Toward a Theory of Social Conflict Ralf Dahrendorf the Journal Of

Toward a Theory of Social Conflict

Ralf Dahrendorf

The Journal of Conflict Resolution, Vol. 2, No. 2. (Jun., 1958), pp. 170-183.

Stable URL: http://links.jstor.org/sici?sici=0022-0027%28195806%292%3A2%3C170%3ATATOSC%3E2.0.CO%3B2-5

The Journal of Conflict Resolution is currently published by Sage Publications, Inc..

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/sage.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Fri Jul 13 13:50:49 2007 Toward a theory of social conflict1

RALF DAHRENDORF Akademie fiir Gemeinwirtschaft, Hamburg, Germ,

sons in 1937 established a certain conver- After an interval of almost fifty years, a gence in the sociological theories of Alfrecl theme has reappeared in sociology which hfarshall, fimile Durkheim, Vilfredo Pareto, has determined the origin of that discipline and Max Weher,* he no longer had in mind more than any other subject area. Fronl an analysi~of social conflict; his was an at- Mzrx and Cointe to Simmel and Sorel, social tempt to solve the problem of integration conflict, especially revolutions, was one of of so-called "social systems" by an organon the central themes in social research. The of interrelated categories. The new question same is true of many early Anglo-Saxon was now "What holds societies together?"- sociologists (although in their work the no longer "What drives them on?" The in- problem of revolution has been character- fluence of the Parsonian posing of the ques- istically somewhat neglected), for ex?mple, tion on the more recent sociology (and by the Webbs in England, Surnner in the no means only on American sociology) can United States. However, when Talcott Par- be hardly overrated. Thus it is possible that the revival of the study of social conflict in ----..---....------.....------...--~~.~.~~~~~~~~~~~~ 1 This paper was translated by Anatol Rapa- the last decades appears to many not so port, Mental Health Research Unit, University much a continuation of traditional research of hlichigan. paths as a new thematic discovery-an in- The following presentation is an attempt to stance of dialectic irony in the development depict in a systematic form ths fundamental of science. ideas of nly book Soziale Klossen uncl Klassen- confilkt it1 der inclustriellen Gesellschaft (Stutt- At this time, approaches toward a ~77s- grrt, 1957). Iiowever, the presentation departs tematic study of social conflict are still rela- significantly in its organization and thematic tively isolated, compared with the innumer- scope from that given in my book: ( 1) whereas able worlcs on social stratification or on the book binds together theoretical consider- st~uctureand fuilction of specific institutions, ations and empirical analysis, the present exposition is essentially limited to the theoretical as- organizations, and societies. Still the thesis pects; (2) whereas in the book I have developed of a revival of the study of social conflict the tlleoretical orientations in a critical dialogue can be justified with regard to the works with other authors, particularly with Marx, the of Aion, Philip, Brinton, Kerr, Coser, Brink- prescrltation in the following exposition is sys- mann, Geiger, Gluckmann, and others,"~ tematic. It need hardly be elaborated that n~uch of what is expressly developed in the book could (New York, be only formally treated here and often with Wf. Structure of Social Action dogmatic brevity. Nevertheless, it may be noted 1937; 2d ed., Glencoe, 1949). tllat the present exposition, especially in the "aymond Aron, "Social Structure and the first and fourth sections, contains in certain re- Kuling Class," in Class Status and Power, ed. spects formulations beyond the scope of the fieinhard Bendix and Seymour 'fartin Lipset book. ( London, 1954); Andre Philip, Le Socialisme

CONFLICT RESOLUTION VOLUME 11 NUMBER 2 TOWARD A THEORY OF SOCIAL CONFLICT 171 well as an attempt to determine a systematic attention for the moment-for reasons which locus and a specific framework for a theory will presentl;~be given-to endogenous con- of conflict in sociological analysis. flicts. Then further subdivisions are directly perceived: slaves versus freemen in Rome, TYPES AND VARIETIES OF Negroes versus whites in the United States, SOCIAL CONFLICT Protestants versus Catholics in the Nether- To begin with a commonplace observa- lands, Flemings versus Walloons in Belgium, tion: The problem of conflict is no less coin- Conservatives versus Laborites in England, plex than thzt of integriition of societies. We u!licns versus employers in many countries. now know that the attempt to reduce all All these are opposing groups in well-knovrm actually occurring conflicts among social conflicts. Perhaps each of these examples groups to a common principle, say that of does not fall into a separate category; but classes, is sterile. It leads either to empty certainly they cannot all be subsumed under gencralizatic~is (such as "Every society ex- a single type of social conflict. Whatever perie:lces social conflicts") or to empirically criterion one chooses for classification-for unjustifiable oversimplfications (such as exnniple, the objects of contention, the "The history of all societies so far has been structural origin of the conflictiug groups, a history of class struggles"). It seems ad- the forms of conflict-several distinct types visable, first, to sort out and to classify the result. problems which are conceived under the general heading of "social conflict." Even THE LIMITS AND GOALS OF A THEORY a superficial reflection leads to the distinc- OF SOCIAL CONFLICT tion of a series of types. An ideal sociology cannot, in principle, There are wars, and there are conflicts exclude any of these categories and types zrnong political parties-evidently two dif- of cor~fiictfrom analysis. Nevertheless, the ferent kinds of struggle. LVith regnrd to a types mentioned do not all have the same given society, A, one could say there are i~nporbancefor sociological analysis. A brief exogenozrs conflicts brought upon or into A recollection of the intent of a sociological from the outside, and there arc enclogenous theory of conflict reveals that the contribu- conflicts generated within A. Of these two tion of sociology to the understailding of categories, which, at least analytically, can conflict (as \\-ell as the contribution of con- be re1:itively precisely distinguished, there flict to the social process) is in specific in- are agai~lseveral types. Let us confine our stances grcater in solne cases than in others. The intent of a sociological theory of con- trcdli (Paris, 1957); Crane Brinton, Tl~eAnatomy flict is to overcome the predo~ninatingly of Rccolution (2d ed.; New York, 1952); Clark nrbitr:.ry nature of unexplained historical Kerr, ''Ind~lstrial Collflict and Its hlediatiotl," events by deriving these events froin social Anz~riconJo~~TI~(I/of Sciciology, Vol. XL, No. 3 structu;.al elements-in other words, to ex- ( No\ ember, 1054) ; Lewis Coscr, Tl~eFunctions of Social Corlflict ( London, 1956 ), and "Social pl.iiil certain processes by prognostic co:l- ConYict and Social Cilatlge," British Journal of nections. Certainly it is important to describe Si:c;olog!l, i70i.1'111, No. 3 (September, 1957); the cdilflict bet\i,ec,n workers and einplo?,ers C:irl Ijri~ikm,lnn,Soziologische Tlceorie der Re~o- purely as such; but it is more imporlunt to llltioli (Tihingcn, 1948); Theodor Geiger, Klas- p:.oduce a proof that s~cha conflict is b'lsed se;~gcsclls~,hcijtill. Sci'cnteI~tie~el(Koln-Hagen, 1949); hfvlax Gluckmann, Cu.~tomaizd Conflict in oil c?:.tain social structural arrangements Africa (London, 1957). and hence is bound to arise xvherever such RALF DAHRENDORF

structural arrangements are given. Thus it is ered from the point of view of their struc- the task of sociology to derive conflicts from tural significance, fall into two groups. On specific social structures and not to relegate the one hand, they point to conflicts which these coilflicts to psychological variables arise only in specific societies on the basis of ("aggressiveness") or to descriptive-his- special historical conditions (Negroes or torical ones (the influx of Negroes into the whites in the United States, Protestants ver- United States) or to chance. sus Catholics in the Netherlands; Flemings In the sense of strict sociological analysis, versus Walloons in Belgium); on the other conflicts can be considered explained if they hand, however, there are conflicts which can be shown to arise from the structure of can be understood as expressions of general social positions indepcvidently of the orien- structural features of societies, or of societies tation of populations and of historical [lei ex in the same stage of development (Con- nlacl~ina.This is necessarily a very abstract servatives versus Laborites in England; formulation; instead of elaborating it, it may unioiis versus employers' associations) .Wer- be advisable to illustrate its meaning by the tainly in both cases an analysis leading to following treatment of a form of social con- generalization is possible: a theory of minor- flict. First, however, let us draw a conse- ity or religious conflict is as meaningful as quence of this foi~nulationwhich will help that of class conflict. Nevertheless, their to make our problem more precise. respective weights within a general theory Since the recognition of the inadequacy of society are evidently distinguishable. It of the Marxist-Leninist theory of imperial- is not surprising that the "classical" theory ism, the explanation of exogenous conflicts of conflict-I mean here primarily the class on the basis of the structuie of a given theory of conflict-has, above all, called at- society is once again an open problem, the tention to such social frictions which can be treatment of which has scarcely begun. It derived from the structure of societies inde- seems, moreover, that the explanation of pendently of stlvcturally incidental histori- exogenous conflicts4 by the tools of socio- cal data. logical structure analysis is possible only in The follo\ving approaches toward a the- a metaphorical sense-namely, only in case ory of conflict also relate themselves to con- the entire societies (or less comprehensive flicts based on structure. So far, we are by "social systems") are taken to be the units no means considering a general theory of of n new structure,Qhat is, when C is ana- lyzed in te~nlsof the structure of its ele- Valcott Parsons and the political scientist ments A and B without consideration of the David Easton (The Political System [New York, inner structure of A and B. On these grouncis 19531) are currently working on an attempt to it seems sensible to exclude exogenous con- analyze international conflicts by means of a inodel in which entire societies, such as the flict for the tlnle being from a theory of United States and the U.S.S.K., appear as ele- social conflicts. nleuts and are treated as if they had no inner On the other hand, the above-mentioned structure. This procedure is methodologically en- examples of endogenous conflict, if consid- tirely legitimate. It remains to be seen what results it can achieve and how it may he connected Werecall here that a conflict which, from to the analysis of intrasocietal conflicts. the point of view of Society A, appears as exoge- G 'rile conf1ic.t between free men ancl slaves in nous is represented froxn another point of view ancient Ilome possibly belongs to this second as a conflict between two societies or systems, group, although not on the same level of gener- A and B. ality.

CONFLICT RESOLUTION VOLUME II NUMBER 2 TOWARD A THEORY OF SOCIAL CONFLICT 173 social conflict, although I would undertake representation of society as a social sys- to defend the assertion that we are dealing tem is the last point of reference. We are here with one of the most important, if not therefore faced with the task of determin- the most important, type of social conflict. ing certain associations, institutions, or proc- However important as problems of social esses within this balanced ~vhole, that is conflict St. Bartholome~v's Night, Crvseal -in Merton's definition-of determining the Night, and Little Rock may be, the French intentional or uiliilte~ltionalconsequences of Revolution and the British General Sirike of these associations for the functioning and 1926 and June 17, 1953, seem to me more the preservation of the system. In this way, germane for structural analjsis. To put it we come to contentions such as "the educa- less dramatically, the sociological theory of tional system functioils as a mechanism of conflict would do well to confine itself for assigning social positions," or "religion func- the time being to an explanation of the iric- tions as an agent of integrating dominant tions between the rulers and the mled in values." The majority of sociological investi- given social structural organizations. gations in the last years moves in this area of analysis. However, such an approach leads to dif- The explanation of motion requires two ficulties, if we put a question of a different separate attacks. We must know the point sort. What was the function of the English of departure and the direction of motion or, trade unions in the General Strike of 1926? better yet, the moving force. No theory of What was the function of the construction social change or of conflict can forego the worker in Stalin Allee on June 17, 1953? description of the structural entity which Without doubt, it can be argued in many undergoes change or within which con- cases that militant trade unions or oppo- flicts occur. Such a description is offered by sition political groups and parties also con- the integration theory of society. However, tribute to the functioning of the existing it is erroneous to assume that a description system.5 But even when this is the case- of how the elements of a structure are put and in the two cases cited it would be dif- together into a stable whole offers, as such, ficult to establish this-such a conclusion a point of departure for a structural anal- viould say little about the role of the group ysis of conflict and change. So far, the in question. Moreover, it is clear that the claim of the so-called "structural-functional" intentional, as well as the unintentional, theory of modern sociology to the status of effects of such oppositioilal groups are in a general theory of society is demonstrably the contribution toward an abolition or de- unjustified. struction of the existing system. The structural-functional position has n comfortable TOWARD A CRITIQUE OF A STRUC- TURAL-FUNCTIONAL THEORY 7 David Lockwood, "Sorne Notes on 'The So- cial System,' " British Joo~izalof Sociology, Vol. This critique has been led in recent VII, No. 2 ( 1956). Altllough Lockwood's argu- times repeatedly, most effectively by D. ment leads to the same conclusion, it proceeds Lockwood.7 It is based on a relatively sim- somewhat differently (cf. my Social Classes and ple argument. As long as we orient our the Class Conflict, pp. 159 ff. ). analysis tox~ardthe question as to how the 8 This aspect of social conflict is, in fact, central in the analysis of Lewis Coser (continuing elements of a society are combined into that of Simmel) in his work on the functions of a co-ordinated functioning whole, then the social conflict (cf. n. 3). 174 RALF DANRENDORF label for such cases: they are "dysfunctional" takes the diametrically opposite position on olganizations, institutions, or processes. But all the four points above: this designation again tclls us less than 1. Every society is subjected at every nothing. It not only fails to explain the moment to change: social change is ubiqui- place of these things in the process but tous. actually hinders such explanation by a ter- 2. Every society experiences at every minology which seems to be congruent with moment social conflict: social conflict is the s>stem but which, up;tn closer exami- ubiquitous. nation, reveals itself as a residual category. 3. Every element in a society contributes TVhatever does not fit is cor~juiedout of the to its change. world by word magic. 4. Every society rests on constraint of In every science, residual categories are some of its members by others. a fruitful point of departure for new de- The remarkable nature of our vantage velopiiie:lts. It seems to me that a careful point becomes evident when we examine analysjs of problems which the term "dys- the two groups of postulates with respect functior;" hides in the structural-fu!lctional to their truth content, that is, if we ask theory nutonlatically puts us on the trace of ourselves which of the two models promises a l~leaningful sociological theory of con- greater utility for cognition of reality. It flict, At the same time, it offers a remarkable appears that the juxtaposed pairs of postu- vantage point associated with an attempt lates are in no way mutually exclusive with of a scientific analysis of society. respect to social reality. It is impossible to decide by an empirical investigation .cvhich TWO hIODELS OF SOCIETY of the two models is more nearly correct; If we extrapolate the analytical ap- the postulates are not hypotheses. Moreover, pronches of the structural-functional theory it seems meaningful to say that both models somewhat beyond their boundaries and in- are in a certain sense valid and analytically vestigate their implicit postulates, we can fruitful. Stability and change, integration co:lstruct a model of society which lies at and conflict, function and "dysfunction," the base of this theory and determines its perspectives. The essential elements of this 9 There is much controversy over this impli- societal model are these: cation of the structural-functional approach. Most functionalists deny that they make such an 1. Every society is a relatively persisting assunytion. Indeed, assertions to the contrary configuration of elements." are found in the works of Parsons, Merton, and 2. Every society is a well-integrated con- others. Nevertheless, it can be shown that these figuration of elements. assertions are, from the point of view of struc- 3. Every element in a society contributes tural-functional theory, mere declarations. The notion of equilibrium and the concept of a sys- to its functioning. tem would have little sense if they did not make 4. Every society rests on the consensus the assumption of stability of societies. However, of its members. two limitations are to be observed: ( 1)we have It should be clear that a theory based on to do here (also in the i~nplicationswhich fol- this model does not lend itself to the ex- low) not with a metaphysical postulate but rather wit11 an assumption made for the purpose of p!nnation, not even the description, of the analysis; and (2)stability does not mean statics phellomena of social conflict and change. in the sense of complete absence of processes For this purpose, one needs a model which within tlie "system."

CONFLICT RESOLUTION VOLUME 11 NUMBER 2 TOWARD A THEORY OF SOCIAL CONFLICT 175 consensus and constraint are, it would seem, the same foci but is open in many direc- hvo equally valid aspects of every imagi- tions and appears as a tension field of the nable society. They are dialectically sepa- determining forces. rated and are exhaustive only in combina- tion as a description of the social problems. THE TASKS OF A THEORY OF Possibly a more general theory of society SOCIAL CONFLICT may be thought of which lifts the equi- The double aspect of society and the validity of both models, the coexistence of dialectics of the two types of sociological the uncombinable, onto a higher level of theory are in themselves a most fruitful generality. As long as we do not have such object of reflection. Nevertheless, another a theory, we must content ourselves with the problem seems to be more urgent. The finding that society presents a double as- theory of social integration has receiitly de- pect to the sociological understanding, each veloped to a flourishing state as the struc- no better, no more valid, than the other. tural-functional approach in eihnclogy and It follows that the criticism of the unap- sociology. Our theory of coi~flict,hcwever, plicability of the structural-functional theory is still in a very rudimentary state. It is for the analysis of conflict is directed only an approach based on postulating ubiqui- against a claim of generality of this theory tous social change and social conflict, the but leaves untouched its competence with "dysfunctionality" of all the elements of so- respect to the problem of integration. It cial structure, and the constraining char- follotvs, on the other hand, also that the acter of social unity. Our considerations put theory of conflict and change is not a gen- us in a position to formulate some require- eral theory. Comparisons between natural ments of such a theory: and social sciences always carry the danger 1. It should be a scientific thecry (as is of misunderstanding. However, it may be the theory of social integration), that is, maintained, without attributing to this anal- it should be formulated tvith reference to a ogy more than a logical meaning, that the plausible and demonstrable explanation of situation of the sociologists is not unlike empirical phenomena. that of the physicists with respect to the 2. The elements of the theory should not theory of light. Just as the physicists can contradict the conflict model of society. solve certain problems only by assuming 3. The categories emp1o)-ed should, the wave character of light and others, on whenever possible, agree with those of the the contrary, only by assuming a corpus- integration theory or at least correspond to cular or quantum theory, so there are prob- them. lems of sociology which can be adequately 4. A conflict theory should enable us to attacked o111y with an integration theory and derive social conflicts from structural others which require a conflict theory for a arrangements and thus show tl~csc:coilflicts meaningful analysis. Both theories can work systematically generated. extensively with the same categories, but 5. It should account both for the IIILI~- they emphasize different aspects. While the tiplicity of forms of conflict and for their integration theory likens a society to an degrees of intensity. ellipse, a rounded entity which incloses all The last goal of a social theor!, is the of its elements, conflict theory sees society explanation of social change. The integra- rather as a h1,perbola which, it is true, has tion theory gives us a tool for determinii~s 176 RALF DAHRENDORF

the point of departure of the process. To which prevail within certain units of social find the locus of the forces which drive the organization. For these units I will use Max process and social change is the task of a Weber's concept of "imperatively co-ordi- theory of conflict. It must develop a model nated group." The thesis is not new; it is which makes understandable the structural found (however often with important modi- origin of social conflict. This seems possible fications) in the formulation of many social only if we understand conflicts as struggles scientists before and after Marx. But we among social groups, that is, if we make shall make no attempt to trace the history our task precise to the extent that it reduces of this thesis. to the structural analysis of conflicting groups. Under this supposition three ques- AUTHORITY AND AUTHORITY tions come especially to the forefront, which STRUCTURES conflict theory must answer: The concepts of power and authority 1. How do conflicting groups arise from are very complex ones. Whoever uses them the structure of society? is likely to be accused of lack of precision 2. What forms can the struggles among and of clarity to the extent that he tries such groups assume? to define them "exhaustively." Is the influ- 3. How does the conflict among such ence of a father on his children, the influ- groups effect a change in the social struc- ence of an industrial combine on the gov- tures? ernment, or the influence of a demagogue on his followers an instance of an authority relation? Here, as in most other cases, it is Wherever men live together and lay basically not a question of a definition but foundations of forms of social organization, rather a question of an "operational defini- there are positions whose occupants have tion," as it is often called today: a method powers of command in certain contexts and of determination which allo\vs us to identify over certain positions, and there are other as such the state of affairs when we are positions whose occupants are subjected to actually confronted with it. However, for such corn~nandc;.The distinction between the purpose of analysis and identification, "up" and "don7n"-or, as the English say, Weber's determination of authority is sufFi- "Them" and "Usm-is one of the funda- cient: "The likelihood that a command of a mental experiences of most men in society,lO certain content will be obe),ed by gi17cn and, moreover, it appears that this distinc- persons7'l' This determination contains the tion is intimately connected nith unequal following elements: distribution of power. The main thesis of 1. Authority denotes a relation of supra- the following attempt to construct a model and subordination. for the structural analysis of coilflict is that 2. The supra-ordinated side prescribes we should seek the structural origin of so- to the subordinated one certain behavior cial conflict in the dominance relations in the form of a command or a prohibition. 3. The supra-ordinated side has the right 10 Empirical corroborations for these general- to make such prescriptions; authority is a izations are found in two significant publications legitimate relation of supra- and subordina- of last year: Heinrich Popitz et a!., Das Gesell- schuftsbild des Arbeiters ("Thc Worker's Image 11 Xlax Txleber, "Wirtschaft und Gesellschaft," of Society") (Tubingen, 1957) ; Richard Hog- in Gruncliiss cler Sozialiikonomik, I11 (3d ed.; gart, The Uses of Literacy (London, 1957). Tubingen, 1947), 28.

CONFLICT EFSOLUTION VOLUME 11 NUMBER 2 TOWARD A THEORY OF SOCIAL CONFLICT 177

tion; authority is not based on personal or cnly general ("civil") basic rights and those situational chance effects but rather on an which have authority rights over the former. expectation associated with social position. In contrast to prestige and income, a con- 4. The right of authority is limited to tinuum of gradual transition cannot be con- certain contents and to specific persons. structed for the distribution of authority. 5. Failure to obey the prescriptions is Rather, there is a clear dichotomy. Every sanctioned; a legal system (or a system of position in an imperatively co-ordinated quasi-legal customs) guards the effective- group can be recognized as belonging to ness of authority. one \vho dominates or one who is domi- This determination of authority makes nated. Sometimes, in view of the bureau- possible the identification of a cabinet min- cratic large-scale organization of modern ister, an employer, and a party secretary as societies-under the influence of the state occupants of authority positions-in con- -this assu~nptionmay at first sight seem trast to an industrial syndicate or a dema- problematic. However, a sharper analysis gogue, neither of which satisfies condition leaves no doubt that here also the split 3 above.12 into the dominating and dominated is valid, It is not the intention of our "definition" even though in reality a considerable meas- of authority to solve all analytical and em- ure of differentiation is discernible among pirical problems of this category.13 In fact, those in the dominating group.14 the very first step of our model leads us deep into these problems: in each impera- THE COhTFLICT-THEORY ?\IODEL tively co-ordinated group, two aggregates The dichotomy of social roles within im- can be distinguished: those which have peratively co-ordinated groups,lg the divi- sion into positive and negative dominance 1"llis third condition, that of legitimacy, de- roles, is a fact of social structure. If and in- notes the distinction between power (as an ac- sofar as social conflicts can be referred to tual command relationship) and authority (cf. Weber's "Definitionen," op. cit.). this factual situation, they are structurally explained. The model of analysis of social 1'3 Thus it is clear that thc uheilomenon of authority is here deliberately treated unilaterally. 14T11e position of authority of the bureaucrat The double aspect of society may be illustrated was already of concern to hfax R7eber and to in this category, as in practically any other. Inte- many sociologists since. Here there seems to be gration theory, too, treats of authority. However, indeed a differenti'ltion of authority. However, this theory emphasizes not the polemical, con- it is a differentiation of a special kind. In modern fiict-generating aspect of this social relation but, . - bureaucratic administration. the exercise of au- on the contrary, the integrative, unifying aspect. thority has undergone to a certain degree a di- Parsons is doubtless right when he says that au- vision of labor; hence the multiplicity of po- thority "is the capacity to mobilize the resources sitions, distinguishable by the number of "as- of the society for the attainment of goals for ~vhicha gcneral 'public' comnlitment has been signable pcrsons" and the scope of "specific content" to which authority pi\-ileges are attached. made, or may be n1:ide. It is mobilization, above In the sense of our analysis, there can be no all, of the action of persons and groups, which is birrrlirig on them by virtue of thcir position in doubt tllcit the entire bureaucracy belongs (at times! ) to the ruling side. societv" ("The Distribution of Power in Ameri- , ~ can Society," World Politics, X, No. 1 [October, 15 111 what follows, I shall designate the roles 1'3571, 140). Ho~~,ever,in a way C. Wright hlills, to tvhicll the cspcctation of the exercise of au- 1~110is criticized by Parsons, is also right when thority is attaclled as "positive donlinance roles" 11e emphasizes, as we do, the "prcsumpti~~eille- and, con\-ersely, the rolcs \vithout authority privi- gitimacy" and "dysfunctionality" of all autl~ority. legcs as "ncgati\,e dominance roles." RALF DAHRENDORF conflict which is developed against a back- a problem area, to identify the factors perti- ground of an assumption of such a dichot- nent to it, to put them into order-that is, to omy involves the following steps: propose fruitful questions-and at the same 1. In every imperatively co-ordinated time to fix precisely their analytical focus. group, the carriers of positive and negative We have delimited our problem area by dominance roles determine two quasi-groups viewing social conflict as a conflict among with opposite latent interests. We call them groups which emerge from the authority "quasi-groups" because we have to do here structure of social organizations. We have with mere aggregates, not organized units; identified pertinent factors in the conditions we speak of "latent interests," because the of organization, of conflict, and of change. opposition of outlook need not be co~lscious Their order, however, can be expressed on on this level; it may exist only in the form the basis of the model in three functions: of expectations associated with certain posi- interest groups (for example, parties) are a tions. The opposition of interests has here function of conditions of organization if an a quite formal meaning, namely, the expec- imperatively co-ordinated group is given; tation that an interest in the preservatio~lof specific forms of conflict (e.g., parliamen- the status quo is associated with the positive tary debates) are a function of the concli- dominance roles and an interest in the tions of conflict if the interest groups are change of the status quo is associated with given; specific forms of change (e.g., revo- the negative dominance roles. lutions) are a function of the conditions 2. The bearers of positive and negative of change if the conflict among interest dominance roles, that is, the members of the groups is given. Thus the task of the theory opposing quasi-groups, organize themselves of conflict turns out to be to identify the into groups with manifest interests, unless three sets of conditions and to determine as certain empirically variable conditions (the sharply as possible their respective weight- condition of organization) intervene. Inter- ideally, by quantitative measure.16 The est groups, in contrast to quasi-groups, are following remarks are hardly more than a organized entities, such as parties, trade tentative indication of the sorts of variables unions; the manifest interests are formulated in question. programs and ideologies. 3. Interest groups which originate in this EA6PIRICAL CONDITIONS OF manner are in constant conflict concerned SOCIAL CONFLICT with the preservation or change in the status As far as the conditions of organization quo. The form and the intensity of the con- are concerned, three groups of factors come flict are determined by empirically variable to mind. First, we have certain effective conditions (the conditions of conflict). social conditioi3s: for example, the possibi9ity 4. The conflict among interest groups in of communication among the members of the sense of this model leads to changes in the quasi-group and a certain method of the structure of the social relations in ques- recruitment into the quasi-groups. Next tion through changes in the dominance rela- there are certain political conditions which tions. The kind, the speed, and the depth ...... -...... --.-.---...... -..-..-.-..-....---- of this development depend on empirically 16 By this remark is meant ( 1) a mathemnti- variable conditions (the conditions of struc- cal formulation of the functions, (2)a develop- ~nrntof meas~iremcntscales for each of the con- tural change). ditions, and (3)the adjustment of the combined The intent of such a model is to delimit sKllcs to groups of conditions. TOWAliD A 1'IJEORY OF SOCIAL COhlFLlCT 179

must be fulfilled if interest groups are to retically oriented investigations. And in the emerge. Were, above all, a guaranty of free- enlpirical sphere, the systematic association dom of coaliti~nis important. Finally, cer- of factors in such an investigation redirects tain technical conditions must be fulfilled: our questions from a haphazard search for :In organization must have material means, ad Izoc relations in the world of coincidences a founder, a leader, and an ideology. to a meaningful study of specific interde- Under conditions of conflict, two kinds pendencies, whose locus and meaning are are immediately conspicuous: the degree of fixed by a general perspective. By the na- social mobili", of individuals (or of families) ture of the subject, our expositioil up to this and the presence of effective mechailisms point had to remain somewhat abstract in for regulating social confiicts. If we imagine form. n continuum of intensity of social conflict In spite of the tentative nature of the among icterest groups, ranging from demo- above-mentioned. franie of reference, it is cratic debate to civil war, we mny conjec- nevertheless possible to test its resolving ture that the preseuce or absence of social power on some empirical problems. mobility and of regulating mechanisms has considerable illflueace on the position of specific given cc~nflicts on this continuum. Strictly spea!cing, every form of differen- Here, as with the other conditions, the de- tiated social organization may also be deter:nin:.,tion of the exact weights of the f:~- scribed as an imperatively co-ordinated tors is a task of empirical investigA t'1011. group-a state and an industrial enterprise, Finally, a third group of conditions or a chess club and a university, a party and a variables determilxs tlie form and the extent church. Thus, strictly speaking, the theory of social structural changes wl~icharise from of conflict is applicable to all these cases. tlie conflict of interest groups. Probably a Our decision to single out two of tl~eseim- relati\.ely intiinate connection exists between peratively co-ordinated groups-the stnte :he intensity of the conflict and the change, :und the indu~trialenterprise-for purposes thut is, also beineen the conditions of con- of analysis is, in principle, arbitrary, al- flict and of the structural changes. However, though the special empirical meaning of aclditional iactors coi:le into play, such as these two forms of social organizaticn in so- the capacity of the iulers to stay in power called industrialized society certainly needs and the pressure potential of the domii~ated no justification. In its application to the interest group. The sociology of revolutions allalysis of industrial and political conflict, and especially the unwritten sociology of the theory of conflict comes very near to the u~icompleted revolutioi~s should contribute positions of the traditional, especially hlarx- considerably to m-king these factors precise. ist, theory of classes. At the same time, it It need hardly be re-emphasized that becomes evident that the theory of classes is these unsystematic observations can, as such, only a special case of the theory of conflict. llardly lay a foundation of a theory of conflict. Nevertheless, w7e put ourselves in a SOCIAL CONFLICT IN INDUSTRIAL position to ask meaningful questions both on ENTERPRISE the theoretical level and with respect to The approach to a theory of social conflict empirical problems. Each of the conditions t'iken here can be tellingly illustrated i11 the mentioned offers a fruitful object of theo- ex:unple of an industrial enterprise. A11 in- RALF DAHREWDORF dustrial enterprise is, among other things,l7 less, even this illustration is not entirely triv- an imperatively co-ordinated group. It con- ial when \i7e realize two of its implications: tains positions with which are associated an if conflict theory is useful, then it follows expectation and a right of exercising author- that industrial conflict exists regardless of ity ancl other positions whose occupants are whether the managers are owners-entrepre- subjected to authority. There are managers neurs or whether they are agenis elected by of many grades, and there are workers. The bodies of stockholders, or whether they are authority of managers is institutionalized govelnment officials; that is, relationships of and legitimate. It is guaranteed by legalistic ownership in principle do not affect either and quasi-legalistic sanctions (disciplinary the existence or the intensity of industrial fines, demotion, dismissal, etc.). A conflict conflicts. Furthermore, it follows that indus- of (latent) interests between managers and trial conflict is present even if the complete workers is thus structurally unavoidable. system of its regulation has been realized. Therefore, we can formulate the assumption Regulation influences, it is true, the intensity that, from these quasi-groups, interest groups of the conflict, but no mechanism is imag- emerge as soon as the conditions of organi- inable which abolishes conflicts altogether. zation (communication within the quasi- Consequences of this sort suggested by ap- groups, regulated recruiting into the quasi- plications of conflict theory are by no means groups, freedom of coalition, leaders and trivial. In the face of two burning problems ideologies, technical means) are on hand. of sociological analysis, this assertion should The emerging interest groups are employers' be justified, at least in outline. associt:tions and trade unions. The conflict between these interest groups varies in its THE PROBLEM OF CO-DETERMINATION intensity in direct relation to the conditions It is now evident that co-determination in of conflict, especially to the degree of mo- the German coal and steel industry has not bility from one group to another and to the led to the abolition, not even to alleviation, presence of effective mechanisms for regu- of industrial conflict. On the other hand, it lating conflicts (channels for collective bar- is not to be doubted that all the involved gaining, arbitration institutions, etc.) . This groups espected that co-determination would conflict leads-either through negotiations or bring such a result about. How can we ex- through strikes-finally to changes in the plain this discrepancy? Under the assump- structure of industrial organizations and in tions and in the light of conflict theory, an the position of the involved groups. explanation is indeed possible: industrial or- This sort of analysis evidently tells us little ganization is an imperatively co-ordinated that we already do not know; it seems, after group. Social conflicts between the bearers decades of industrial conflict, almost trivial. of positive and negative dominance roles are Therefore, I have expressly designated it as unavoidable in it. Co-determination means, an illustration of conflict theory. Neverthe- above all, the elevation of workers' repre-

17 It should be recalled that the description sentatives into management positions, that of a social organization as an imperatively co- is, a change of certain persons from negative ordinated group is not exhaustive, nor should it to positive dominance positions (Aufsichts- be so. Rather, this description singles out for rat, Arbeitsdirekto~), These changes leave analysis one aspect of social organizations. For the authority structure of industrial organi- this reason, the statement "This social organization is an imperatively co-ordinated group" is not zation as a structure of positions with com- a tautology. mand functions unchanged. Co-determina-

CONFLICT RESOLUTION VOLUME 11 NUMBhR 2 TOWAXD A TIlEORY OF SOCIAL CONFLICT 181 tion has created a new authority position, pants of positive dominance positions and, that of the Arbeitsdirektor; but it has not as such, stand, as a result of structural ne- abolished the contrast between up and cessity, on the other side of the barrier that down, nor could it abolish it. It bypasses the separates up from down. Somewhat pointed- pcssibilities of effective regulation of social ly expressed: not only has co-determination contradictions and thus has neither annihi- failed to make industrial conflict milder, but lated industrial conflict nor contributed to it has at the same time robbed the occupants its regulation. of the negative dominance positions of their Conflict theory allows us to go a step representation, that is, it has blocked a farther and to formulate the assumption that channel of expression of the conflict. There co-determination not only is useless as an is thus the danger that the existing latent instrument for regulating industrial conflicts conflict will create new, completely unregu- but also, in the long run, threatens to lead lated, forms of expression and will assume to a sharpening of such conflicts. In this more radical forms when the representatives connection, I do not have in mind the much- of the workers perceive their task as repre- discussed problem of "loyalty conflict" with sentatives of interests in an unambiguous which the Arbeitsdirektor is faced.18 Struc- and radical manner.19 turally, another fact is more important. The The above analysis, because of its almost Arbeitsdirektor and the Aufsichtsratsz;ertre- superficial brevity, is wide open to many ter in the industries affected by the right of kinds of critical objections. Therefore, we co-determination are defined as the repre- shall break it off at this point, with the as- sentatives of the employees. Their rise to re- sertion that a strict and detailed application sponsible positions appears, therefore, as a of conflict theory to the problem of co-deter- rise of a new group to authority. However, mination enables us to make the indicated this group consists not of the totality of work- assumptions sufficiently precise that they can ers but of workers' representatives. Thus a be subjected to empirical tests. The same situation results that those whose task it is to holds for a second problem of outstanding represent the interests of the occupants of actual importance, which will be briefly ana- negative dominant positions in an industrial lyzed here from the perspective of conflict enterprise have themselves become occu- theory: the problem of conflict and of change in totalitarian states. 1s The social role of the Arbeitsdirektor is complex in the legal formulation of its rights 1Vhe dogmatic brevity of the present analy- and duties. The law prescribes that the Arbeits- sis can give rise to misunderstandings. What is direktor ( 1) is a representative of the employees, meant here is that conflict theory offers a con- or should not be appointed to the Aufsichts~at clusion to the effect that the intensity of social (supervisory council) against the votes of the conflict is at a minimum where the conflict as workers' representatives; (2) shall have equal such is taken seriously and is pursued most ener- rights and duties with the other members of the getically, for example, in United States industry. board of directors of the enterprise. However, Converselv.2, all attemots to erase the lines of the conflict which results from the incompatible conflict by institutions such as co-determination role expectations is less a sociological structural threaten, contrary to their intent, to sharpen con- one than a psychological problem for the indi- flicts. The oft repeated question of Sombart, vidual Arbeitsdirektor. For, structurally, only the "Why is there no socialism in the United States?" latter expectation can hold realistically: the Ar- finds an answer not in the vague notion of the beitsdirelctor is, first and foremost, a member of "American way of life" but in the generally posi- the board of directors, that is, of the manage- tive value attached to conflict in the United States. 182 RALF DAHRENDORF

TI-IE PROBLEM OF TIIE ating social conflicts). In the same way the TOTALITARIAN STATE systematic requirement of discussion with Since June 17, 1953, and with greater the aim of deciding the political "platforms" certainty since the events in Hungary and within and outside the state party may be Poland in the autumn of 1956, we know an effective mechanism of regulation.20 Still that social conflict (and social change!) there seems to be an inherent tendency in have by no means disappeared in the totali- most totalitarian states to isolate socially the tarian states. Conflict theory raises this leadership laycr and to prevent discussions, knowledge to the status of law. The state, that is, to disregard the mechanisms for that is, society in its political aspect, is an regulating conflicts. If this is the case, social imperatively co-ordinated group. There are conflicts threaten to increase in potential in it mere citizens (voters) and occupants intensity and to take on a revolutionary of positions equipped with command oppor- character. tunities. Therefore, political conflict is a From the point of view of conditions of structural fact of society under every im- structural change, this means that political aginable condition. This conflict can assume conflicts in totalitarian states aim more and mild or severe forms; it can even disappear more at sudden replacement of the ruling for limited periods from the field of vision class. The important variable which deter- of a superficial observer; but it cannot be mines the probability of realizing a radical abolished. Now one of the aspects of a to- change- is the resistance of the rulers to the talitarian state is an attempt to suppress the pressures making for change. Perhaps it is opposition, that is, to suppress social conflict. meaningful to make the empirical generali- The question then arises, against the back- zation that this resistance does increase to ground of conflict theory, In which way do a certain degree with increasing pressure, social frictions become manifest under such but then gives way to a relatively speedy circumstances? We can analyze totalitarian dissolution and so promotes change. states from the point of view of conditions Central for the analysis of conflicts in to- of organization of interest groups-that of talitarian states, however, is our third set conflict and of structural change-and hope of conditions (first, as listed in the theory) : to arrive in this way at meaningful expla- the condition of organization. It follows in a way from the "definition" of a totalitarian nations of historical events and to testable state that there are no conditions in it for predictions. Again it is possible here to the organization of opposing interest groups. make only a few indications. More specifically, although the social and Let us begin-for reasons which will soon technical conditions are often present, the become evident-with the conditions of con- political conditions are 1acking;zl there is no flict. The intensity of social conflicts de- freedom of coalition. At this point, the re- pends on the measure of social mobility and on the presence of mechanisms for regulat- 20 I would suspect the significant part of an ing the conflicts. Both mobility and regula- explanation of the remarkable stability of the tion can be present in totalitarian states. U.S.S.R. in arguments of this sort (and not in the assumption of the unlimited power of totali- One could argue that the regular "purges" tarian leaders). in Communist states-that is, a replacement 21 For certain technical conditions of organ- of the bearers of authority-function as a ization, this is valid only within limits. Thus the guaranty of stability (in the sense of allevi- liquidation of potential leaders of the opposition

CONFLICT RESOLUTION VOLUME 11 NUMBER 2 TOWARD A TI-IEORY OF SOCIAL CONFLICT 183

sistance of the German Eastern Zone gov- of the rulers that a permitted organization, ernment to free elections becomes clear, as even the state party itself, may become the does the general threat of violent, possibl~7 root of an opposition movement and of revo- revolutionary, cordict in totalitaria11 states. lutionary conflict. When-as expressly in Hungary or virtually Again our analysis will be broken off at on Junc 17, 1953 in Bellin-an opportunity the point where it promises testable results. for organizaiioii occurs to latent conflict It was not the intent of this discussion to groups, the total edifice of the totalitarian treat exhaustively some empirical problem. state collapses. Moreover, it seems very Rather, we wanted to show that conflict the- probable that this possibility can become ory puts us in a position to formulate more redlized at any moment in every totalitarian sharply urgent problems of empirical in- state.2Vn modern totalitalian societies vestigation, to bring within our grasp un- founded on icleological state parties, there is explained events, to see what is known from a constant danger from the point of view additional points of view, and to transform tentative questions into a systematic search- is a central component of totalitarian authority. that is, to do precisely what a scientific In a way, both the East German and the Hun- theo~vshould accomplish. It needs hardly to garian events call be taken as corroborations of be said explicitly that conflict theory in the the effectiveness of this policy. form here presented is almost as incomplete 22 Relevant here is the well-known slight de- as the two empirical analyses indicated in crease of pressure which seerns to precede every this section. In spite of a11 progress, the re\folution. Insofar, for example, as a certain re- theory of social conflict is still more a chal- laxation of police control makes possible only an ad hoc organization, the emergence of open con- lenge to the sociologist than a result of his flict beconles acute. researches.