But it re- mains the case that schemata are not images which become con- cretely fixed at the moment of depiction; they are merely rules for the
repetition
of operations (which then are concrete again).
Luhmann-Niklas-the-Reality-of-the-Mass-Media
cognition alone, this would lead to every system demanding more of itself than it could possibly achieve operationally.
This is neces- sarily the case if only because given the complexity of the environ- ment, the system does not have the 'requisitive variety' (Ashby).
And even the concept of observation is meant to register that the world can never be observed, let alone understood, because every observation generates with an 'unwritten cross' an 'unmarked space' which it does not observe.
3 It is hard to see how systems of con- sciousness or communication-based social systems might break free from this disparity of system and environment.
The question can only be what share an environment-related cognition has in the evolutionary opportunities of particular kinds of systems.
But what must first be ensured is that the environment tolerates the autopoiesis of the system.
In the case of the social system society, then, it must first be ensured that communication connects onto communication and that not every transition from one communication to another would have to keep a check on the entirety of environmental con- ditions necessary for this, that is, would have to communicate, amongst other things, about whether the participants are still alive.
Under these conditions, therefore, cognition is primarily deployed in such a way that it is oriented to the inside.
The first thing to be sure of is that one communication fits onto another.
4 What is im- portant, then, is adequate behaviour - and not, for example, whether there is enough air to carry a sound from one organism to another.
If, unexpectedly, conditions are no longer given, this will be regis- tered as a disturbance and ways out will be sought (again by means of communication).
This leads to the fundamental question of how communication must be, in order that it can not only reproduce itself but also take on cognitive functions and separate reproductive or informational components. The answer is that communication only comes about at all by being able to distinguish utterance and information in its self-observation (in understanding). Without this distinction, com- munication would collapse, and participants would have to rely on perceiving something which they would only be able to describe as behaviour. 5 The difference of utterance and information corresponds precisely with the requirement of not making the progress of com- munication to communication dependent upon information being
? complete and relevant. And only because this primary, constitutive difference exists can communication code itself in a binary form (for example, with regard to acceptable/not acceptable, relevant/ not relevant) and in this way feel its way around the environment with a distinction for which there is no correlate whatever in the environment itself. Without this distinction, which has been en- tered into its own operation, the system would not be capable of constituting any recognizable identities or developing any memory. Nor could it evolve, or build up its own complexity, or test the possibilities for structuration positively/negatively and thus meet the minimum condition for the continuation of its own autopoiesis. 6 Society as we know it would be impossible.
For the same reasons, no great expectations can be placed on the understanding of communication. Expectations can certainly be raised forcibly, but they then require special differentiated dis- courses. Normally, ambivalences and misunderstandings are borne along as well, as long as they do not block communication; indeed, understanding is practically always a misunderstanding without an understanding of the mis.
It is a big jump from these general systems-theoretical and social- theoretical considerations to the mass media of modern society. The function of the mass media lies after all that in the directing of self-observation of the social system7 - by which we do not mean one specific object amongst others, but a way of splitting the world into system (that is, society) and environment. What is involved is a universal, not an object-specific observation. We have already spoken, in another context,8 of the function of the system's memory which provides a background reality for all further communica- tions, which in turn is constantly reimpregnated by the mass me- dia. What is also involved is an observation which itself generates the conditions of its own possibility and in this sense occurs autopoietically. For the uncertainty as well as the distinctions used for observation are products of the system and are not simply pre- given attributes of the world or ontologically or transcendentally provable decomponates ('categories') of the unity of the world. This means also that the impetus for further communication is repro- duced within the system itself and cannot be explained anthropo- logically, as a drive for knowledge, for example.
? Therefore, one cannot comprehend the 'reality of the mass me- dia' if one sees its task in providing relevant information about the world and measuring its failure, its distortion of reality, its ma- nipulation of opinion against this - as if it could be otherwise. The mass media realize in society precisely that dual structure of repro- duction and information, of continuation of an always already adapted autopoiesis and cognitive willingness to be irritated. Their preference for information, which loses its surprise value through publication, that is, is constantly transformed into non-informa- tion, makes it clear that the function of the mass media consists in the constant generation and processing of irritation - and neither in increasing knowledge nor in socializing or educating people in conformity to norms. The descriptions of the world and of society to which modern society orients itself within and outside the sys- tem of its mass media arise as a factual effect of this circular perma- nent activity of generating and interpreting irritation through information tied to a particular moment (that is, as a difference which makes a difference).
Of course, it should not be implied that irritation happens only in the system of the mass media and not, for example, in marriages, in school lessons or in other interactions; just as power is present not only in the political system, standardizations not only in the law, truth not only in science. Irritability is the most general struc- tural characteristic of autopoietic systems, which in modern de- scription occupies that place once accorded to nature and to the essence of things defined as nature. 9 Irritability arises from the sys- tem having a memory that is actively involved in all operations and therefore being able to experience and balance out inconsistencies - which means nothing other than being able to generate reality. This points to a recursive constitutive context of memory, irritabil- ity, information processing, reality construction and memory. The differentiation of a function system specialized in this serves to improve and simultaneously normalize a means of communication likewise specialized in this. Only from the mass media do we ex- pect this special performance every day, and only thus is it possible to arrange modern society in its execution of communication in an endogenously restless way like a brain and thereby prevent it hav- ing too strong a link to established structures.
? In contrast to the function system of the mass media, science can be specialized in cognitive gains, that is, in social learning proc- esses, whilst the system of law takes on the ordering of expectation which is normative, held onto in spite of the facts and to this extent unwilling to learn. However, the cognitive/normative division be- tween science and law can never divide up among itself and thereby cover the entire orientational requirement of social communica- tion. Under normal circumstances social communication is oriented towards neither science nor the law. But neither can it be left in modern global society to the merely local everyday knowledge that is only found in the nearest vicinity. Accordingly, it seems to be the function of the mass media to remedy this neither cognitively nor normatively specified requirement. The mass media guarantee all function systems a present which is accepted throughout society and is familiar to individuals, and which they can take as given when it is a matter of selecting a system-specific past and establish- ing decisions about future expectations important to the system. Other systems, depending on their own requirements, can then adapt themselves to the past reference of their anticipation; for example, the economy can adapt itself to new circumstances in companies or in the market, and on this basis establish their own connections between their past and their future.
It was Parsons who saw that the particular contribution of the mass media to the 'interchanges' of modern society lies in the in- crease in levels of freedom of communication - analogous to the function of money in the economy. 10 This diagnosis can be broad- ened if one additionally takes into consideration the increase in society's capacity for irritation and the recursive interweaving of mass media communication with everyday communication in the interactions and organizations of society. On the one hand, the mass media draw communication in, on the other, they stimulate ongoing communication. 11 So they continuously apply new com- munication to the results of previous communication. In this sense they are responsible for the production of modern society's 'Eigenvalues' - those relatively stable orientations in the cognitive, the normative and the evaluative domain which cannot be given ab extra but rather arise out of operations being applied recursively to their own results. 12
? It seems that a centuries-old tradition has led us astray, with the result that mass media appear in an unfavourable light. The tradi- tion says that the stability of the social system rests upon consensus - or even on an explicitly/implicitly agreed social contract, and if no longer upon a commonly held religion, then at least on consen- sually accepted background convictions, encapsulated in Jiirgen Habermas's concept of lifeworld. Were this not the case, the mass media would be a destabilizing factor, only out to destroy these presuppositions and to replace them with something the French might call symbolic violence.
In fact, however, the stability (= reproductive capacity) of soci- ety is based in the first instance on the generation of objects, which can be taken as given in further communication. 13 It would be much too risky to rely primarily on contracts or on consensuses that can be called for as a normative requirement. Objects arise out of the recursive functioning of communication without prohibiting the opposing side. And they only leave residual problems for deciding the issue of whether one wants to agree or disagree. Modern soci- ety owes it to the mass media that such objects 'exist', and it would be hard to imagine how a society of communicative operations that extends far beyond individual horizons of experience could func- tion if this indispensable condition were not secured through the communication process itself.
This merely serves to re-confirm the fact that communication has a problem of time to solve in the first instance, and this also applies to the mass media in particular which operate under pres- sures of acceleration. The problem is how one gets from one com- munication to the next, especially if the social system has become highly complex and non-transparent to itself and takes on an enor- mous variety every day which it has to transfer as irritation over to communication. It is impossible to make this dependent upon a previously secured consensus that is to be made sure of operation- ally. On the contrary: every explicit communication poses the ques- tion of acceptance and rejection anew, puts consensus at stake, knowing full well that it is still possible to communicate further even and especially where dissent exists. Under modern conditions, this risking of dissent, this testing of communication by communi- cation, is more or less freed of any inhibitions. This is precisely
? vVhv communication has to be run alongside objects constituted by itself which can be treated as topics. It is therefore incumbent upon the mass media in the first instance to generate familiarity and vary it from moment to moment so that in the following communica- tion one can risk provoking either acceptance or rejection.
This analysis can be summarized in a theory of the memory of society. A system which is able to observe the system/environment difference generated by its operations needs a temporal double ori- entation for its observing operations (or, with Spencer Brown, for bringing about the re-entry of this difference into the system). This double orientation, comprising a memory on the one hand and an open future on the other, maintains the possibility of oscillating between the two sides of any distinction. 14 The problem which is posed for the social system and is essentially solved through the mass media is as follows: how memory function and oscillator func- tion can be combined if only the present, that is, practically no time at all, is available to do so. 15 And that is just another form of the old question as to how a complex system can secure sufficient re- dundancy and sufficient variety at the same time.
If one wants to describe the function of memory with regard to the future right from the start, one must let go of the psychologi- cally plausible idea that memory has the task, only needed occa- sionally, of recalling past events. Rather, memory is performing a constantly co-occurring discrimination of forgetting and remem- bering that accompanies all observations even as they occur. The main part of this activity is the forgetting, whereas only exception- ally is something remembered. For without forgetting, without the freeing up of capacities for new operations, the system would have no future, let alone opportunities for oscillating from one side to the other of the distinctions used in each instance. To put it an- other way: memory functions as a deletion of traces, as repression and as occasional inhibiting of repression. It recalls something, however short- or long-term, when the current operations offer an occasion to repeat, to 'reimpregnate' freed capacities. 16 It does not follow from this that memory operates with reference to the envi- ronment, serving the ongoing adaptation of the system to changing circumstances in its environment. It may indeed look that way to an external observer (with a memory of his or her own). However,
? in the system itself all that is going on is a constantly re-activated internal test of consistency, in which the memory performs recur- sions and organizes the system's resistance to surprising new de- mands placed on meaning. And as we have already said, it is through resistance of the system's operations to the system's operations that the system generates reality.
The feats of memory of communicative systems in general and of the mass media in particular are furnished by topics of communi- cation. For only that which can organize a sequence of contribu- tions and is open for future yes or no options will coagulate around a topic. Topics are extracts of communicative relevances, 'local' modules, as it were, which can be swapped and changed as re- quired. As a result they make possible a highly differentiated memory that can tolerate and indeed facilitate a rapid change of topic with the proviso of return to topics put aside at that moment.
All function systems have a memory specific to them. Thus, for example, the money economy has a memory that is designed to forget the origin of amounts of money paid in each instance, so that turnovers may occur more easily. 17 The memory of the mass media likewise functions internally to the system, but additionally produces functions appropriate for the entire social system. Obvi- ously this social use of the mass media constantly to link past and future is connected to the extremely high expectations of redun- dancy and variety which modern society poses and which it must attribute temporally and take account of via the distinction of past and future. For without this temporal, dimensional stretching, on- going reconstructed reality would collapse due to internal contra- dictions. And it is not least this which explains that this feat requires strong selectors which in turn must be protected by differentiation and operational closure.
? 14
The Public
It may be gathered from the preceding observations what kind of questions need to be asked about the 'function' of the mass media. They make a contribution towards society's construction of real- ity. Part of this includes a constant reactualization of the self-de- scription of society and its cognitive world horizons, be this in a form marked by consensus or dissent (for example, when the real causes of the 'dying of the forests' are at issue). The mass media may not have an exclusive claim on constructing reality. After all, every communication contributes to constructing reality in what it takes up and what it leaves to forgetting. However, the involvement of the mass media is indispensable when the point at issue is widespread dissemination and the possibility of anonymous and thus unpredictable uptake. As paradoxical as it may sound, this means not least, when it is a matter of generating non- transparency in reactions to this uptake. The effect if not the func- tion of the mass media seems to lie, therefore, in the reproduction of non-transparency through transparency, in the reproduction of non-transparency of effects through transparency of knowledge. This means, in other words, in the reproduction of future.
This at first paradoxical thesis, only resolvable through the distinc- tion of past and present that is present in each instance, can be treated further if one distinguishes between the system of the mass media and the public. In order to do this, we must first introduce a concept of the 'public' which differs clearly enough from the system of the mass media as well as from the concept of 'public opinion'.
? It seems that there has always been an element of unpredictability built into the concept of the 'public'. In classical juridical discourse, 'public' is defined by accessibility for everyone, that is, by the inad- missibility of control over access. In this sense, the printed prod- ucts and programmes of the mass media are public because there is no control over who pays attention to them. But from the point of view of this conceptual scheme, this is only part of the public. Pub- lic toilets are neither opinions nor a product of the mass media. The concept of accessibility refers in a real or metaphorical under- standing to space and to action. This limitation can be corrected if one switches from action to observation. Then, following Dirk Baecker's suggestion, one can define the public as a reflection of every system boundary internal to society,1 or again, as the envir- onment, internal to the system, of social subsystems, that is, of all interactions and organizations, but also of social function systems and social movements. The advantage of this definition is that it can be transferred onto social function systems. The 'market' would then be the environment, internal to the economic system, of eco- nomic organizations and interactions;2 'public opinion' would be the environment, internal to the political system, of political or- ganizations and interactions. 3
It still holds that system boundaries cannot be crossed over op- erationally. But it is also the case that every observing system can reflect this. It sees on the inside of its boundary that there must be an outside, otherwise the boundary would not be a boundary. If specific experiences of irritation repeatedly crop up internally, the system can assume that there are other systems in the environment which are responsible. If, on the other hand, the system reflects that it is being observed from outside, without it being established how and by whom, it conceives itself as observable in the medium of the public. This can, but need not, lead to an orientation to- wards generalizable (publicly defensible) points of view. Function- ally equivalent strategies are those of secrecy and hypocrisy.
Thematic groups around secrecy, simulation, dissimulation, hy- pocrisy come to be worked out especially in the (printed! ) litera- ture of the sixteenth and seventeenth centuries, and this occurs by no means merely as political theory going by the name of state reason, but is also exemplified in theatre, with reference to the
? market and for social behaviour per se. 4 In the eighteenth century, the emphatic demand of public life as a means for establishing rea- son came to be directed against this stress upon the necessity of social intercourse. But this is a rather narrow, as it were constitutionalized, concept of public life with demands such as free- dom of opinion, freedom of the press, abolition of censorship. The polemic itself is based on a much more general concept of the pub- lic, which forms the background to precisely such strategies as se- crecy and hypocrisy and later the effort to protect a 'private sphere'. Public life is therefore a general social medium of reflection which registers the non-exceedability of boundaries and, thus inspired, the observing of observations.
Even before an emphatic concept of public opinion arose towards the end of the eighteenth century, the printing press had been used to achieve public resonance for politically ambitious communica- tions and thus to expose decision-making authorities to the dual grasp of writing directed towards them and of its public resonance. In England, petitions directed at the crown and at parliament were printed as early as the seventeenth century, although they retained the form of a letter with address and deferential politeness. In France, the courts began in the eighteenth century to have their remonstra- tions directed at the king printed in order to play the public off against the sole acknowledged 'public person', the monarch. 5 Pub- lic accessibility of communications in the political apparatus of domination is thus expanded with the aid of the printing press, and only afterwards does the idea emerge of public opinion as the ulti- mate authority for the judging of political affairs. Although, in- deed because, the public cannot decide politically, but rather lies to an extent outside the boundaries of the system of politics, it is used politically in politics and copied into the system.
The function of the mass media would therefore be not the pro- duction but the representation of the public. And what is meant here is 'representation' in a 'contrasting', reductive sense. Precisely because the 'public' always describes the other, inaccessible side of the boundaries of all systems, including the mass media, and can- not be specified in the direction of particular partner systems, it is necessary to represent them in the form of constructions of reality in which all subsystems, indeed, all people, can have a part, with-
? out any obligation arising to go about it in a particular way. Thus the representation of the public by the mass media simultaneously guarantees transparency and non-transparency as events continu- ously happen, that is, particular thematic knowledge in the form of objects that are made concrete in each instance, and uncertainty in the issue of who is reacting to them and in what way.
As we have already noted repeatedly, this is an 'autological' con- cept. It applies also to the mass media themselves. By reproducing themselves as a system, they too generate boundaries with an in- side and an outside that is inaccessible to them. They too reflect their outside as public life, so long as specific external relation- ships, such as to politics or to the advertisers, are not in question. This reflection has for them, however, a different status, because their function becomes recognizable here. So there is largely no recourse to the functionally equivalent strategies of secrecy and hy- pocrisy, even if ultimately it is said hypocritically that there is no hypocrisy. A metaphorical redescription6 can attach to this - for example, in the form of professional ethics, which allows journal- ists to understand their efforts as a service to the public, and this as a justification for claims to autonomy and as a reason for neutral- ity of interests, and to institutionalize critical standards and profes- sional consensus for it. There is good reason for the restriction to journalism/profession/ethics if it is the self-regulation of the system of the mass media that is at stake. It also offers a starting point for an independence, however Utopian, from the desires of the audi- ence or of particular interest groups. But these achievements have to be bought with a severely restricted concept of autonomy. Here and here alone, therefore, is there reason to speak, in what is in itself a paradoxical sense, of 'relative autonomy'.
? 15
Schema Formation
The discussion thus far has opted decisively and exclusively for the system references 'society' and 'mass media' and has banished eve- rything else in their 'environment'. This involved disregarding indi- viduals as living bodies and as systems of consciousness. True, we were able to speak of individuals, and in fact no system of the mass media can get by without naming names or conveying images of people. But those are obviously only topics of communication or objects that have been depicted, and in every case it is due to deci- sions in the system of the mass media, that is, to communications, whether or not they are named or shown. It is not the individuals themselves. It is only persons, only 'Eigenvalues', which every com- munication system has to generate in order to be able to repro- duce itself. 1
Obviously, the theory of operational closure of autopoietic sys- tems does not say that these systems could exist without any envi- ronment. The suspicion of 'solipsistic' existences was always an absurd one and says more about whoever formulates it as an objec- tion than about the theory being attacked itself. Certainly, cogni- tive systems are unable to reach their environment operationally, and so they cannot know it independently from their own struc- tural formations. Nonetheless, there are structural couplings be- tween autopoietic systems and systems in their environment which are compatible with autopoiesis. They do not bring about any de- termination of systems' conditions through conditions or events in the environment. Systems can only determine themselves, and this
? they can only do through self-generated structures. But massive and repeated irritations can still arise, each of which is then proc- essed into information within the system. Viewed over the longer term, structural development is thus explained by the constant sup- ply of irritations from certain sources - and by the lack of stimuli on the part of other segments of the environment. Maturana called this evolutionary tendency 'structural drift'.
Of course, this coupling in the relationship of individuals and social systems presupposes that individuals are able to perceive, that is, externalize an environment worked out internally. More- over, it depends upon the perceiving of others' perceiving - other- wise no individual could generate anything that was meant to be perceived by others. Equally without doubt, language must be avail- able, for perception as well as for communication. But these pre- suppositions do not offer us any hypotheses about the direction taken by structural drift when knowledge of the world is generated almost exclusively by the mass media. We still lack a concept, for example, which might explain (or which might lead to hypotheses which might explain) how knowledge of the world arising from life in the family households of traditional society is suppressed or cov- ered over by participation in the output of the mass media. For this issue, a repertoire that has provoked broad discussion with terms such as schema, cognitive map, prototype, script, frame might be useful. 2
These are psychological terms, but ones that are increasingly being used to explain social coordinations or so-called 'collective' behav- iour. 3 Their starting point is memory's need to discriminate con- stantly in the torrent of operations which occupy a system between forgetting and remembering, because without forgetting, the ca- pacities of the system for further operations would very quickly be blocked and, to put it another way, one would only ever be able in future to experience or do the same thing. Forgetting sets you free. But since for its part forgetting cannot be remembered, one needs a schema that regulates what is retained and can be reused. These may be schemata of perception which enable the gaze to be focused and the unfamiliar to be recognized by setting it against what is familiar. But they may also be more abstract categoriza- tions, or both at the same time if, for example, people's qualities or
? behaviour are inferred from racial characteristics. Schemata do not force repetitions to be made, neither do they specify action. In fact, their function is precisely to generate space for freely chosen be- haviour in a system which with its own past has put itself in the state (and in no other) in which it currently finds itself. This is what abstraction (not necessarily conceptual) is for, the disregarding of . . . , the repression of the countless details which mark situations as unique and unrepeatable. But abstraction also means that new situ- ations can modify the schema. The schema allows for supplements and replenishments; it cannot be applied 'schematically'. 4 Devia- tions come as a surprise because of the schema; they become con- spicuous and thus imprint themselves on the memory. Schemata are instruments of forgetting - and of learning; they are limitations to flexibility which make flexibility within prestructured barriers possible in the first place.
As Kant taught us,5 schemata are not images but rather rules for accomplishing operations. The circle schema, for example, is not the depiction of any circle, but the rule for drawing a circle. The diversity given to the inner meaning in the form of time differences can only be reconstructed as procedures (also presupposing time) for purposes of knowledge. For Kant, this copying of time from the empirical over into the transcendental sphere was the reason why a relationship of similarity could be assumed in spite of the radical difference of objects and ideas. This problem does not arise if one takes on board a radically constructivist epistemology.
But it re- mains the case that schemata are not images which become con- cretely fixed at the moment of depiction; they are merely rules for the repetition of operations (which then are concrete again). Thus, memory does not consist of a supply of images which one can look at again whenever necessary. Rather, it is a question of forms which, in the ceaseless temporal flow of autopoiesis, enable recursions, retrospective reference to the familiar, and repetition of operations which actualize it.
Schemata can refer to things or to persons. The utility meaning of things is one schema, the hierarchies among people or standard- ized role expectations are another. 'Script' refers to the special case where temporal successions are stereotyped (for example, the fact that we are supposed to buy a ticket before getting on a train). The
? observation of causal relationships typically follows a script be- cause it cuts out other, equally realistic possibilities for causal attri- bution. 6 It is only by way of a script that one comes to attribute effects to actions. A script is therefore an already fairly complex schema which also cuts out many things and presupposes both a stereotyping of events and a standardized coupling of their succes- sion. If thing or person schemata are linked to a script, it also means that the observer is no longer free to choose between object schema and time schema or to let his or her gaze oscillate, but that object schema and time schema enter a relationship of mutual depend- ence where the one cannot be chosen without consideration of the other. We have already considered such a case using the example of the narrative structure of novels: the sequence of actions charac- terizes the people whose motives then make the sequence of actions understandable again - with sufficient scope for surprises.
Now, we assume that the structural coupling of mass media com- munication and psychically reliable simplifications uses, and indeed generates, such schemata. The process is a circular one. The mass media value comprehensibility. But comprehensibility is best guar- anteed by the schemata which the media themselves have already generated. They use a psychic anchoring7 for their own workings which can be assumed to be the result of consumption of mass media representations, and indeed can be assumed to be such with- out any further tests. Let us elucidate this using two examples: the production of causal scripts in domains that are inaccessible to in- dividual experience and thus very typical in the case of ecological problems; and the presupposition of different person schematizations depending upon whether it is oneself or other peo- ple who are involved.
Communication about ecological problems is a particularly good example for our purposes,8 because it goes far beyond the individu- al's world of experience. (Who could say from their own knowl- edge what would have happened to the contents of the Brent Spar platform, given the pressure operating on the sea bed, if it had been sunk? ) The mass media too are unequal to the task, and when they turn to science, they will typically be given more knowledge and more ignorance at the same time. So, we are dependent upon schema formation. It might be normative sentences which are set against a
? 'virtual reality' and are very typically fashioned metaphorically. For example, the ocean should not be used as a rubbish dump. This is self-evident, so to speak. If one asks further, more scripts are brought to bear. Out of innumerable possible causal constellations, one is picked out which can be made plausible. Usually the points at issue are the effects of actions, not nature's own course. Effects can then be coupled onto this which are sufficiently worrying to prevent people from asking any further as to how likely they actually are. To put it another way, what is involved are schemata of change which correspond to the selection criteria of news and in-depth reporting (for example: new, action, drama, morality). Environ- mental pollution changes the living conditions of people on earth to the point of conditions which make the continuance of life im- possible. There is no coming up against difficulties with individu- als' memories or their world of experience here. They have not yet experienced such things or can at best, if the script is offered, acti- vate experiences of their own that fit (the layer of filth on the car parked outside). So it is not a case of the 're-education' of individu- als, of them unlearning, in a more or less difficult process, some- thing that had been thought of as knowledge. The ecological imagery, its schemata, its scripts are developed on a greenfield site, so to speak, they form a terrain that is not yet occupied.
People speak of a 'transformation of values'. The question, how- ever, is whether the reorientation with newly recommended values is beginning, or whether it is the causal scripts which impress us first; whether, that is, it is the change which we find fascinating and which then leads to values being associated with it. Salancik and Porac speak of 'distilled ideologies' and mean by this, 'values de- rived from causal reasonings in complex environments'. 9
Anyone who adheres to ideas such as 'objective truth' or psychi- cally binding 'consensus' will not be able to accept this analysis and will accuse the mass media of superficiality, or even manipula- tion. If, on the other hand, one takes the individuality and the op- erational closure of autopoietic systems seriously, one will see that it cannot be otherwise. From the point of view of society, struc- tural coupling mediated via schemata has the benefit of accelerat- ing structural changes in such a way that, if this acceleration is successful, it will not break the structural coupling of media and
? individuals but will simply link up to other schemata. From the point of view of the individual, the advantage of schemata is that they structure memory but do not determine action. At the same time, they offer liberation from burdens that are too concrete as well as a background against which deviations, opportunities for action and constraints can be recognized. Individuals are still at liberty in this instance to get involved or to leave it be. They can allow feelings to arise and identify with them, or they can observe this in others and think of it as strange or even as dangerous. And with that, we have arrived at our second theme, a complementary hypothesis about relationships between mass media and individu- als.
In psychology, it has long been common to distinguish the schematization of one's own person (that is, answers to the ques- tion: who am I? ) from the schematization of other persons. 10 The distinction is interesting in various respects - qua distinction. First of all, every human being is given as a concrete individual, that is, different from others in terms of appearance, name and other char- acteristics. Why then is it not sufficient and since when has it no longer been sufficient to distinguish oneself from others just as every individual does from every other? Why is it not sufficient to use the same list of objective characteristics (age, sex, family, good- looking or not so good-looking, place of residence, virtues, vices etc. ) and to concretize the person being referred to only by a com- bination of these? There would be unlimited possibilities which could be supplemented as required. Furthermore, why, when one is dealing with concrete individuals, is schema formation necessary at all? As in the case of ecology, we have to assume non-transpar- ency, which is what offers the occasion in the first place for simplifications or, as we also call it nowadays, 'identity'. But why is someone non-transparent to himself, that is, in need of a schema, even though, according to Descartes, he cannot doubt his thinking existence?
We can be certain of the fact that the difference of one's own I from other individuals is given from the start, meaning as early as a few days after birth. The newborn child has to practise comple- mentary behaviour, not an imitative one, such as reversing right/ left perceptions. 11 Infant socialization after this presupposes what
? Stein Braten calls 'dialogic closure', that is, systems that can be fenced off to the outside, in which there is provision for a place for a 'virtual other', that is, for effective occupation. 12 This position of the virtual other can only be occupied with the aid of schemata, since it requires recognition, that is, memory. On the other hand, one does not need a 'virtual ego'. One is who one is from the start. But how then does a secondary need for self-schematizations arise? And what happens when the requirement of a direct 'dialogical closure' is overstepped and the occupation of the position of 'vir- tual other' no longer occurs effectively (in the sense of virtus), but is 'enriched' by fictional components?
We can assume that effects of the mass media become visible at this point. Early modern theatre in particular will have introduced this new development first. It offered the possibility of making ac- tors' inner processes of opinion formation, conflicts and uncertain- ties visible on stage through language. It might be that the actors would address themselves directly to the audience in forms which implied that the other players on the stage could not hear it (but how does one learn this unusual, counterfactual implication? ); or else it might occur in the form of monologues or soliloquies. The audience could then observe how the actors on the stage motivate themselves and deceive themselves and others, and that this pro- cess initially remains invisible to other participants in the play. 13 In the finely honed dialogues of Vienna theatre (for example, Schnitzler's Liebelei ('Light-O'-Love') or Hofmannsthal's Der
Unbestechliche ('The Incorruptible One')) the sentences themselves are constructed in such a way that the audience is able to observe more than those being addressed. The emergence of this compli- cated, as it were highly charged, cultural form of observation of observers and the development of suitable schematizations is there- fore not a direct product of the printing press or of the mass media. But once this specific form of second-order observation with its schemata of motives (love, criminality, sincerity/insincerity etc. ) is practised and can be presupposed as a way of observing, it can then be used in other contexts as well, such as in the novel and ultimately even in philosophy. And then the viewer or the reader is tempted as well to take a second look at his or her own way of observing and its motives.
? Shaftesbury seems to have been one of the first to retreat to a private conversation with himself in order to gain clarity about himself, in spite of having clear misgivings about the printing press and its commercial publishers, of which, of course, he himself makes use. 14 Rousseau likewise has his confessions printed, even though he explicitly exempts himself from the criteria of judgement which apply also to others. 15 The Romantic era plays with doppelgangers, twins, reflections, in order to represent the transformation of iden- tity into communication. Towards the end of the nineteenth cen- tury, William James, Georg Simmel and many others would speak of the need for a 'social self' or an 'identity' which is to be, or is to pretend to be, a fragmentary, turbulent, chaotic individual, in or- der to be something for others which it itself is not by itself. 16 And now the 'search for meaning' begins - at least in printed texts. We arrive at a time in which literature and life in literature can no longer be separated. The problem of 'self-realization' is invented and is taken up and disseminated by the mass media. Individuals are encouraged to believe that, although they have without doubt been really alive since conception, and certainly since birth, they must become even more real (or unreal? ) than they already are.
This semantic ambiguity can be understood if we read it as an indication of a need for a schema that cannot, however, be admit- ted. We can recognize schemata, in others and in ourselves, if we take them to be cognitive routines, abbreviations for something that might be elucidated. But this itself would also be a schema which conceals the issues that are ultimately involved. In view of the unobservability of the world and the non-transparency of indi- viduals to themselves and to others, schema formation is unavoid- able. Without it there would be no memory, no information, no deviation, no freedom. One can also, with Spencer Brown, under- stand this as the necessity of a form which marks a distinction, one side of which must be marked if one wants to observe and to for- mulate more operations. This does not stop us from asking about the social conditions of the plausibility of such schemata. In the age of the mass media, they are virtually unthinkable without the par- ticipation of the media. Like theatre, the mass media also put the individual into a scene that is outside the scene set on the stage. We have described this as a technical condition for the differentiation
? of a media system. This distance has to seem ambivalent to the individuals: on the one hand they are not themselves the text being performed for them; and if, like Rousseau, they have written and published it, they are it no longer. Neither do they see themselves on television, and if in an exceptional case they do, it is with special pleasure in the self-recognition only found in exceptions. On the other hand, the mass media produce the world in which individu- als find themselves. This is true of all programme sectors: of news, advertising, entertainment. What is presented to them affects them too, since they have to lead their lives in this world; and it affects them even when they know very well that they will never get into the situations or play the roles presented to them as factual or fictional. Instead, they can still identify with the cult objects or the motives which the scripts of the mass media offer them. When indi- viduals look at media as text or as image, they are outside; when they experience their results within themselves, they are inside. They have to oscillate between outside and inside, as if in a paradoxical situation: quickly, almost without losing any time, and undecidably. For the one position is only possible thanks to the other - and vice versa.
The consequence must be that the individual must resolve this paradox for herself and construct her identity or her 'self' herself. The materials used for this can be the usual ones. But there is no possibility of taking on an 'I' by analogy from outside. No one can be like someone else. No one sees himself as the reflection of an- other. The only point of agreement is the necessity of using sche- mata for sustaining a memory. But self-schematization cannot relieve the strain on itself through the illusion of an 'objective' (even if disputed) reality. On the one hand it (self-schematization) is indis- putable, for no one can perform it for another, and on the other it is under threat of constant dissolution. This is because no one can know whether he will remain who he had thought he was. He can- not know because he himself decides the issue.
The structural couplings between individuals and society affect the whole of reality. This is true of all social formations. However, the mass media vary the structural conditions of these structural couplings because they change the need for schemata as well as what they offer. The schemata and scripts of ecological concerns
? and the necessity of schematization of one's own person are only extreme examples chosen to illustrate this. And perhaps it is no coincidence that these two environments of social communication, the complexity of non-human nature and the auto-dynamic and non-transparency of human individuals, are dependent in a par- ticular way upon schemata and therefore upon structural couplings to the system of the mass media.
? Second-order Cybernetics as Paradox
The second-order cybernetics worked out by Heinz von Foerster is rightly held to be a constructivist theory,1 if not a manifesto for operational constructivism. The reverse does not apply, however. Constructivist epistemologies do not necessarily have the rigour of a cybernetics of cybernetics. One can observe cognitions as con- structions of an observer, without linking with this the theory that the observing observer observes himself or herself as an observer. This difference is so crucial that we must devote a final chapter to it.
The discussion thus far has been guided by two points of depar- ture. The first is that the mass media, like any broadcasting system, are an operationally closed and, in this respect, autopoietic system. The second emphasizes that this is also true of cognitions, because cognitions are also operations and can therefore only be produced in the system. This remains the case even when one considers that in society communication can take place with the system of the mass media from out of the latter's environment, for these commu- nications too are possible only on the basis of the knowledge that the mass media have provided. Furthermore, the mass media un- derstand what is uttered to them only on the basis of their own network of reproduction of information. Every communication in and with the mass media remains tied to the schemata which are available for this purpose.
This theoretical description is designed in the mode of second- order observation. It observes and describes observers. But it does
? not presuppose that the mass media observe themselves in the mode of second-order observation. The media designate what they are communicating about and must therefore distinguish it. For exam- ple, they inform people about scandals and in doing so must pre- suppose that non-scandalous behaviour would have been possible as well. What is not reflected here, however, is that one could pose the question (which a sociologist might pose) why something is even being observed in the schema scandalous/non-scandalous at all, and why the frequency of use of this schema is clearly increas- ing. In other words, the media remain (for good reason, as we shall presently see) invisible to themselves as an observer. They are turned towards the world in their operations and do not reflect that this turning itself generates an unmarked space in which they find them- selves.
We can reformulate this statement by splitting our concept of autonomy. First, there is autopoietic autonomy which is based on operational closure and means that the system can only reproduce its own structures and operations with its own operations, that is, from its own products. This is to be distinguished from cognitive closure, and, correspondingly, cognitive autonomy. This says that along with all its cognitions the system is also observing that these are only its own observations. Only having reached this point do we find ourselves on the terrain in which second-order cybernetics in the strict sense is interested. 2 Here, the question 'who is the ob- _ server? ' is asked universally and is also applied to the observing system. Questions about the observer take the place of questions about reasons, which would necessarily result in an infinite regress. And therefore, whoever wishes to give reasons for his own experi- ence or actions must observe himself as an observer and, in doing so, allow access to the choice of the distinctions which guide his observing. But how is that possible?
Obviously, from an empirical point of view, the system of the mass media does not operate at the cognitively closed level of sec- ond-order cybernetics. It does distinguish self-reference and other- reference. In its attitude of other-reference it reports on facts and opinions. This includes the possibility of observing observers. The second-order observation common in modern society comes about in this respect. But this merely leads into the infinite regress of the
? question as to which observer is observing this. In the system itself, there is no final figure of the ambiguous 'observing system',3 no autological realization that whatever is true for observers is also true for the system which is observing them. Thanks to the distinc- tion of self-reference and other-reference, the system of the mass media can also mark itself in contrast to everything else. It can make its own structures and operations into a topic as though they were objects. But it does not additionally ask: how am I operating as an observer and why do I make distinctions in this way and not another? With every distinction it uses it places itself in the unob- served, unmarked space, and this is even so when it marks itself in contrast to other things. Every distinction makes the observer in- visible - but this is precisely what we can still know. If she wanted to de-invisibilize herself, she would have to mark herself, that is, distinguish herself. And then one would again have the question, who is the observer who distinguishes thus and not otherwise?
This is also true of modern society, and also in conditions which some people describe as 'postmodern'. It even applies if one re- nounces absolute demands for validity which in the tradition went under names such as God or nature or reason. This renunciation is presented as relativism or historicism. One accepts the contingency of all criteria and of all possible observer positions. But that only means that one is able to switch from any distinction to another, that, for example, one can take into account fashions or transfor- mations of values. In fact, these are now accepted schemata. The problem of transformation and of contingency has been digested and can be expressed with the normal schematisms of the mass media. The system may then be operating at a level of greater un- certainty, but that is also true of the other function systems, of the money economy, art, science, politics. In accepting this character- istic postmodern style the mass media are merely following what the form of social differentiation suggests. But with a constant change of perspectives, the observer who is performing this trans- formation with the before/after distinction still cannot be grasped. 'God is dead', they said - and meant: the last observer cannot be identified.
As a reaction to this finding, attempts have been evident for some years now to shift the problem onto ethics. This is true throughout
? society and thus also in the mass media. For example, a code of ethics for journalists can be drawn up and the attempt made to apply it via the profession's self-regulatory procedures. The fact that this cannot be an ethics of reasoning in the academic style is easy to see if one follows the academic debate about transcendental ethics, utilitarian ethics or value ethics. In none of these cases have radical deductive steps towards decisions succeeded. We can know this. Therefore, these can only be conventions which continually find themselves confronted with new situations. Nor does this eth- ics, if it is not condensed into norms of law, contain any indication of how deviants are to be treated.
The position of a second-order cybernetics offers an opportunity to reflect this flight into ethics as a displacement of the problem. After all, whatever else it is understood to be in concrete terms, ethics too is a distinguishing practice. It distinguishes standards and ways of behaving, it distinguishes conforming and deviant be- haviour and usually even in a moral sense good and bad, or evil, behaviour. Moreover, it is a part of its presuppositions that devia- tions are attributed to behaviour and not to inappropriately cho- sen standards or, as critical sociologists thought for a while, to 'labelling'. 4 Even if strong ties and highly charged emotions are to be expected in heavily moralized domains, second-order cybernet- ics can still ask: why are you distinguishing in this way and not in another? Or again: who is the observer who is trying to impose these schemata here?
Standard authors of constructivist epistemology, such as Humberto Maturana and Heinz von Foerster, have attempted to develop a new ethics on this basis. However, they have not gone beyond making a few suggestions,5 and it is doubtful whether this venture can succeed. For an ethics would sabotage itself if what was demanded of it was that it make distinctions and simultane- ously reflect that it is itself making these distinctions.
Even in the face of numerous efforts to find ethical foundations, second-order cybernetics can only ever repeat the question: who is the observer?
This leads to the fundamental question of how communication must be, in order that it can not only reproduce itself but also take on cognitive functions and separate reproductive or informational components. The answer is that communication only comes about at all by being able to distinguish utterance and information in its self-observation (in understanding). Without this distinction, com- munication would collapse, and participants would have to rely on perceiving something which they would only be able to describe as behaviour. 5 The difference of utterance and information corresponds precisely with the requirement of not making the progress of com- munication to communication dependent upon information being
? complete and relevant. And only because this primary, constitutive difference exists can communication code itself in a binary form (for example, with regard to acceptable/not acceptable, relevant/ not relevant) and in this way feel its way around the environment with a distinction for which there is no correlate whatever in the environment itself. Without this distinction, which has been en- tered into its own operation, the system would not be capable of constituting any recognizable identities or developing any memory. Nor could it evolve, or build up its own complexity, or test the possibilities for structuration positively/negatively and thus meet the minimum condition for the continuation of its own autopoiesis. 6 Society as we know it would be impossible.
For the same reasons, no great expectations can be placed on the understanding of communication. Expectations can certainly be raised forcibly, but they then require special differentiated dis- courses. Normally, ambivalences and misunderstandings are borne along as well, as long as they do not block communication; indeed, understanding is practically always a misunderstanding without an understanding of the mis.
It is a big jump from these general systems-theoretical and social- theoretical considerations to the mass media of modern society. The function of the mass media lies after all that in the directing of self-observation of the social system7 - by which we do not mean one specific object amongst others, but a way of splitting the world into system (that is, society) and environment. What is involved is a universal, not an object-specific observation. We have already spoken, in another context,8 of the function of the system's memory which provides a background reality for all further communica- tions, which in turn is constantly reimpregnated by the mass me- dia. What is also involved is an observation which itself generates the conditions of its own possibility and in this sense occurs autopoietically. For the uncertainty as well as the distinctions used for observation are products of the system and are not simply pre- given attributes of the world or ontologically or transcendentally provable decomponates ('categories') of the unity of the world. This means also that the impetus for further communication is repro- duced within the system itself and cannot be explained anthropo- logically, as a drive for knowledge, for example.
? Therefore, one cannot comprehend the 'reality of the mass me- dia' if one sees its task in providing relevant information about the world and measuring its failure, its distortion of reality, its ma- nipulation of opinion against this - as if it could be otherwise. The mass media realize in society precisely that dual structure of repro- duction and information, of continuation of an always already adapted autopoiesis and cognitive willingness to be irritated. Their preference for information, which loses its surprise value through publication, that is, is constantly transformed into non-informa- tion, makes it clear that the function of the mass media consists in the constant generation and processing of irritation - and neither in increasing knowledge nor in socializing or educating people in conformity to norms. The descriptions of the world and of society to which modern society orients itself within and outside the sys- tem of its mass media arise as a factual effect of this circular perma- nent activity of generating and interpreting irritation through information tied to a particular moment (that is, as a difference which makes a difference).
Of course, it should not be implied that irritation happens only in the system of the mass media and not, for example, in marriages, in school lessons or in other interactions; just as power is present not only in the political system, standardizations not only in the law, truth not only in science. Irritability is the most general struc- tural characteristic of autopoietic systems, which in modern de- scription occupies that place once accorded to nature and to the essence of things defined as nature. 9 Irritability arises from the sys- tem having a memory that is actively involved in all operations and therefore being able to experience and balance out inconsistencies - which means nothing other than being able to generate reality. This points to a recursive constitutive context of memory, irritabil- ity, information processing, reality construction and memory. The differentiation of a function system specialized in this serves to improve and simultaneously normalize a means of communication likewise specialized in this. Only from the mass media do we ex- pect this special performance every day, and only thus is it possible to arrange modern society in its execution of communication in an endogenously restless way like a brain and thereby prevent it hav- ing too strong a link to established structures.
? In contrast to the function system of the mass media, science can be specialized in cognitive gains, that is, in social learning proc- esses, whilst the system of law takes on the ordering of expectation which is normative, held onto in spite of the facts and to this extent unwilling to learn. However, the cognitive/normative division be- tween science and law can never divide up among itself and thereby cover the entire orientational requirement of social communica- tion. Under normal circumstances social communication is oriented towards neither science nor the law. But neither can it be left in modern global society to the merely local everyday knowledge that is only found in the nearest vicinity. Accordingly, it seems to be the function of the mass media to remedy this neither cognitively nor normatively specified requirement. The mass media guarantee all function systems a present which is accepted throughout society and is familiar to individuals, and which they can take as given when it is a matter of selecting a system-specific past and establish- ing decisions about future expectations important to the system. Other systems, depending on their own requirements, can then adapt themselves to the past reference of their anticipation; for example, the economy can adapt itself to new circumstances in companies or in the market, and on this basis establish their own connections between their past and their future.
It was Parsons who saw that the particular contribution of the mass media to the 'interchanges' of modern society lies in the in- crease in levels of freedom of communication - analogous to the function of money in the economy. 10 This diagnosis can be broad- ened if one additionally takes into consideration the increase in society's capacity for irritation and the recursive interweaving of mass media communication with everyday communication in the interactions and organizations of society. On the one hand, the mass media draw communication in, on the other, they stimulate ongoing communication. 11 So they continuously apply new com- munication to the results of previous communication. In this sense they are responsible for the production of modern society's 'Eigenvalues' - those relatively stable orientations in the cognitive, the normative and the evaluative domain which cannot be given ab extra but rather arise out of operations being applied recursively to their own results. 12
? It seems that a centuries-old tradition has led us astray, with the result that mass media appear in an unfavourable light. The tradi- tion says that the stability of the social system rests upon consensus - or even on an explicitly/implicitly agreed social contract, and if no longer upon a commonly held religion, then at least on consen- sually accepted background convictions, encapsulated in Jiirgen Habermas's concept of lifeworld. Were this not the case, the mass media would be a destabilizing factor, only out to destroy these presuppositions and to replace them with something the French might call symbolic violence.
In fact, however, the stability (= reproductive capacity) of soci- ety is based in the first instance on the generation of objects, which can be taken as given in further communication. 13 It would be much too risky to rely primarily on contracts or on consensuses that can be called for as a normative requirement. Objects arise out of the recursive functioning of communication without prohibiting the opposing side. And they only leave residual problems for deciding the issue of whether one wants to agree or disagree. Modern soci- ety owes it to the mass media that such objects 'exist', and it would be hard to imagine how a society of communicative operations that extends far beyond individual horizons of experience could func- tion if this indispensable condition were not secured through the communication process itself.
This merely serves to re-confirm the fact that communication has a problem of time to solve in the first instance, and this also applies to the mass media in particular which operate under pres- sures of acceleration. The problem is how one gets from one com- munication to the next, especially if the social system has become highly complex and non-transparent to itself and takes on an enor- mous variety every day which it has to transfer as irritation over to communication. It is impossible to make this dependent upon a previously secured consensus that is to be made sure of operation- ally. On the contrary: every explicit communication poses the ques- tion of acceptance and rejection anew, puts consensus at stake, knowing full well that it is still possible to communicate further even and especially where dissent exists. Under modern conditions, this risking of dissent, this testing of communication by communi- cation, is more or less freed of any inhibitions. This is precisely
? vVhv communication has to be run alongside objects constituted by itself which can be treated as topics. It is therefore incumbent upon the mass media in the first instance to generate familiarity and vary it from moment to moment so that in the following communica- tion one can risk provoking either acceptance or rejection.
This analysis can be summarized in a theory of the memory of society. A system which is able to observe the system/environment difference generated by its operations needs a temporal double ori- entation for its observing operations (or, with Spencer Brown, for bringing about the re-entry of this difference into the system). This double orientation, comprising a memory on the one hand and an open future on the other, maintains the possibility of oscillating between the two sides of any distinction. 14 The problem which is posed for the social system and is essentially solved through the mass media is as follows: how memory function and oscillator func- tion can be combined if only the present, that is, practically no time at all, is available to do so. 15 And that is just another form of the old question as to how a complex system can secure sufficient re- dundancy and sufficient variety at the same time.
If one wants to describe the function of memory with regard to the future right from the start, one must let go of the psychologi- cally plausible idea that memory has the task, only needed occa- sionally, of recalling past events. Rather, memory is performing a constantly co-occurring discrimination of forgetting and remem- bering that accompanies all observations even as they occur. The main part of this activity is the forgetting, whereas only exception- ally is something remembered. For without forgetting, without the freeing up of capacities for new operations, the system would have no future, let alone opportunities for oscillating from one side to the other of the distinctions used in each instance. To put it an- other way: memory functions as a deletion of traces, as repression and as occasional inhibiting of repression. It recalls something, however short- or long-term, when the current operations offer an occasion to repeat, to 'reimpregnate' freed capacities. 16 It does not follow from this that memory operates with reference to the envi- ronment, serving the ongoing adaptation of the system to changing circumstances in its environment. It may indeed look that way to an external observer (with a memory of his or her own). However,
? in the system itself all that is going on is a constantly re-activated internal test of consistency, in which the memory performs recur- sions and organizes the system's resistance to surprising new de- mands placed on meaning. And as we have already said, it is through resistance of the system's operations to the system's operations that the system generates reality.
The feats of memory of communicative systems in general and of the mass media in particular are furnished by topics of communi- cation. For only that which can organize a sequence of contribu- tions and is open for future yes or no options will coagulate around a topic. Topics are extracts of communicative relevances, 'local' modules, as it were, which can be swapped and changed as re- quired. As a result they make possible a highly differentiated memory that can tolerate and indeed facilitate a rapid change of topic with the proviso of return to topics put aside at that moment.
All function systems have a memory specific to them. Thus, for example, the money economy has a memory that is designed to forget the origin of amounts of money paid in each instance, so that turnovers may occur more easily. 17 The memory of the mass media likewise functions internally to the system, but additionally produces functions appropriate for the entire social system. Obvi- ously this social use of the mass media constantly to link past and future is connected to the extremely high expectations of redun- dancy and variety which modern society poses and which it must attribute temporally and take account of via the distinction of past and future. For without this temporal, dimensional stretching, on- going reconstructed reality would collapse due to internal contra- dictions. And it is not least this which explains that this feat requires strong selectors which in turn must be protected by differentiation and operational closure.
? 14
The Public
It may be gathered from the preceding observations what kind of questions need to be asked about the 'function' of the mass media. They make a contribution towards society's construction of real- ity. Part of this includes a constant reactualization of the self-de- scription of society and its cognitive world horizons, be this in a form marked by consensus or dissent (for example, when the real causes of the 'dying of the forests' are at issue). The mass media may not have an exclusive claim on constructing reality. After all, every communication contributes to constructing reality in what it takes up and what it leaves to forgetting. However, the involvement of the mass media is indispensable when the point at issue is widespread dissemination and the possibility of anonymous and thus unpredictable uptake. As paradoxical as it may sound, this means not least, when it is a matter of generating non- transparency in reactions to this uptake. The effect if not the func- tion of the mass media seems to lie, therefore, in the reproduction of non-transparency through transparency, in the reproduction of non-transparency of effects through transparency of knowledge. This means, in other words, in the reproduction of future.
This at first paradoxical thesis, only resolvable through the distinc- tion of past and present that is present in each instance, can be treated further if one distinguishes between the system of the mass media and the public. In order to do this, we must first introduce a concept of the 'public' which differs clearly enough from the system of the mass media as well as from the concept of 'public opinion'.
? It seems that there has always been an element of unpredictability built into the concept of the 'public'. In classical juridical discourse, 'public' is defined by accessibility for everyone, that is, by the inad- missibility of control over access. In this sense, the printed prod- ucts and programmes of the mass media are public because there is no control over who pays attention to them. But from the point of view of this conceptual scheme, this is only part of the public. Pub- lic toilets are neither opinions nor a product of the mass media. The concept of accessibility refers in a real or metaphorical under- standing to space and to action. This limitation can be corrected if one switches from action to observation. Then, following Dirk Baecker's suggestion, one can define the public as a reflection of every system boundary internal to society,1 or again, as the envir- onment, internal to the system, of social subsystems, that is, of all interactions and organizations, but also of social function systems and social movements. The advantage of this definition is that it can be transferred onto social function systems. The 'market' would then be the environment, internal to the economic system, of eco- nomic organizations and interactions;2 'public opinion' would be the environment, internal to the political system, of political or- ganizations and interactions. 3
It still holds that system boundaries cannot be crossed over op- erationally. But it is also the case that every observing system can reflect this. It sees on the inside of its boundary that there must be an outside, otherwise the boundary would not be a boundary. If specific experiences of irritation repeatedly crop up internally, the system can assume that there are other systems in the environment which are responsible. If, on the other hand, the system reflects that it is being observed from outside, without it being established how and by whom, it conceives itself as observable in the medium of the public. This can, but need not, lead to an orientation to- wards generalizable (publicly defensible) points of view. Function- ally equivalent strategies are those of secrecy and hypocrisy.
Thematic groups around secrecy, simulation, dissimulation, hy- pocrisy come to be worked out especially in the (printed! ) litera- ture of the sixteenth and seventeenth centuries, and this occurs by no means merely as political theory going by the name of state reason, but is also exemplified in theatre, with reference to the
? market and for social behaviour per se. 4 In the eighteenth century, the emphatic demand of public life as a means for establishing rea- son came to be directed against this stress upon the necessity of social intercourse. But this is a rather narrow, as it were constitutionalized, concept of public life with demands such as free- dom of opinion, freedom of the press, abolition of censorship. The polemic itself is based on a much more general concept of the pub- lic, which forms the background to precisely such strategies as se- crecy and hypocrisy and later the effort to protect a 'private sphere'. Public life is therefore a general social medium of reflection which registers the non-exceedability of boundaries and, thus inspired, the observing of observations.
Even before an emphatic concept of public opinion arose towards the end of the eighteenth century, the printing press had been used to achieve public resonance for politically ambitious communica- tions and thus to expose decision-making authorities to the dual grasp of writing directed towards them and of its public resonance. In England, petitions directed at the crown and at parliament were printed as early as the seventeenth century, although they retained the form of a letter with address and deferential politeness. In France, the courts began in the eighteenth century to have their remonstra- tions directed at the king printed in order to play the public off against the sole acknowledged 'public person', the monarch. 5 Pub- lic accessibility of communications in the political apparatus of domination is thus expanded with the aid of the printing press, and only afterwards does the idea emerge of public opinion as the ulti- mate authority for the judging of political affairs. Although, in- deed because, the public cannot decide politically, but rather lies to an extent outside the boundaries of the system of politics, it is used politically in politics and copied into the system.
The function of the mass media would therefore be not the pro- duction but the representation of the public. And what is meant here is 'representation' in a 'contrasting', reductive sense. Precisely because the 'public' always describes the other, inaccessible side of the boundaries of all systems, including the mass media, and can- not be specified in the direction of particular partner systems, it is necessary to represent them in the form of constructions of reality in which all subsystems, indeed, all people, can have a part, with-
? out any obligation arising to go about it in a particular way. Thus the representation of the public by the mass media simultaneously guarantees transparency and non-transparency as events continu- ously happen, that is, particular thematic knowledge in the form of objects that are made concrete in each instance, and uncertainty in the issue of who is reacting to them and in what way.
As we have already noted repeatedly, this is an 'autological' con- cept. It applies also to the mass media themselves. By reproducing themselves as a system, they too generate boundaries with an in- side and an outside that is inaccessible to them. They too reflect their outside as public life, so long as specific external relation- ships, such as to politics or to the advertisers, are not in question. This reflection has for them, however, a different status, because their function becomes recognizable here. So there is largely no recourse to the functionally equivalent strategies of secrecy and hy- pocrisy, even if ultimately it is said hypocritically that there is no hypocrisy. A metaphorical redescription6 can attach to this - for example, in the form of professional ethics, which allows journal- ists to understand their efforts as a service to the public, and this as a justification for claims to autonomy and as a reason for neutral- ity of interests, and to institutionalize critical standards and profes- sional consensus for it. There is good reason for the restriction to journalism/profession/ethics if it is the self-regulation of the system of the mass media that is at stake. It also offers a starting point for an independence, however Utopian, from the desires of the audi- ence or of particular interest groups. But these achievements have to be bought with a severely restricted concept of autonomy. Here and here alone, therefore, is there reason to speak, in what is in itself a paradoxical sense, of 'relative autonomy'.
? 15
Schema Formation
The discussion thus far has opted decisively and exclusively for the system references 'society' and 'mass media' and has banished eve- rything else in their 'environment'. This involved disregarding indi- viduals as living bodies and as systems of consciousness. True, we were able to speak of individuals, and in fact no system of the mass media can get by without naming names or conveying images of people. But those are obviously only topics of communication or objects that have been depicted, and in every case it is due to deci- sions in the system of the mass media, that is, to communications, whether or not they are named or shown. It is not the individuals themselves. It is only persons, only 'Eigenvalues', which every com- munication system has to generate in order to be able to repro- duce itself. 1
Obviously, the theory of operational closure of autopoietic sys- tems does not say that these systems could exist without any envi- ronment. The suspicion of 'solipsistic' existences was always an absurd one and says more about whoever formulates it as an objec- tion than about the theory being attacked itself. Certainly, cogni- tive systems are unable to reach their environment operationally, and so they cannot know it independently from their own struc- tural formations. Nonetheless, there are structural couplings be- tween autopoietic systems and systems in their environment which are compatible with autopoiesis. They do not bring about any de- termination of systems' conditions through conditions or events in the environment. Systems can only determine themselves, and this
? they can only do through self-generated structures. But massive and repeated irritations can still arise, each of which is then proc- essed into information within the system. Viewed over the longer term, structural development is thus explained by the constant sup- ply of irritations from certain sources - and by the lack of stimuli on the part of other segments of the environment. Maturana called this evolutionary tendency 'structural drift'.
Of course, this coupling in the relationship of individuals and social systems presupposes that individuals are able to perceive, that is, externalize an environment worked out internally. More- over, it depends upon the perceiving of others' perceiving - other- wise no individual could generate anything that was meant to be perceived by others. Equally without doubt, language must be avail- able, for perception as well as for communication. But these pre- suppositions do not offer us any hypotheses about the direction taken by structural drift when knowledge of the world is generated almost exclusively by the mass media. We still lack a concept, for example, which might explain (or which might lead to hypotheses which might explain) how knowledge of the world arising from life in the family households of traditional society is suppressed or cov- ered over by participation in the output of the mass media. For this issue, a repertoire that has provoked broad discussion with terms such as schema, cognitive map, prototype, script, frame might be useful. 2
These are psychological terms, but ones that are increasingly being used to explain social coordinations or so-called 'collective' behav- iour. 3 Their starting point is memory's need to discriminate con- stantly in the torrent of operations which occupy a system between forgetting and remembering, because without forgetting, the ca- pacities of the system for further operations would very quickly be blocked and, to put it another way, one would only ever be able in future to experience or do the same thing. Forgetting sets you free. But since for its part forgetting cannot be remembered, one needs a schema that regulates what is retained and can be reused. These may be schemata of perception which enable the gaze to be focused and the unfamiliar to be recognized by setting it against what is familiar. But they may also be more abstract categoriza- tions, or both at the same time if, for example, people's qualities or
? behaviour are inferred from racial characteristics. Schemata do not force repetitions to be made, neither do they specify action. In fact, their function is precisely to generate space for freely chosen be- haviour in a system which with its own past has put itself in the state (and in no other) in which it currently finds itself. This is what abstraction (not necessarily conceptual) is for, the disregarding of . . . , the repression of the countless details which mark situations as unique and unrepeatable. But abstraction also means that new situ- ations can modify the schema. The schema allows for supplements and replenishments; it cannot be applied 'schematically'. 4 Devia- tions come as a surprise because of the schema; they become con- spicuous and thus imprint themselves on the memory. Schemata are instruments of forgetting - and of learning; they are limitations to flexibility which make flexibility within prestructured barriers possible in the first place.
As Kant taught us,5 schemata are not images but rather rules for accomplishing operations. The circle schema, for example, is not the depiction of any circle, but the rule for drawing a circle. The diversity given to the inner meaning in the form of time differences can only be reconstructed as procedures (also presupposing time) for purposes of knowledge. For Kant, this copying of time from the empirical over into the transcendental sphere was the reason why a relationship of similarity could be assumed in spite of the radical difference of objects and ideas. This problem does not arise if one takes on board a radically constructivist epistemology.
But it re- mains the case that schemata are not images which become con- cretely fixed at the moment of depiction; they are merely rules for the repetition of operations (which then are concrete again). Thus, memory does not consist of a supply of images which one can look at again whenever necessary. Rather, it is a question of forms which, in the ceaseless temporal flow of autopoiesis, enable recursions, retrospective reference to the familiar, and repetition of operations which actualize it.
Schemata can refer to things or to persons. The utility meaning of things is one schema, the hierarchies among people or standard- ized role expectations are another. 'Script' refers to the special case where temporal successions are stereotyped (for example, the fact that we are supposed to buy a ticket before getting on a train). The
? observation of causal relationships typically follows a script be- cause it cuts out other, equally realistic possibilities for causal attri- bution. 6 It is only by way of a script that one comes to attribute effects to actions. A script is therefore an already fairly complex schema which also cuts out many things and presupposes both a stereotyping of events and a standardized coupling of their succes- sion. If thing or person schemata are linked to a script, it also means that the observer is no longer free to choose between object schema and time schema or to let his or her gaze oscillate, but that object schema and time schema enter a relationship of mutual depend- ence where the one cannot be chosen without consideration of the other. We have already considered such a case using the example of the narrative structure of novels: the sequence of actions charac- terizes the people whose motives then make the sequence of actions understandable again - with sufficient scope for surprises.
Now, we assume that the structural coupling of mass media com- munication and psychically reliable simplifications uses, and indeed generates, such schemata. The process is a circular one. The mass media value comprehensibility. But comprehensibility is best guar- anteed by the schemata which the media themselves have already generated. They use a psychic anchoring7 for their own workings which can be assumed to be the result of consumption of mass media representations, and indeed can be assumed to be such with- out any further tests. Let us elucidate this using two examples: the production of causal scripts in domains that are inaccessible to in- dividual experience and thus very typical in the case of ecological problems; and the presupposition of different person schematizations depending upon whether it is oneself or other peo- ple who are involved.
Communication about ecological problems is a particularly good example for our purposes,8 because it goes far beyond the individu- al's world of experience. (Who could say from their own knowl- edge what would have happened to the contents of the Brent Spar platform, given the pressure operating on the sea bed, if it had been sunk? ) The mass media too are unequal to the task, and when they turn to science, they will typically be given more knowledge and more ignorance at the same time. So, we are dependent upon schema formation. It might be normative sentences which are set against a
? 'virtual reality' and are very typically fashioned metaphorically. For example, the ocean should not be used as a rubbish dump. This is self-evident, so to speak. If one asks further, more scripts are brought to bear. Out of innumerable possible causal constellations, one is picked out which can be made plausible. Usually the points at issue are the effects of actions, not nature's own course. Effects can then be coupled onto this which are sufficiently worrying to prevent people from asking any further as to how likely they actually are. To put it another way, what is involved are schemata of change which correspond to the selection criteria of news and in-depth reporting (for example: new, action, drama, morality). Environ- mental pollution changes the living conditions of people on earth to the point of conditions which make the continuance of life im- possible. There is no coming up against difficulties with individu- als' memories or their world of experience here. They have not yet experienced such things or can at best, if the script is offered, acti- vate experiences of their own that fit (the layer of filth on the car parked outside). So it is not a case of the 're-education' of individu- als, of them unlearning, in a more or less difficult process, some- thing that had been thought of as knowledge. The ecological imagery, its schemata, its scripts are developed on a greenfield site, so to speak, they form a terrain that is not yet occupied.
People speak of a 'transformation of values'. The question, how- ever, is whether the reorientation with newly recommended values is beginning, or whether it is the causal scripts which impress us first; whether, that is, it is the change which we find fascinating and which then leads to values being associated with it. Salancik and Porac speak of 'distilled ideologies' and mean by this, 'values de- rived from causal reasonings in complex environments'. 9
Anyone who adheres to ideas such as 'objective truth' or psychi- cally binding 'consensus' will not be able to accept this analysis and will accuse the mass media of superficiality, or even manipula- tion. If, on the other hand, one takes the individuality and the op- erational closure of autopoietic systems seriously, one will see that it cannot be otherwise. From the point of view of society, struc- tural coupling mediated via schemata has the benefit of accelerat- ing structural changes in such a way that, if this acceleration is successful, it will not break the structural coupling of media and
? individuals but will simply link up to other schemata. From the point of view of the individual, the advantage of schemata is that they structure memory but do not determine action. At the same time, they offer liberation from burdens that are too concrete as well as a background against which deviations, opportunities for action and constraints can be recognized. Individuals are still at liberty in this instance to get involved or to leave it be. They can allow feelings to arise and identify with them, or they can observe this in others and think of it as strange or even as dangerous. And with that, we have arrived at our second theme, a complementary hypothesis about relationships between mass media and individu- als.
In psychology, it has long been common to distinguish the schematization of one's own person (that is, answers to the ques- tion: who am I? ) from the schematization of other persons. 10 The distinction is interesting in various respects - qua distinction. First of all, every human being is given as a concrete individual, that is, different from others in terms of appearance, name and other char- acteristics. Why then is it not sufficient and since when has it no longer been sufficient to distinguish oneself from others just as every individual does from every other? Why is it not sufficient to use the same list of objective characteristics (age, sex, family, good- looking or not so good-looking, place of residence, virtues, vices etc. ) and to concretize the person being referred to only by a com- bination of these? There would be unlimited possibilities which could be supplemented as required. Furthermore, why, when one is dealing with concrete individuals, is schema formation necessary at all? As in the case of ecology, we have to assume non-transpar- ency, which is what offers the occasion in the first place for simplifications or, as we also call it nowadays, 'identity'. But why is someone non-transparent to himself, that is, in need of a schema, even though, according to Descartes, he cannot doubt his thinking existence?
We can be certain of the fact that the difference of one's own I from other individuals is given from the start, meaning as early as a few days after birth. The newborn child has to practise comple- mentary behaviour, not an imitative one, such as reversing right/ left perceptions. 11 Infant socialization after this presupposes what
? Stein Braten calls 'dialogic closure', that is, systems that can be fenced off to the outside, in which there is provision for a place for a 'virtual other', that is, for effective occupation. 12 This position of the virtual other can only be occupied with the aid of schemata, since it requires recognition, that is, memory. On the other hand, one does not need a 'virtual ego'. One is who one is from the start. But how then does a secondary need for self-schematizations arise? And what happens when the requirement of a direct 'dialogical closure' is overstepped and the occupation of the position of 'vir- tual other' no longer occurs effectively (in the sense of virtus), but is 'enriched' by fictional components?
We can assume that effects of the mass media become visible at this point. Early modern theatre in particular will have introduced this new development first. It offered the possibility of making ac- tors' inner processes of opinion formation, conflicts and uncertain- ties visible on stage through language. It might be that the actors would address themselves directly to the audience in forms which implied that the other players on the stage could not hear it (but how does one learn this unusual, counterfactual implication? ); or else it might occur in the form of monologues or soliloquies. The audience could then observe how the actors on the stage motivate themselves and deceive themselves and others, and that this pro- cess initially remains invisible to other participants in the play. 13 In the finely honed dialogues of Vienna theatre (for example, Schnitzler's Liebelei ('Light-O'-Love') or Hofmannsthal's Der
Unbestechliche ('The Incorruptible One')) the sentences themselves are constructed in such a way that the audience is able to observe more than those being addressed. The emergence of this compli- cated, as it were highly charged, cultural form of observation of observers and the development of suitable schematizations is there- fore not a direct product of the printing press or of the mass media. But once this specific form of second-order observation with its schemata of motives (love, criminality, sincerity/insincerity etc. ) is practised and can be presupposed as a way of observing, it can then be used in other contexts as well, such as in the novel and ultimately even in philosophy. And then the viewer or the reader is tempted as well to take a second look at his or her own way of observing and its motives.
? Shaftesbury seems to have been one of the first to retreat to a private conversation with himself in order to gain clarity about himself, in spite of having clear misgivings about the printing press and its commercial publishers, of which, of course, he himself makes use. 14 Rousseau likewise has his confessions printed, even though he explicitly exempts himself from the criteria of judgement which apply also to others. 15 The Romantic era plays with doppelgangers, twins, reflections, in order to represent the transformation of iden- tity into communication. Towards the end of the nineteenth cen- tury, William James, Georg Simmel and many others would speak of the need for a 'social self' or an 'identity' which is to be, or is to pretend to be, a fragmentary, turbulent, chaotic individual, in or- der to be something for others which it itself is not by itself. 16 And now the 'search for meaning' begins - at least in printed texts. We arrive at a time in which literature and life in literature can no longer be separated. The problem of 'self-realization' is invented and is taken up and disseminated by the mass media. Individuals are encouraged to believe that, although they have without doubt been really alive since conception, and certainly since birth, they must become even more real (or unreal? ) than they already are.
This semantic ambiguity can be understood if we read it as an indication of a need for a schema that cannot, however, be admit- ted. We can recognize schemata, in others and in ourselves, if we take them to be cognitive routines, abbreviations for something that might be elucidated. But this itself would also be a schema which conceals the issues that are ultimately involved. In view of the unobservability of the world and the non-transparency of indi- viduals to themselves and to others, schema formation is unavoid- able. Without it there would be no memory, no information, no deviation, no freedom. One can also, with Spencer Brown, under- stand this as the necessity of a form which marks a distinction, one side of which must be marked if one wants to observe and to for- mulate more operations. This does not stop us from asking about the social conditions of the plausibility of such schemata. In the age of the mass media, they are virtually unthinkable without the par- ticipation of the media. Like theatre, the mass media also put the individual into a scene that is outside the scene set on the stage. We have described this as a technical condition for the differentiation
? of a media system. This distance has to seem ambivalent to the individuals: on the one hand they are not themselves the text being performed for them; and if, like Rousseau, they have written and published it, they are it no longer. Neither do they see themselves on television, and if in an exceptional case they do, it is with special pleasure in the self-recognition only found in exceptions. On the other hand, the mass media produce the world in which individu- als find themselves. This is true of all programme sectors: of news, advertising, entertainment. What is presented to them affects them too, since they have to lead their lives in this world; and it affects them even when they know very well that they will never get into the situations or play the roles presented to them as factual or fictional. Instead, they can still identify with the cult objects or the motives which the scripts of the mass media offer them. When indi- viduals look at media as text or as image, they are outside; when they experience their results within themselves, they are inside. They have to oscillate between outside and inside, as if in a paradoxical situation: quickly, almost without losing any time, and undecidably. For the one position is only possible thanks to the other - and vice versa.
The consequence must be that the individual must resolve this paradox for herself and construct her identity or her 'self' herself. The materials used for this can be the usual ones. But there is no possibility of taking on an 'I' by analogy from outside. No one can be like someone else. No one sees himself as the reflection of an- other. The only point of agreement is the necessity of using sche- mata for sustaining a memory. But self-schematization cannot relieve the strain on itself through the illusion of an 'objective' (even if disputed) reality. On the one hand it (self-schematization) is indis- putable, for no one can perform it for another, and on the other it is under threat of constant dissolution. This is because no one can know whether he will remain who he had thought he was. He can- not know because he himself decides the issue.
The structural couplings between individuals and society affect the whole of reality. This is true of all social formations. However, the mass media vary the structural conditions of these structural couplings because they change the need for schemata as well as what they offer. The schemata and scripts of ecological concerns
? and the necessity of schematization of one's own person are only extreme examples chosen to illustrate this. And perhaps it is no coincidence that these two environments of social communication, the complexity of non-human nature and the auto-dynamic and non-transparency of human individuals, are dependent in a par- ticular way upon schemata and therefore upon structural couplings to the system of the mass media.
? Second-order Cybernetics as Paradox
The second-order cybernetics worked out by Heinz von Foerster is rightly held to be a constructivist theory,1 if not a manifesto for operational constructivism. The reverse does not apply, however. Constructivist epistemologies do not necessarily have the rigour of a cybernetics of cybernetics. One can observe cognitions as con- structions of an observer, without linking with this the theory that the observing observer observes himself or herself as an observer. This difference is so crucial that we must devote a final chapter to it.
The discussion thus far has been guided by two points of depar- ture. The first is that the mass media, like any broadcasting system, are an operationally closed and, in this respect, autopoietic system. The second emphasizes that this is also true of cognitions, because cognitions are also operations and can therefore only be produced in the system. This remains the case even when one considers that in society communication can take place with the system of the mass media from out of the latter's environment, for these commu- nications too are possible only on the basis of the knowledge that the mass media have provided. Furthermore, the mass media un- derstand what is uttered to them only on the basis of their own network of reproduction of information. Every communication in and with the mass media remains tied to the schemata which are available for this purpose.
This theoretical description is designed in the mode of second- order observation. It observes and describes observers. But it does
? not presuppose that the mass media observe themselves in the mode of second-order observation. The media designate what they are communicating about and must therefore distinguish it. For exam- ple, they inform people about scandals and in doing so must pre- suppose that non-scandalous behaviour would have been possible as well. What is not reflected here, however, is that one could pose the question (which a sociologist might pose) why something is even being observed in the schema scandalous/non-scandalous at all, and why the frequency of use of this schema is clearly increas- ing. In other words, the media remain (for good reason, as we shall presently see) invisible to themselves as an observer. They are turned towards the world in their operations and do not reflect that this turning itself generates an unmarked space in which they find them- selves.
We can reformulate this statement by splitting our concept of autonomy. First, there is autopoietic autonomy which is based on operational closure and means that the system can only reproduce its own structures and operations with its own operations, that is, from its own products. This is to be distinguished from cognitive closure, and, correspondingly, cognitive autonomy. This says that along with all its cognitions the system is also observing that these are only its own observations. Only having reached this point do we find ourselves on the terrain in which second-order cybernetics in the strict sense is interested. 2 Here, the question 'who is the ob- _ server? ' is asked universally and is also applied to the observing system. Questions about the observer take the place of questions about reasons, which would necessarily result in an infinite regress. And therefore, whoever wishes to give reasons for his own experi- ence or actions must observe himself as an observer and, in doing so, allow access to the choice of the distinctions which guide his observing. But how is that possible?
Obviously, from an empirical point of view, the system of the mass media does not operate at the cognitively closed level of sec- ond-order cybernetics. It does distinguish self-reference and other- reference. In its attitude of other-reference it reports on facts and opinions. This includes the possibility of observing observers. The second-order observation common in modern society comes about in this respect. But this merely leads into the infinite regress of the
? question as to which observer is observing this. In the system itself, there is no final figure of the ambiguous 'observing system',3 no autological realization that whatever is true for observers is also true for the system which is observing them. Thanks to the distinc- tion of self-reference and other-reference, the system of the mass media can also mark itself in contrast to everything else. It can make its own structures and operations into a topic as though they were objects. But it does not additionally ask: how am I operating as an observer and why do I make distinctions in this way and not another? With every distinction it uses it places itself in the unob- served, unmarked space, and this is even so when it marks itself in contrast to other things. Every distinction makes the observer in- visible - but this is precisely what we can still know. If she wanted to de-invisibilize herself, she would have to mark herself, that is, distinguish herself. And then one would again have the question, who is the observer who distinguishes thus and not otherwise?
This is also true of modern society, and also in conditions which some people describe as 'postmodern'. It even applies if one re- nounces absolute demands for validity which in the tradition went under names such as God or nature or reason. This renunciation is presented as relativism or historicism. One accepts the contingency of all criteria and of all possible observer positions. But that only means that one is able to switch from any distinction to another, that, for example, one can take into account fashions or transfor- mations of values. In fact, these are now accepted schemata. The problem of transformation and of contingency has been digested and can be expressed with the normal schematisms of the mass media. The system may then be operating at a level of greater un- certainty, but that is also true of the other function systems, of the money economy, art, science, politics. In accepting this character- istic postmodern style the mass media are merely following what the form of social differentiation suggests. But with a constant change of perspectives, the observer who is performing this trans- formation with the before/after distinction still cannot be grasped. 'God is dead', they said - and meant: the last observer cannot be identified.
As a reaction to this finding, attempts have been evident for some years now to shift the problem onto ethics. This is true throughout
? society and thus also in the mass media. For example, a code of ethics for journalists can be drawn up and the attempt made to apply it via the profession's self-regulatory procedures. The fact that this cannot be an ethics of reasoning in the academic style is easy to see if one follows the academic debate about transcendental ethics, utilitarian ethics or value ethics. In none of these cases have radical deductive steps towards decisions succeeded. We can know this. Therefore, these can only be conventions which continually find themselves confronted with new situations. Nor does this eth- ics, if it is not condensed into norms of law, contain any indication of how deviants are to be treated.
The position of a second-order cybernetics offers an opportunity to reflect this flight into ethics as a displacement of the problem. After all, whatever else it is understood to be in concrete terms, ethics too is a distinguishing practice. It distinguishes standards and ways of behaving, it distinguishes conforming and deviant be- haviour and usually even in a moral sense good and bad, or evil, behaviour. Moreover, it is a part of its presuppositions that devia- tions are attributed to behaviour and not to inappropriately cho- sen standards or, as critical sociologists thought for a while, to 'labelling'. 4 Even if strong ties and highly charged emotions are to be expected in heavily moralized domains, second-order cybernet- ics can still ask: why are you distinguishing in this way and not in another? Or again: who is the observer who is trying to impose these schemata here?
Standard authors of constructivist epistemology, such as Humberto Maturana and Heinz von Foerster, have attempted to develop a new ethics on this basis. However, they have not gone beyond making a few suggestions,5 and it is doubtful whether this venture can succeed. For an ethics would sabotage itself if what was demanded of it was that it make distinctions and simultane- ously reflect that it is itself making these distinctions.
Even in the face of numerous efforts to find ethical foundations, second-order cybernetics can only ever repeat the question: who is the observer?
