Continuing Relevance of Georg Simmel: Staking Out Anew the Field of Sociology

The Four Sub-Processes of Sociation

Filtered through different ways of experiencing, acting and receiving, (primary and secondary) contents and form stimulate each other reciprocally and crosswise, thus setting into motion four different sub-processes. In their combination, they constitute the overall process of sociation. Going beyond Simmel’s terminology, we call the first process ‘externalization’; it is related to the concept of ‘acting.’ The second process is called ‘internalization’; it is related to the concept of ‘receiving.’ The third process, ‘institutionalization,’ refers to the concept of ‘form’ and means the process of constructing, shaping and reshaping social institutions. The fourth process, finally, ‘interest-formation,’ is related to the concept of ‘contents’ and means the dynamics of shaping and reshaping social interests, needs, wishes or emotions. I have to limit myself to making a few remarks about what these four sub-processes of sociation mean more precisely and how they are linked to each other.

Externalization and Internalization

The processes of ‘externalization’ and ‘internalization’ are the dynamic aspects of interaction, of mutually reciprocating effects. In stressing these two processes, Simmel looks at the individual (or collective) actor from two different points of view. He, first, sees him or her as the ‘creator’ of processes of sociation, and, secondly, as the receiver or addressee of the social effects emerging from previous interaction sequences. Externalization refers to processes of social productivity, internalization to the elaboration of emergent social effects through experience. ‘What happens to men,’ Simmel (1976: 64) asks in The Field of Sociology, ‘… in so far as they form groups and are determined by their group existence because of interaction?’ To look at the agent from these two aspects must not be confounded with emphasizing an active and a passive aspect. Both processes demand active individuals, but there are different kinds of activity demanded from them. Simmel’s example from the Metropolis (GSG 7: 116-31) may help explain what is meant by experience as an activity. The attitude of aloofness (blasé) results from the very fact that the inhabitant of the metropolis is exposed to a permanent overstimulation of his (or her) senses. To display the attitude of aloofness vis-à-vis the co-citizens results from consciously filtering the effects received from the urban environment and stylizing them into a distinct social attitude. Different ways of experiencing the world do not simply emanate from passively absorbing the environmental effects. They have to be selected and culturally stylized under the guidance of both individual and supraindividual categories. In shaping externally visible ways of experience, the interiority of the individuality functions, as it were, as a battleground between the everconflicting principles of individuality and generality. Aesthetic, erotic, ethical and other ways of experience are characterized by Simmel by the proportions in which these two conflicting principles are represented in them.

At this point, it may be advisable to take into consideration a warning made by Margaret S. Archer (1995) against conflationary theorizing. Although both processes, externalization and internalization, are linked together empirically, they have to be treated as analytically independent, producing different effects in different interaction sequences and for different actors. It is exactly this type of non-conflationary theorizing Simmel demonstrates in his empirical work. Just take as an example his sociology of culture: ‘Objective culture’ is the emergent product of previous interaction sequences in which individuals have externalized their cultural interests by creating cultural forms. The longer these forms exist, the more they produce emergent properties conditioning in their turn the individuals’ cultural interests. Following Archer’s advice, it is important to make clear which actors are exposed to these emergent effects at which point in time. Different groups or generations of actors may and, as a rule, do react differently to being conditioned in their interest realization. As Simmel described, some may protest against the overwhelming powers of objective culture (for example, against the institution of marriage) and revolt against ‘forms’ as such, in the futile attempt at trying to realize their sexual desires outside any social institutions. Other types and actors may at other points in time adapt more easily to being culturally conditioned, limiting themselves to reforming the given cultural institutions. Bringing back this Simmelian sociological approach to the ongoing theoretical debate may help overcome the controversy between individualism and collectivism and strengthen the arguments presented in Archer’s morphogenetic theory against ‘conflationism’ and ‘elisionism.’

Institutionalization

The third and fourth processes, institutionalization and interest-formation, are related to Simmel’s famous (but often misunderstood) distinction between form and contents. The investigation of the forms of sociation is the core task Simmel assigns to sociology. His ‘Great Sociology’ is an attempt at demonstrating the scientific fertility of this type of formal or institutional analysis. Departing from Levine’s (1971: xxiv-xxvii) distinction between four different forms in Simmel’s work—(1) the forms of elementary social interaction; (2) institutionalized structures; (3) the generic forms of society itself, and (4) autonomous ‘play’ forms—the ‘Great Sociology’ can also be re-read as a contribution to the sociology of institutions and institutionalization. The decisive criterion of progressive institutionalization is the degree in which forms are ‘condensed’ (verdichtet) or crystallized. Increasing condensation means that social forms become increasingly autonomous in relation to their original creators. Accordingly, three different levels of institutionalization can be distinguished. The first level of institutionalization is made up of patterned everyday mundane interactions. On this level, the actors autonomously negotiate and control the ways in which they realize their interests. Individual deviations from the negotiated patterns of interaction can be made at a relatively low social cost. The more these elementary forms crystallize and start defining their own laws of behavior, the more the individuals feel restricted in their ability to realize their interests spontaneously. The higher the degree of institutionalization, the higher the social costs for individual freedom.

The second level of institutionalization is characterized by interactions within institutionalized structures. The very quality of a higher degree of crystallization is felt by the actors as an increasing constraint on their individual choice. On this level of institutionalization, we find the essential characteristic of social institutions proper, defined by Jepperson (1991: 146) in a Simmelian manner as ‘freedom within constraint’: individual action is free within the constraints of institutionalized structures. Throughout his ‘Great Sociology,’ Simmel pays special attention to the question of how the tension emanating from this duality between freedom and constraint expresses itself socially. He emphasizes that the very social quality of forms results only from the struggle between opposite poles, such as freedom and constraint. This is one reason why institutionalized structures are socially productive. Another reason for the social productivity of social institutions consists in the very fact that institutions offer action alternatives which individuals could not have chosen without institutions. This fundamental insight into the social productivity of institutions helps us correct the widespread prejudice held among many contemporary social theorists against institutional analysis as being biased against individual freedom. Among the themes to be researched further in institutional theory today is the social productivity of institutionalized structures.

The third level of institutionalization is made up by the ‘generic forms of society itself. Rephrased in Weberian terms as ‘social spheres,’ they can be characterized by the dominance of one criterion of rationality (for example, maximization of profit in the sphere of economics), thus putting a relatively high price on those individually chosen action alternatives which deviate from this criterion of rationality. Deviations from the institutionalized criterion of rationality are followed up by sanctions and control mechanisms, making it increasingly difficult, although not impossible, to act against the logics implicit in institutionalized social spheres.

Applying Simmel’s dynamic perspective, the three types of institutions can be understood as a continuum of increasing institutionalization (or de-institutionalization), in which the costs of individual freedom (in the sense of deviations from the institutionalized criterion of social action) are gradually increasing (or decreasing). Contemporary institutional analysis could largely profit from this kind of dynamic reinterpretation of Simmel’s contribution to the sociology of social forms. The increasing costs for individual choice can be seen as an emergent property of increasing institutionalization, institutional change thus becoming a more and more ‘expensive’ enterprise. Contemporary critiques of institutional analysis have stressed the bias towards stability and constraint and the neglect of the individual actor in conventional sociological approaches to institutional analysis. Reinterpreting institutions in the Simmelian dynamic way helps correct this bias. It contributes to enriching contemporary institutional analysis in yet another sense. We have said that condensation also refers to a gradual process in which social forms become increasingly abstract. As Neil J. Smelser puts it, patterns of interaction become increasingly ‘imagined,’ meaning ‘that they are not “;seen” in any immediate sense, in the way that neighbors, policemen on the beat, the corner grocery store, and the local school are seen’ (Smelser, 1997: 46). In this process of increasing ‘imaginedness,’ individuals do not disappear from the stage of institutional interaction, but they take over more and more the role of representatives of the institution in question. As institutional representatives, ‘institutionalized individuals’ become the very carriers of institutions. We can only agree with Smelser (1997: 47) that this aspect of institutional representation deserves more systematic investigation in future.

The fourth type of form, autonomous ‘play forms,’ is perhaps Simmel’s most original contribution to the analysis of social institutions. Simmel scholars have either overlooked this type or treated it as a rare, or even exotic, social form. The importance of Simmel’s analysis of sociability (GSG 16: 103-121) goes beyond its being a brilliant example of ‘pure’ sociology, as Simmel calls it. I rather classify sociability as a vertical type of institution, a type which deserves being integrated systematically into contemporary sociological theories of institutions. Whereas the three types of institutions mentioned so far refer to horizontal processes of institutionalization, this fourth type is related to a vertical dimension of patterned interaction. What do we mean by patterned vertical processes of interaction? In play forms, the forms represented in ‘real’ life (such as super- and subordination, conflict and secrecy) are transformed into objects of playful discourse. To play, as it were, with the forms of social life presupposes the individual’s ability to take a distance from the seriousness of ‘real’ life and to reflect about it lightly on a meta-societal level. In sociable encounters, the real ‘heavy’ society and the playful ‘light’ society are related to each other vertically. In picking out society and its forms of interaction as a central theme of sociable discourse, sociable people playfully turn the social forms characteristic of society upside down: in sociable encounters, powerholders play as if they were equals among equals, subordinates jokingly imitate their superiors, or men and women flirt with love. As Simmel stresses, sociable encounters are not limited to the sphere of sociability proper. Just as the form of, for example, super- and subordination permeates many (if not all) social phenomena (GSG 8: 180), so play forms also permeate almost all social spheres. Taken together, they constitute what could be called the light social superstructure of social life. It contributes to smoothing the functioning of the heavy social substructure. So, for example, political decision-making is eased by sociable meetings among politicians; decisions over life and death in hospitals are transformed into routine with the help of joking rituals (as Erving Goffman has described so brilliantly); and the ‘iron cage’ (Weber) of bureaucracy is felt as a lighter burden if clients teasingly allude to its built-in problems when dealing with functionaries. Vertical processes of institutionalization function as supporting mechanisms of horizontal processes of institutionalization in yet another sense: they are a socializing mechanism in so far as they playfully teach their participants how social forms function in ‘real’ life. Taking the supportive and socializing functions of sociability together, it could be said that vertical playforms contribute to increasing the ‘flexibility management’ (Nedelmann, 1995a: 21) of social institutions. An institution that is suffused by playforms increases the range of interpretation within which it has to control deviations from its rules and norms. A more systematic integration of this vertical type of institution into the analysis of the other three types of horizontal institutionalization would certainly enrich present theories in this field of research. It would also help include a range of empirical phenomena which, so far, have lived in isolation from ‘real’ institutional analysis, that is, joking rituals and social games.

Interest-Formation

Let us now turn to the fourth process that is related to the concept of secondary contents, interest-formation. Simmel gives a prominent example of this type of process in his excursus on Treue und Dankbarkeit (GSG 11: 652-70). The emotions of fidelity and gratitude emerge from interactions within social institutions. They constitute the emotional bond between individuals and institutions, thus making up the ‘cement of society’ (Elster, 1989). Why is it of relevance for contemporary sociological theory to pay attention to such processes of emerging secondary contents, or, interest-formation? The great success rational choice theory enjoys at present among sociologists has as its consequence the narrowing of our awareness for processes of interest-formation and transformation. Looking through Simmel’s sociological glasses, interests are permanently formed and transformed within and through social institutions. In the example mentioned, Simmel analyses the dialectics between love and fidelity in marriage: love (primary content), first, giving rise to fidelity (secondary content), which, then, reactivates feelings of love, which appear now as institutionally generated emotions. Institutionally mediated contents can, accordingly, follow a dialectical process of succession of primary and secondary contents, the institution itself functioning both as generator and terminator of contents (Nedelmann, 1984: 99-101).

The problem of how institutions can generate interests and feelings is a highly important one in many areas of institutional analysis today. Just take the example of institutional design in politics. How can, for example, the institutional setting of the European Union generate feelings of belongingness among us Europeans, just as the old institution of the nation-state once gave rise to feelings of belonging to the same ‘community of fate’ (Weber)? Having introduced the European Monetary Union, how can we trust the new currency, the Euro? Or how can people trust the newly designed democratic institutions in Eastern Europe (Sztompka, 1998)? The Simmelian approach certainly does not give us a readymade answer to these questions. But his explicit awareness of institutionally generated interests and feelings helps us formulate a sociological problematic which is highly important for contemporary institutional analysis: how are processes of institutionalization linked to processes of interest-formation? And how do they feed back upon each other?

The Importance of Ambivalence

Let us now turn to the set of variables Simmel repeatedly used in his empirical research. The broad spectrum of empirical subjects Simmel researched is held together not only by a leitmotiv to which he returns over and over again (and to which we will turn later), but also by his interest in studying them with reference to recurrently reappearing variables. In their combination, they make up a coherent research program which Simmel followed himself. For us contemporary Simmel scholars this set of variables can function as a scheme for systematically reinterpreting his sociological oeuvre. Although impressive work has already been undertaken with relation to some of these variables, each of them deserves more systematic exploration. In this context, I wish to highlight especially variable (4), dualism.

The meaning of dualism goes beyond being an empirical variable. It characterizes Simmel’s way of thinking throughout his sociological work. After having followed one way of interpretation, Simmel typically starts all over again by unfolding the opposite direction of interpretation. Witnesses of his university lectures report that these shifts from one line of thought to the opposite were marked by a ‘silent second of self-oblivion in which he inwardly annihilated’ what he had just said before (Marcuse, 1958: 191; my translation). This ‘silent second’ is translated in his written texts into phrasings with which he typically opens the next paragraph, such as ‘on the other hand …,’ ‘the opposite direction …,’ or ‘a totally different picture presents itself ….’ They only insufficiently prepare the reader to meet an author whose body of thought is based on unfolding contradictions, dualisms and opposites which he stubbornly refuses to merge into one predominant statement. As frustrating as this reading experience may be for the student who yearns for salvation by discovering one truth and one theory, it is stimulating for the enlightened scholar who looks for progress by accumulating refutations.

Dualism has yet another, more substantive meaning. As has been hinted at above, it is Simmel’s conviction that forms of interaction acquire a social quality only if dualistic or opposite social forces are at work. ‘It was Simmel’s repeatedly expressed view that the condition for the existence of any aspect of life is the coexistence of a diametrically opposed element’ (Levine, 1971: xxxv). The coexistence of dualistic social forces gives rise to social tension, which, in its turn, causes the dynamics of social forms. So, for example, processes of change in forms of super- and subordination are based on the permanent conflict between obedience and opposition; intimate relations get their momentum from the very fact that feelings of both attraction and repulsion, love and indifference are at work. These examples already illustrate what Simmel means: without the co-presence of dualistic forces, there is neither social life, nor social life.

Simmel distinguishes different social phenomena according to the way in which these dualistic forces are related more precisely to each other. Here, we wish to emphasize only one type, ambivalence. Levine (1985), Robert K. Merton (1976), and Ann-Mari Sellerberg (1994) belong to those scholars who, departing from Simmel, already have done systematic research in this area. But in spite of their efforts, contemporary sociologists are still reluctant to integrate sociological ambivalence systematically into social theory. The importance of ambivalence has, therefore, to be highlighted again.

Generally speaking, ambivalence can be defined in the following way: in contrast to social phenomena which are structured by a contradiction between ‘A’ and ‘non-A’ (such as, freedom and constraint in super- and subordination), ambivalent social forces are at work if ‘A’ and ‘non-A’ are present simultaneously. Simmel’s highly complicated study on Flirtation (1984) can be read as a model for ambivalence as a form of interaction. Mutually reciprocating ambivalent messages (‘yes’ and ‘no’) in one and the same action unit are not only characteristic of gender relations, but also of interactions between other types of actors and in other social areas, such as politics, law or business. In my own research, I have distinguished between ambivalence (a) as & form and (b) as a norm of interaction, the latter one referring to Merton’s (1976: 6) core type of sociological ambivalence. The difference between these two types consists in the very fact that ambivalence does not necessarily have to be integrated into the role structure and thus be normatively prescribed. Ambivalence can also be a form of interaction which is chosen deliberately as an action alternative without being normatively expected. The preference for ambivalence as a form of interaction increases both present and future alternatives of action. This is especially so in decision-making on highly sensitive political issues. The German law on abortion is an excellent case in point. After unification, it became necessary to rewrite legislation on abortion. Whereas abortion was defined as an illegal act in the old Federal Republic, it was not in the former DDR. In the new legislation, abortion is, on the one hand, defined as a criminal act to be prosecuted legally; but, it is, on the other hand and in the very same paragraph, not considered criminal and not prosecutable, if certain conditions are fulfilled. Ambivalent decision-making is a widely used strategy of conflict management without which peace agreements, resolutions on the labor market, or family conflicts could not be handled. In these areas, ambivalence is not normatively prescribed, but deliberately chosen as a way out of the restrictions implied in any unambiguous decision-making. Combining this Simmelian type of ambivalence as a form of interaction with the Mertonian type of ambivalence as a norm of interaction, we arrive at a third type of ambivalence, that is, normative forms of ambivalence. Professional politics, especially diplomacy, is an excellent area in which this third type could be studied further (Nedelmann, 1997). By making more systematic investigations in these social areas, we could not only hope for gaining new empirical insights, but also for correcting the bias of one-dimensionality still dominating contemporary theories of social action. The argument I am stressing here differs from the one often held against the onesidedness of present theories of social action, namely, that they have to take into consideration both rational and emotional motives. Instead, I am arguing for a theory of social action that is built on the actors’ simultaneous orientation towards two, equally valent criteria of action (be they rational, affective, traditional or whatever). The Weberian types of social action and the Parsonian pattern alternatives are excellent starting points for elaborating further such a theory of ambivalent social action. They can be reinterpreted as possible ‘mixes’ of ambivalent action orientations in which the extreme poles of action orientation are fused into equally valent action alternatives. Social interaction based on ambivalence can be understood as an exchange process in which the actors reciprocate ambivalences.

Stopping the Flight from Sociology by Reintroducing Simmel’s Leitmotiv

At the turn of this millennium, there are two different theories enjoying particular attraction among sociologists—individualization and globalization. The followers of each theory are ‘fleeing the iron cage’ (Scaff, 1989) of ‘society’ in two opposite directions: individualization theorists into the direction of the individual, globalization theorists into the opposite direction of imagined global networks. In a truly Simmelian manner, this flight from society could be described as a double process of centripetally fleeing theorists of individualization, on the one hand, and centrifugally fleeing theorists of globalization, on the other. There could hardly be a stronger manifestation of one-dimensional thought in present social theory-building than the fashionable upswing of these two diametrically opposed theoretical streams. Together they leave behind a vacuum in contemporary sociology. Both evade, although in different ways, the encounter with the ‘vexacious fact of society’ (Archer, 1995: 2), that is, in Simmelian terms, with the fact that supraindividual social forms (or social institutions in the four meanings discussed above) emerge from interactions between individuals. Whereas globalization theorists evade the encounter with social institutions by escaping into the meta-societal atmosphere of imagined global networks, individualization theorists crawl back under the institutional level finding comfort from the ‘vexacious fact of society’ in the closeness of face-to-face interactions. One of the main puzzles our discipline has tried to solve since its beginning, the emergence of individually caused supraindividual institutions, is precipitously abandoned even before it has been solved satisfactorily. As a consequence, both theoretical directions contribute in their way to deconstructing sociology. Globalization theorists have started a frenetic clearing-out activity of some of the basic sociological concepts replacing those ‘which reflected an older order, such as society, class, state’ (Albrow et al, 1994: 371) with new ones, such as ‘community, socioscapes and milieux’ (Albrow, 1996: 155-9). Individualization theorists have declared the beginning of the so-called ‘second modernity’ (Beck, 1997a), before having fully profited from the contributions our classics made to the analysis of the ‘first modernity.’ Johannes WeiB (1998: 418-19) rightly observes that the way in which Ulrich Beck talks about ‘individuality’ or ‘individualization’ is surprisingly undifferentiated, reflecting even a stubborn unwillingness to differentiate it further. At best, Simmel’s highly differentiated theories of individualization are alluded to, but no systematic efforts are made at integrating them into their theorizing. Thus, Simmel’s own prophecy has come true that he will not leave behind any heirs—but it has come true not because he did not leave behind any heritage, but, on the contrary, because we contemporary sociologists have not fully discovered his heritage, which offers us ‘one of the most sophisticated perspectives on social interaction that we possess’ (Levine 1997b: 202).

I believe that the double flight from the core problem of sociology could be slowed down, if not stopped altogether, by remembering the leitmotiv of Simmel’s sociology. Given the broad range of empirical subjects he dealt with, Simmel scholars have asked themselves if it is at all possible to identify a general problem in Simmel’s sociological opus. I claim that this is indeed possible. As has already been mentioned above, Simmel’s sociological program was to investigate the forms of sociation. Over and over again he asked himself how autonomous social forms emerging from interactions between individuals could be mediated with their creators’ interests and desires. Dealing with such highly different topics as the Metropolis (GSG 7: 116-31) or The Picture-Frame (GSG 7: 101-8), Simmel was intrigued by the question of how to find a balanced relationship between the single individual and cultural institutions. Let me just take the example of The Picture Frame. For Simmel, the relations between the painting, the frame and the cultural environment function as a metaphor for discussing the fundamental sociological problematic of how to mediate between the individual and culture. It is the frame, says Simmel, which has to take over the function of mediating between the piece of art and the individual in his or her role as art consumer, by both separating and uniting, ‘the task on which, in analogy to history, the individual and the society crush each other’ (GSG 7: 108; my emphasis and translation). The leitmotiv reappears here as the problematic of ‘social framing’ (Nedelmann, 2000), that is, of finding intermediary institutions between the ‘individual’ and objectified social institutions. Also the study of the metropolis is marked by Simmel’s deep concern about how the city dweller can avoid being crushed in his or her individuality by the overwhelming objective power of modern urban life. One mediating instance is here the development of stylized attitudes towards fellow citizens, such as aloofness. Without being able to go into further detail, Simmel’s leitmotiv could be summarized in the following question: how is it possible to mediate between the single individual and social institutions without either of the two sides having to sacrifice its underlying principles of existence? His diagnosis of the modern era was that such mediating instances were missing. As a consequence, the relationship between the individual and the cultural world was distorted, ‘individualization,’ ‘exaggeration’ and ‘paralyzation’ representing only three of the problems of modern culture (Nedelmann, 1991). It is exactly because this leitmotiv forces us to look at both sides of sociation, to individual interaction and emerging social institutions, that it lends itself self-evidently as a sociological problematic bridging, and perhaps even integrating, individualization and globalization theories.

Overcoming Moralizing and Trivializing Sociology by Learning Simmel’s Lessons

The present internal fragmentation of our discipline has not only increased the uncertainty of how to legitimize sociology internally, but also externally. There are two extreme ways in which contemporary sociologists typically react to the weakening of external support. They either trivialize or moralize the role of sociology in contemporary society. Whereas the first type of reaction underestimates the task of our discipline, the second one overestimates it by deliberately taking over political and moral competences we sociologists are not trained for. Both trivialization and moralization are unprofessional scientific strategies for filling the legitimacy gap.

It is not only by his merits, but also by his obvious weaknesses that Simmel can teach we contemporary sociologists how to react professionally in such a situation of legitimacy crisis. As founders not only of sociology but also of the German Association of Sociology, Simmel and his contemporaries took special care in defining sociology as an autonomous academic field obeying its own set of rules. Not moral rules, but only the ethics of science itself were considered by them as the basis of sociology. When Weber, Simmel, Sombart and others jointly left the German Association of Sociology in 1912, they did it in protest against attempts at moralizing sociology and thereby deconstructing our field as an independent self-regulating academic discipline. Their declaration of Werturteilsfreiheit was a commitment towards sociology as a discipline that refrains from making value-judgements, simply because it is beyond the competence of professional sociologists to decide moral values.

As is well known, Simmel broke with this professional rule when he, in 1914, started lecturing about the war and propagated militarism as a solution to the cultural problems of modernity. His Kriegsschriften (GSG 16: 7-58), published in 1917, are a warning of how even the most brilliant and sharp sociological thought can be deconstructed by one-sided political judgements. Therefore, and only therefore, his writings on the war deserve rereading whenever voices are raised today for moralizing or politicizing our discipline. The fact that Simmel wrote The Field of Sociology in the same year (1917) shows that his commitment to the professional role dominated his temporary aberrations in the field of political judgements. Let us, therefore, by way of conclusion, return to Simmel the professional sociologist, in order to see what positive lessons we can learn from him.

The first lesson to be learned could be summarized as consisting in the principle of theoretical homelessness. The greater the fragmentation of our discipline, the greater the felt need among sociologists to find, as it were, a theoretical home. The belongingness to one theoretical home helps identify ourselves and others in a situation in which it is more and more difficult (if not impossible) to get an overview of our field. The security of belonging to a theoretical home has a high price. We limit our perspective of how to look at the empirical world by wanting the latter to fit our theory and to stabilize our theoretical fortress. Not seeking shelter under any theoretical roof, Simmel can teach us to take higher risks as theoretically homeless sociologists. Instead of only looking for empirical facts confirming his findings, he also searched for those contradicting them. He systematically practiced (in the ‘silent second’ mentioned above) a procedure which, following Popper, could be called the procedure of ‘conjecture and self-refutation.’ By refusing to integrate his contradictory findings into a closed theoretical system, he expressed his resistance against any kind of dogmatic thinking. He consciously exposed himself to the risk that goes hand in hand with self-refutation and theoretical homelessness, that is, of not being easily identified in the midst of a sociological landscape made up of closed theory buildings. The present tendencies towards dogmatic self-closure and fragmentation could be overcome by taking higher professional risks and following Simmel’s principle of theoretical homelessness and the procedure of self-refutation.

There is a second lesson to be learned from Simmel. As a problem-finder, he presents the counter-type to present theory builders who consider their main role as consisting in solving problems. Simmel’s sociology does not teach us how to solve problems, but rather how to find problems (Kráhnke, 1999: 100). As we have argued here, Simmel’s problem-finding activities are systematically guided by his conceptual frame of reference and by a set of recurrently used variables. In his ‘Notes on Problem-Finding in Sociology,’ Merton (1959: ix) quotes the experience of scientists according to which ‘it is often more difficult to find and formulate a problem than to solve it.’ To this it could be added that it is also more difficult for the student to follow problem-finding than problem-solving scientific activities. Readymade theory constructions are easier to pass on to next generations of scholars than ever-so systematic procedures of problem-finding. Problem-finding activities are generally less gratifying than problem-solving activities. The way in which empirical research is institutionalized today has as a natural consequence to put a premium on solving problems others have defined for us. This is why skills in problem-finding are less cultivated in contemporary sociology. Rereading Simmel could correct this shortcoming and bring back the figure of the sociologist as a problem-finder. In autonomously finding and defining our problems, we could raise the autonomy of our discipline and, as a consequence, also its internal and external legitimacy.

The third lesson to be learned by Simmel is his sociological glance. In his concluding paragraph of The Self-Maintenance of Social Groups (GSG 5: 371/2; my translation), Simmel stresses: ‘What is most important is to sharpen the glance for that which is sociological in a particular phenomenon and that which belongs into the realm of other sciences—in order for sociology to finally stop digging in an already occupied territory.’ To sharpen the sociological glance today is more important than ever, unless sociology wants to run the risk of being dissolved in the great melting pot of the social sciences. If our sociological glance is obscured by moral concepts, we dig in fields already occupied by professional politics; if we trivialize our professional role, our eyes can only grasp a small fraction of sociologically relevant phenomena. To sharpen the sociological glance means, following Simmel, to look at the empirical world through the perspective of interaction (including the research implications of relationality, self-referentiality and process-analysis). This may seem a modest, perhaps too modest, sociological perspective. But, as Simmel writes in concluding his study on the Metropolis (GSG 7: 131), it is neither our task to accuse, nor to forgive, but only to understand. Simmel’s Goethe epigraph quoted at the outset can also function as a motto for Simmel scholars in the next century: ‘The century has run its course, but the single individual has to start all over again.’ German-reading students are encouraged to start their Simmel research all over again by the fact that the edition of Simmel’s oeuvre will soon be completed. However, they are well advised to integrate into their work, more than they have done before, the foundations laid especially by Levine in the past three decades for a better understanding of Simmel’s sociology. Concerning students who are dependent upon English translations of Simmel’s texts, Alan Sica (1997: 294) may be right when saying: ‘The chance that a hermeneutically competent rendering of Simmel will glide into the canon … is small, … until Levine or a similarly knowledgeable expert displaces the earlier definitive texts with new ones more reflective of the “complete” Simmel that begins to emerge from Levine’s characterization.’