelings by Henry Day, and attitudes by Charles Morris.

These remarks concerning nonsymbolic interaction are tantamount to declaring that in group life there is a collective interplay of feeling which constitutes a milieu for the affective life of each one of us, and so for the development of our social attitudes. It is inside of such a texture of expressive behavior that our social feelings are nurtured - its absence leads to their impoverishment or decay. Our attitudes, or their affective side, are sustained through the reinforcement we receive from the disclosures of feeling in the expressive conduct of others. (Blumer 1936: 522)

Recent White House Office request for information on artificial intelligence contained the idea (from Respondent 31) that "Humans update their memory with every interaction" - thus highlighting the fact that "common memory" is an important (though frequently ignored) component of human communication, and that human-computer interactions are lacking common memory because our "conversations" with technology are not between mutually comprehending intelligences but that between a master and an unthinking slave. Respondent 31 is emphasizing the dynamic quality of human-computer relations, where much is to be wished for. Likewise, Blumer is here putting forth the idea that the affective side of our attitudes are socially updated through nonsymbolic interactions. There's clearly a pullulation of similar thinking here. But where the excrement figuratively contacts the ventilation is the "nonsymbolic" part of this formulation. While some emphasize the technology's ability to understand human motives, plans, and actions, others are working on deep learning application on multimodal media, which must - if mutually intelligible dialogues between us is to become natural - include our nonsymbolic interactions, bringing up the paradox of how to codify something so dynamic that uncodifyability is part of its logic. This would include not only human emotional communication, but also our aesthetic inclinations and semiotic creativity.

To refer to the expressive behavior of others as forming a collective texture is not to speak in idle metaphor. I should like to point out that expressive behavior is regularized by social codes much as is language or conduct. There seems to be as much justification and validity to speak of an affective structure or ritual in society as of a language structure or pattern of meanings. Almost every stabilized social situation in the life of a group imposes some scheme of affective conduct on individuals, whose conformity to it is expected. (Blumer 1936: 522)

Social Text is an actual journal title. The problem of codifying expressive behaviour is dynamic in this particular casue due to its multiple sources of influence: it can be regularized by linguistic traditions and innovations, everyday behaviour, cultural material, memes, and any of the thousands of channels through which "the objects of our minds" travel. The problem is partly social, in intergroup variety, but partly inexplicably reflexive and enigmatic.

At a funeral, in a church, in the convivial group, in the polite assemblage, in the doctor's office, in the theater, at the dinner table, to mention a few instances, narrow limits are set for the play of expressive conduct and affective norms are imposed. In large measure, living with others places a premium on skill in observing the affective demands of social relations; similarly, the socialization of the child and his incorporation into the group involves an education into the niceties of expressive conduct. These affective rules, demands, and expectations form a code, etiquette, or ritual which, as suggested above, is just as much a complex, interdependent structure as is the language of the group or its tradition. (Blumer 1936: 522-523)

Emotional expressions are regulated. Social relations impose affective demands. Socialization involves learning to express emotions in a socially regulated way.

The view which I am suggesting in this discussion is that social life in human groups can be viewed in one of its aspects as a network of affective relations, operating in the form of expressive stimulation and impressive response. It is this nonsymbolic interaction which seems to form the setting for the formation of the feelings which are intrinsic to and basic to social attitudes. My foregoing remarks are chiefly as a series of conjectures, but they will suffice, I think, to call attention to a primary phase of social attitudes which seems to be unduly ignored in current theoretical discussions. (Blumer 1936: 523)

Affective networks? This paper is 80 years removed from present, yet some parts of it read like something first published this year.