Introduction

Ever since Georg Simmel’s seminal works, social relations have been a central building block of sociological theory. Relational approaches to sociology assume that the primary building blocks of the social world are not actors or objects, but social relations. By prioritising social relations over actors and entities, relational sociology aims at overcoming long-standing dualisms, such as those between the individual and society, or between action and structure (Emirbayer, 1997). From a relational perspective, actors are not considered as ‘given’, but seen as socially constructed identities shaped by the experience of and interaction with other identities. With reference to White (1992), relational sociology uses the concept of identity not only for human individuals and their personal and social identities, but more generally for any entity with the capability for action to which observers can attach meaning and significance (White, 1992: 2).

This relational genesis of identities can, for example, proceed on the basis of positions in a stratified social space, on the basis of role relations and expectations, or by means of cultural descriptions of identities, as found in narratives or stories. The essence of this view is that identities are constituted in specific structures of interaction through relations to other actors. Relations thus become hallmarks for identities; they determine the interpretation of situations and people’s intentions and motivations. In this way, a person becomes the bearer of a multitude of identities that are borne from different structures of interaction that make up, for example, social networks and fields.

In this relational framework, symbolic interactionist approaches to the concept of meaning play a central role. As in the sociology of knowledge, the intersubjective processes that constitute meaning are inextricably linked to larger social structures. Subjective meaning making and culturally shared patterns of meaning are thus mutually constitutive and mediated by (structures of) social interactions (Fine and Kleinman, 1983; Fuhse, 2012). In this tradition, meaning is often reconstructed from stories and narratives by using qualitative methods, focusing on different forms of declarative knowledge: in other words, knowledge that is based on concepts, symbols and theories, which can be represented symbolically and thus analysed empirically (cf. Fuhse and Mützel, 2011: 1076).

In our view, an exclusive emphasis on declarative knowledge and symbolic representations unjustifiably privileges explicit and reflexive forms of meaning over implicit and non-conceptual ones, although these can also well be accessed using qualitative methods. In recent years, pioneering approaches in social theory and cognitive science have emphasised the relevance of implicit, bodily and non-declarative processes of meaning making for social action and interaction (Semin and Cacioppo, 2008; Vaisey, 2009; Turner, 2013; Patterson, 2014; Lizardo, 2017). Based on these works, we argue that affect is a concept that accounts for such forms of implicit and bodily meaning, and that affect can be assessed, methodologically, as relations between symbolic representations. To overcome the above-mentioned dualisms, the concept of affect is particularly useful for three reasons. First, it explicitly bears bodily and non-representational implications, which are nevertheless meaningful and motivational. Second, although affect is primarily a concept that reflects processes, relations and situations, it also encompasses an historical dimension of the individual and collective ‘sedimentation’ of socialisation and past experiences. This dimension is reflected, for example, on the level of symbols, as the affective connotations of signs and linguistic expressions. Third, affect emphasises that the co-constitution of identities in social interaction is not just a cognitive or linguistic process (as sometimes assumed in symbolic interactionism), but fundamentally a bodily and material one.

It therefore seems theoretically plausible that, in addition to explicit and denotative forms of meaning, implicit and affective forms are also essential to the relational constitution of identities. Nonetheless, it is not obvious how these identity-constituting affective meanings emerge and to what extent they can be regarded as socially shared (or institutionalised) or as rather idiosyncratic forms of meaning based on individual – or even collective (for example, generational) – experiences and understandings. The present study addresses precisely this question with an exploratory study based on a survey of the affective meanings of a variety of linguistic concepts. Drawing on data from a nationwide survey (N=2,849) and the theoretical and methodological framework of affect control theory (ACT) (Heise, 2010), we show that linguistic concepts designating identities are linked to (relations of) affective meanings that correspond to established and institutionalised relations between these identities: for instance, students–teachers and grandparents–grandchildren. ACT generally assumes that the affective meanings of linguistic concepts are, similar to denotative meanings, widely shared within a society, that these meanings are constitutive for identities, that they are instrumental to explain social action in a given situation, and that they can be reliably measured using surveys and semantic differential rating scales. Although our analysis fundamentally draws on the principles of ACT, it also moves beyond this approach by pointing out significant interindividual differences in the affective meanings of concepts and between the relations between these meanings, which we show to be associated with individuals’ positions in the social space. We interpret these differences as resulting from subjective experiences – as opposed to institutionalised knowledge – arising from often prototypical situational contexts, in which, however, markers of social differentiation (such as class or gender) become relevant. Our contribution therefore advances ACT while at the same time translating ideas from this paradigm across a wide range of other theoretical traditions, in particular relational and cultural sociology.

Social relations, interaction and meaning

Relational sociology has become increasingly relevant in recent years. The concept of ‘relation’, moreover, enjoys popularity not only within sociology but also, for example, in philosophy and cultural anthropology (for example, DeLanda, 2006; Barad, 2007). Although relational sociology is characterised by a remarkable plurality of theoretical and methodological approaches and therefore cannot necessarily be called a unified research paradigm (cf. Prandini, 2015; Dépelteau, 2018), its basic arguments and the associated criticism of established approaches in social theory nevertheless have a wide audience.

First, from the perspective of relational sociology, social order is not regarded as an already established structural or symbolic form but is conceived of as a process or a set of processes produced in interactions (Dépelteau, 2018). Second, within relational sociology, actors and identities do not exist independently of their relationships to others but are only mutually constituted through these very relations (Emirbayer, 1997: 296). Relations are not understood as static, but rather as dynamic and process-related connections between entities (Emirbayer, 1997). The link between actors and social order is thus in principle construed dialectically, since both actors (or identities) and social order are constantly and mutually reconfigured. These constant reconfigurations take place primarily through interaction and communication and are mediated by more or less stable interaction structures, for example networks, fields, institutions or arrangements.

To address this mediation between interaction and order, relational sociology often relies on symbolic interactionist understandings of meaning (Fine and Kleinman, 1983). Classically, the subjective meanings guiding social action are central to symbolic interactionism, and identities in this perspective are constituted through intersubjective meaning. In the tradition of the Chicago School, the emphasis usually lies on the situational encounters, in which meanings are constantly negotiated and renegotiated between actors. Although this view, in its orthodox interpretation, does not assign much relevance to ‘exogenous’ factors (for example, social structure, public culture) in the process of meaning making, relational sociology has attempted to relate both levels – structure and meaning – to each other. In this attempt, social order is understood both structurally and symbolically (or ‘meaning related’) and arises from meaningful actions as well as from different forms of alter-ego relations (Fine and Kleinman, 1983: 97f). Structures in this view emerge both at the level of interaction – in the sense of the aforementioned interaction structures – and at the level of meanings, when repeated interactions and the associated attributions of meaning become stabilised and (institutionally) reproduced. The ‘generalised other’ (G.H. Mead) is a perfect example of this (cf. also Berger and Luckmann, 1969).

In this respect, relational sociology comes conceptually close to structural symbolic interactionism (Stryker, 1980; Stets and Burke, 2003; Heise, 2007). This approach emphasises the largely stable, socially shared and trans-situational structures of meaning that are crucial for the constitution of identity. In line with (Fine and Kleinman, 1983: 100), we assume that meanings certainly have a latent structure, but are also constantly changing to the extent that patterns of interaction and experience change. In any event, relational sociology emphasises, more strongly than most symbolic interactionists, the relevance of ‘exogenous’ interaction structures through which meanings are constituted that guide interpretations of the self, of situations and of others. Insofar as meanings are constituted in interactions, this relational-interactionist approach can clearly be distinguished from both atomistic individualism and holistic collectivism (Crossley, 2015). Meanings are not transferred unto actors in a ‘top-down’ fashion from exogenous macrosocial structures, but rather are formed on the level of concrete communication and interaction processes (Fuhse, 2012; Crossley, 2015).

Arguably, this perspective puts a double burden on the concept of meaning. On the one hand, it must, precisely – in the sense of symbolic interactionism – capture the individual and situational-relational dynamics of the genesis of meaning in specific interactions. On the other hand, however, it must also account for the socially shared and institutionalised dimension of meaning in structural terms: for example, in relation to status positions, networks or role identities. In the following, we argue that this burden can be alleviated by an expanded understanding of meaning that is able to reflect both the socially shared and the individual and situational aspects of meaning.

In doing so, we draw on recent developments in sociological and socio-psychological theory that distinguish between reflexive (deliberative) and pre-reflexive (automatic) processes of meaning making. Practice theory, in particular, has pointed out the relevance of implicit and bodily knowledge, which plays a decisive role in the genesis of meaning and action (Reckwitz, 2002; Vaisey, 2009; Adloff et al, 2015; Brekhus, 2015; Lizardo, 2015). On the one hand, implicit knowledge and meaning is often associated with cognitive concepts such as frames, schemas, metaphors and associations, which refer to non-conscious and automatic forms of meaning making (Turner, 2013; Smith and Queller, 2002). On the other hand, many have argued that implicit (as well as explicit) knowledge and meaning is rooted in bodily and physiological processes (Lakoff and Johnson, 1980; Ignatow, 2007; Barsalou, 2008). Recent findings from various disciplines also suggest that affective processes play a central role in the genesis of meaning (Semin and Cacioppo, 2008; Johnson, 2014: 148).

In the sections below, we discuss these theories and findings to develop a perspective on affective meanings as a genuinely relational complement to denotative understandings of meaning. Affective meanings, we argue, are particularly useful for depicting the situational and relational dynamics of meaning which feed from a combination of (a) the individual and idiosyncratic horizons of affective meaning and experience and (b) the socially shared and institutionalised denotative aspects of meaning. In doing so, we combine two lines of theorising on affect: approaches from cultural studies that put a strong emphasis on the ontological relationality of affect, and affect control theory, which accounts for situational and structured aspects of meaning making and the relational constitution of identities.

Affects as social relationality

The concept of affect has recently been elaborated, especially in cultural studies, as a genuinely relational construct that relates bodies of different kinds to each other (Seigworth and Gregg, 2010; Mühlhoff, 2015). The concept oscillates between strong ontological assumptions (for example, Massumi, 2002) and more phenomenological perspectives (for example, Ahmed, 2004). Compared to understandings of affect in psychology and sociology (for example, Russell and Barrett, 1999; Heise, 2010), this approach is strict in that affect is never only a ‘state’ or ‘characteristic’ of an individual, but can only be understood as genuinely relational.

Our own previous work has attempted to reconcile these views (see von Scheve, 2017). Accordingly, affect refers to a specific form of world-relatedness, an attunement towards the world that is meaningful and can be experienced in a meaningful way.¹ This affective world-relatedness is not primarily based on symbolic representation, conceptual knowledge or propositional thought, but on the basic perceptive and evaluative capacities and dispositions of biological bodies (cf. Clough, 2007). Affect can be understood as a continuous relational and evaluative process between bodies (and other objects or ideas), characterised by constant fluctuations, which is essential for the meaningful experience of the social world (for example, in the form of a subjective feeling). Although fluctuations in affect can be triggered by ‘higher’ cognitive processes, such as memory recall and imagination, they are often based on changes in the environment that are not consciously perceived. Animal bodies continually register, through different perceptual systems, their environment and changes in the environment, which produce shifts in the affective modes of bodies. Importantly, affect modulates a body’s potential for action: for example, through changes in sensory perception, in the endocrine system, or the activity of the autonomous and peripheral nervous system (Brennan, 2004).

Human bodies in particular are subject to various socialisation processes and are persistently integrated into cultural worlds and interaction structures such as networks, fields, milieus or communities. Discourses, norms, interactions, values and practices thus all determine the ways in which bodies affect each other (Seyfert, 2012; Wetherell, 2012). A good example is the sense of smell. A scent of gingerbread (or madeleines, as in Proust’s In Search of Lost Time) – which one may not be consciously aware of or even able to identify immediately – might be associated with meaningful past experiences and trigger fluctuations in affect. This, in turn, may change a person’s affective world-relatedness and produce shifts in the disposition for perception and action.

Affect therefore gives rise to different forms of sociality, since it integrates bodies into social formations via affective relations with other bodies, ideas or objects. Actors can therefore be understood as elements or nodes in networks of affective relations. Importantly, these affective relations are meaningful and these meanings have evolved over time or been incorporated over the course of socialisation. They contour the people’s mental and emotional lives, their perceptions and actions. This relationality of affect shows clear parallels to current developments in relational sociology, since affect primarily relates different types of bodies to each other. These relations can be constituted through social exchange or communication. However, they can also be established through physiological and involuntary or unintended forms of interaction, such as mimicry, contagion, touch or chemical communication (cf. Brennan, 2004).

Although a main characteristic of affect in this proposed view is the capacity of bodies to relate to one another through non-conceptual channels, it is important to underline that affect is by no means independent of language, discourse and symbolic representations (see Wetherell, 2013; von Scheve, 2017). A key argument for this interdependence is that (discursive) practices leave ‘footprints’ on bodies (for example, Wacquant, 2004) and, in this way, language and discourse contribute significantly to the potential of (human) bodies to affect others and to be affected.

The interdependence of affect and language is also evident in the cognitive science literature, which is increasingly recognised in sociology when it comes to the cultural foundations of thought and action, especially when meaning is constituted by means of implicit and bodily processes (Cerulo, 2009; Brekhus, 2015; Lizardo, 2015). On the one hand, key words like ‘embodiment’ and ‘grounded cognition’ draw attention to the physiological roots of mental processes, conceptual knowledge and the genesis of meaning (cf. Lakoff and Johnson, 1980). The central line of thought is that mental processes are in principle closely connected to bodily perceptions and experiences (Ignatow, 2007; 2009). On the other hand, there is a fundamental concern with how actors reproduce socially shared meanings. In recourse to dual process models of information processing, Vaisey (2009) suggests two. The first mode is a comparatively time-consuming, reflexive-deliberative mode, which refers to conceptual knowledge and propositional representations and reproduces these primarily through symbolic interaction and communication. The second mode assumes the automatic-associative and essentially pre-conscious information processing, which is based on implicit and non-representational forms of knowledge that produce habitualised evaluations and actions (for example, as components of practices) (Vaisey, 2009).

Following this view, affective world relations are always expressions of past experiences and closely associated with corresponding conceptual knowledge. Affect is thus not only operative in concrete situational – that is, spatially and temporally specific contexts (online embodiment) – but as a general potential and disposition (Mühlhoff, 2019) it is always also operative beyond concrete situations. For instance, it can be activated or recalled by discursive practices through its connection to linguistic concepts (offline embodiment) (cf. Ignatow, 2007).

Affect control theory

In sociology, this view of affect comes close to structuralist theories of symbolic interactionism and their understanding of identity (cf. Stryker, 1980; Stryker and Burke, 2000), especially to affect control theory (ACT) (Heise, 1979; 2010). This theory proposes that people act on the basis of the meanings they ascribe to situations, including affective or connotative meanings. In the ACT framework, the term ‘sentiments’ denotes affective-evaluative world relations and connotative meanings (as opposed to conceptual knowledge and denotative meanings) that actors attribute to themselves and entities in their environment: for example, objects, relationships, ideas, actions or other actors (MacKinnon, 1994: 22; Robinson et al, 2006: 186). Fundamental sentiments are trans-situative affective meanings that are relatively stable and resistant to change (Robinson et al, 2006: 182). Transient impressions, on the other hand, denote situational and dynamic affective meanings that may, but need not, deviate from fundamental sentiments (Robinson et al, 2006). Fundamental sentiments thus represent the structures of affective meaning based on past experiences of interaction, in which the expectations of concrete situations, of other actors and of associated transient impressions (which can either be confirmed or refuted by transient impressions), are embodied at the same time.

In line with structural symbolic interactionism, ACT assumes that affective meanings – in particular fundamental sentiments – are outcomes of cultural practices and thus widely shared and agreed on within a given society (Robinson et al, 2006: 180). Empirical research has indeed shown that affective meanings are substantially shaped by culture, which is demonstrated by the extensive agreement over fundamental sentiments (for example, regarding identities or actions) within a given society and notable variation across societies (Heise, 2010; Heise et al, 2015). In principle, ACT therefore prioritises the shared, intersubjective aspects of meaning (for example, regarding roles and institutions) over its subjective, idiosyncratic facets which provide potential for change, deviance and divergence in meaning making (Stryker, 2008). Recent studies, however, have shown that despite this high degree of consensus in affective meanings within society, there are also notable differences in affective meanings: for example, between social classes (Ambrasat et al, 2014), milieus and lifestyles (Ambrasat et al, 2016), and social networks (Wisecup, 2011; Rogers, 2013). Insofar as interaction experiences depend on one’s socialisation and position in the social space, affect can also be understood as culturally differentiated and socially stratified within a given society (Ambrasat et al, 2016; Ambrasat, 2017). There are obvious parallels here: for example, to Bourdieu’s (1982) concept of taste as a socially stratified pattern of (affective) meaning and valuation.

In the ACT framework, identities are constituted by the mutual attribution of meaning in concrete interaction contexts. In contrast to classical symbolic interactionism, however, ACT concentrates on the affective (both fundamental and transient) rather than the denotative facets of meaning that are empirically assessed on three dimensions: evaluation (pleasant vs. unpleasant,), potency (strong vs. weak), and activity (exciting vs. calming) (for details, see ‘Data and methods’ section). Identities can thus be understood as the denotative and affective meanings associated with subjects and objects. This applies both to the genesis of personal identity (the self) and to the attribution of meaning to other (prototypical) identities.

For example, in the Federal Republic of Germany, the identity of the familiar is perceived as very pleasant and strong (powerful), but not particularly exciting. In contrast, the identity of the troublemaker is perceived as unpleasant and exciting (cf. Ambrasat et al, 2014: SI Appendix). Identities like that of the familiar or the troublemaker are therefore determined, among other things, by their widely shared affective meanings, which reflect corresponding expectations and motivations. Within these constellations of affective meaning, each actor (ego) is at the same time always the object of the perception and thus affective meaning making by others. In this respect, identities are constituted on the basis of the mutual perception and attribution of affective meaning.

From the perspective of relational sociology, the reference to ACT seems particularly useful because it establishes links between the relations and interactions between actors and the meanings that determine and guide these interactions. People interact with each other according to the specific identities they assume in specific social contexts, such as fields or networks: for example, as students and teachers, men and women, parents and children, or doctors and patients. Since these relations and interactions are often institutionalised, they are accompanied by typical status and power relations, expectations and opportunities for action. Affect plays a central role in this respect in that it structures relations and interactions at the bodily and non-conceptual dimension of meaning – as a veritable ‘body knowledge’ – and thus as a disposition for perception and action. This kind of bodily and non-conceptual, experiential knowledge is reflected in the affective and connotative meanings that actors associate with linguistic expressions and concepts rather than in their denotative meanings (for a similar consideration, see Bourdieu, 1991: 38f). Even if only a fraction of this bodily ‘archive’ of affective experience becomes visible (and effective) on the conceptual and symbolic level, we can nevertheless assume that interindividual differences in meaning are precisely what allow us to gain insights into previous patterns of experience.

Possible differences between socially shared and institutionalised meaning, on the one hand, and individual, idiosyncratic meaning on the other, are especially relevant here. While the institutionalised meanings on a structural level represent typical – that is, empirically frequent and culturally categorised relations between identities (for example, student–teacher or man–woman) – the individual and socially differentiated deviations represent the subjective horizon of perception and experience of individuals. Whether we look up to someone, or whether someone makes us nervous, or whether we instead meet them in an easygoing manner, always hinges on our respective identities and their relative positions in the social space. A pupil perceives a teacher differently from the expert who leads teacher training. This is, in a nutshell, what the ACT paradigm teaches us about the power of roles and role identities. Here, we propose to go a little further, claiming that not every pupil, nor every teacher is the same; in fact, no role identity is the same because it is embodied by individuals with different backgrounds and experiences, which are based, for example, on class, race or gender. The perception and interpretation of other actors and identities – in concrete interaction situations and as a linguistic concepts – therefore depends on ego’s position in social space (which shapes ego’s fundamental sentiments) and ego’s relational position to other identities (characterised by institutionalised role relations and their fundamental sentiments). Therefore, with regard to the affective meanings of identities, we would expect interindividual differences that reflect the relational position of these identities within a given culture and social structure.

Empirical analysis

To determine the relationship between social relations and affective meanings more precisely, we rely on data from a previously conducted survey of affective meanings (Ambrasat et al, 2014). This survey established the affective meanings of a variety of linguistic concepts, including a wide range of identity concepts. In the present study, we are interested in examining two issues in particular. First, assuming that the meanings of identities are socially shared, we are concerned with the nature of these meanings and how they relate to the structure of the relations between specific identities. As a proof of concept, we first tested our approach on prototypical constellations between identities – in particular, role-relations such as mother, father, child, and on ‘necessary relations’ (Archer, 2013): for example, students and teachers. Second, we are concerned with the subjective perspectives of actors and the assumption of interindividual variation in affective meanings with respect to the positioning of survey respondents in the social space, as measured through a range of sociodemographic variables. We focus on differences that are known to be associated with noticeable differences in implicit knowledge, such as between genders, persons with different social statuses or from different social and cultural backgrounds (Heise 2010; Heise et al, 2015).

Data and methods

In the tradition of ACT, affective meanings are measured using semantic differentials. This type of measurement originates from Osgood and colleagues’ (1957; 1975) classical studies of affective perception and meaning, which are predicated on three core dimensions of the constitution of meaning. These three dimensions are regarded as cross-culturally universal and are able to explain a significant portion of the variance in affective meanings (see Heise, 2010, for details). The dimension evaluation (E) refers to the perception of a concept as pleasant or unpleasant; potency (P) refers to the perception of it as strong or weak; and the activity (A) reflects whether a concept is perceived as exciting or calming. These three dimensions (EPA) are also considered to be the standard for determining affective perception and meaning in psycholinguistics (Schmidtke et al, 2014) and the psychology of perception (Cuddy et al, 2008) and are usually measured using a nine-point bipolar scale (see Heise, 2010). Respondents are presented with individual words that refer to social concepts with the question: ‘What sentiments do you associate spontaneously with the following word: [—]?’

Although the measurement of affective meaning here is facilitated through language, it is based less on reflexive-propositional and more on spontaneous-associative modes of assessment, which is supported by psycholinguistic rating studies at the lexical level. Consequently, this procedure does not capture opinions or attitudes towards concepts, but rather their associated connotations: that is, their affective meanings. Typically, concepts are measured randomised at the level of single words without any contextual information. This allows the collection of affective meanings (to be more precise, fundamental sentiments) in large surveys for a large number of linguistic concepts.

For our analyses, we draw on data collected in 2011, which contains the affective meanings of 909 words from the German language (see Ambrasat et al, 2014). For the present analysis, we focus on words that refer to prototypical social relations and interactions, especially on social roles and identities. Family identities such as ‘mother’, ‘father’, ‘child’, or occupational identities such as ‘doctor’, ‘caretaker’ or ‘nanny’ refer to concepts that may serve to represent both the self (that is, the respondent) and other actors as subjects or objects in interaction situations. In addition to the affective meanings of the 909 words, the data also contain a range of sociodemographic characteristics that allow differences in affective meanings between different categories of respondents to be identified. The total of 2,849 respondents were recruited nationwide through a commercial online access panel and the sample was stratified according to sociodemographic characteristics (age, gender, area of residence, income and education) to reflect the distribution of the German population (cf. in detail Ambrasat et al, 2014).

Affective meanings of identities and relations

In a first step, we examine how affective meanings of identities reflect the prototypical interactions and relations of these identities. We concentrate above all on role identities that reflect relations which are institutionalised to a substantial degree (Parsons, 1951; 1955). Examples include the mother-father-child triad as well as relations between men and women and between pupil and teacher. Although we measure affective meanings using EPA ratings, we refrain from presenting them in a three-dimensional space of Euclidean distances. Instead, we focus on two-dimensional representations in which we relate two dimensions of affective meaning to each other. The means represent all respondents in our sample.

Figure 1 shows the affective meanings of the identities mother, father and child on the dimensions evaluation-potency and evaluation-activity. In view of the conceptual, institutionalised meanings of the terms, the identities mother and father are affectively perceived as much stronger by the respondents than the identity child, which, by contrast, is perceived as much more exciting. Moreover, mother is perceived as more pleasant than father, although the latter is not – as one might have expected – perceived more strongly than the identity of mother.

View Full Size

Mother-father-child triad

Citation: Emotions and Society 4, 2; 10.1332/263169021X16340311969719

• Download Figure
• Download figure as PowerPoint slide

Figure 2 shows the affective meanings of different identities that constitute ‘necessary relations’ (Archer, 2013). Here, we focus on the dimensions of potency and activity. The identity of teacher is perceived as stronger than that of student, while the latter is perceived as more active or exciting. We find an analogous affective relation for professor and student, although this relation is more distinct in terms of potency and less pronounced in terms of activity. Figure 2 thus shows differences in the relations of the affective meanings of professor and teacher, on the one hand, and student and pupil, on the other hand, in terms of potency and activity. The affective dimension ‘potency’ reflects aspects of competence and status, whereas the dimension ‘activity’ is more associated with youth and, possibly, attractiveness. Accordingly, the word grandparents is perceived more as calming and the word grandchildren more as exciting. The affective perception of ‘activity’ can be interpreted here, for instance, in terms of age stereotypes.

View Full Size

Necessary relations

Citation: Emotions and Society 4, 2; 10.1332/263169021X16340311969719

• Download Figure
• Download figure as PowerPoint slide

These first descriptive findings suggest, that the affective meanings of identities – and, unsurprisingly, across our entire sample – are largely consistent with a socially established and institutionalised everyday understanding of these identities and their typical interactions and relations. At the same time, these affective meanings are also likely to reflect the positions and relations of identities in a differentiated and structured social space that are not necessarily part of their denotative meanings. In the following, we explore these possible position-relative differences in affective meanings in more detail by comparing different subgroups of our sample.

Affective meanings and social difference

In this second step, we are less concerned with the relations between identities, but rather with the question to what extent affective meanings also reflect individual, subjective experiences according to respondents’ positions in the social (and cultural) space. We assume that these more fine-grained differences in the affective meanings of identities and their relations (and probably also of any other social concepts) can be shown: for example, for respondents of different status groups (high vs. low occupational status),² genders (male vs. female), and cultural socialisation contexts (East vs. West Germany).³ We focus on the meanings of those identities that we suspect are particularly sensitive to the respective contexts.

As shown in Figure 3, respondents in lower professional status groups perceive the professional identities of lawyer, police officer, doctor, teacher and scientist similarly in terms of potency and evaluation. Respondents in high occupational status groups show more pronounced differences in their perceptions of affective meanings: doctor is perceived as much stronger and more pleasant than lawyer and policeman. Scientists are perceived as much stronger and much more pleasant compared to teachers by people of high occupational status.

View Full Size

Perception of professional identities according to the professional status of the respondents

Citation: Emotions and Society 4, 2; 10.1332/263169021X16340311969719

• Download Figure
• Download figure as PowerPoint slide

Cultural differences, such as those between occupational status groups, cannot be solely explained to one’s social structural positioning, but rather to the circumstances of socialisation, which we assume to also reflect shared affective meanings. We test this assumption by comparing respondents from East and West Germany, again with regard to professional identities (see Figure 4).

View Full Size

Perception of professional identities, separately for East and West German respondents

Citation: Emotions and Society 4, 2; 10.1332/263169021X16340311969719

• Download Figure
• Download figure as PowerPoint slide

While the identities of lawyer and policeman are perceived by West Germans as similarly strong and policeman as more pleasant, East Germans perceive the identity of policeman as significantly weaker and less pleasant. In comparison to West Germans, East Germans devalue the identity of pastor much more: that is, they perceive it as less strong and much more unpleasant. On the other hand, scientists are more appreciated by East Germans than by West Germans: that is, they are perceived as more pleasant and stronger. Such differences in affective meanings can be interpreted as the expression of individual as well as collective experiences. In accordance with ACT, we assume that these meanings, on the one hand, reflect attributions to other identities and, on the other hand, that they are also informative with regard to the self-identities of our respondents, given that identities are co-constituted in social interactions.

We expected similar contrasts for gender differences and gender relations. The average of all (female and male) respondents shows that the identity of man is perceived as much stronger than the identity of woman, whereas woman is perceived as more exciting and pleasant. A comparison of the affective attributions of meaning by female and male respondents also highlights gender-specific affective perceptions (see Figure 5). First of all, it appears that both genders perceive man as stronger than woman. This, however, does not apply to the relations between mother and father or husband and wife. Female respondents perceive husband and wife to be roughly equal in strength, while mother is perceived to be significantly stronger than father. Conversely, male respondents perceive father and mother as being about equally strong, while they perceive wife as being significantly stronger than husband. Differences in perception are also evident in the dimension evaluation. Male respondents perceive all female identities as significantly more pleasant than male identities. On the other hand, female respondents perceive woman as not much more pleasant than man, and husband as even more pleasant than wife (see Figure 5).

View Full Size

Perception of gender identity concepts according to the gender of the respondents

Citation: Emotions and Society 4, 2; 10.1332/263169021X16340311969719

• Download Figure
• Download figure as PowerPoint slide

It should be noted that all affective meanings were measured not in the relationship contexts discussed here, but randomly and without any contextual reference. Thus, affective meanings measured this way refer only to the respective linguistic concept of the single word (fundamental sentiment) and not to any of the relations portrayed here. Only by juxtaposing identities that are related through common and frequent interactions does it become clear that the relations between these identities are reflected in their characteristic affective meanings.

Conclusion and discussion

The aim of this contribution was to make the concept of affect and affective meaning useful for a relational understanding of the social, especially from a symbolic-interactionist perspective. We have argued that a concept of meaning that focuses primarily on explicit knowledge and denotative meanings, as it is usually employed in relational sociology, neglects the bodily and non-conceptual dimension of meaning. With reference to symbolic interactionist theories, and in particular ACT, we have argued that affect is a suitable concept for addressing this dimensions.

Drawing on theories of affect in cultural studies and ACT, we have proposed an understanding of affect that does not capitalise on individual traits or states, but rather on social relations that create meaning and guide action on a bodily and non-representational level. Affect in this view is conceived of as a form of bodily world-relatedness, not one that depends on propositional contents and representations. Despite this emphasis, we have argued that affective meanings can indeed be accessed at a linguistic level, thus allowing for the measurement of affect using established methods of ACT. Against this background, we have investigated: first, the question of how these affective meanings are constituted and how they correspond to established, culturally categorised social relations; and second, the extent to which these meanings can be regarded as socially shared and institutionalised, or as subjective forms of experience. Our explorative empirical analysis reveals four key implications.

First, our analysis showed that affective meanings and the relations of meaning between identities are to a large extent socially shared, despite the conjecture that affect is rooted in bodily experience (see also Heise, 2010; Ambrasat et al, 2014). On the one hand, these socially shared meanings reflect stable structural patterns of interactions and relations between identities and are closely related to the symbolic orders and cultural practices associated with identities and their prototypical relations. In our view, this correspondence of affective meanings and prototypical role relations reflects the far-reaching institutionalisation of precisely these meanings. We suspect, further, that the latter go hand in hand with a certain regularity and structure of actual interaction experiences within a society. Such a pattern is consequently reflected in the form of socially shared, bodily modes of affect.

Nonetheless, these affective meanings are by no means perfectly shared or even ‘homogeneous’ in a given society. They instead leave room for interindividual differences, which, we argued, is nurtured by subjective as well as collective experiences of interaction. Moreover, in this context, our findings show that affective meanings of identities are socially stratified along relevant criteria of social differentiation. There are systematic differences in the affective meanings and relations of meaning between different occupational status groups, East and West Germans, and male and female respondents. We interpret these differences, on the one hand, as an expression of individual or collective experiences and understandings and, on the other hand, as an indicator of the identity and social positions of the actors within the social space, which is always a relative position to other identities.

In general, choosing affective meanings as the starting point for investigating social relations amounts to linking a structural level of meaning and a subjective level of meaning. How individuals relate to the world is determined, initially, by the symbolic order of a common culture and associated interaction structures, which is empirically reflected in the grand means of the affective meanings of concepts and identities. At the same time, individual deviations from these means reflect actors’ relative positions in the social space and their subjective experiences of relations and interactions. This should also have profound implications for how ACT deals with transient impressions, which not only result from the widely shared meanings of identities and their relations, but also from the social differences of these meanings across, for example, class, race and gender divides.

Second, for relational sociology, this theoretical and empirical approach can open up new perspectives for the conceptualisation of social relations. The analysis of affect promises complementary insights: for example, into the motivational aspects of identities, self-understandings, and individuals’ evaluative world-relatedness. As a result, theories that focus on routine and habitualised modes of action can be linked to theories that capitalise on networks, fields and social structures (Schütz, 1974; Bourdieu, 1993; Reckwitz, 2016). On a conceptual level, our empirical findings emphasise the importance of affect for understanding social order. They open a new perspective insofar as relations are understood, not only as symbolic, normative and propositional patterns of positions in the social space, as expressed, for example, in roles and role relationships, but also as fundamental patterns of affective perception and interpretation. Finally, our approach to operationalising and measuring affective meaning enables comparative analyses between social categories, milieus, networks, communities, and even across societies, facilitating the reconciliation of interactionist and structural perspectives in relational sociology.

Third, we also see much potential to complement our analytical approach with qualitative methods, which could provide more fine-grained insights into how processes of affective meaning making operate on an everyday basis and their impact on – or even how they constitute – relations between identities linked to class, race and gender. Ethnographic approaches, as well as textual analyses focusing on the affective dynamics of language use, might be adequate techniques to consider (for example, Berg et al, 2019).

Fourth, our exploratory study also opens up avenues for informing affective inquiries in the tradition of cultural studies. The field of affect studies is traditionally concerned with the role of affect in power relations and issues of inequality, subjugation, and the potential for emancipation. Our approach may be used to inform these lines of research by providing an analytical tool that precisely taps into the affective dynamics of the sorts of power relations that are essential to the field.

In concluding, it should be noted that the present study is only an exploratory empirical approach, much more to be seen as a proof of concept rather than a definitive and hypothesis-testing approach. More substantial future analyses should, for instance, develop standardised measures for relational distances and their statistical significance and analyse more specific and problem-focused semantic fields. These may include complete networks or empirically documented positions in social fields: for instance, in politics, the economy, religion or the family. Also, longitudinal surveys of affective meanings would demonstrate variability over time and thus reveal the dynamic aspects of the affective meanings of structures and social relations.

Conflict of interest

The authors declare that there is no conflict of interest.