Globalization or World Society? How to conceive of modern society

Niklas Luhmann

International Review of Sociology March 1997, Vol. 7 Issue 1, p. 67. Text version corrected and adapted from the Library of Russia. Annotations by Erik Empson

No one, I think, will dispute the fact of a global system. Whether we watch the BBC news in Brisbane, Bangkok or Bombay, its programme preview indicates Hong Kong time and other times so that we can calculate what to see and when to see it wherever we are. And the news comes from all over the world, not just from England. Wherever people have money to spend, they find supermarkets and boutiques aptly named to remind us of an American or a French background, whether or not the items on display retain any connection with American or French culture. One may, of course, mention the volatility of the financial market with its new derivative instruments for simultaneously maximizing security and risk with unpredictable effects. One may think of the international concern with events in the former Yugoslavia, in Somalia, in South Africa, in Azerbeidjan and not just with events close to the borders of one's own country. 'International', indeed, no longer refers to a relation between two (or more) nations but to the political and the economic problems of the global system. And last but not least, science is not differentiated into regional, ethnic or cultural sciences but into disciplines and research fields.

Moreover, the simultaneity of changes all over the world deserves attention. Everywhere new problems in planning and controlling innovations in organizations and in production technology arise. Religious, ethnic and other types of 'fundamentalisms' emerge all over the world and show that those conflicts of interest to which the state apparatus became adapted while developing into a constitutional state and a welfare state, are just trivial compared with what we have to expect in the future. The economic system has shifted its bases of security from property and reliable debtors (such as states or large corporations) to speculation itself. He who tries to maintain his property will loose his fortune, and he who tries to maintain and increase his wealth will have to change his investments one day to the next. He can either use new derivative instruments or must trust some of the many funds that do this for him. This leads to unsolvable problems in all kinds of 'socialist' policies. And intellectuals are developing their own derivative instruments as well, describing what others are describing under the common denominator of 'postmodernity'. There is no possible regional explanation for these facts. They do not have an 'origin', and one may doubt whether or not they have a 'function'. Apparently, society reacts to itself, but what do we mean by society? What do we do with the everyday knowledge that we take for granted? How can we conceptualize it? What do these facts indicate? Is the global system a society, or is it a system of societies, as Parsons would have it?(1)

They facilitate communication, but they do not serve as concepts. They have to be avoided in the construction of theories? Using names prevents explanations and, above all, comparisons. It is nothing but 'political' talk.(2)

This is not simply a terminological question. It touches upon the very concept of society, the most difficult concept sociology has inherited from the past. What is the core meaning of this ambiguous concept and what are its essential features? Can sociology following Max Weber avoid it altogether? Can we conserve its traditional 'civil' (= political) meaning or are we compelled by the emergence of a global system to change the concept?

My main point will be that, throughout the tradition and in modern times as well, the concept of society proclaims a specific combination of difference and identity, of differentiation and reconstructed unity, or, in traditional language, of the parts and the whole. In all traditional societies, whether antique, medieval or early modern, the principle of differentiation has been stratification, or hierarchy, although the secularization and de-cosmologization of this concept changed the semantical context. In order for society to count as such, this and only this form of differentiation has to be recognized and accepted. On this basis one could then try to find a corresponding reconstruction of unity.

Looking back several centuries, we can observe an increasing de-naturalization of the idea of human society, and this semantic change seems to correlate, on the structural level, with an increasing importance of functional differentiation. With the secularization of hierarchy, the principle of unity had to become a secular principle. From the late 17th century to the 18th century, from Moliere to Alexander Pope, this principle was human happiness.(3) If one were satisfied with the status or condition acquired at one's birth, one could be happy in all walks of life. The world offers more chances to be happy than any status group can realize for itself. And happiness is no zero-sum affair. Even if the world were not designed to make humans happy, even if it were a test ground for salvation only, it could be considered the best of all possible worlds (Leibniz). Every individual could make himself or herself happy by adapting to the conditions; or at least could enjoy the idea that high status by itself does not provide for happiness. The higher classes, in addition, may have 'taste' (Pittock, 1973).

Pope still seems to be convinced that 'Some are, and must be, greater than the rest more rich, more wise; but who infers from hence that those are happier shocks all common sense?

But if one finds this in writing and print one may come to a different opinion. The printing press evokes critique. Malthus' Essay on Population (1926) [1798] marks the end of the idea of a society perfectible in terms of human happiness. The successor to this fragile unity of happiness (for all) and taste (for a few) was, in England and Scotland at least, a new concept of culture as cultivation--for the new commercial society (Williams, 1961).

But the prospect of happiness did not take into account the new conflicts, of the new society. The 17th and 18th centuries brought about an increasing commercialization and market orientation--first of agriculture (on the base of inherited real estate) and, then, of production (on the base of alienable capital). For the 19th century, from Fourier to Durkheim, the keyword was not culture but solidarity.(4) The industrial revolution and the coming to power of the new Bourgeoisie had replaced the old 'natural' order of social estates with a new structure of social classes, depending not upon origin but on career and therefore being visible as contingent. Solidarity was conceived as a kind of moral obligation or, at least, as a binding collective consciousness. But the background assumption remained stratification, now .in the form of a class society producing wealth by division of labour and, thereby, multiplying vertical and horizontal differences. The concept of society included an injunctive component 'that the very name of society implies that it shall not be a mere race, but that its object is to provide for the common good of all.(5) At the same time, we find, as public knowledge, a theory of thermodynamic trends towards entropy that predicted the unavoidable 'heat death' and seemed to prove that the moralities of the day (solidarity or struggle for survival), capital formation and industrial organization, are but temporary and contingent crystallizations?

The enormous increase in diversity within the boundaries of the global systems and the increase in possibilities set free by functional differentiation and by technological development leads to a response at the semantic level of societal self-descriptions. Relativism generates the quest for legitimation.(7) Within the frame of the possible, society needs a narrower frame of the permissible. It produces a variety of devices to enclose what can then be regarded as meaningful expectations: a frame within the frame of the possible. This internal frame may be described in terms of institutions, such as ethics, culture, the canon, recognized heroic action, masterpiece, or the classics. The master discourse of modern society, its 'incomplete project' (Habermas, 1981), uses a humanistic framework. Upon close inspection the project 'modern society' shows a paradoxical face: freedom and equality, self-realization and solidarity. But the paradox is called 'reason', and the project is, to use an 18th-century slogan, pregnant with future. The future, however, remains future and can never become present. It contains the prospect of oscillation between the two sides of the paradox. But in view of the many urgent problems confronting us today, is there any guarantee that this self-contained paradox of modernity will remain our paradox and the future will remain the unlimited horizon of resolving this paradox?

The 20th century has brought about neither happiness nor solidarity. Solidarity, in fact, has become a euphemistic term for social movements (Poland), tax increases (Germany), or for demonstrating public spending in the countryside (Mexico). Society now pretends to be an 'active society' (Etzioni, 1968), heading towards an increasing 'similarity of living conditions' in each country and all over the world. No one should be better off than the others or, at least, no one should be forced by the circumstances of his life to live far below a decent average standard. This aspiration is reproduced by the mass media and the mass markets--as an aspiration. But there are no signs of realization. Obviously, society cannot live up to its own promises. The discrepancy is even more obvious with regard to living conditions than with regard to solidarity or human happiness. And again, similarity of living conditions is opposed to stratification as unity is to differentiation. It is meant to level off stratification as if this were the problem of the human condition.

At the end of the 20th century we have to learn this lesson. In vain we try to use the leftover vocabularies of a tradition whose ambition it was to define the unity, or even the essence, of the social. Our problem is to define difference and to mark off a space in which we can observe the emergence of order and disorder.

We have to come to terms, once and for all, with a society without human happiness and, of course, without taste, without solidarity, without similarity of living conditions. It makes no sense to insist on these aspirations, to revitalize or to supplement the list by renewing old names such as civil society or community. This can only mean dreaming up new utopias and generating new disappointments in the narrow span of political possibilities. These desirabilities serve as a central phantom that seems to guarantee the unity of the system. But one cannot introduce the unity of the system into the system. We may well recognize the hardships and the injustice of stratification, but this is no longer the main problem of society. For its scheme of difference and identity is no longer framed by stratificatory (or hierarchical) differentiation. Stratification would mean that we could know the addresses of influential people and the ropes, and that we would be able to change the structure of society by appealing to reason, by critique, by reforming institutions, or by revolution. But this has become more than doubtful.

If we look at the huge masses of starving people, deprived of all necessities for a decent human life, without access to any of the function systems, or if we consider all the human bodies, struggling to survive the next day, neither 'exploitation? nor 'suppression'--terms that refer again to stratification--are adequate descriptions. It is only by habit and by ideological distortion that we use these terms. But there is nothing to exploit in the favelas; nor are there, at the higher levels of society, actors or dominant groups that use their power to suppress these people. (There are of course individuals, families or groups which, like everyone else, use their networks to their own advantage.) 'Exploitation' and 'suppression' are outdated mythologies, negative utopias suggesting an easy way out this situation, e.g. by 'revolution'. The predominant relation is no longer a hierarchical one, but one of inclusion and exclusion; and this relates not to stratification but to functional differentiation.

Traditional societies included and excluded persons by accepting or not accepting them in family households, and families (not individuals) were ordered by stratification.(8) Modern society includes and excludes persons via function systems, but in a much more paradoxical way. Function systems presuppose the inclusion of every human being, but, in fact, they exclude persons that do not meet their requirements. Many individuals have to live without certified birth and identity cards, without any school education and without regular work, without access to courts and without the capacity to call the police. One exclusion serves as an excuse for other exclusions. At this level, and only at this level, society is tightly integrated, but in a negative way. And modern values, such as equality and freedom, serve as cover terms to preserve an illusion of innocence--equality as equal opportunity and freedom as allowing for individual (and not societal) attribution.

It is one thing to describe modern society as a functionally differentiated system that generates social classes as a useless byproduct of the selective operations of its function systems. It is quite another thing to define society as a social system that can change its form of primary internal differentiation. Functional differentiation is a specific historical arrangement that has developed since the late Middle Ages and was recognized as disruptive only in the second half of the 18th century. One cannot define the concept of society by one of its possible realizations. If' one restricts the concept to particular aspects of modern society, the temptation becomes irresistible, to include in the concept, ideological or normative assumptions such as human happiness, solidarity, similarity of living conditions, or communal integration. The theoretical decision then becomes the source of misdirected dissatisfaction, critique and protest. To date, however, we do not have a convincing meta-concept that would encompass all possible dominant forms of differentiation from segmentation to centre/periphery differentiation to hierarchy and, finally, to functional differentiation. Therefore, we need to rephrase the problem and replace the humanistic approach and its affectionately 'social' concern by the question: what does it mean and how is it possible that a system can change its dominant form of internal differentiation?

We can conceive of differentiation as the process of reproducing systems within systems, boundaries within boundaries and, for observing systems, frames within frames, and distinctions within the distinguished.(9) This presupposes the stability of boundaries as a result and as a condition of evolution. Protected only by' boundaries, and only inside its boundaries, can a system grow in complexity; for only within its boundaries, can a system operate, build up, change, or forget: structures. A 'double closure' or 'double framing' by external and internal boundaries that separate the external environment from the internal environments. of subsystems is a necessary condition for maintaining stability in spite of an evolution toward an ever increasing improbability of structures and 'evolutionary universals' (Parsons, 1964) such as advanced forms of differentiation. How, then, and this again is our question, can a society survive changes in its forms of double closure, its forms of stability, how can it survive a 'catastrophe' in the sense of Rene Thom or, perhaps better, an evolutionary 'anastrophe' toward forms of differentiation that involve higher complexity, more opportunities, more structural contingencies, shorter time periods (acceleration), and more risks of unpredictable breakdowns?

If all this adds up to a point, the concept of society has to be defined not by an idealized state with compensatory functions but by a boundary, that is, by a boundary-drawing operation. Such an operation produces the difference between the system and its environment and thereby produces the possibility of observing the system, that is, the distinction between the system and its environment. This distinction can re-enter the system,(11) it can be copied in the system and then allows for the stability of the system, for referential oscillation between observations, respectively indicating external and internal states and events.

Systems that operate at the level of a re-entry of their form into their form are non-trivial machines in the sense of von Foerster (1984). They cannot compute their own states. They use their own output as input. They are 'autopoietic' systems, and that means that they are their own product. In contradistinction to all traditions that teach that one can only understand what one has made oneself (Bacon, Hobbes, Vico etc.), a re-entry leads to an unresolvable indeterminacy. The system cannot match its internal observations with its reality, nor can external observers compute the system. Such systems need a memory function (i.e. culture) that presents the present as an outcome of the past. But memory means forgetting and highly selective remembering, it means constructing identities for re-impregnating recurring events. In addition, such systems need an oscillator function to be able to cross the boundaries of all distinctions they use, such as, being/not-being, inside/ outside, good/bad, male/female, true/false etc. To be able to separate memory and oscillation, the systems constructs time, that is, a difference of past and future states, by which the past becomes the realm of memory and the future the realm of oscillation. This distinction is an evolutionary universal. It is actualized by every operation of the system and thus gives time the appearance of a dimension of the 'world'. And if there are sufficient cultural guarantees for conceptualizing time, the distinction of time re-enters itself with the effect that past and future presents, too, have their own temporal horizons, their own pasts and futures. The European description of time reaches this reflection of time within time in the 18th century. Since the 18th century, we have to live with the historical relativity of all cultural forms and with a lack of 'origins', that are binding for the present. 'Origins' are now considered self-made origins, e.g. works of art, positive laws, scientific theories or political decisions. The public description of time conceives of the present as the differential of the past and the future, that is, as the time for decision, and this leads to new, highly organized forms of recursivity. Memory and oscillation, selectivity of reconfirmations and uncertainty of the future, are now unavoidable facts of social life.

Now we are ready to settle the issue of the global system. Under modern conditions, the global system is a society, in which all internal boundaries can be contested and all solidarities shift. All internal boundaries depend upon the self-organization of subsystems and no longer on an 'origin' in history or on the nature or logic of the encompassing system. Solidarities are in a process of deconstruction and reconstruction that requires the self-distinguishing capacity of social movements or of ethnic or fundamentalist religious groups. And this means, that solidarity consolidates :itself within society against others. Solidarity, accepting its own genetic conditions, does not and cannot want truce.

Society generates its external boundaries by its elementary operations. These are boundaries between the recursive, self-referring network of communications on the inside, and everything else (including human bodies and minds) on the outside. Communication cannot be sufficiently explained in terms of Saussure's distinction of langue and parole, that is, as an application of language in concrete cases. Each communication identifies itself by referring to past communications and by opening a limited space for further communications. It cannot happen as one single event, it cannot be recognized as communication outside of its own recursive network. 'Communication is recursivity' (yon Foerster, 1993). It has to reproduce both the memory function and the oscillator function, the past and the future of the system. This requires selective operations and, therefore, the drawing of boundaries. These boundaries are relatively clear because the use of language requires the distinction between words and things. Remaining ambiguities (e.g. whether something is communication or only behaviour) can be clarified by communication. Regional boundaries do not have this operational quality. They are political conventions, relevant for the segmentary differentiation of the political subsystem of the global society. They designate places to show passports and, occasionally, generate reasons for war. It does not make any sense to say that they separate societies.

Society constructs its environment around a basic distinction, that between human individuals (bodies and minds) and other environmental facts, which nowadays are called 'ecological' conditions. This distinction is drawn by society itself, by its communicative processes. It has no fundamentum in re but varies its meaning according to changing historical circumstances. To distinguish between individuals and other ecological conditions is a projective reflection of communication; it mirrors the requirements of the autopoietic reproduction of the societal system. For only the consciousness of individuals is structurally coupled with the autopoiesis of the societal system (Luhmann 1988). Only consciousness can irritate communication in a way that is compatible with the autopoiesis and the operational closure of the social system. All other environmental changes (physical, chemical, biological, e.g. death) can only have destructive effects.

Conscious states in the environment of the system have to be presupposed at any moment in time, in every single communicative operation. They have to be presupposed, not only for the time being, but also in the form of a possibility of future communication on the one hand, which links up with what has been said or written before, and, on the other hand, in the form of a past that has already successfully reduced uncertainty and in which individuals have committed themselves to continue communication. This structural coupling of consciousness and society does not determine system states on both sides of the boundary. On the contrary, it presupposes the reciprocal inaccessibility ofconsciousness for communication and of communication for consciousness. The other side cannot be reached, it can only be imagined; for no system can operate outside of its own boundaries. The structural coupling depends upon language as linking device, but there is no supersystem organizing this coupling. Language is not a system.

Moreover, this is the only direct coupling that connects the societal system with its outside. Only consciousness can produce the noise necessary for the emergence and evolution of social order. Only conscious operations can perturbate the communicative system and create preconditions of sense-making within this system. Everything else--say, death, fire, earthquakes, climatic shifts, technological catastrophes--can only destroy communication. Such events can, of course, be observed, that is, thematized by the social system, but to do so requires communication and, as its external condition, consciousness.

The extraordinary importance of individuals with respect to the ongoing reproduction of the societal operation is due to their external (environmental), not to their internal (social) status; it is due to their own self-reproduction, to their own 'autopoietic' closure as minds and as living bodies. Individuals are not and cannot: be 'parts' of society, and it makes no sense to speak of 'participation' (if we: remember the medieval connotations of this term). Given this importance of individual reproducing consciousnesses other ecological concerns might be of minor consequence. Individuals, however, are easy to replace, they die anyway and they live in great numbers. We have greater problems with fresh air and fresh water, with oil and with nourishment, with pollution and with the ozone layer depletion. Besides, ecological interrelations are much more complicated than relations between individuals, which are almost exclusively mediated by society itself, i.e. by communication. I cannot go into details, at this point, but it would be worthwhile'. to compare these two environments from the viewpoint of the reproduction of the societal system. We have to remember, however, that the distinction between human and ecological conditions is quite an artificial construction, reflecting the internal operative necessities of society, whereas individuals as far as their organic: life is concerned, are, in fact, part of the ecological environment, contributing to and suffering from its deterioration.

In our context, where we have to decide between assuming a global system of regional societies or a world society, we have now clear and theoretically consistent arguments for a single world society. The autopoietic system of this society can be described without any reference to regional particularities. This certainly does not mean that these differences are of minor importance. But a sociological theory that wants to explain these differences, should not introduce them as givens, that is, as independent variables; it should rather start with the assumption of a world society and then investigate, how and why this society tends to maintain or even increase regional inequalities. It is not very helpful to say that the Serbs are Serbs and, therefore, they make war. The relevant question is rather, whether or not the form of the political state forced upon all regions on earth fits to all local and ethnic conditions, or, whether or not the general condition, not of exploitation or suppression but of global neglect stimulates the search for personal and social, ethnic or religious identities. One may further raise the question of whether the modern way to describe conflicts as conflicts of interests and values is still adequate in view of a global condition that suggests the emergence of fundamentalist identifications, that is 'against-identities' and not 'career-identities'. If we include utopian schemes such as human happiness or integrative community into the definition of society, we will sensitize our theory to regional differences. But then, even Manhattan would have to be considered a plurality of societies. If, on the other hand, we use the distinction between system and environment as our scheme, we enable ourselves to see, on a world-wide level, the impact of societal operations and structures on individuals as well as on ecological conditions. That the ecological impacts of societal operations cannot be restricted to regional territories does not need any further argument. But their impact on individuals, too, seems to be a universal phenomenon and this will become more evident in the near future. Increasingly, individuals are permitted, or even required, to declare their own identity, their own preferences, interests, beliefs, aspirations and to refer to themselves in communication as if this could legitimate expectations. Even in intimate relations, experience shows that love, not hate, separates, because it provides special hopes and chances for identity development.(12) Apparently, individuals do not see any reason not to use their external positions to produce demands, claims and disappointments. And if this is true for love relations, it is even more so for all social contexts in which careers contribute to identity formation.

Obviously, no autopoietic system can 'adapt' to its environment. It only operates as if it were adapted. This is the reason why modern society slides into more and more problems with its individuals and its ecological conditions. Autopoietic systems are operationally closed systems. But they can observe, that is, communicate about whatever comes into their span of attention. They oscillate between external references and self-reference by focusing on the constative and the performative aspects of communication, on information and on utterance. Sociology may well see a task in correcting its own tradition and in shifting its attention from the outworn themes of stratification and compensatory social ideas to the more urgent external problems.

Let us now return to the question of whether, under modern conditions this primary form of differentiation is hierarchy or functional differentiation. Each type has its special calamities. If we see stratification we will tend to see, as I have said before, injustice, exploitation and suppression; and we may wish to find corrective devices or at least to formulate normative schemes and moral injunctions that stimulate a rhetoric of critique and protest. If, on the other hand, we see functional differentiation, our description will point to the autonomy of the function systems, to their high degree of indifference, coupled to high sensitivity and irritability, in very specific respects that vary from system to system. Then, we will see a society without top and without centre; a society that evolves but cannot control itself. And then, the calamity is no longer exploitation and suppression but neglect. This society makes very specific distinctions with respect to its environment, e.g. usable and not usable resources with respect to ecological questions or (excluded) bodies and (included) persons with respect to human individuals.

Today, the problem is much worse than before. We may continue with our habits and resort to moral claims that are as justified as ever. But who will hear these complaints and who can react to them, if the society is not in control of itself?.

And what can we expect when we know that the very success of the function systems depends upon neglect? When evolution has differentiated systems whose very complexity depends upon operational closure (and the paradigmatic case is, of course, the human brain), how can we expect to include all kinds of concerns into the system?

But this is a question and not an answer, and the question is meant to redirect sociological research. We can observe an enormous amount of structural flexibility and we can presuppose a certain capacity for making the distinction between system and environment within systems. We may redefine system rationality as a re-entry of the distinction between system and environment into systems (Luhmann, 1993). Functional differentiation, then, means that we can expect very different solutions for the problem of rationality in different function systems, but any solution will depend upon complexity, that is, on boundaries and on neglect.

Functional differentiation, too, has produced self-justifying or at least complementary descriptions. The first idea referred to division of labour, transferring it from the role level to the system level (Durkheim, 1930). This could justify the costs and disadvantages of modern life-conditions by an overwhelming surplus of welfare. After World War II, a new conception of 'modernization' emerged. It distinguished between different function systems and proclaimed, under the name of 'development', their modernization by way of a market orientation of the economy, a democratization of politics, equal access to school education, the establishment of constitutional legality (rule of law) all over the world, a political control of the military, a free press, self-directed scientific research and so on. There was no question that all this would add up to an improved state of society and, once more, to better life-conditions for human beings. But how could one expect to integrate the effects of modernization or guarantee the reciprocal support of modernizations in different function systems?

Marxist critique and the dependencia theory both missed the point. They profited from obvious miseries and disappointments, but they returned to ideas of stratification that were already outdated at this time. They assumed centre of power, whether the capitalist class or regional centre of wealth, knowledge, and power, and placed their hopes, of course, again on revolution. They invented theological arguments against 'international corporations' or 'monetaristic policies'--and this no longer in the contemptus mundi style of the past but with the hope and a strong demand for 'liberation'. Liberal and Marxist traditions, as well, seemed to promise that less coercion would mean more freedom.

Recent developments in systems theory suggest a very different picture. If function systems are operationally closed systems, their differentiation will produce more independencies and more dependencies at the same time--more independencies because of their operational closure and their highly selective structural couplings, and more dependencies because society can maintain its present achievements only if all the function systems operate and reproduce themselves at an adequate level. The world society has reached a higher level of complexity with higher structural contingencies, more unexpected and unpredictable changes (some people call this 'chaos') and, above all, more interlinked dependencies and interdependencies. This means that causal constructions, (calculations, plannings) are no longer possible from a central and therefore 'objective' point of view. They differ. depending upon observing systems, that attribute effects to causes and causes to effects, and this destroys the ontological and the logical assumptions of central guidance. We have to live with a polycentric, polycontextural society.

Given these conditions, there is no longer a quasi cosmological guarantee that structural developments within function systems remain compatible with each other. Science does not add knowledge to power but uncertainty and risk to decisions. Physics made it possible to produce the atomic bomb, the economy finds it profitable to use high risk technologies--both with enormous impacts on the political system. The free press changes politics into a turmoil of scandals and enforces and reveals hypocrisy as the typical style of political talk, and this leads to a widespread critique of the 'political class' and to a decline of political trust. The highly efficient modern medicine has demographic consequences. The new centrality of international financial markets, the corresponding marginalization of production, labour and trade, and the transfer of economic security from real assets and first rate debtors to speculation itself, leads to a loss of jobs and seduces politicians to 'promise' jobs (without markets?). The welfare state produces completely new problems for the legal supervision of politics and leads to deformations of legal doctrine that undermine the predictability of legal decisions. On the other hand, the corresponding judicial 'legislation' of constitutional courts affects politics in a way that can hardly be called 'democratic' (the degree of centralization of the emerging European Union will not be decided by the governments in London, Paris or Berlin but by the European Court in Luxembourg).

It would be easy to add further items to this list. The point is that we are not in a phase of 'posthistoire' but, on the contrary, in a phase of turbulent evolution without predictable outcome. In classical perspectives, one could compare the 'degree of modernization' - say, of Japan and China - and explain their differences by different structural preconditions and semantic traditions. But when we want to observe the evolution of society there is no other choice than to focus on the social system of the world society.

Looking ahead to our future, we cannot see any other form of differentiation. Regression to earlier forms, say stratification or segmentary (tribal) differentiation, may be possible, but is probable only after some large scale catastrophe. We cannot close the list of possible types of differentiation on ontological or logical grounds, but we cannot conceive of another type either. (Likewise, the stratified societies of the past could think of functional differentiation only at the role level and not as primary differentiation of the societal system itself.) The worst imaginable scenario might be that the society of the next century will have to accept the metacode of inclusion/exclusion. And this would mean that some human beings will be persons and others only individuals; that some are included into function systems for (successful or unsuccessful) careers and others are excluded from these systems, remaining bodies that try to survive the next day; that some are emancipated as persons and others are emancipated as bodies; that concern and neglect become differentiated along this boundary; that tight coupling of exclusions and loose couplings of inclusions differentiate fate and fortune: and that two forms of integration will compete: the negative integration of exclusions and the positive integration of inclusions.

In some places, e.g. the favelas or other forms of ghettoization in large cities, we can already observe this condition, and it is not unrealistic to expect that demographic developments and migrations will feed this kind of differentiation, even in Europe. And again, this is not a regional problem that could be avoided by political regulations and public spending; it is a problem in the relation between the social system of the world society and its human environment.

All of these considerations apply to the social system of sociology as well. Sociologists are not supposed to play the role of the lay-priests of modernity; nor should they satisfy their theoretical curiosity by a post mortera examination of their classics. But the discipline may improve on constructing its object by conceiving of its own re-entry into its object. The distinction between the observed and the observing system (or, in classical terms, between the object and the subject) may re-enter the system as a condition of cognitive rationality. This requires a double re-entry: The contributions of sociological research to the self-description of society become a topic of sociological theory and a problem for its logic and its methodology, or in other words, the re-entry of the observer into the observed re-enters the observer.

At present, the unsolved problems surrounding the concept of society seem to prevent theoretical progress. The idea of a good, or, at least, a better society still dominates the field. Sociologists, interested in theory, continue to explore the old mazes with diminishing returns instead of moving into new ones. It might be rewarding, however, not to look for better solutions of problems - of problems that are constructed by the mass media - but to ask 'what is the problem?' in the first place. For the definition and elaboration of problems and not the proposal of problem solutions is the point around which theoretical stages revolve. In this sense we have to face the choice whether to retain the notion of homeostasis with prospects of improved integration (happiness, solidarity and the like) or whether to see the problem as a problem of complexity (contingency, intransparency, risk, and the like) produced by differentiation. The first option will lead us to accept a regional concept of society as a frame for improvements, the second would recommend starting from a concept of world society in order to define the problems that regions may have to solve by political or other means.

A sociological theory of society is a scientific task, and a very special one. As a science, it specializes in cognitive operations. It follows the binary code of true and false propositions and is, in fact, identified by this distinction (and not, for instance, teleologically, by an end).[13] It has to find and to confirm truth, and to avoid falsity. However, if we as observers and sociologists change the system reference and focus, not on science but on society itself, another function comes in view. A theory of society, whether true or false, and this makes no difference here, contributes to the self-description of the society. It is communicated within society to convey a description of society, including the describing of the description. It refers to its object but also to itself as part of its object---as a subsystem of a subsystem of the society. Any communication about society is an autological operation. It produces a text that combines, even confuses, autoreference and heteroreference, a text that implies a collapse of the distinction between a subject and its object, between the observer and the observed, on which science has to rely for methodological and logical reasons. It can be scientific and non-scientific, depending upon which system reference has been chosen. And by whom? By sociology, of course.

This is no longer a question of social and political responsibilities, not to mention of ethical concerns. It is not a question of whether or not sociology as a science has to commit itself to a value-free stance (which could only mean avoiding selection). Nor is it, by any means, a question of making a decision. Rather, given the structure of its object, society, sociology cannot follow a rule of self-exemption. In this sense (but only in this sense), we find a close parallel between sociological and moral reasoning. Both forms of social communication cannot avoid self-implication. If they try to avoid it they simply fail in what they intend and pretend to do. Now, we shall re-formulate the problem of whether or not we have to accept the fact of a world society for the last time. The problem of autological reference is a universal problem. It cannot be avoided by transferring a centre of research from Bielefeld to Berkeley, and even Paris is no exception. It is not a problem of contesting boundaries or shifting loyalities within the societal system. The question is rather, whether or not a sociological theory is capable of satisfying all the technical requirements of the subsystem science and at the same time, and with the same set of texts, can contribute to the self-description of the society. Can sociology, in other words, operate as science and simultaneously observe the society in which it operates as observer? Can it observe itself as the observer?

We cannot give an 'objective' and definite answer to this question. For the question itself implies a re-entry of the observer/observed distinction into itself. And this means that we shall have to face unresolvable indeterminacies, temporalization, oscillation, memory function and above all that must replace the computation of all possible statements by a feedback reference to the historical situation from which we have to start.

Notes

1. See the title of Parsons' book The System of Modern Societies (1971). Most more recent authors follow this lead. See, among others, Peters (1993) for pragmatic reasons!

2. One may discuss this also in terms of the names of authors of theories and the damages produced by using names such as 'Luhmann' in theoretical discussions.

3. Alexander Pope (1950), 'Essay on Man' Epistle 3, pp. 50-52; cf. Mauzi, 1960.

4. The strong component of 'feeling' in the British concept of culture/cultivation may have been, for some time at least, a substitute for solidarity.

5. So Thomas Arnold (Matthew Arnold's father) as cited by Williams (1961, p. 123).

6. See Hayles (1990, p. 39), based on Crosby (1976).

7. Truth has been replaced by the twins "Relativity" and "Legitimation", writes Burgin (1986, p. 49).

8. There were, of course, important exceptions, e.g. the legal form of corporations (universitates) such as the church or monasteries that collected not families but individuals.

9. In formal terms this would require a 're-entry' of a form in itself in the sense of Spencer Brown (1979, p. 56 ff:) with all its mathematical consequences such as bistability, oscillation, memory functions, temporalization and above all the irresolvable indeterminacy for internal and external observers.

10. For a more general use of this concept, see von Foerster (1981, p. 304 ff).

11. To make the complicated architecture of this theory clear: not only the distinction but also the diffrence between system and environment can re-enter the system. But then we have differentiation.

12. See already Schlegel (1980) [1799] vol. 2, p. 74]) 'Nicht der HaB, wie die Weisen sagen, sondern die Liebe trennt die Wesen und bildet die Welt.

13. As, for example, Edmund Husserl would have it during the difficult years of the Nazi regime. See his Viennese lecture 'Die Philosophie und die Krise des europaischen Menschentums' (7. and 10.5.1935), printed in: Husserl (1954, pp. 314-348).

References

Alexander, J. C. and Colomy, P. (eds) (1990) Differentiation Theory and Social Change, New York, Columbia UP.

Burgin, V. (1986) The End of Art Theory: Criticism and Postmodernity, London, Macmillan.

Crosby, S. (1976) 'Natural philosophy and thermodynamics: William Thompson and the "dynamical theory of heat" ', British Journal of the Philosophy of Science, Vol. 1, pp. 293-319.

Durkheim, E. (1930) [1893] De la Division du Travail Social, Paris, Presses universitaires de France.

Etzioni, A. (1968) The Active Society, New York, Free Press.

Habermas, J. (1981) 'Die Moderne--ein unvollendetes Projekt', Kleine Politische Schriften, Frankfurt, Suhrkamp, pp. 446-464.

Hayles, K. N. (1990) Chaos Bound: Orderly Disorder in Contemporary Literature and Science, Ithaca, Cornell UP.

Husserl, E. (1954) 'Die Philosophie und die Krise des europaischen Menschentums', in ibidera, Die Krisis der europaischen Wissenschaften und die Transzendentale Phanomenologie.--Husserliana, Vol. VI, pp. 314 348. Haag, Nijhoff.

Luhmann, N. (1988) 'Wie ist bewuBtsein an kommunikation beteiligt?', in Gumbrecht H. U. and Pfeiffer K. L. (eds), Materialitat der Kommunikation, Frankfurt, Suhrkamp. pp. 884-905.

Luhmann, N. (1993) 'Observing re-entries', Graduate Faculty Philosophy Journal, Vol. 6, pp. 485-498.

Malthus, T. R. (1926) [1798] An Essay on the Principle of Population, as It Affects the Future Improvement of Society, with Remarks on the Speculations of Mr Godwin, M. Condorcet, and Other Writers, London, MacMillan.

Mauzi, R. (1960) L'idee du Bonheur dans la Litterature et la Pensde Franfaise au XVIIIe Siecle, Paris Parsons, T. (1964) 'Evolutionary universals in society', American Sociological Review, Vol. 29, pp. 339 357.

Parsons, T. (1971) The System of Modem Societies, Englewood Cliffs, NJ, Prentice Hall.

Peters, B. (1993) Die Integration moderner Gesellschaften, Frankfurt, Suhrkamp.

Pittock, J. (1973) The Ascendency of Taste: The Achievement of Joseph and Thomas Warton, London, Routledge.

Pope, A. (1950) The Poems of Alexander Pope, London.

Schlegel, F. (1980) [1799] Lucinde. Werke in zwei Banden, Vol. 2. Berlin, Aufbau Verlag.

Spencer Brown, G. (1979) Laws of Form, 2nd edn. reprint, New York, Dutton.

Ulrich, H. and Probst, G.J.B. (eds) (1984) Self-Organization and Management of Social Systems: Insights, Promises, Doubts and Questions, Berlin, Springer.

von Foerster, H. (1981) Observing Systems, Seaside, CA, Intersystems.

Von Foerster, H. (1984) 'Principles of self-organization in a socio-managerial context', in Ulrich, H. and Probst, G.J.B. (eds), Self-Organization and Management of Social Systems: Insights, Promises, Doubts, and Questions, Berlin, pp. 2-24.

Von Foerster, H. (1993) 'Fur Niklas Luhmann: Wie Rekursiv lst Kommunikation?' Teoria Sociologica, Vol. 2, pp. 61- 85.

Williams, R. (1961) Culture and Society 1780-1950, Harmondsworth, Penguin edition.