Herbert George Blumer’s theory of symbolic interactionism proposed relations between human beings and society and between actors and society. These relations were completely antithetical to the relation between human beings and society entailed by the structural-functionalist school of sociology associated with Talcott Parsons. The latter’s position, briefly stated, was that society shaped human beings in a unidirectional fashion. Against such a relation—that is, against the unidirectional formation of human beings by society—Blumer’s symbolic interactionism proposed an inverse theory and was concerned with the formation of society by human beings.

This paper seeks to reconsider the relations between human beings and society and between actors and society outlined according to Blumer’s symbolic interactionism. In particular, based on its firm links with the self-interaction concept, which serves as the cornerstone of Blumer’s theory, it attempts to unravel the reasons for society’s dynamic and processual character within the theoretical context of symbolic interactionism.

The findings of this paper suggest, first of all, that within the theoretical context of Blumer’s symbolic interactionism, the relation between actors and society has been considered to be established through the former’s self-interactions. However, such self-interactions are performed in correspondence with two a priori schemas obtained from actors’ worlds and their constituent objects (i.e., other actors), rather than being freely performed.

Actors define and forge certain relations with the world according to these two schemas. However, the world does not simply acquiesce to these unidirectional acts of definition; it is an empirical world that can resist or ‘talk back’ to such attempts. Moreover, it is by using such resistance or ‘talking back’ as a clue that we may perceive the validity of—and if necessary, modify—our own acts of definition, thereby reconfiguring the existing relations that connect us to the world. Thus, in the context of Blumer’s symbolic interactionism, the relation between the actor and the world is not dictated by a unidirectional act of definition on the actor’s part but can rather be perceived as something that is (re)constructed through the interaction or mutual influence of the actor’s attempts at definition and the world’s ‘talking back’ against these attempts.

Given this understanding, I reconsider Blumer’s theory of joint action and provide an explanation for the society’s dynamic and processual character based on the theory’s firm links with the concept of self-interaction.

In our consideration, multiple actors participating in a social interaction (ego and alter) are also actively engaged in acts of ‘taking account of taking account’ as a particular form of self-interaction; when they thus take action in relation to each other on this basis, in the context of Blumer’s symbolic interactionism, a joint action (i.e., society itself) is considered to be formed. Here, ‘taking account of taking account’ refers to how two actors, by taking account of each other ‘do not merely take account of each other but also take account of each other as entities taking account of each other’. Moreover, as a consequence, another actor’s assumptions about oneself shape one’s own assumptions. Blumer argued that when both parties performed this properly, it leads to the smooth formation of a ‘joint action’.

In such a situation, properly performing ‘taking account of taking account’ leads to the act of common definition between the two actors, thus further enabling the repetition of a stable form of joint action. Furthermore, this common definition can also be perceived in Blumer’s symbolic interactionism as something that can only be sustained through the continued performance of an identical interpretive scheme—in other words, a mutually compatible ‘taking account of taking account’—by participants involved in a social interaction. For such joint action (i.e., society) to be dynamic and processual, the mechanism by which it alters its form must be explicable. In other words, the mechanism by which participants in a social interaction alter their interpretive schemes, or the manner of their ‘taking account of taking account’, must be explained.

In the context of the dyadic model of two actors (alter and ego) who shape a joint action together, each actor will be an object for the other; therefore, each performs a ‘taking account of taking account’ in relation to the other. Furthermore, to say that they are objects for each other also indicates that they are both entities that are interpreted and defined as mutual partners and are simultaneously a part of an empirical world that can resist or talk back against such interpretation or definition. Therefore, if one actor (alter) were to reconfigure an action that is witnessed by the other (ego), it would signify ‘talk back’ for the latter. Moreover, if ego were to seize on this ‘talk back’ and consider it an opportunity to reconfigure that action and if this were then witnessed by alter, it would signify ‘talk back’ for alter. In other words, when ‘talk back’ is oriented toward either, the party against whom it is directed will attempt to reconfigure his or her actions and thus encourage the reconfiguration of actions on the part of the other. When such reconfiguration of actions occurs on both sides, it changes the formation of the joint action that is realised between the two actors. Moreover, we have revealed that the possibility of such changes is ever-present. This is because, within the relation between these two actors, although each of them attempts to perceive the other accurately, the other is always an aspect of an empirical world ‘that might go entirely unperceived by human beings, and which, even were it perceived, might be perceived entirely incorrectly’.

Thus, joint action (i.e., society) alters its form according to changes in the format of human beings’ self-interactions. However, thus stated, Blumer’s formula has, from time to time, been criticized as being overly micro-oriented. Its critics have pointed to its lack of a social structural viewpoint and the difficulties associated with it when attempting to account for collective-level phenomena. In response to such criticism, even within the theoretical framework of Blumer’s symbolic interactionism, various commentators adhering to the view that macro-level analysis is possible have counter-argued that, within Blumer’s symbolic interactionism, the term ‘actor’ also finds expression as an ‘acting unit’ and that the referent of this expression extends not only to individuals but also to groups. In other words, when analysing the micro-sphere, it perceives society (i.e., joint action) as being formed through interactions between individuals; on the other hand, when analysing the macro-sphere, it already perceives this as being formed by interactions between groups. However, thus stated, Blumer’s theory, as far as I can see, still suffers from two limitations. The first is that when conducting macro-analysis, the concept of self-interaction, which is a cornerstone of Blumer’s symbolic interactionism, retreats into the background of the analysis. The second is that envisioning Blumer’s theory in such a way makes any approach from the actor’s position—the methodological iron rule of Blumer’s symbolic interaction—impossible to implement.

How then might we perform a macro-analysis that positions the concept of self-interaction at the root of social theory and complies with the methodological iron rule of approaching analysis from the actor’s position? This is the greatest challenge that has been left by Blumer’s symbolic interactionism.