4- Systems Theory as an Alternative to Action Theory the Rise of Stichweh

7/29/2019 4- Systems Theory as an Alternative to Action Theory the Rise of Stichweh

1/10

Systems Theory as an Alternative to Action Theory? The Rise of 'Communication' as aTheoretical OptionAuthor(s): Rudolf StichwehReviewed work(s):Source: Acta Sociologica, Vol. 43, No. 1 (2000), pp. 5-13Published by: Sage Publications, Ltd.

Stable URL: http://www.jstor.org/stable/4201178 .Accessed: 17/12/2012 03:09

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

.

Sage Publications, Ltd. is collaborating with JSTOR to digitize, preserve and extend access toActa

Sociologica.

http://www.jstor.org

This content downloaded on Mon 17 Dec 2012 03:09:52 AM
http://www.jstor.org/action/showPublisher?publisherCode=sageltdhttp://www.jstor.org/stable/4201178?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/4201178?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=sageltd


2/10

ACTAOCIOLOGICA000

Systems Theory as an Alternative to Action

Theory? The Rise of 'Communication' as aTheoretical Option

Rudolf StichwehDepartment of Sociology, University of Bielefeld, Germany

ABSTRACTThe argument of the essay has two main parts. First, it reflects on the presumed conflictbetween action theories and systems theories in sociology. Looking at authors such asJames Coleman, Talcott Parsons and Niklas Luhmann, the essay tries to show that thereis a natural complementarity of action and systems theories, and therefore thepresumed disjunction of 'action' and 'system' is not based on the empirical reality oftheory-building ventures. But then another line of conflict becomes visible. Since theinformation theories of the late 1940s, 'communication theory* has become a viableand universalistic option in social theory, one that indeed conflicts with action theory.In its second part, the essay first gives a brief sketch of the conceptual career ofcommunication theory since Shannon and Weaver. It then presents the sociologicaltheory of Niklas Luhmann as the first major sociological theory that opts forcommunication as the constitutive element of society and other social systems. Causesand reasons for this theoretical decision are reconstructed, first in terms of problemsinternal to Niklas Luhmann's social theory (the distinction of psychic and socialsystems; the distinction of action and experience; formal properties of the concept ofcommunication; the implications of autopoiesis) and secondly in terms of processes ofsocietal change (the rise of the information society; the genesis of world society), whichfavour the switch towards a communication-based (instead of action-based) systemstheory.Rudolf Stichweh, Department of Sociology, University of Bielefeld, D-33501 Bielefeld PE100131, Germany? Scandinavian Sociological Association 2000

1. Systems theory versus action theorySystems theory and action theory are normallysupposed to be alternative sociological ap-proaches.1 Therefore, my first question is ifthis description is really true. Are systemstheory and action theory complete options insocial theory? closed in on themselves andcompeting as such? The answer this essayfavours is clearly 'no': There is no such thingas a disjunction of systems and action theories.This may easily be seen in looking, forexample, at James Coleman, probably the mostinfluential action theorist in present-day sociol-

ogy. Coleman's social theory, as presented in itsdefinitive form in Foundations of Social Theoryfrom 1990, is first of all an exchange theory,that is, social exchange is considered to be theelementary transaction constitutive of society(Coleman 1990; cf. Clark 1996; M?ller &Schmid 1998). But how to combine socialexchange and social action? One can exchangeresources or property, or perhaps even informa-tion, although I doubt this last possibility of so-called 'information exchange'. But obviouslythere is no sense in saying that actors exchangeactions. How could they do it? An action is notsomething I can hand over to another actor,

This content downloaded on Mon, 17 Dec 2012 03:09:52 AMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp


3/10

6 ACTASOCIOLOGICA000 VOLUME3

saying to him or her, 'Now it is your action,please give me an action of yours in return'.Therefore, action theory and exchange theoryare incompatible, or even incommensurable, touse Thomas Kuhn's term. James Coleman hasfound an interesting solution to this problem. Inhis view, social exchange is focussed on rights ofcontrol Actors typically exchange rights ofcontrol over actions, and Coleman's book is anextended essay that shows the near universalityof this formula. To take one suggestive examplefrom his book: if while escaping in a panic froma burning cinema or theatre I decide to follow aspontaneously emerging leader who tries to stopthe running and to lead the crowd quietly to anexit, in Coleman's interpretation I transfer theright of control over my actions to this leader,and I do this because I hope to acquire anenhanced probability of survival in return(Coleman 1990:203-215).What I find remarkable is Coleman'scentral formula: rights of control over actions.Where do these rights of control come from? Ifone does not want to argue in terms of naturalrights (lex naturae), one will have to accept thatthese rights must be created, institutionalizedand legitimized in social processes. Thus, actiontheory even in its most basic terms presupposesan encompassing social system, which func-tions as the context of creating, institutionaliz-ing and legitimizing rights of control overactions (cf. Fararo 1996; Stichweh 1998a).This conceptual situation is akin to a well-known problem in Talcott Parsons' theorizing.Parsons wanted to prove the possibility of socialorder by analysing an elementary social situa-tion he called double contingency (Parsons & Shils1951:16; Parsons 1968:167-168). There arealways at least two actors, alter and ego. If insuch an elementary situation the choices ofeach one of these two actors are contingent onthe choices of the other one, then there arises acircular situation of double contingency inwhich no action at all may happen. Parsonstried to show that a shared symbolic system forthese two actors must be presupposed if onewants to evade the conceptual consequence ofaction being blocked by reciprocal uncertainty.Only if such a shared symbolic system exists canuncertainty about the probable choices andreactions of the other one be reduced. Only thendoes it become possible for one of the twoparticipants to begin acting. In other words: theprobability of social order can only be demon-strated if an already existing social order ispresupposed.

The same logic seems to reign in Coleman'sargument. Only if a pre-existent distribution ofrights of control over actions is presupposed canelementary processes of social exchange - i.e.processes of exchanging these rights of control -start. What this reconstruction of parallelproblems in Parsons and Coleman is intendedto show is that action and system are notalternative or competing versions of socialtheory. They are more realistically to bedescribed as complementary aspects of socialtheorizing - and if you read Coleman's massivetome from 1990, you register an extensive useof the words 'system' and 'social system'.2 Whatmakes the crucial difference in theory design isthe fact that the degree of conceptual elabora-tion of the concept of system is minimal inColeman, whereas he invested much effort inthe concept of action (Coleman 1986).I want to introduce a further argument forcomplementarity by looking at Parsons again.In my opinion there is no sense in classifyingTalcott Parsons as either an action or systemstheorist, or in postulating different stages in thedevelopment of his theory which gradually shiftthe focus from action to system. Regarding theinterrelation of action and system, Parsons'point simply seems to be that action is system.This point is already present in The Structure ofSocial Action from 1937, where he decomposesthe unit act into its constituent components (i.e.ends, conditions, means, norms and, finally,the actor).3 None of these components can beconsidered the final determinant or cause of aunit act. Therefore, even in the early Parsons,you make a categorical mistake if you attributea unit act to an actor as its originator. The actoror the personality is only one of five compo-nents, which are always necessary for produ-cing a unit act. There must always be an actor,but he is not privileged in relation to the othercomponents, such as conditions, means andnorms.

The same analytical strategy is to be foundin the later Parsons. Now the classification ofthe components of the unit act, which even in1937 was called the 'action frame of reference',has been replaced by four subsystems of action -behavioural system, personality, social systemand culture.4 All of these four subsystems mustalways be involved if a single action is to arise.This allows a remarkable symmetry in analys-ing action and system. On the one hand, actionis always a system, because it is not a final,irreducible entity, but a complex emergencefrom plural structural components. On the



4/10

Systems Theoryand ActionTheory 7

other hand, the social system - being theprototypical candidate for systemness in manyrespects - is only a lower-level system in relationto action. The action frame of referenceencompasses the social system as one of itssubsystems, not the other way around.

2. Luhmann as an action theoristI could finish my paper here and conclude thatthere is no problem at all regarding action andsystems theories. Obviously it is very difficult tosynthesize exchange theory and action theory,as we saw in the case of James Coleman. Thesame difficulty would arise if we tried tocompare or synthesize network and actiontheories (cf. White 1992). On the other hand,there seems to be a natural complementarity ofaction theory and systems theory. There aredifferences in perspective, different ways ofsolving or evading the micro/macro problem,of course. But there is no such thing as analternative of action theory versus systemstheory, so far. Systems theories seem to pre-suppose a microstructure of actions and actiontheories presuppose a macrostructure of sys-tems. And there is, as a special case, the originaland unorthodox solution by Parsons in whichaction is considered to be the more generalphenomenon and as such presupposes aninfrastructure of contributing systems.If one looks at the more recent develop-ments of sociological systems theory, primarilyrepresented in the writings of Niklas Luhmann,one might for the time being come to exactly thesame conclusion. In Luhmann's writings fromthe 1970s, one often finds the term actionsystems. But it is used as an obvious term, not asa problem in theory building and not as a focusof his conceptual interests. In the essays fromthe late 1960s/early 1970s,5 in which NiklasLuhmann presents the foundations of hisversion of systems theory for the first time -which are still valid in most relevant respects -one finds no discussion of a supposed alternative'action versus systems' theory. For Luhmann itwas obvious that action is an elementary termin describing social systems.Not until 1978 did Luhmann publish anessay called 'Handlungstheorie und System-theorie', reprinted in Soziologische Aufkl?rung 3in 1981 (Luhmann 1978a). It is a polemicalessay directed against authors such as AlanDawe or Michel Crozier and Erhard Friedbergwho try to construct 'the actor' and 'the system'

as two paradigms for two alternative socio-logies. Luhmann decrees, in contrast, that therecan be no such disjunction of action versussystem after Durkheim, Weber and Parsons. Theonly question in which there is a real con-ceptual option at hand is how action and systemare related to one another in sociologicaltheories. In Luhmann's essay there arises nodoubt that actions are considered to be the elementsof social systems. Luhmann goes on to debatequestions such as the temporality of actions andthe hypothesis that social systems processattributions by which it is decided of whoseactions we speak. But these attributions arecontestable and there may arise disagreementsregarding the question as to whom we hold to beresponsible for an action.

There are some related essays in the yearsfrom 1978 to 1980, among which the mostoriginal is, in my opinion, 'Time and Action. AForgotten Theory' from 1979 (Luhmann1979). There again, actions are treated aselements of social systems, and the decisivepoint of the essay is to establish the specifictemporality of actions. Actions are events,vanishing the same moment they happen, andtherefrom arises a peculiar property of socialsystems. Social systems operate against aconsiderable probability of void. The emergenceof ever-new actions may be attributed to a kindof horror vacui.

From this well-defined and therefore appar-ently stable conceptual situation arose a sharptheoretical shift in the early 1980s. I am notquite sure if this instability and subsequent shiftmight be anticipated from a careful reading ofthe essays I just mentioned. But I remember wellthat in the late 1970s/early 1980s Luhmannrepeatedly said in lectures and seminars that hedid not yet know how to take a major theoreticaldecision: if one looks for the constitutiveelements of social systems, which is the bestcandidate for element status, actions or commu-nications? Some years later in Soziale Systemefrom 1984, the decision is taken. Systemstheory is reformulated as communicationstheory, with the concept of action relegated toa secondary status. Therewith arises a real andconsequential alternative in constructing socio-logical theories: one can either formulate themas communication theories or opt for continu-ing action theories.In the following I look for the causes andreasons for this theoretical discontinuity. First Idescribe trends and shifts in science and socialscience after World War ?, which are respon-



5/10


sible for communication theory becoming aviable option in social theory (section UJ). ThenI look for the more specific context of systemstheory in the writings of Niklas Luhmann.Which conceptual tensions inherent to Luh-mann's writings motivate the switch fromaction theory to communication theory (sectionIV)? And are there any trends in the evolution ofthe structures of modern society on which anargument might be based that changing fromaction to communication theory reflects onsocietal change and increases the adequacy ofsociological theory to the emergent structures ofmodern society (section V)?

3. The rise of communication theoryFirst of all I will look for global trends in scienceand social science since World War ?. It shouldbe noted that not a single sociological theorybased on communication theory existed beforeNiklas Luhmann. The subject of communica-tion was surely relegated to special sociologiessuch as mass communication, public opinionresearch, etc. Communication theory had beenaround since antiquity in the rhetorical tradi-tion (cf. Craig 1999). But this was receivedneither into sociology nor into other socialscientific disciplines. Only after World War ? didthe concept of communication and the closelyrelated concept of information enter science andsocial science as fundamental scientific con-cepts; this entry into the foundations of sciencewas based on technologies of informationprocessing arising at the same time.

My hypothesis is that there is a relativelydirect lineage from the early information andcommunication theories of the late 1940s and1950s to the adoption of communicationtheories in sociological theory, and especiallyin Niklas Luhmann. There are, first of all, theideas of Claude E. Shannon, well presented inthe co-authored book with Warren Weaver TheMathematical Theory of Communication from1949, of which it is often unjustly said that itrestricts itself to machine communication(Shannon & Weaver 1949; Shannon 1970).But the most interesting point in Shannon andWeaver, which has nothing to do with machinecommunication, regards the relation of infor-mation and selection. Information is defined viathe number of states from which it selects, andtherefore information is related to unpredict-ability. This allows the famous analogy ofinformation and entropy, of which Ruesch and

Bateson, two years later in Communication: TheSocial Matrix of Psychiatry, enthusiastically saythat they consider it the most importantscientific discovery since Aristotle (Ruesch &Bateson 1951). The book by Ruesch andBateson, which is still a very remarkable texttoday, is, as far as I know, the first essay to basean entire scientific discipline (psychiatry) in itsfundamentals on the new concept of commu-nication. Regarding Bateson, in later writingshe added the very apt formula 'information is adifference which makes a difference', whichcouples a concept of information based onselectivity with the idea that there are alwaystwo systems involved that are operationallyclosed and therefore differ in their selectivities(Bateson 1973:286ff. et passim). Thereforeinformation is always related to the selectivityoperative in a system. There is one furtherdecisive point in Ruesch and Bateson. In theirkey chapter called 'Information and Codifica-tion', they incessantly try to level the differencebetween fact and value, the difference ofmforming and evaluating (Ruesch & Bateson1951:168-211). From this levelling effortarises the idea of distinguishing communicationand metacommunication as two componentsalways inextricably entangled in any act ofcommunication. It is easily seen that Luh-mann's distinction between 'Information' and'Mitteilung', information and utterance, derivesfrom this.

To these antecedents a number of newdevelopments were added in the 1960s. There isthe Palo Alto school of Watzlawick and otherspopularizing and extending the Bateson line ofargument and exploring the pragmatics andparadoxes of communication (Watzlawick et al.1967). Speech act theory arises via Austin andSearle and in its sociological reception transmitsthe so-called 'linguistic turn', which is regis-tered in many disciplines to the discipline ofsociology (see Searle 1969). At last there areEthnomethodology and Conversation Analysis,the first theories-cum-methodologies germaneto sociology that allow exploration of thedomain of communication (Sacks 1992). Thepoint I want to establish is simply that thesedevelopments seem to make it nearly unavoid-able that a major sociological theorist shoulddraw the consequences, synthesize these ratherdiverse conceptual strategies and then no longerbase sociology on the concept of action, butinstead on the concept of communication.There was no chance of Parsons becoming thetheorist to take on this role, his thinking being



6/10


rooted firmly in the action theory of the 1930s.Even Parsons' ingenious idea of generalizedsymbolic media, which inevitably had to beconceived as macrosocietal mechanisms, was,via the language Parsons used for it, subsumedunder the concept of interaction. Thus, it wasLuhmann, in choosing communication theoryover action theory, who took on the role of thefirst major sociological communication theorist,a role which had to be taken by someoneanyway.

4. Tension in Luhmann's writings: causesfor a practical switchWhy did in Niklas Luhmann's theories arise aneed for deciding between communication andaction as alternative constitutive elements ofsociety? To this question I give an answer in twoparts. First, I discuss tensions inherent toLuhmann's theory and its conceptual develop-ments, which in the long run favoured theswitch to communication theory. Secondly, Iwill complement this list of causes and reasonsby pointing to structural changes in modernsociety, which, being reflected in Luhmann'stheories, again privileged communication overthe rival concept of action.Psychic and social systemsOne of the earliest and most enduring distinc-tions in Luhmann's theory is that betweenpsychic and social systems. These are two levelsof system formation, autonomous as self-orga-nizing entities, but related because socialsystems are dependent on psychic systemsoperating in their environments, and psychicsystems are incessantly being socialized anddisturbed by ongoing processes in social sys-tems. If one regards this distinction as funda-mental, one will soon perceive that it is noteasily compatible with action theory. Actionnormally is closely related to an actor and his/her goals, intentions, motives, will and effort.One can then introduce a distinction of actionand social action, as has often been done sinceMax Weber. But how to distinguish psychic andsocial systems in these terms? If one refersactions to the domain of psychic systems andsocial actions to the domain of social systems,one obviously argues in terms of an analyticaltheory, which attributes different aspects of oneand the same action event to the two differentlevels of system formation. Parsons is a goodillustration of this, as he combined a commit-

ment to action theory with a clear demarcationbetween psychic and social systems, and this onthe basis of an analytical systems theory (whichhe called 'analytical realism'). But there are nosuch things as analytical systems in NiklasLuhmann, who introduced the distinctionbetween psychic and social systems as referringto concrete, 'real' systems. Therefore there wasalways an in-built, latent bias against actiontheory in Niklas Luhmann, which had tobecome more explicit as soon as he perceivedaction only as a conceptual option with variousconceptual alternatives.Action and experienceMy second point refers to another distinctioncentral to systems theory: action and experi-ence. This is a distinction Luhmann introducedearly in the 1970s.6 In German it is Handeln andErleben. In my view this is one of the mostoriginal of Luhmann's ideas, for which I see noantecedents in the history of sociological theory.In distinguishing action and experience, Luh-mann claims first of all that the most generaldescription of social systems would describethem as processing selections.7 In social systemsthere are two ways of processing selections. Onemay interpret selections as actions, attributingthem to a concrete acting system which isthought to be responsible for the genesis of theseselections. But this is only one of two possibi-lities. There are other cases in which one takes aselective event as information about states of theworld. Then there is no need and no motive toattribute these selections to concrete actors.These selections are not actions, and they arenot causally related to actors, but they functionin Luhmann's terminology as experience (Erle-ben). This distinction is not ontological. Selec-tions are not actions or experiences due toinherent properties that they possess. Classifica-tion as action or experience is an achievement ofthe participants in social processes, whoseclassifications are contestable. There may arisedisagreement as to whether a relevant eventshould be attributed to an actor whose respon-sibility can then be postulated, or if it simplyrepresents a state of the world not having beencaused by actors involved in the presentsituation.

What we learn from this distinction is thatthere has always been one difficulty for anysociological action theory. No matter whichsocial entity is considered the constitutiveelement of social systems, it has to be denomi-



7/10


nated by a more general term than either'action' or 'experience'. This disqualifies theconcept of action as the constitutive element, asaction cannot function as the generic term forthe distinction between action and experience.Communication is a much better candidate fortwo reasons. First, the concept of communica-tion is more specific than the concept ofselection, which does not discriminate betweensocial and biological systems. Secondly, theconcept of communication is more generalthan either action or experience. It seems to beplausible that the processes in which attribu-tions are made, contested and remade arecommunication processes. Therefore manyyears before Luhmann adopted communicationtheory as the basis of his sociology, there was asecond in-built bias favouring this choice.Formal properties of the concept ofcommunicationMy third point regards what may be calledformal properties of the concept of communicationin contradistinction to the concept of action. Inaction theory relations of action and actor aresymmetrical. There exists always one individualor collective actor for one action. Symmetryimplies that this relation can be interpreted inboth directions: the action may be caused by theactor; the actor may be constituted or selectedby the action, which only then is his or heraction.8 There is no need for a third term inaddition to action and actor. In communicationtheory it is wholly different. There, one alwaysneeds at least three terms: sender, receiver andthe information or communication that relatesthem. Or, in Luhmann's version, which focuseson the internal structure of the communicativeact: one needs an observing system whichunderstands communication by projecting thedifference between information and utterance onthe system observed, and by doing this inferscommunication (Luhmann 1984. ch. 4).

This three-term structure - observingsystem, observed system, communication -has two more interesting formal properties. Itis asymmetrical and it is bidirectional. Bidir-ectionality means that one can read a commu-nication forwards and backwards. One reads itforwards when one looks at a sequence ofcommunications, at communication as anongoing process in time. On the other hand,one must read it backwards, too, as a commu-nication only begins with the second participantwho understands and in the act of understandingprojects the difference between information and

utterance on the first participant. In this respectany communicative event is retrospective; itdepends on the projection of differences on pastevents. From bidirectionality immediately fol-lows the other formal property: asymmetry.Whereas in action theory action and actor canexchange their roles - the actor producing theaction, the action constituting the actor - thesame is not true in communication theory Onlyin the next step, in the next communicativeevent, can the observed system become anobserver itself and observe the previous observerin assuming that his or her behaviour may beinterpreted as communication and as a reactionto the first communicative event.

One further remark regarding formalproperties of communication. It has alreadybecome obvious that any communicative eventis distributed over at least two participatingsystems. One cannot say that communication isdone by the observer or alternatively is effectedby the observed system. Both are involved, andthe three-component structure of communica-tion (information, utterance, understanding)refers to and includes both of them. This iswell adapted to systems theory and its thesis ofoperationally closed systems. It refutes anyreductive strategy that tries to decompose asystem by reducing its constitutive elements tocausative agents in the environment of thesystem. In this atmosphere of indisputableirreducibility, communication theory is muchbetter adapted to the premises of systems theorythan any action theory might be. Communica-tion theory is clearly incompatible with meth-odological individualism. Systems theory, on theother hand, still has to find its own methodo-logical agenda, for which the distinctionbetween social macro-order as self-organizationand an elementary level characterized bymicrodiversity gives some suggestive hints.9AutopoiesisMy fourth and perhaps most important pointregards a theoretical shift effected in the samebook in which communication theory wasintroduced: autopoiesis (Luhmann 1984). Inthe shift from a cybernetic theory of selectivesystem/environment relations to a Maturana-styie theory of operationally closed systems, agreat number of concepts had to be adapted.Autopoiesis, as defined by Maturana and V?rela,demands a system that produces all its compo-nents via the interaction of these same compo-nents, which are recursively involved in thenetwork of production of components by which



8/10

SystemsTheoryand ActionTheory 11

they themselves were produced (Maturana &V?rela 1980). To prove that a specific system isautopoietic, by this definition, one needs a moreprecise concept of the elements of a system thanone might have had before. In my opinion this isthe proximate cause for Luhmann's change to acommunication theory.10 In Luhmann's writ-ings of the 1970s there is a certain ambiguity inhis designations of the elements of society. Forexample, his standard definition of 'worldsociety' said that this system consists of actionsthat reach towards one another via commu-nications ('Handlungen, die kommunikativ f?rein-ander erreichbar sind', Luhmann 1984:755).One might add that there is a strategic use ofsuch ambiguities in Luhmann's work in dealingwith dilemmata and bifurcation points, atwhich a conceptual alternative does not yetenforce an obvious decision on him. Butautopoiesis does not allow ambiguity in desig-nating the constitutive element of society, whichwas obviously one of the motives behind thedecision between action and communication.

One may furthermore suppose that autop-oiesis favours communication as the element ofsociety. It is not at all simple to imagine adescription of society as an autopoietic systemclosed on the basis of actions as its constitutiveelements. Recursive closure of a system isprobably more easily established for a commu-nication system than for an action system.11Actions are very much individualized. Eachsingle action introduces a discontinuity intosocial process. Either something finishes orsomething new begins. An action is somehowisolated from its antecedents and its conse-quences; therefore it is very difficult to imaginerecursive closure and the production of some-thing from its own products for an actionsystem.12 It is wholly different with communi-cations, where it is much simpler to imaginea continuous flow of communications, recur-sively returning to its somehow modified start-ing point and thereby closing in on itself.

5. Communication theory as description ofmodern societyIn the last part of this paper I extend the list ofarguments, motives and causes for Luhmann'sswitching to communication theory by pointingto aspects of societal change that favourconceiving communication as the elementaryconstituent of society. I discuss these aspects

under two main headings: (1) Information/Information Society and (2) World Society.

information/information societyHow well is sociological theory able to deal withinformation and knowledge processes? RichardEmerson, one of the most interesting exchangetheorists of recent decades, has said thatexchange theory is well adapted to studyingthe flow of resources in social processes, but forstudying the flow of information you needanother theory, for example, symbolic interac-tionism (Emerson 1981). Information transfercannot be reduced to exchange, as the informa-tion is not lost to the person who hands it overto someone else. If the diagnosis of the margin-ality of information in the intellectual core ofindividualist sociology is true, it describes anunsatisfactory state. In this regard again thebalance sheet of communication theory looksmore promising. Whichever formulation ofcommunication theory one chooses, the con-cept of information is always a strategic part ofit. The unidirectional flow of information frompoint to point; a diffusion process; an epidemio-logica! process of information dispersal; all areclearly analysable phenomena in any commu-nication theory. On the other hand, a pairwisecoupling of two communications may wellprove to be exchange, or it may exhibit theproperties of social conflict. Thus, communica-tion theory is universal in its ability toreconstruct the core concepts of alternativesociological approaches. Finally, in Niklas Luh-mann's communication theory, information andaction are included as indispensable componentsof any single act of communication (see esp.Luhmann 1984, ch. 4). As far as I can see, thereare no categorical exclusions inherent to theconcept of communication.One may resume this point in terms of thesociology of knowledge or the sociology ofsociological knowledge. Then one might classifyaction theory as an intellectual phenomenongermane to industrial society. Its cognitive focusis on processes of producing and processinggoods and resources, and on exchanging thegoods and resources produced. Communication,theory can then be classified as a kind ofsociology adequate to information or knowledgesocieties.13 Sociological paradigm shifts thuswould be perceived to reflect the societaltransformations which the same theories tryto understand.



9/10


World societyThe last point is related to the previous one. Itrefers to world society. I already quoted Luh-mann's early conception of world society devel-oped when he was still an action theorist. Thisconception said: 'World society is the systemwhich consists of all actions that can reachtowards one another via communications'.14This definition already pointed in its inconclu-sive parallelization of actions and communica-tions to the difficulty of formulating a concept ofmacrosociety in action theoretical terms. Onemay illustrate this in looking at one of thefunction systems of modern society. I take theglobal system of science as an example.Science studies have been dominated in thelast ten or fifteen years by a school for which onename is 'laboratory studies'.15 That is, manyempirical studies have been microsociologies ofconcrete places of scientific research, mainlylaboratories in natural science and the smalleror larger experiments which inhabit them. Theimplicit or explicit social theory of these studieswas mostly action theory. This action theoreti-cal tendency was supported by processes ofscientific research often being confrontationswith nature, other social actors only being ofindirect relevance. Such lone 'conversations'with nature obviously are not communicativeacts. There were, of course, many valuableinsights in laboratory studies, but no socio-logical view of the global system of science wasever articulated. In my view the reason for thisfailure is that the global system of science canonly be identified via communications (Stich-weh 1987, 1990). Global processes of generat-ing scientific hypotheses, of validating andfalsifying theories, of informing about researchfindings, are communication processes that arevery selective about which action events (i.e.research acts) in science come to light. Manyresearch processes in science are never docu-mented in print, never reported about atconferences or otherwise made public. But theglobal system of science - and I think it is one ofthe few undisputed global systems - consistsonly of these communicative events and theresearch actions they refer to via attributionprocesses (cf. Stich weh 1996). I assert that amacrosociology of science must be written viacommunication theory.I take this example as a paradigm. Mythesis says once more that communicationtheory as a foundation of sociology reflectssocietal change. Not only is world society in onerelevant respect brought about by communica-

tion technologies (cf. L?bbe 1996; Stichweh1999). World society is also an unforeseensocietal circumstance, which disprivilegesaction theories. This does not mean that it isimpossible to analyse world society via actiontheory. One thinks of Norbert Elias, who speaksof the prolongation of action chains as a specifictrait of global civilization (Elias 1969). But suchan action theory of globalization may have thedisadvantage - again found in the writings ofElias - that it describes the global circumstanceas mainly consisting of unintended effects. Thatpoints to the discontinuity that separatesactions from their consequences. If we onlywait long enough, most structures in worldsociety will be explained as resulting fromunintended effects. Again, my hypothesis isthat a continuous modelling of change pro-cesses and a systematic description of worldsociety are only possible in terms of commu-nication theory.

First version received May 1999Final version accepted August 1999

Notes1 See fora recentstatementNolte (1999).2 Cf.Coleman(1994:166): 'Rationalchoice theory is not

theory designedto account for action, despiteits name. It istheory designedto account for the functioningof social andeconomicsystems'.3 Parsons(1937. esp.pp. 731-737. The ActionFrameofReference'and pp. 737-748. 'Systems of Action and TheirUnits').4 Forthis last statementsee Parsons 1978).5 Collectedn Soziologische ufkl?rung and 2 (Luhmann1970. 1975).6 Therepresentativeext is Luhmann 1978b).7 This descriptionis. of course, true for biological andphysicalsystems,too. whichmeans that a generalized electiontheoryfunctionsas the basis of generalsystemstheory.8 Cf. White (1992:3): 'Social action is induced beforeactors, who derivefromthe action andneed not be persons'.9 Luhmann 1997): some interestingremarks n Hodgson(1998).10For the distinction between proximate and ultimatecausation, see Mayr 1983); cf. Durham(1991:36-37).11Cf. heopeningpassagesof thechapter Kommunikationund Handlung' 19Iff.) in Luhmann 1984).12Cf. very interestingdiscussion in Tyrell (1998. esp.115??).13On information society in sociological theory, seeStichweh(1998b).14Cf.the definitionLuhmanngives in a dictionaryentryfrom 1973: Das umfassendsteSystem menschlichenZusam-menlebens (Gesellschaft)nur welteinheitlichgebildetwerdenkann, nachdem alle Menschen f?reinander kommunikativerreichbar ind und durchFolgenihrerHandlungenbetroffenwerden' Luhmann1973:755).15Seeforan overviewKnorr-Cetina1995).



10/10


ReferencesBateson.G.1973. Steps oanEcology fMind.Collectedssays nAnthropology,sychiatry,EvolutionndEpistemologyondon:PaladinBooks.Clark.J.(ed.) 1996. James . Coleman. ondon/Washington.DC:FalmerPress.Coleman. J. S. 1986. Social Theory,Social Research,and aTheoryof Action. AmencanJournal f Sociology, 1. 1309-1335.Coleman,J. S. 1990. Foundationsf SocialTheoryCambridge,MA:HarvardUniversityPress.Coleman, .S. 1994. A RationalChoicePerspectiven EconomicSociology. nN. J.Smelser& R.Swedberg eds.).TheHandbookofEconomic ociologypp.166-180. Princeton,NJ/NewYork:PrincetonUniversityPress/RussellSageFoundation.Craig.R. T. 1999. CommunicationTheoryas a Field. Commu-nicationTheory, . 119-161.Durham,W. H. 1991. Coevolution. enes,Culture, nd HumanDiversityStanford:StanfordUniversityPress,Elias. ?. [1969] 1993-94. ?ber den Proze?der Zivilisation.Soziogenetischendpsychogenetischentersuchungen.ol. 1-2.Frankfurt .M.:Suhrkamp.Emerson.R. M. 1981. SocialExchangeTheory. n M.Rosenberg& R. H. Turner(eds.).SocialPsychology: ociological erspec-tives,pp. 30-65. New York:BasicBooks.Fararo. T. J. 1996. Foundational Problems in TheoreticalSociology.In J.Clark(ed.), JamesS. Coleman, p. 263-384.London/Washington,DC:FalmerPress.Hodgson.G. M. 1998. TheApproachof InstitutionalEconomics.Journal fEconomicLiterature,6, 166-192.Knorr-Cetina,K. 1995. Laborstudien.Der kultursoziologischeAnsatz n der Wissenschaftsforschung.n R.Martinsen ed.).DasAugederWissenschaftZurEmergent onRealit?t, p.101-135. Baden-Baden:Nomos.L?bbe.H. 1996. Netzverdichtung.ZurPhilosophieindustrie-gesellschaftlicherEntwicklungen.Zeitschrift?rphilosophischeForschung, 0. 133-150.Luhmann. N. 1970. SoziologischeAufkl?rung1. Opladen:WestdeutscherVerlag.Luhmann.N. 1973, Weltgesellschaft.n W.Fuchs,R.Klima,R.Lautmann.0. Rammstedt& H. Wienold (eds.).Lexikon urSoziologie, . 755. Opladen:WestdeutscherVerlag.Luhmann, N. 1975. SoziologischeAufkl?rung . Aufs?tze urTheone erGesellschaft. pladen:WestdeutscherVerlag.Luhmann.N. [1978a] 1981. Handlungstheorieund System-theorie. In SoziologischeAufkl?rung3. Soziales System,Gesellschaft, rganisation,p. 50-66. Opladen:WestdeutscherVerlag.Luhmann. N. [1978b] 1981. Erleben und Handeln. InSoziologischeAufkl?rung3. Soziales System, Gesellschaft,Organisation,p. 67-80. Opladen:WestdeutscherVerlag.Luhmann. N. [1979] 1981. Zeit und Handlung - EinevergesseneTheorie. In SoziologischeAufkl?rung . SozialesSystem,Gesellschaft,Organisation, p. 101-125. Opladen:WestdeutscherVerlag.Luhmann,N. 1984. SozialeSysteme:Grundn? inerallgemeinenTheorie. rankfurt .M.:Suhrkamp.Luhmann,N. 1997. Selbstorganisation ndMikrodiversit?t:urWissenssoziologie es neuzeitlichenIndividualismus.SozialeSysteme,3. 2 3-32.Maturana,H. &V?rela.F.J.1980. AutopoiesisndCognition: heRealizationf theLiving.Dordrecht/Boston:Reidel.

Mayr.E. 1983. TheGrowth fBiological hought. ambridge,MA:Belknap.M?ller.H.- P. & Schmid.M. (eds.) 1998. Norm,Herrschaft ndVertrauen.Beitr?gezu James S. ColemansGrundlagen erSozialtheorie. pladen/Wiesbaden:WestdeutscherVerlag.Nolte, H. 1999. Ann?herungen zwischen Handlungstheorienund Systemtheorien.Ein Review ?ber einige Integration-strends.Zeitschriftur Soziologie,8, 93-113.Parsons,T. 1937. TheStructure fSocialAction.New York:FreePress.Parsons,T. [1968] 1977. Social Systems.InT. Parsons.SocialSystemsand heEvolution fActionTheory,p. 177-203. NewYork:Free Press.Parsons,T. 1978. ActionTheory ndtheHumanCondition. ewYork:Free Press.Parsons.T. & Shils, E. (eds.) [1951] 1962. TowardA GeneralTheory f Action.New York:Harper& Row.Ruesch.J.&Bateson.G.[1951] 1968. Communication:heSocialMatrixof PsychiatryNew York:Norton.Sacks, H. 1992. Lectureson Conversation. ol. 1-2. Oxford:Blackwell.Searle,J.R. 1969. SpeechActs.Cambridge: ambridgeUniversityPress.Shannon. C. E. 1970. InformationTheory.In EncyclopaediaBritannica,Vol.11, pp.246-249. Chicago:Benton.Shannon, C. E. & Weaver.W. [1949] 1969. TheMathematicalTheoryof Communication. rbana, IL:Universityof IllinoisPress.

Stichweh, R. 1987. Die Autopoiesis der Wissenschaft. In R.Stichweh. Wissenschaft,niversit?t, rofessionen.oziologischeAnalysen, p. 52-83. Frankfurt .M.:Suhrkamp1994.Stichweh, R. 1990. Self-Organization nd Autopoiesis n theDevelopmentof ModernScience.InW Krohn.G. K?ppers&H. Nowotny (eds.), Self-organization.ortraitof a ScientificRevolution, p. 195-207. Dordrecht:Kluwer.Stichweh. R. 1996. Science in the System of WorldSociety.SocialScience nformation,5, 327-340.Stichweh,R. 1998a. Generalisierter ausch und Kommunika-tionsmedienbeiJamesColeman. nH.-P.M?ller& M. Schmid(eds.).Norm,Herrschaft ndVertrauen,p.92-102. Opladen/Wiesbaden:WestdeutscherVerlag.Stichweh, R. 1998b. Die Soziologie und die Informationsge-sellschaft.InJ.Friedrichs,M. R.Lepsius& K. U.Mayer eds.).DieDiagnosef?higkeiterSoziologie= Sonderheft38. K?lnerZeitschrift ?r Soziologie und Sozialpsychologie),pp. 43 3-443. Opladen:WestdeutscherVerlag.Stichweh,R. 1999. Zur Geneseder Weltgesellschaft Innova-tionen und Mechanismen.In M. Bauschulte,V. Krech & H.Landweer eds.), 'Wege Bilder Spiele. Festschrift um 60.Geburts tag von J?rgen Frese, pp. 289-302. Bielefeld:Aisthesis.

Tyrell. H. 1998. Handeln, Religion und KommunikationBegriffsgeschichtlichend systematische?berlegungen. InH. Tyrell, V. Krech & H. Knoblauch (eds.), ReligionalsKommunikation,p.83-134. W?rzburg:Ergon.Watzlawick. .,Beavin.J.H. &Jackson.D. D. 1967. PragmaticsfHuman Communication. Study of interactionalPatterns,Pathologiesnd Paradoxes. ew York:Norton.White.H. C. 1992. Identityand Control:A StructuralTheory fSocialAction.Princeton.NJ:PrincetonUniversityPress.

Thi t t d l d d M 17 D 2012 03 09 52 AM
http://www.jstor.org/page/info/about/policies/terms.jsp

Documents

4- Systems Theory as an Alternative to Action Theory the Rise of Stichweh