talcottparsons_on the concept of power.pdf

7/28/2019 TalcottParsons_On the concept of power.pdf

1/32

On the Concept of Political Power

Author(s): Talcott ParsonsReviewed work(s):Source: Proceedings of the American Philosophical Society, Vol. 107, No. 3 (Jun. 19, 1963), pp.232-262Published by: American Philosophical SocietyStable URL: http://www.jstor.org/stable/985582 .

Accessed: 07/06/2012 10:22

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 ofcontent 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].

American Philosophical Society is collaborating with JSTOR to digitize, preserve and extend access to

Proceedings of the American Philosophical Society.
http://www.jstor.org/action/showPublisher?publisherCode=ampshttp://www.jstor.org/stable/985582?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/985582?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=amps


2/32

ON THE CONCEPT OF POLITICAL POWERTALCOTTPARSONS

Professorof Sociology,HarvardUniversity(Read November, 1963)

POWERs one of the key concepts in the greatWestern tradition of thought about political phe-nomena. It is at the same time a concept onwhich, in spite of its long history, there is, onanalytical levels, a notable lack of agreementboth about its specific definition, and about manyfeatures of the conceptual context in which itshould be placed. There is, however, a corecomplex of its meaning, having to do with the ca-pacity of persons or collectivities "to get thingsdone" effectively, in particular when their goalsare obstructed by some kind of human resistanceor opposition. The problem of coping with re-sistance then leads into the question of the roleof coercive measures, including the use of physicalforce, and the relation of coercion to the volun-tary and consensual aspects of power systems.The aim of this paper is to attempt to clarifythis complex of meanings and relations by placingthe concept of power in the context of a generalconceptual scheme for the analysis of large-scaleand complex social systems, that is of societies.In doing so I speak as a sociologist rather thanas a political scientist, but as one who believes thatthe interconnections of the principal social disci-plines, including not only these two, but especiallytheir relations to economics as well, are so closethat on matters of general theory of this sort theycannot safely be treated in isolation; their interre-lations must be made explicit and systematic. Asa sociologist, I thus treat a central concept ofpolitical theory by selecting among the elementswhich have figured prominently in political the-ory in terms of their fit with and significance forthe general theoretical analysis of society as awhole.There are three principal contexts in which itseems to me that the difficulties of the concept ofpower, as treated in the literature of the last gene-ration, come to a head. The first of these con-cerns its conceptual diffuseness, the tendency, inthe tradition of Hobbes, to treat power as simplythe generalized capacity to attain ends or goals insocial relations, independently of the media em-

ployed or of the status of "authorization"to makedecisions or impose obligations.'The effect of this diffuseness, as I call it, is totreat "influence"and sometimes money, as well ascoercion in various aspects, as "forms" of power,thereby making it logically impossible to treatpower as a specific mechanism operating to bringabout changes in the action of other units, individ-ual or collective, in the processes of social interac-tion. The latter is the line of thought I wish topursue.Secondly, there is the problem of the relationbetween the coercive and the consensual aspects.I am not aware of any treatment in the literaturewhich presents a satisfactory solution of this prob-lem. A major tendency is to hold that somehow"in the last analysis" power comes down to oneor the other, i.e., to "rest on" command of coercivesanctions, or on consensus and the will to volun-

tary cooperation. If going to one or the otherpolar solution seems to be unacceptable,a way out,taken for example by Friedrich, is to speak of eachof these as different "forms" of power. I shallpropose a solution which maintains that bothaspects are essential, but that neither of the abovetwo ways of relating them is satisfactory, namelysubordinating either one to the other or treatingthem as discrete "forms."

Finally the third problem is what, since theTheory of Games, has widely come to be called the"zero-sum" problem. The dominant tendency inthe literature, for example in Lasswell and C.Wright Mills, is to maintain explicitly or im-plicitly that power is a zero-sum phenomenon,

1ThusE. C. Banfield,PoliticalInfluence New York,The Free Pressof Glencoe, 962), p. 348,speaksof con-trol as the abilityto causeanother o give or witholdaction,andpoweras the ability o establish ontroloveranother. Similarly Robert Dahl, "The Concept ofPower," Behavioral Scientist 2 (July, 1957), says that"A has poweroverB to the extentthathe canget B todo something hat B wouldnot otherwisedo." C. J.Friedrichtakes a similar positionin his forthcomingbook, hetentativeitleof which s "Man ndhis Govern-ment."

PROCEEDINGS OF THE AMERICAN PHILOSOPHICAL SOCIETY, VOL. 107, NO. 3, JUNE, 1963232


3/32

ON THE CONCEPT OF POLITICAL POWERwhich is to say that there is a fixed "quantity"of power in any relational system and hence anygain of power on the part of A must by definitionoccur by diminishing the power at the disposal ofother units, B, C, D .... There are, of course, re-stricted contexts in which this condition holds,but I shall argue that it does not hold for totalsystems of a sufficient level of complexity.

SOME GENERAL ASSUMPTIONSThe initial assumption is that, within the con-ception of society as a system, there is an essen-tial parallelism in theoretical structure between theconceptual schemes appropriate for the analysisof the economic and the political aspects of so-cieties. There are four respects in which I wishto attempt to work out and build on this parallel,showing at the same time the crucial substantivedifferences between the two fields.First "political theory" as here interpreted,which is not simply to be identified with themeaning given the term by many political sci-entists, is thought of as an abstract analyticalscheme in the same sense in which economic the-

ory is abstract and analytical. It is not the con-ceptual interpretation of any concretely completecategory of social phenomena, quite definitely notthose of government, though government is thearea in which the political element comes nearestto having clear primacy over others. Politicaltheory thus conceived is a conceptual schemewhich deals with a restricted set of primary vari-ables and their interrelations, which are to befound operating in all concrete parts of socialsystems. These variables are, however, subject toparametric conditions which constitute the valuesof other variables operating in the larger systemwhich constitutes the society.Secondly, following on this, I assume that theempirical system to which political theory in thissense applies is an analytically defined, a "func-tional" subsystem of a society, not for examplea concrete type of collectivity. The conceptionof the economy of a society is relatively well de-fined.2 I should propose the conception of thepolity as the parallel empirical system of directrelevance to political theory as here advanced.The polity of a given society is composed of theways in which the relevant components of the2 Cf. Talcott Parsons and Neil J. Smelser, Economyand Society (Illinois, The Free Press of Glencoe, 1956),chapterI, for a discussion of this conception.

total system are organized with reference to oneof its fundamental functions, namely effective col-lective action in the attainment of the goals of col-lectivities. Goal-attainment in this sense is theestablishment of a satisfactory relation between acollectivity and certain objects in its environmentwhich include both other collectivities and cate-gories of personalities, e.g. "citizens." A totalsociety must in these terms be conceived, in oneof its main aspects, as a collectivity, but it is alsocomposed of an immense variety of subcollectivi-ties, many of which are parts not only of this so-ciety but of others.3A collectivity, seen in these terms, is thus clearlynot a concrete "group" but the term refers togroups, i.e. systematically related pluralities ofpersons, seen in the perspective of their interestsin and capacities for effective collective action.The political process then is the process by whichthe necessary organization is built up and oper-ated, the goals of action are determined and theresources requisite to it are mobilized.These two parallels to economic theory can beextended to still a third. The parallel to col-lective action in the political case is, for the eco-nomic, production. This conception in turn mustbe understood in relation to three main operativecontexts. The first is adjustment to the condi-tions of "demand"which are conceived to be ex-ternal to the economy itself, to be located in the"consumers" of the economic process. Secondly,resources must be mobilized, also from the en-vironment of the economy, the famous factors ofproduction. Thirdly, the internal economic proc-ess is conceived as creatively combinatorial; it is,by the "combination" of factors of production inthe light of the utility of outputs, a process ofcreating more valuable facilities to meet the needsof consuming units than would be available tothem without this combinatorial process. I wishmost definitely to postulate that the logic of "valueadded"applies to the political sphere in the presentsense.4

3 E.g. the American medical profession is part ofAmerican society, but also it is part of a wider medicalprofession which transcends this particular society, tosome extent as collectivity. Interpenetration in member-ship is thus a feature of the relations among collectivities.4 For discussions of the conception of "valued-added"in spheres of application broader than the economic alone,cf. Neil J. Smelser, Social Changein the Industrial Revo-lution (Glencoe, Illinois, The Free Press of Glencoe,1959), chapter II, pp. 7-20, and Neil J. Smelser, Theoryof Collective Behavior (New York, The Free Press ofGlencoe, 1963), chapter II, pp. 23-47.

233OL. 107, NO. 3, 1963]


4/32

TALCOTT PARSONSIn the political case, however, the value refer-ence is not to utility in the economic sense but toeffectiveness, very precisely, I think in the senseused by C. I. Barnard.5 For the limited purposesof political analysis as such the givenness of the

goal-demands of interest groups serves as thesame order of factor in relation to the politicalsystem as has the corresponding givenness of con-sumers' wants for purposes of economic analysis-and of course the same order of qualifications onthe empirical adequacy of such postulates.Finally, fourth, political analysis as here con-ceived is parallel to economic in the sense that acentral place in it is occupied by a generalizedmedium involved in the political interaction proc-ess, which is also a "measure" of the relevantvalues. I conceive power as such a generalizedmedium in a sense directly parallel in logicalstructure, though very different substantively, tomoney as the generalized medium of the economicprocess. It is essentially this conception of poweras a generalized medium parallel to money whichwill, in the theoretical context sketched above,provide the thread for guiding the following analy-sis through the types of historic difficulty withreference to which the paper began.THE OUTPUTS OF POLITICAL PROCESS ANDTHE FACTORS OF EFFECTIVENESS

The logic of the combinatorial process which Ihold to be common to economic theory and thetype of political theory advanced here, involvesa paradigm of inputs and outputs and their rela-tions. Again we will hold that the logic is strictlyparallel to the economic case, i.e. that there shouldbe a set of political categories strictly parallel tothose of the factors of production (inputs) on theone hand, the shares of income (outputs) on theother.In the economic case, with the exception of land,the remaining three factors must be regarded asinputs from the other three cognate functionalsubsystems of the society, labor from what we callthe "pattern-maintenance" system, capital fromthe polity and organization, in the sense of AlfredMarshall, from the integrative system.6 Further-more, it becomes clear that land is not, as a factorof production, simply the physical resource, but

5 C. I. Barnard, The Functions of the Executive (Cam-bridge, Harvard University Press, 1938), chapter V,pp. 46-64.6 On the rationale of these attributions,see Economyand Society, op. cit., chapter II.

essentially the commitment, in value terms, of anyresources to economic production in the systemindependent of price.In the political case, similarly the equivalent ofland is the commitment of resources to effectivecollective action, independent of any specifiable"pay-off" for the unit which controls them.7 Par-allel to labor is the demands or "need" for collec-tive action as manifested in the "public"which insome sense is the constituency of the leadershipof the collectivity in question-a conception whichis relatively clear for the governmental or otherelectoral association, but needs clarification inother connections. Parallel to capital is the con-trol of some part of the productivity of the econ-omy for the goals of the collectivity, in a suffi-ciently developed economy through financial re-sources at the disposal of the collectivity, acquiredby earnings, gift, or taxation. Finally, parallel toorganization is the legitimation of the authorityunder which collective decisions are taken.It is most important to note that none of thesecategories of input is conceived as a form ofpower. In so far as they involve media, it is themedia rooted in contiguous functional systems,not power as that central to the polity-e.g. con-trol of productivity may operate through money,and constituents' demands through what I call'influence." Power then is the means of acquiringcontrol of the factors in effectiveness; it is not it-self one of these factors, any more than in theeconomic case money is a factor of production;to suppose it was, was the ancient mercantilistfallacy.Though the analytical context in which theyare placed is perhaps unfamiliar in the light oftraditional political analysis, I hope it is clearthat the actual categories used are well established,though there remain a number of problems ofexact definition. Thus control of productivitythrough financing of collective action is very fa-miliar, and the concept of "demands"in the senseof what constituents want and press for, is alsovery familiar.8 The concept legitimation is used inessentially the same sense in which I think MaxWeber used it in a political context.9

7 "Pay-off"may be a deciding factor in choice betweenparticularcontexts of use, but not as to whether the re-source shall be devoted to collective effectiveness at all.8 I have in fact adoptedthe term "demands" rom theusage of David Easton, "An Approachto the Analysis ofPoliticalSystems,"WorldPolitics9(1957): 383400.9Cf.MaxWeber,TheTheoryof SocialandEconomicOrganization (New York, Oxford University Press,

234 [PROC. AMER. PHIL. SOC.


5/32

ON THE CONCEPT OF POLITICAL POWERThe problem of what corresponds, for thepolitical case, to the economist's "shares of in-come" is not very difficult, once the essential dis-tinction, a very old one in economic tradition, be-tween monetary and "real" income is clearly taken

into account. Our concern is with the "real" out-puts of the political process-the analogue of themonetary here is output of power.There is one, to us critically important revisionof the traditional economic treatment of outputswhich must be made, namely the bracketing to-gether of "goods and services," which then wouldbe treated as outputs to the household as, in ourtechnical terms, a part of the "pattern-mainte-nance" system. The present position is thatgoods, i.e., more precisely property rights in thephysical objects of possession, belong in this cate-gory, but that "services," the commitment of hu-man role-performances to an "employer," or con-tracting agent constitute an output, not to thehousehold, but to the polity, the type case (thoughnot the only one) being an employing organiza-tion in which the role-incumbent commits himselfto performance of an occupational role, a job,10asa contribution to the effective functioning of thecollectivity.There is, from this consideration, a conclusionwhich is somewhat surprising to economists,namely that service is, in the economic sense the"real" counterpart of interest as monetary incomefrom the use of funds. What we suggest is thatthe political control of productivity makes it pos-sible, through combinatorial gains in the politicalcontext, to produce a surplus above the monetaryfunds committed, by virtue of which under speci-fied conditions a premium can be paid at the mone-tary level which, though a result of the combina-torial process as a whole, is most directly relatedto the output of available services as an economicphenomenon, i.e. as a "fluid resource." Seen alittle differently, it becomes necessary to make aclear distinction between labor as a factor of pro-duction in the economic sense and service as anoutput of the economic process which is utilized ina political context, that is one of organizationalor collective effectiveness.

Service, however, is not a "factor" in effective-1947), p. 124. Translation by A. M. Henderson andTalcott Parsons; edited by Talcott Parsons.10The cases of services concretelyrenderedto a house-hold will be consideredas a limiting case where the rolesof consumerand employerhave not becomedifferentiatedfrom each other.

ness, in the sense in which labor is a factor ofproduction, precisely because it is a category ofpower. It is the point at which the economicutility of the human factor is matched with itspotential contribution to effective collective ac-tion. Since the consumer of services is in prin-ciple the employing collectivity, it is its effective-ness for collective goals, not its capacity to satisfythe "wants" of individuals, which is the vantagepoint from which the utility of the service is de-rived. The output of power which matches theinput of services to the polity, I interpret to bethe "opportunity for effectiveness" which employ-ment confers on those employed or contract of-fers to partners. Capital in the economic senseis one form of this opportunity for effectivenesswhich is derived from providing, for certain typesof performances, a framework of effective organi-zation.ll

The second, particularly important context of"real" output of the political process is the cate-gory which, in accord with much tradition, Ishould like to call capacity to assume leadershipresponsibility. This, as a category of "real" out-put also is not a form of power, but this time ofinfluence.12 This is an output not to the economybut to what I shall call the integrative system,which in its relevance to the present context isin the first instance the sector of the "public"which can be looked on as the "constituencies" ofthe collective processes under consideration. Itis the group structure of the society looked at interms of their structured interests in particularmodes of effective collective action by particularcollectivities. It is only through effective organi-zation that genuine responsibility can be taken,hence the implementation of such interest de-mands responsibility for collective effectiveness.1311In the cases treated as typical for economicanalysisthe collective element in capital is delegated through thebindingnessof the contractsof loan of financial resources.

To us this is a special case, employmentbeing another,of the binding obligation assumed by an organization,whether it employs or loans, by virtue of which the re-cipient can be more effective than would otherwise bethe case. It is not possible to go further into these com-plex problems here, but they will, perhaps,be somewhatilluminated by the later discussion of the place of theconceptof bindingness n the theory of power.12See my paper "On the Conceptof Influence,"to bepublished in the Public Opinion Quarterly 27(Spring,1963).13Here again Barnard's usage of the concept of re-sponsibility seems to me the appropriate one. SeeBarnard, op. cit.

235OL. 107, NO. 3, 1963]


6/32

TALCOTT PARSONSAgain it should be made quite clear that leader-ship responsibility is not here conceived as an out-put of power, though many political theorists (e.g.Friedrich) treat both leadership and, more broadlyinfluence, as "forms" of power. The power cate-gory which regulates the output of leadership in-fluence takes this form on the one side of bindingpolicy decisions of the collectivity, on the other ofpolitical support from the constituency, in the typecase through franchise. Policy decisions we wouldtreat as a factor in integration of the system,not as a "consumable" output of the politicalprocess.14

Finally, a few words need to be said about whatI have called the combinatorial process itself. Itis of course assumed in economic theory that the"structures" of the factors of production on theone hand, the "demand system" for real outputson the other hand, are independent of each other."Utility" of outputs can only be enhanced, to saynothing of maximized, by processes of transforma-tion of the factors in the direction of providingwhat is wanted as distinguished from what merelyis available. The decision-making aspect of thistransformative process, what is to be produced,how much and how offered for consumption, iswhat is meant by economic production, whereasthe physical processes are not economic but"technological"; they are controlled by economicconsiderations, but are not themselves in ananalytical sense economic.The consequence of successful adaptation ofavailable resources to the want or demand sys-tem is an increment in the value of the resource-stock conceived in terms of utility as a type ofvalue. But this means recombination of the com-ponents of the resource-stock in order to adaptthem to the various uses in question.The same logic applies to the combinatorialprocess in the political sphere. Here the resourcesare not land, labor, capital, and organization, butvaluation of effectiveness, control of productivity,structured demands and the patterning of legiti-mation. The "wants" are not for consumption inthe economic sense, but for the solution of "in-terest" problems in the system, including both

14 In order not to complicatethings too much, I shallnot enter into problemof the interchange ystemin-volving legitimationhere. See my paper "Authority,Legitimation, nd Political Process," n Nomos 1, re-printed as chapter V of my Structure and Process inModernSocieties (Glencoe,Illinois, The Free Press,1960),chapterV, pp.170-198.

competitive problems in the allocative sense andconflict problems, as well as problems of enhance-ment of the total effectiveness of the system ofcollective organization. In this case also the"structure" of the available resources may not beassumed spontaneously to match the structure ofthe system of interest-demands. The increment ofeffectiveness in demand-satisfaction through thepolitical process is, as in the economic case, ar-rived at through combinatorial decision-processes.The organizational "technology" involved is not inthe analytical sense political. The demand-refer-ence is not to discrete units of the system con-ceived in abstraction from the system as a whole-the "individual" consumer of the economist-but to the problem of the share of benefits andburdens to be allocated to subsystems of variousorders. The "consumption" reference is to theinterest-unit's place in the allocative system ratherthan to the independent merits of particular"needs."

THE CONCEPTOF POWERThe above may seem a highly elaborate settingin which to place the formal introduction of themain subject of the paper, namely the conceptof power. Condensed and cryptic as the exposi-tion may have been, however, understanding of itsmain structure is an essential basis for the spe-

cial way in which it will be proposed to combinethe elements which have played a crucial part inthe main intellectual traditions dealing with theproblems of power.Power is here conceived as a circulating me-dium, analogous to money, within what is calledthe political system, but notably over its bound-aries into all three of the other neighboring func-tional subsystems of a society (as I conceivethem), the economic, integrative, and pattern-maintenance systems. Specification of the proper-ties of power can best be approached through anattempt to delineate very briefly the relevantproperties of money as such a medium in theeconomy.Money is, as the classical economists said, botha medium of exchange and a "measure of value."It is symbolic in that, though measuring and thus"standing for" economic value or utility, it doesnot itself possess utility in the primary consump-tion sense-it has no "value in use" but only "inexchange," i.e. for possession of things havingutility. The use of money is thus a mode of com-



7/32

ON THE CONCEPT OF POLITICAL POWERmunication of offers, on the one hand to purchase,on the other to sell, things of utility, with and formoney. It becomes an essential medium onlywhen exchange is neither ascriptive, as exchangeof gifts between assigned categories of kin, nortakes place on a basis of barter, one item ofcommodity or service directly for another.In exchange for its lack of direct utility moneygives the recipient four important degrees of free-dom in his participation in the total exchange sys-tem. (1) He is free to spend his money for anyitem or combination of items available on the mar-ket which he can afford, (2) he is free to shoparound among alternative sources of supply fordesired items, (3) he can choose his own time topurchase, and (4) he is free to consider termswhich, because of freedom of time and sourcehe can accept or reject or attempt to influence inthe particular case. By contrast, in the case ofbarter, the negotiator is bound to what his par-ticular partner has or wants in relation to whathe has and will part with at the particular time.The other side of the gain in degrees of freedomis of course the risk involved in the probabilitiesof the acceptance of money by others and of thestability of its value.Primitive money is a medium which is still veryclose to a commodity, the commonest case beingprecious metal, and many still feel that the valueof money is "really" grounded in the commodityvalue of the metallic base. On this base, however,there is, in developed monetary systems, erected acomplex structure of credit instruments, so thatonly a tiny fraction of actual transactions is con-ducted in terms of the metal-it becomes a "re-serve" available for certain contingencies, and isactually used mainly in the settlement of interna-tional balances. I shall discuss the nature of creditfurther in another connection later. For the mo-ment suffice it to say that, however important incertain contingencies the availability of metallicreserves may be, no modern monetary systemoperates primarily with metal as the actual me-dium, but uses "valueless" money. Moreover,the acceptanceof this "valueless" money rests on acertain institutionalized confidence in the monetarysystem. If the security of monetary commitmentsrested only on their convertibility into metal, thenthe overwhelming majority of them would beworthless, for the simple reason that the totalquantity of metal is far too small to redeem morethan a few.One final point is that money is "good," i.e.

works as a medium, only within a relatively de-fined network of market relationships which to besure now has become world-wide, but the main-tenance of which requires special measures tomaintain mutual convertibility of national cur-rencies. Such a system is on the one hand a rangeof exchange-potential within which money may bespent, but on the other hand, one within whichcertain conditions affecting the protection andmanagement of the unit are maintained, both bylaw and by responsible agencies under the law.The first focus of the concept of an institu-tionalized power system is, analogously, a rela-tional system within which certain categories ofcommitments and obligations, ascriptive or volun-tarily assumed-e.g. by contract-are treated asbinding, i.e. under normatively defined conditionstheir fulfillment may be insisted upon by the ap-propriate role-reciprocal agencies. Furthermore,in case of actual or threatened resistance to "com-pliance," i.e. to fulfillment of such obligationswhen invoked, they will be "enforced" by thethreat or actual imposition of situational negativesanctions, in the former case having the functionof deterrence, in the latter of punishment. Theseare events in the situation of the actor of referencewhich intentionally alter his situation (or threatento) to his disadvantage, whatever in specific con-tent these alterations may be.Power then is generalized capacity to secure theperformance of binding obligations by units in asystem of collective organization when the obliga-tions are legitimized with reference to their bear-ing on collective goals and where in case ofrecalcitrance there is a presumption of enforce-ment by negative situational sanctions-whateverthe actual agency of that enforcement.It will be noted that I have used the concep-tions of generalization and of legitimation in de-fining power. Securing possession of an object ofutility by bartering another object for it is not amonetary transaction. Similarly, by my defini-tion, securing compliance with a wish, whether itbe defined as an obligation of the object or not,simply by threat of superior force, is not an exer-cise of power. I am well aware that most politicaltheorists would draw the line differently andclassify this as power (e.g. Dahl's definition), butI wish to stick to my chosen line and explore itsimplications. The capacity to secure compliancemust, if it is to be called power in my sense, begeneralized and not solely a function of one par-ticular sanctioning act which the user is in a posi-

237OL. 107, NO. 3, 1963]


8/32

TALCOTT PARSONStion to impose,15 and the medium used must be"symbolic."Secondly, I have spoken of power as involvinglegitimation. This is, in the present context, thenecessary consequence of conceiving power as"symbolic,"which therefore, if it is exchanged forsomething intrinsically valuable for collective ef-fectiveness, namely compliance with an obligation,leaves the recipient, the performer of the obliga-tion, with "nothing of value." This is to say, thathe has "nothing"but a set of expectations, namelythat in other contexts and on other occasions, hecan invoke certain obligations of the part of otherunits. Legitimation is therefore, in power sys-tems, the factor which is parallel to confidence inmutual acceptability and stability of the monetaryunit in monetary systems.The two criteria are connected in that ques-tioning the legitimacy of the possession and useof power leads to resort to progressively more"secure" means of gaining compliance. Thesemust be progressively more effective "intrinsi-cally," hence more tailored to the particular situ-ations of the objects and less general. Further-more in so far as they are intrinsically effective,legitimacy becomes a progressively less importantfactor of their effectiveness-at the end of thisseries lies resort, first to various types of coercion,eventually to the use of force as the most intrin-sically effective of all means of coercion.'6I should like now to attempt to place bothmoney and power in the context of a more generalparadigm, which is an analytical classification ofways in which, in the processes of social interac-tion, the actions of one unit in a system can, in-tentionally, be oriented to bringing about a changein what the actions of one or more other unitswould otherwise have been-thus all fitting intothe context of Dahl's conception of power. It isconvenient to state this in terms of the conventionof speaking of the acting unit of reference-indi-vidual or collective-as ego, and the object onwhich he attempts to "operate"as alter. We maythen classify the alternatives open to ego in termsof two dichotomous variables. On the one handego may attempt to gain his end from alter either

15 There is a certain element of generality in physicalforce as a negative sanction,which gives it a specialplacein power systems. This will be taken up later in thediscussion.16There are complicationshere deriving from the factthat power is associated with negative sanctions andhence that, in the face of severe resistance, their ef-fectiveness is confinedto deterrence.

by using some form of control over the situationin which alter is placed, actually or contingently tochange it so as to increase the probability of alteracting in the way he wishes, or, alternatively,without attempting to change alter's situation, egomay attempt to change alter's intentions, i.e. hemay manipulate symbols which are meaningfulto alter in such a way that he tries to make alter"see" that what ego wants is a "good thing" forhim (alter) to do.The second variable then concerns the type ofsanctions ego may employ in attempting to guar-antee the attainment of his end from alter. Thedichotomy here is between positive and negativesanctions. Thus through the situational channela positive sanction is a change in alter's situationpresumptively considered by alter as to his advan-tage, which is used as a means by ego of havingan effect on alter's actions. A negative sanctionthen is an alteration in alter's situation to thelatter's disadvantage. In the case of the inten-tional channel, the positive sanction is the ex-pression of symbolic "reasons" why compliancewith ego's wishes is "a good thing" independentlyof any further action on ego's part, from alter'spoint of view, i.e. would be felt by him to be"personally advantageous," whereas the negativesanction is presenting reasons why noncompli-ance with ego's wishes should be felt by alter tobe harmful to interests in which he had a signifi-cant personal investment and should therefore beavoided. I should like to call the four types of"strategy" open to ego respectively (1) for thesituational channel, positive sanction case, "in-ducement"; (2) situational channel negative sanc-tion, "coercion"; (3) intentional channel, posi-tive sanction "persuasion," and (4) intentionalchannel negative sanction "activation of commit-ments" as shown in the following table:

Sanctiontype

Positive IntentionalPersuasion

Channel

3

Negative Activation of 4Commitments

Situational1 Inducement

2 Coercion

A further complication now needs to be intro-duced. We think of a sanction as an intentionalact on ego's part, expected by him to change his



9/32

ON THE CONCEPT OF POLITICAL POWERrelation to alter from what it would otherwisehave been. As a means of bringing about achange in alter's action, it can operate most obvi-ously where the actual imposition of the sanctionis made contingent on a future decision by alter.Thus a process of inducement will operate in twostages, first contingent offer on ego's part that, ifalter will "comply" with his wishes, ego will"reward" him by the contingently promised situ-ational change. If then alter in fact does comply,ego will perform the sanctioning act. In the caseof coercion the first stage is a contingent threatthat, unless alter decides to comply, ego willimpose the negative sanction. If, however, altercomplies, then nothing further happens, but, if hedecides on noncompliance, then ego must carryout his threat, or be in a position of "not meaningit." In the cases of the intentional channel ego'sfirst-stage act is either to predict the occurrence,or to announce his own intention of doing some-thing which affects alter's sentiments or interests.The element of contingency enters in in that ego"argues" to alter, that if this happens, on the onehand alter should be expected to "see" that itwould be a good thing for him to do what egowants-the positive case-or that if he fails to doit it would imply an important "subjective cost"to alter. In the positive case, beyond "pointingout" if alter complies, ego is obligated to deliverthe positive attitudinal sanction of approval. Inthe negative case, the corresponding attitudinalsanction of disapproval is implemented only fornoncompliance.It is hence clear that there is a basic asymmetrybetween the positive and negative sides of thesanction aspect of the paradigm. This is that, inthe cases of inducement and persuasion, alter'scompliance obligates ego to "deliver" his promisedpositive sanction, in the former case the promisedadvantages, in the latter his approval of alter's"good sense" in recognizing that the decisionwished for by ego and accepted as "good" byalter, in fact turns out to be good from alter's pointof view. In the negative cases, on the other hand,compliance on alter's part obligates ego, in thesituational case, not to carry out his threat, in theintentional case by withholding disapproval toconfirm to alter that his compliance did in factspare him what to him, without ego's intervention,would have been the undesirable subjective con-sequences of his previous intentions, namely guiltover violations of his commitments.

Finally, alter's freedom of action in his de-

cisions of compliance versus noncompliance is alsoa variable. This range has a lower limit at whichthe element of contingency disappears. That is,from ego's point of view, he may not say, if youdo so and so, I will intervene, either by situationalmanipulations or by "arguments"in such and sucha way, but he may simply perform an overt actand face alter with a fait accompli. In the case ofinducement a gift which is an object of value andwith respect to the acceptance of which alter isgiven no option is the limiting case. With respectto coercion, compulsion, i.e. simply imposing adisadvantageous alteration on alter's situation andthen leaving it to alter to decide whether to "dosomething about it" is the limiting case.The asymmetry just referred to appears here aswell. As contingent it may be said that the pri-mary meaning of negative sanctions is as meansof prevention. If they are effective, no further ac-tion is required. The case of compulsion is thatin which it is rendered impossible for alter toavoid the undesired action on ego's part. In thecase of positive sanctions of course ego, for ex-ample in making a gift to alter, cuts himself outfrom benefiting from alter's performance which ispresumptively advantageous to him, in the par-ticular exchange.Both, however, may be oriented to their effecton alter's action in future sequences of interaction.The object of compulsion may have been "taughta lesson" and hence be less disposed to noncom-pliance with ego's wishes in the future, as well asprevented from performance of a particular un-desired act and the recipient of a gift may feel a"sense of obligation" to reciprocate in some formin the future.

So far this discussion has dealt with sanction-ing acts in terms of their "intrinsic" significanceboth to ego and to alter. An offered inducementmay thus be possession of a particular object ofutility, a coercive threat, that of a particularfearedloss, or other noxious experience. But just as, inthe initial phase of a sequence, ego transmits hiscontingent intentions to alter symbolically throughcommunication, so the sanction involved may alsobe symbolic, e.g. in place of possession of certainintrinsically valuable goods he may offer a sum ofmoney. What we have called the generalizedmedia of interaction then may be used as types ofsanctions which may be analyzed in terms of theabove paradigm. The factors of generalizationand of legitimation of institutionalization, how-

VOL. 107, NO. 3, 1963] 239


10/32

TALCOTT PARSONSever, as discussed above, introduce certain compli-cations which we must now take up with referenceto power. There is a sense in which power maybe regarded as the generalized medium of coercionin the above terms, but this formula at the veryleast requires very careful interpretation-indeedit will turn out by itself to be inadequate.I spoke above of the "grounding" of the valueof money in the commodity value of the monetarymetal, and suggested that there is a correspondingrelation of the "value," i.e. the effectiveness ofpower, to the intrinsic effectiveness of physicalforce as a means of coercion and, in the limitingcase, compulsion."'In interpreting this formula due account mustbe taken of the asymmetry just discussed. Thespecial place of gold as a monetary base rests onsuch properties as its durability, high value insmall bulk, etc., and high probability of acceptabil-ity in exchange, i.e. as means of inducement, in avery wide variety of conditions which are not de-pendent on an institutionalized order. Ego's pri-mary aim in resorting to compulsion or coercion,however, is deterrence of unwanted action onalter's part.l8 Force, therefore, is in the first in-stance important as the "ultimate" deterrent. Itis the means which, again independent of any in-stitutionalized system of order, can be assumedto be "intrinsically" the most effective in the con-text of deterrence, when means of effectivenesswhich are dependent on institutionalized orderare insecure or fail. Therefore, the unit of anaction system which commands control of physicalforce adequate to cope with any potential counterthreats of force is more secure than any otherin a Hobbesian state of nature.l9But just as a monetary system resting entirelyon gold as the actual medium of exchange is a veryprimitive one which simply cannot mediate a com-plex system of market exchange, so a power sys-tem in which the only negative sanction is thethreat of force is a very primitive one which can-not function to mediate a complex system of or-ganizational coordination-it is far too "blunt" aninstrument. Money cannot be only an intrinsi-

17 I owe the insight into this parallel to Professor KarlW. Deutsch of Yale University (personal discussion).18 "Sadistic" infliction of injury without instrumentalsignificanceto ego does not belong in this context.19I have attemptedto develop this line of analysis ofthe significanceof force somewhat more fully in "SomeReflectionsof the Role of Force in Social Relations,"inHarry Eckstein, ed., The Problem of Internal War (NewJersey, PrincetonUniversity Press, 1963).

cally valuable entity if it is to serve as a general-ized medium of inducement, but it must, as wehave said, be institutionalized as a symbol; itmust be legitimized, and must inspire "confi-dence" within the system-and must also withinlimits be deliberately managed. Similarly powercannot be only an intrinsically effective deterrent;if it is to be the generalized medium of mobilizingresources for effective collective action, and forthe fulfillment of commitments made by collectivi-ties to what we have here called their constituents;it too must be both symbolically generalized, andlegitimized.There is a direct connection between the con-cept of bindingness, as introduced above, anddeterrence. To treat a commitment or any otherform of expectation as binding is to attribute aspecial importance to its fulfillment. Where itis not a matter simply of maintenance of an es-tablished routine, but of undertaking new actionsin changed circumstances, where the commitmentis thus to undertake types of action contingent oncircumstances as they develop, then the risk to beminimized is that such contingent commitmentswill not be carried out when the circumstances inquestion appear. Treating the expectation or ob-ligation as binding is almost the same thing assaying that appropriate steps on the other sidemust be taken to prevent nonfulfillment, if possi-ble. Willingness to impose negative sanctions is,seen in this light, simply the carrying out of theimplications of treating commitments as binding,and the agent invoking them "meaning it" orbeing prepared to insist.On the other hand there are areas in interactionsystems where there is a range of alternatives,choice among which is optional, in the light of thepromised advantageousness, situational or "in-tentional," of one as compared to other choices.Positive sanctions as here conceived constitute acontingent increment of relative advantageous-ness, situational or intentional, of the alternativeego desires alter to choose.If, in these latter areas, a generalized, symbolicmedium, is to operate in place of intrinsic ad-vantages, there must be an element of binding-ness in the institutionalization of the medium it-self-e.g. the fact that the money of a society is"legal tender" which must be accepted in thesettlement of debts which have the status of con-tractual obligations under the law. In the caseof money, I suggest that, for the typical actingunit in a market system, what specific under-



11/32

ON THE CONCEPT OF POLITICAL POWERtakings he enters into is overwhelmingly optionalin the above sense, but whether the money in-volved in the transactions is or is not "good" isnot for him to judge, but his acceptance of it isbinding. Essentially the same is true of the con-tractual obligations, typically linking monetaryand intrinsic utilities, which he undertakes.I would now like to suggest that what is in acertain sense the obverse holds true of power.Its "intrinsic" importance lies in its capacity toensure that obligations are "really" binding, thusif necessary can be "enforced" by negative sanc-tions. But for power to function as a generalizedmedium in a complex system, i.e. to mobilize re-sources effectively for collective action, it must be"legitimized" which in the present context meansthat in certain respects compliance, which is thecommon factor among our media, is not binding,to say nothing of being coerced, but is optional.The range within which there exists a continuoussystem of interlocking binding obligations is es-sentially that of the internal relations of an or-ganized collectivity in our sense, and of the con-tractual obligations undertaken on its behalf at itsboundaries.The points at which the optional factors come tobear are, in the boundary relations of the col-lectivity, where factors of importance for collectivefunctioning other than binding obligations are ex-changed for such binding commitments on thepart of the collectivity and vice versa, nonbindingoutputs of the collectivity for binding commit-ments to it. These "optional" inputs, I have sug-gested above, are control of productivity of theeconomy at one boundary, influence through therelations between leadership and the public de-mands at the other.20This is a point at which the dissociation of theconcept of polity from exclusive relation to gov-ernment becomes particularly important. In asufficiently differentiated society, the boundary-relations of the great majority of its importantunits of collective organization (including someboundaries of government) are boundaries wherethe overwhelming majority of decisions of com-mitment are optional in the above sense, thoughonce made, their fulfillment is binding. This,however, is only possible effectively within therange of a sufficiently stable, institutionalizednormative order so that the requisite degrees of

20Thus,if control of productivity peratesthroughmonetaryfunds, their possessor cannot "force"e.g.prospective mployeeso acceptemployment.

freedom are protected, e.g. in the fields of em-ployment and of the promotion of interest-de-mands and decisions about political support.This feature of the boundary relations of aparticular political unit holds even for cases oflocal government, in that decisions of residence,employment, or acquisition of property within aparticular jurisdiction involve the optional ele-ment, since in all these respects there is a rela-tively free choice among local jurisdictions, eventhough, once having chosen, the citizen is, for ex-ample, subject to the tax policies applying withinit-and of course he cannot escape being subjectto any local jurisdiction, but must choose amongthose available.In the case of a "national" political organiza-tion, however, its territorial boundaries ordinarilycoincide with a relative break in the normativeorder regulating social interaction.21 Henceacross such boundaries an ambiguity becomes in-volved in the exercise of power in our sense. Onthe one hand the invoking of binding obligationsoperates normally without explicit use of coercionwithin certain ranges where the two territorialcollectivity systems have institutionalized their re-lations. Thus travelers in friendly foreign coun-tries can ordinarily enjoy personal security andthe amenities of the principal public accommoda-tions, exchange of their money at "going" rates,etc. Where, on the other hand, the more generalrelations between national collectivities are at is-sue, the power system is especially vulnerable tothe kind of insecurity of expectations which tendsto be met by the explicit resort to threats of co-ercive sanctions. Such threats in turn, operatingon both sides of a reciprocal relationship, readilyenter into a vicious circle of resort to more andmore "intrinsically" effective or drastic measuresof coercion, at the end of which road lies physicalforce. In other words, the danger of war isendemic in uninstitutionalized relations betweenterritorially organized collectivities.There is thus an inherent relation between boththe use and the control of force and the territorialbasis of organization.22 One central condition of

21 This, of course, is a relative difference. Somehazards increase the moment one steps outside his ownhome, police protection may be better in one local com-munity than the next, and crossing a state boundarymaymean a considerable differencein legal or actual rights.22Cf. my paper "The Principal Structures of Com-munity,"Nomos 2 and Structure and Process, op. cit.,chapter8. See also W. L. Hurst, Law and Social Proc-

241OL. 107, NO. 3, 1963]


12/32

TALCOTT PARSONSthe integration of a power system is that it shouldbe effective within a territorial area, and a crucialcondition of this effectiveness in turn is themonopoly of control of paramountforce within thearea. The critical point then, at which the insti-tutional integration of power systems is mostvulnerable to strain, and to degeneration into re-ciprocating threats of the use of force, is betweenterritorially organized political systems. This,notoriously, is the weakest point in the normativeorder of human society today, as it has beenalmost from time immemorial.In this connection it should be recognized thatthe possession, the mutual threat, and possible useof force is only in a most proximate sense theprincipal "cause" of war. The essential point isthat the "bottleneck"of mutual regression to moreand more primitive means of protecting or ad-vancing collective interests is a "channel" intowhich all elements of tension between the collec-tive units in question may flow. It is a questionof the many levels at which such elements oftension may on the one hand build up, on theother be controlled, not of any simple and un-equivocal conception of the "inherent" conse-quences of the possession and possible uses oforganized force.It should be clear that again there is a directparallel with the economic case. A functioningmarket system requires integration of the mone-tary medium. It cannot be a system of N inde-pendent monetary units and agencies controllingthem. This is the basis on which the main rangeof extension of a relatively integrated marketsystem tends to coincide with the "politically or-ganized society," as Roscoe Pound calls it, over aterritorial area. International transactions re-quire special provisions not required for domestic.The basic "management" of the monetary sys-tem must then be integrated with the institutional-ization of political power. Just as the latter de-pends on an effective monopoly of institutionallyorganized force, so monetary stability depends onan effective monopoly of basic reserves protectingthe monetary unit and, as we shall see later, oncentralization of control over the credit system.

THE HIERARCHICAL ASPECT OFPOWER SYSTEMSA very critical question now arises, which maybe stated in terms of a crucial difference between

ess in the United States (Ann Arbor, University ofMichigan Law School, 1960).

money and power. Money is a "measure ofvalue," as the classical economists put it, in termsof a continuous linear variable. Objects of utilityvalued in money are more or less valuable thaneach other in numerically statable terms. Simi-larly, as medium of exchange, amounts of moneydiffer in the same single dimension. One actingunit in a society has more money-or assets ex-changeable for money-than another, less than,or the same.Power involves a quite different dimensionwhich may be formulated in terms of the con-ception that A may have power over B. Of coursein competitive bidding the holder of superior fi-nancial assets has an advantage in that, as econo-mists say, the "marginal utility of money" is lessto him than to his competitor with smaller assets.But his "bid" is no more binding on the potentialexchange partner than is that of the less affluentbidder, since in "purchasing power" all dollars are"createdfree and equal." There may be auxiliaryreasons why the purveyor may think it advisableto accept the bid of the more affluent bidder;these, however, are not strictly economic, but con-cern the interrelations between money and othermedia, and other bases of status in the system.The connection between the value of effective-ness-as distinguished from utility-and bind-ingness, implies a conception in turn of the focus-sing of responsibility for decisions, and hence ofauthority for their implementation.23 This impliesa special form of inequality of power which inturn implies a priority system of commitments.The implications of having assumed binding com-mitments, on the fulfillment of which spokesmenfor the collectivity are prepared to insist to thepoint of imposing serious negative sanctions fornoncompliance, are of an order of seriousnesssuch that matching the priority system in the com-mitments themselves there must be priorities inthe matter of which decisions take precedence overothers and, back of that, of which decision-makingagencies have the right to make decisions at whatlevels. Throughout this discussion the crucialquestion concerns bindingness. The reference isto the collectivity, and hence the strategic signifi-cance of the various "contributions" on the per-formance of which the effectiveness of its action

23 As already noted, in this area, I think the analysisof Chester I. Barnard, in The Function of the Execu-tive, op. cit., is so outstandingly clear and cogent thatit deservesthe status of a classic of political theory in myspecific sense. See especially chapterX.



13/32

ON THE CONCEPT OF POLITICAL POWERdepends. Effectiveness for the collectivity as awhole is dependent on hierarchical ordering ofthe relative strategic importance of these contribu-tions, and hence of the conditions governing theimposition of binding obligations on the con-tributors.

Hence the power of A over B is, in its legit-imized form, the "right" of A, as a decision-mak-ing unit involved in collective process, to makedecisions which take precedence over those of B,in the interest of the effectiveness of the collectiveoperation as a whole.The right to use power, or negative sanctionson a barter basis or even compulsion to assert pri-ority of a decision over others, I shall, followingBarnard, call authority. Precedence in this sensecan take different forms. The most serious am-biguity here seems to derive from the assumptionthat authority and its attendant power may beunderstood as implying opposition to the wishesof "lower-order" echelons which hence includesthe prerogative of coercing or compelling com-pliance. Though this is implicit, it may be thatthe higher-order authority and power may implythe prerogative is primarily significant as "de-fining the situation" for the performance of thelower-order echelons. The higher "authority"may then make a decision which defines termswithin which other units in the collectivity willbe expected to act, and this expectation is treatedas binding. Thus a ruling by the Commissionerof Internal Revenue may exclude certain tax ex-emptions which units under his jurisdiction havethought taxpayers could claim. Such a decisionneed not activate an overt conflict between com-missioner and taxpayer, but may rather "channel"the decisions of revenue agents and taxpayerswith reference to performance of obligations.There does not seem to be an essential theoreti-cal difficulty involved in this "ambiguity." Wecan say that the primary function of superior au-thority is clearly to define the situation for thelower echelons of the collectivity. The problem ofovercoming opposition in the form of dispositionsto noncompliance then arises from the incompleteinstitutionalization of the power of the higher au-thority holder. Sources of this may well includeoverstepping of the bounds of his legitimate au-thority on the part of this agent. The concept ofcompliance should clearly not be limited to "obedi-ence" by subordinates, but is just as importantlyapplicable to observance of the normative order

by the high echelons of authority and power. Theconcept of constitutionalism is the critical one atthis level, namely that even the highest authorityis bound in the strict sense of the concept bind-ingness used here, by the terms of the normativeorder under which he operates, e.g. holds office.Hence binding obligations can clearly be "in-voked" by lower-order against higher-order agen-cies as well as vice versa.This of course implies the relatively firm in-stitutionalization of the normative order itself.Within the framework of a highly differentiatedpolity it implies, in addition to constitutionalismitself, a procedural system for the granting of highpolitical authority, even in private, to say nothingof public organizations, and a legal frameworkwithin which such authority is legitimized. Thisin turn includes another order of procedural in-stitutions within which the question of the legalityof actual uses of power can be tested.

POWER AND AUTHORITYThe institutionalization of the normative order

just referred to thus comes to focus in the conceptof authority. Authority is essentially the insti-tutional code within which the use of power asmedium is organized and legitimized. It standsto power essentially as property, as an institution,does to money. Property is a bundle of rights ofpossession, including above all that of alienation,but also at various levels of control and use. Ina highly differentiated institutional system, prop-erty rights are focussed on the valuation of utility,i.e. the economic significance of the objects, e.g.for consumption or as factors of production, andthis factor comes to be differentiated from au-thority. Thus, in European feudalism the "land-lord" had both property rights in the' land, andpolitical jurisdiction over persons acting on thesame land. In modern legal systems these com-ponents are differentiated from each other so thelandowner is no longer the landlord; this functionis taken over mainly by local political authority.

Precisely with greater differentiation the focusof the institution becomes more generalized and,while specific objects of possession of course con-tinue to be highly important, the most importantobject of property comes to be monetary assets,and specific objects are valued as assets, i.e., interms of potentials of marketability. Today wecan say that rights to money assets, the ways inwhich these can be legitimately acquired and dis-

VOL. 107, NO. 3, 1963] 243


14/32

TALCOTT PARSONSposed of, the ways in which the interests of otherparties must be protected, have come to constitutethe core of the institution of property.24Authority, then, is the aspect of a status in asystem of social organization, namely its collec-tive aspect, by virtue of which the incumbent isput in a position legitimately to make decisionswhich are binding, not only on himself but on thecollectivity as a whole and hence its other mem-ber-units, in the sense that so far as their impli-cations impinge on their respective roles andstatuses, they are bound to act in accordance withthese implications. This includes the right to in-sist on such action though, because of the generaldivision of labor, the holder of authority veryoften is not himself in a position to "enforce" hisdecisions, but must be dependent on specializedagencies for this.If, then, authority be conceived as the institu-tional counterpart of power, the main differencelies in the fact that authority is not a circulatingmedium. Sometimes, speaking loosely, we sug-gest that someone "gives away his property." Hecan give away property rights in specific posses-sions but not the institution of property. Simi-larly the incumbent of an office can relinquish au-thority by resigning, but this is very differentfrom abolishing the authority of the office. Prop-erty as institution is a code defining rights in ob-jects of possession, in the first instance physicalobjects, then "symbolic" objects, including cul-tural objects such as "ideas" so far as they arevaluable in monetary terms, and of course includ-ing money itself, whoever possesses them. Au-thority, similarly, is a set of rights in status in acollectivity, precisely in the collectivity as actor,including most especially right to acquire and usepower in that status.The institutional stability, which is essential tothe conception of a code, then for property inheresin the institutional structure of the market. At ahigher level the institution of property includesrights, not only to use and dispose of particularobjects of value, but to participate in the systemof market transactions.

24Two particularly important manifestations of thismonetizationof property are, first the general legal un-derstandingthat executors of estates are not obligatedtoretain the exact physical inventory intact pending fullstatement,but may sell various items-their fiduciaryob-ligation is focussed on the money value of the estate.Similarly in the law of contract increasing option hasbeen given to compensatewith money damages in lieuof the specific"performance"riginally contractedfor.

It is then essentially the institutionalized codedefining rights of participation in the power sys-tem which I should like to think of as authority.It is this conception which gives us the basis forthe essential distinction between the internal andthe external aspects of power relative to a particu-lar collectivity. The collectivity is, by our con-ception, the definition of the range within whicha system of institutionalized rights to hold anduse power can be closed. This is to say, the im-plications of an authoritative decision made at onepoint in the system can be made genuinely bind-ing at all the other relevant points through therelevant processes of feed-back.The hierarchical priority system of authorityand power, with which this discussion started can,by this criterion, only be binding within a givenparticular collectivity system. In this sense thena hierarchy of authority-as distinguished fromthe sheer differences of power of other coercivecapacities-must be internal to a collectively or-ganized system in this sense. This will includeauthority to bind the collectivity in its relationsto its environment, to persons and to other col-lectivities. But bindingness, legitimized and en-forced through the agency of this particular col-lectivity, cannot be extended beyond its bound-aries. If it exists at all it must be by virtue ofan institutionalized normative order which tran-scends the particular collectivity, through con-tractual arrangements with others, or throughother types of mutually binding obligation.

POWER, INFLUENCE, EQUALIZATION,AND SOLIDARITYIt is on this basis that it may be held that at theboundaries of the collectivity the closed system of

priorities is breached by "free" exercise, at theconstituency or integrative boundary, of influence.Status in the collectivity gives authority to settlethe terms on which power will be exchanged withinfluence over this boundary. The wielder ofinfluence from outside, on the collectivity, is notbound in advance to any particular terms, and itis of the essence of use of power in the "foreignrelations" of the collectivity, that authority is aright, within certain limits of discretion, to spendpower in exchange for influence. This in turncan, through the offer of accepting leadershipresponsibility in exchange for political support,replenish the expenditure of power by a cor-responding input.



15/32

ON THE CONCEPT OF POLITICAL POWERBy this reasoning influence should be capable ofaltering the priority system within the collectivity.This is what I interpret policy decision as a cate-gory of the use of power as a medium to be, theprocess of altering priorities in such a way thatthe new pattern comes to be binding on thecollectivity. Similarly, the franchise must be re-garded as the institutionalization of a marginal,interpenetrating status, between the main collec-tivity and its environment of solidary groupingsin the larger system. It is the institutionalizationof a marginal authority, the use of which is con-fined to the function of selection among candidatesfor leadership responsibility. In the governmentalcase, this is the inclusion in a common collectivitysystem of both the operative agencies of govern-ment and the "constituencies" on which leadershipis dependent, a grant not only in a given instanceof power to the latter but a status of authority withrespect to the one crucial function of selection ofleadership and granting them the authority ofoffice.In interpreting this discussion it is essential tokeep in mind that a society consists, from thepresent point of view, not in one collectivity, butin a ramified system of collectivities. Because,however, of the basic imperatives of effective col-lective action already discussed, these must inaddition to the pluralistic cross-cutting which goeswith functional differentiation,also have the aspectof a "Chinesebox" relation. There must be some-where a paramount focus of collective authorityand with it of the control of power-though it iscrucial that this need not be the top of the totalsystem of normative control, which may for ex-ample be religious. This complex of territorialityand the monopoly of force are central to this,because the closed system of enforceable binding-ness can always be breached by the interventionof force.25The bindingness of normative orders other thanthose upheld by the paramount territorial collec-tivity must be defined within limits institution-alized in relation to it. So far as such collectivitiesare not "agencies" of the state, in this sense, their25 Since this system is the territorially organized col-lectivity, the state with its government, these consider-ations underlie the critical importance of foreign rela-tions in the sense of the relations to other territoriallyorganized, force-controlling collectivities, since, once in-ternal control of force is effectively institutionalized, thedanger of this kind of breach comes from the outside inthis specific sense of outside. The point is cogently madeby Raymond Aron.

spheres of "jurisdiction" must be defined in termsof a normative system, a body of law, which isbinding both on government and on the non-governmental collectivity units, though in the "lastanalysis" it will, within an institutionalized ordereither have to be enforced by government, orcontrariwise, by revolutionary action againstgovernment.Since independentcontrol of serious, socially or-ganized force cannot be given to "private" collec-tivities, their ultimate negative sanctions tend to beexpulsion from membership, though many othertypes of sanction may be highly important.Considerations such as these thus do not inany way eliminate or weaken the importance ofhierarchical priorities within a collective decision-system itself. The strict "line" structure of suchauthority is, however, greatly modified by theinterpenetration of other systems with the politi-cal, notably for our purposes the importance oftechnical competence. The qualifications of theimportance of hierarchy apply in principle at theboundaries of the particular collective system-analytically considered-rather than internally toit. These I would interpret as defining the limitsof authority. There are two main contexts inwhich norms of equality may be expected tomodify the concrete expectations of hierarchicaldecision-systems, namely on the one hand, thecontext of influence over the right to assumepower, or decision-making authority and, on theother hand, the context of access to opportunityfor status as a contributing unit in the specificpolitical system in question.It is essential here to recall that I have treatedpower as a circulating medium, moving back andforth over the boundaries of the polity. The"real" outputs of the political process, and thefactors in its effectiveness-in the sense corre-sponding to the real outputs and factors of eco-nomic production-are not in my sense "forms"of power but, in the most important cases, offinancial control of economic resources, and ofinfluence, in the meaning of the category of in-fluence, defined as a generalized mechanism ofpersuasion. These are very essential elements inthe total political process, but it is just as im-portant to distinguish them from power as it isto distinguish financially valuable outputs andfactors of production from money itself. Theymay, in certain circumstances, be exchangeablefor power, but this is a very different thing frombeing forms of power.

245OL. 107, NO. 3, 1963]


16/32

TALCOTT PARSONSThe circulation of power between polity andintegrative system I conceive to consist in bindingpolicy decisions on the one hand, which is aprimary factor in the integrative process, andpolitical support on the other, which is a primary

output of the integrative process. Support is ex-changed, by a "public" or constituency, for theassumption of leadership responsibility, throughthe process of persuading those in a position togive binding support that it is advisable to do soin the particular instance-through the use ofinfluence or some less generalized means ofpersuasion. In the other political "market"vis-a-vis the integrative system, policy decisionsare given in response to interest-demands in thesense of the above discussion. This is to say thatinterest groups, which, it is most important to noteas a concept says nothing about the moral qualityof the particular interest, attempt to persuadethose who hold authority in the relevant collectiv-ity, i.e. are in a position to make binding decisions,that they should indeed commit the collectivityto the policies the influence-wielders want. Inour terms this is to persuade the decision makersto use and hence "spend" some of their power forthe purpose in hand. The spending of power is tobe thought of, just as the spending of money, asessentially consisting in the sacrifice of alternativedecisions which are precluded by the commitmentsundertaken under a policy. A member of the col-lectivity we conceive as noted to have authority to"spend" power through making binding decisionsthrough which those outside acquire claims againstthe collectivity. Its authority, however, is in-alienable; it can only be exercised, not "spent."It has been suggested that policies must behierarchically ordered in a priority system andthat the power to decide among policies must havea corresponding hierarchical ordering since suchdecisions bind the collectivity and its constituentunits. The imperative of hierarchy does not,however, apply to the other "market" of the powersystem in this direction, that involving the rela-tions between leadership and political support.Here on the contrary it is a critically importantfact that in the largest-scale and most highlydifferentiated systems, namely the leadership sys-tems of the most "advanced"national societies, thepower element has been systematically equalizedthrough the device of the franchise, so that theuniversal adult franchise has been evolved in all

the Western democracies.26 Equality of the fran-chise which, since the consequences of its exerciseare very strictly binding,27I classify as in fact aform of power, has been part of a larger complexof its institutionalization, which includes in addi-tion the principle of universality-its extension toall responsible adult citizens in good standing andthe secrecy of the ballot, which serves to differen-tiate this context of political action from othercontexts of involvement, and protect it againstpressures, not only from hierarchical superiorsbut, as Rokkan points out, from status-peers aswell.Of course the same basic principle of one mem-ber, one vote, is institutionalized in a vast numberof voluntary associations, including many whichare subassociations of wider collectivities, such asfaculties in a university, or boards and committees.Thus the difference between a chairman or presid-ing officer,and an executive head is clearly markedwith respect to formal authority, whatever it maybe with respect to influence, by the principle thata chairman, like any other member, has only onevote. Many collectivities are in this sense "trun-cated" associations, e.g. in cases where fiduciaryboards are self-recruiting. Nevertheless the im-portance of this principle of equality of powerthrough the franchise is so great empirically thatthe question of how it is grounded in the struc-ture of social systems is a crucial one.It derives, I think, from what I should call theuniversalistic component in patterns of normativeorder. It is the value-principle that discrimina-tions among units of a system, must be groundedin intrinsically valued differences among them,which are, for both persons and collectivities, ca-pacities to contribute to valued societal processes.Differences of power in decision-making whichmobilizes commitments, both outward in relationto the environment of the collectivity and inter-nally, to the assignment of tasks to its members,are ideally grounded in the intrinsic conditions ofeffectiveness. Similarly, differences on the basisof technical competence to fulfill essential roles aregrounded in the strategic conditions of effectivecontribution.These considerations do not, however, apply tothe functions of the choice of leadership, where

26See, on this process, Stein Rokkan, "Mass Suffrage,Secret Voting, and Political Participation," EuropeanJournal of Sociology 2 (1961) : 132-152.27I.e., the aggregate of votes, evaluatedby the electoralrules, determinesthe incumbencyof office.



17/32

ON THE CONCEPT OF POLITICAL POWERthis choice has been freed from ascriptive bases ofright, e.g. through kinship status or some imputed"charismatic" superiority as in such a case as"white supremacy." There is a persistent pres-sure of the sufficiently highly valued functions oroutcomes, and under this pressure there seems tohave been a continual, though uneven, process oferosion of discriminations in this critical field ofthe distribution of power.It may be suggested that the principle of univer-salistic normative organization which is immedi-ately superordinate to that of political democracyin the sense of the universal equal franchise, isthe principle of equality before the law; in thecase of the American Constitution, the principleof equal protection of the laws. I have emphasizedthat a constitutional framework is essential toadvanced collective organization, given of courselevels of scale and complexity which precludepurely "informal" and traditional normative regu-lation. The principle in effect puts the burden ofproof on the side of imposing discriminations,either in access to rights or in imposition of obli-gations, on the side that such discriminations areto be justified only by differences in sufficientlyhighly valued exigencies of operation of thesystem.The principle of equality both at the level ofapplication of the law and of the political fran-chise, is clearly related to a conception of the statusof membership. Not all living adults have equalright to influence the affairs of all collectivitieseverywhere in the world, nor does an Americanhave equal rights with a citizen of a quite differentsociety within its territory. Membership is infact the application to the individual unit of theconcept of boundary of a social system which hasthe property of solidarity, in Durkheim's sense.The equal franchise is a prerogative of members,and of course the criteria of membership can bevery differently institutionalized under differentcircumstances.There is an important sense in which the doubleinterchange system under consideration here,which I have called the "support" system linkingthe polity with the integrative aspect of the society,is precisely the system in which power is mostdirectly controlled, both in relation to moreparticularized interest-elements which seek rela-tively particularized policies-which of course in-cludes wanting to prevent certain potential actions-and in relation to the more general "tone" givento the directionality of collective action by the

character of the leadership elements which as-sume responsibility and which, in exchange, areinvested, in the type case by the electoral process,with authority to carry out their responsibilities.One central feature of this control is coming toterms with the hierarchical elements inherent inpower systems in the aspects just discussed. Cer-tain value systems may of course reinforce hier-archy, but it would be my view that a universal-istically oriented value system inherently tends tocounteract the spread of hierarchical patterns withrespect to power beyond the range felt to be func-tionally necessary for effectiveness.28There is, however, a crucial link between theequality of the franchise and the hierarchicalstructure of authority within collectivities, namelythe all-or-none character of the electoral process.Every voter has an equal vote in electing to anoffice, but in most cases only one candidate is infact elected-the authority of office is not dividedamong candidates in proportion to the numbersof votes they received, but is concentrated in thesuccessful candidate, even though the margin bevery narrow, as in the U. S. presidential electionof 1960. There are, of course, considerablepossible variations in electoral rules, but thisbasic principle is as central as is that of theequality of the franchise. This principle seems tobe the obverse of the hierarchy of authority.The hierarchical character of power systemshas above been sharply contrasted with the linearquantitative character of wealth and monetaryassets. This has in turn been related to the funda-mental difference between the exigencies of effec-tiveness in collective action, and the exigenciesof utility in providing for the requirements ofsatisfying the "wants" of units. In order to placethe foregoing discussion of the relations betweenpower and influence in a comparable theoreticalcontext, it is necessary to formulate the value-standard which is paramount in regulating theintegrative function which corresponds to utility

28 Of course where conditions are sufficiently simple, orwhere there is sufficient anxiety about the hierarchialimplications of power, the egalitarian element may pene-trate far into the political decision-making system itself,with, e.g. insistence that policy-decisions, both externaland internal in reference, be made by majority vote ofall members, or even under a unanimity rule. The re-spects in which such a system-which of course real-istically often involves a sharply hierarchical stratifica-tion of influence-is incompatible with effectiveness inmany spheres, can be said to be relatively clear, especiallyfor large collectivities.

247OL. 107, NO. 3, 1963]


18/32

TALCOTT PARSONSand effectiveness in the economic and politicalfunctions respectively.This is, with little doubt, the famous concept ofsolidarity as formulated by Durkheim.29 Thetwo essential points of reference for present pur-poses concern the two main aspects of member-ship, as outlined above, the first of which concernsclaims on executive authority for policy decisionswhich integrate the total collective interest on theone hand, the "partial" interest of a subgroup onthe other. The second concerns integration ofrights to a "voice" in collective affairs with theexigencies of effective leadership and the corre-sponding responsibility.The principle is the "grounding" of a collectivesystem in a consensus in the sense of the abovediscussion, namely an "acceptance" on the partof its members of their belonging together, in thesense of sharing, over a certain range, commoninterests, interests which are defined both by type,and by considerations of time. Time becomes rele-vant because of the uncertainty factor in all humanaction, and hence the fact that neither benefits norburdens can be precisely predicted and planned forin advance; hence an effective collectivity must beprepared to absorb unexpected burdens, and tobalance this, to carry out some sort of just dis-tribution of benefits which are unexpected and/orare not attributable to the earned agency of anyparticular subunit.Solidarity may then be thought of as the im-plementation of common values by definition ofthe requisite collective systems in which they areto be actualized. Collective action as such wehave defined as political function. The famousproblem of order, however, cannot be solvedwithout a common normative system. Solidarityis the principle by virtue of which the commit-ment to norms, which is "based" in turn onvalues, is articulated with the formation of col-lectivities which are capable of effective collectiveaction. Whereas, in the economic direction, the"problem" of effective action is coping with thescarcity of available resources, including tryingto facilitate their mobility, in the integrative direc-tion it is orderly solution of competing claims, onthe one hand to receive benefits-or minimize

29 It is the centralconceptof The Division of Labor inSociety. For my own relatively recent understandingofits significance, see "Durkheim's Contribution to theTheory of Integrationof Social Systems,"in Kurt Wolff,Ed., Smile Durkheim,1858-1917 (Ohio, Ohio State Uni-versity Press, 1960), pp. 118-153.

losses-deriving from memberships, on the otherto influence the processes by which collective ac-tion operates. This clearly involves some institu-tionalization of the subordination of unit-interestto the collective in cases where the two are in con-flict, actual or potential, and hence the justificationof unit interests as compatible with the more ex-tensive collective interest. A social system thenpossesses solidarity in proportion as its membersare committed to common interests through whichdiscrete unit interests can be integrated and thejustification of conflict resolution and subordina-tion can be defined and implemented. It defines,not the modes of implementation of these commoninterests through effective agency, but the stand-ards by which such agency should be guided andthe rights of various constituent elements to havea voice in the interpretation of these standards.

POWERAND EQUALITYOF OPPORTUNITYWe may now turn to the second major bound-ary of the polity, at which another order of modi-fications of the internal hierarchy of authoritycomes to focus. This is the boundary vis-a-visthe economy where the "political" interest is tosecure control of productivity and services, andthe economic interest lies in the collective controlof fluid resources and in what we may call op-portunity for effectiveness. I shall not attempthere to discuss the whole interchange complex,but will confine myself to the crucial problem ofthe way that here also the hierarchical structure ofpower can, under certain conditions, be modifiedin an egalitarian direction.Productivity of the economy is in principle al-locable among collective (in our sense political)claimants to its control as facilities, in linear

quantitative terms. This linear quantification isachieved through the medium of money, either al-location of funds with liberty to expend them atwill, or at least monetary evaluation of more spe-cific facilities.In a sufficiently developed system, services mustbe evaluated in monetary terms also, both fromthe point of view of rational budgeting and of themonetary cost of their employment. In terms oftheir utilization, however, services are "packages"of performance-capacity, which are qualitativelydistinct and of unequal value as contributions tocollective effectiveness. Their evaluation as fa-cilities must hence involve an estimate of strategicsignificance which matches the general priority



19/32

ON THE CONCEPT OF POLITICAL POWERscale which has been established to regulate theinternal functioning of the collectivity.Services, however, constitute a resource to beacquired from outside the collectivity, as Weberputs it through a "formally free" contract of em-ployment. The contracts thus made are bindingon both sides, by virtue of a normative systemtranscending the particular collectivity, thoughthe obligation must articulate with the internalnormative order including its hierarchical aspect.But the purveyors of service are not, in advance,bound by this internal priority system and hencean exchange, which is here interpreted to operatein the first instance as between strategic signifi-cance expressed as power-potential, and the mone-tary value of the service, must be arrived at.

Quite clearly, when the purveyor of service hasonce entered into such a contract, he is bound bythe aspect of its terms which articulates the serviceinto this internal system, including the level ofauthority he exercises and its implications for hispower position in the collectivity. If the collectiv-ity is making in any sense a rational arrangement,this must be tailored to an estimate of the levelof the value of his strategic contribution, hencehis performance-capacity.Since, however, the boundary interchange is notintegral to the internal system of bindingness, thehierarchical imperatives do not apply to the op-portunity aspect of this interchange on the extra-political side. This is to say that the same orderof pressures of a higher-order universalistic nor-mative system can operate here that we suggestedoperated to bring about equality in the franchise.Again the principle is that no particularistic dis-criminations are to be legitimized which are notgrounded in essential functional exigencies of thesystem of reference.In the case of the franchise there seems to be noinherent stopping place short of complete equality,qualified only by the minim

talcottparsons_on the concept of power.pdf

Documents