Structural Functional Approach
Retrieved from: http://www. cifas. us/smith/chapters. html Title: “A structural approach to comparative politics. ” Author(s): M. G. Smith Source: In Varieties of Political Theory. David Easton, ed. Englewood Cliffs, NJ: Prentice-Hall. p. 113-128. Reprinted in Corporations and Society. p. 91-105. FIVE M. G. SMITH University of California, Los Angeles A Structural Approach to Comparative Politics Comparative politics seeks to discover regularities and variations of political organization by comparative analysis of historical and contemporary systems.
Having isolated these regularities and variations, it seeks to determine the factors which underlie them, in order to discover the properties and conditions of polities of varying types. It then seeks to reduce these observations to a series of interconnected propositions applicable to all these systems in both static and changing conditions. Hopefully, one can then enquire how these governmental processes relate to the wider milieux of which they are part. It would seem that this comparative enquiry may be pursued i~. various ways that all share the same basic strategy, but differ in emphases arid sta~ ing points.
Their common strategy is to abstract one aspect of political reality and develop it as a frame of reference. With this variable held constant, enquiries can seek to determine the limits within which other dimensions vary; as the value of the primary variable is changed, the forms and values of the others, separately or together, can also be investigated. Ideally, we should seek to deduce relevant hypotheses from a general body of theory, and then to check and refine them by inductive analyses of historical and ethnographic data. ActuaJ procedures vary. 113 114 /
A STRUCTURAL APPROACH TO COMPARATIVE POLITICS Initially, we might expect anyone of four approaches to be useful in the comparative study of political systems. These four approaches use respectively the dimensions of process, content, function, and form as the bases for their conceptual frameworks. In fact, cOlIlparative studies based on process and content face insuperable obstacles due to the enormous variability of political systems. In centralized polities, the institutional processes of government are elaborately differentiated, discrete, and easy to identify.
They are often the subject, as well as the source, of a more or less complex and precise body of rules which may require specialists to interpret them. In simpler societies, the corresponding processes are rarely differentiated and discrete. They normally occur within the context of institutional activities with multiple functions, and are often difficult to abstract and segregate for analysis as self-contained processual systems. Before this is possible, we need independent criteria to distinguish the governmental and nongovernmental dimensions of these institutional forms.
The substantive approach rests on the category of content. By the con.. tent of a governmental system, I mean its specific substantive concerns and resources, whether material, human, or symbolic. As a rule, the more differentiated and complex the governmental processes are, the greater the range and complexity of content. This follows because the content and processes of government vary together. Since both these frameworks are interdependent and derivative, both presuppose independent criteria for identifying government. The functional approach avoids these limitations.
It defines government functionally as all those activities which influence “the way in which authoritative decisions are formulated and executed for a society. “l From this starting point, various refined conceptual schemes can be developed. As requisites or implications of these decisional processes, David Easton identifies five modes of action as necessary elements of all political systems: legislation, administration, adjudication, the development of demands, and the development of support and solidarity. They may be grouped as input and output requisites of governmental systems.
According to Almond, the universally necessary inputs are political socialization and recruitment, interest articulation, interest aggregation, and political communication. As outputs, he states that rule making, rule application, and rule adjudication are all universa1. 2 Neither of these categorical schemes specifies foreign relations and defense, which are two very general governmental concerns; nor is it easy to see how these schemes could accommodate political processes in non-societal units. Such deductive models suffer from certain inexplicit assumptions with1
David Easton, “An Approach to the Analysis of Political Systems,” World Politics, IX, No. 3 (1957), 384. 2 Gabriel Almond, “Introduction” to Almond and James S. Coleman, The Politics of the Developing Areas (Princeton: Princeton University Press, 1961). A STRUCTURAL APPROACH TO COMPARATIVE POLITICS / 115 out which the initial exclusive stress on political functions might be im- . possible. But despite their universal claims, it remains to be shown that Bushmen, Pygmies, or Eskimos have governments which are functionally homologous with those of the United States and the Soviet Union.
Legislation, rule adjudication, and interest articulation are categories appropriate to the discussion of complex, modern polities rather than simple, primitive ones. But the problem which faces the student of comparative politics is to develop a conceptual framework useful and applicable to all. To impute the features and conditions of modern polities to the less differentiated primitive systems is virtually to abandon the central problem of comparative politics. The functional approach, as usually presented, suffers from a further defect: It assumes a rather special ensemble of structural conditions.
When “authoritative decisions are formulated and executed for a society,” this unit must be territorially delimited and politically centralized. The mode of centralization should also endow government with “more-or-Iess legitimate physical compulsion. “3 In short, the reality to which the model refers is the modern nation-state. By such criteria, ethnography shows that the boundaries of many societies are fluctuating and obscure, and that the authoritative status of decisions made in and for them are even more so.
Clearly bounded societies with centralized authority systems are perhaps a small minority of the polities with which we have to deal. A structural approach free of these functional presumptions may thus be useful, but only if it can accommodate the full range of political systems and elucidate the principles which underlie their variety. In this paper, I shall only indicate the broad outlines of this approach. I hope to present it more fully in the future. Government is the regulation of public affairs.
This regulation is a set of processes which defines government functionally, and which also identifies its content as the affairs which are regulated, and the resources used to regulate them. It does not seem useful or necessary to begin a comparative study of governmental systems by deductive theories which predicate their minimum universal content, requisites, or features. The critical element in government is its public character. Without a public, there can be neither public affairs nor processes to regulate them.
Moreover, while all governments presuppose publics, all publics have governments for the management of their affairs. The nature of these publics is therefore the first object of study. Publics vary in scale, composition, and character, and it is reasonable to suppose that their common affairs and regulatory arrangements will vary correspondingly. The first task of a structural approach to comparative politics is thus to identify the properties of a public and to indicate the principal varieties and bases of publics. 3 Almond, “Introduction,” p. . 116 / A STRUCTURAL APPROACH TO COMPARATIVE POLITICS As I use the term, public does not include mobs, crowds, casual assemblies, or mass-communication audiences. It does not refer to such categories as resident aliens, the ill, aged, or unwed, or to those social segments which lack common affairs and organized procedures to regulate them-for example, slaves, some clans, and unenfranchised strata such as the medieval serfs or the harijans of India. Such categories are part of one or more publics; they are not separate publics of their own.
For example, in an Indian village, a medieval manor, or a slave plantation, members of the disprivileged categories constitute a public only if they form an enduring group having certain common affairs and the organization and autonomy necessary to regulate them; but the existence of such local publics is not in itself sufficient for the strata from which their memberships are drawn to have the status of publics. For this to be the case, these local publics must be organized into a single group co-extensive with the stratum. With such organization, we shall expect to find a set of common affairs and procedures to regulate them.
The organization is itself an important common affair and a system of institutional procedures. By a public, then, I mean an enduring, presumably perpetual group with determinate boundaries and membership, having an internal organization and a unitary set of external relations, an exclusive body of common affairs, and autonomy and procedures adequate to regulate them. It will be evident that a public can neither come into being nor maintain its existence without some set of procedures by which it regulates its internal and external affairs. These procedures together form the governmental process of the public.
Mobs, crowds, and audiences are not publics, because they lack presumptive continuity, internal organization, common affairs, procedures, and autonomy. For this reason, they also lack the determinate boundaries and membership which are essential for a durable group. While the categories mentioned above are fixed and durable, they also lack the internal organization and procedures which constitute a group. When groups are constituted so that their continuity, identity, autonomy, organization, and exclusive affairs are not disturbed by the entrance or exit of their individual members, they have the character of a public.
The city of Santa Monica shares these properties with the United States, the Roman Catholic Church, Bushman bands, the dominant caste of an Indian village, the Mende Pora, an African lineage, a Nahuatl or Slavonic village community, Galla and Kikuyu age-sets, societies among the Crow and Hidatsa Indians, universities, medieval guilds, chartered companies, regiments, and such “voluntary” associations as the Yoruba Ogboni, the Yako lkpungkara, and the American Medical Association. The units just listed are all publics and all are corporate groups; the governmental process inherent in publics is a feature of all corporate groups.
Corporate groups-Maine’s “corporations aggregate”-are one species of “perfect” or fully-fledged corporation, the other being the “corporation A STRUCTURAL APPROACH TO COMPARATIVE POLITICS / 117 sole” exemplified by such offices as the American Presidency, the British Crown, the Papacy, governorships, chieftaincies, and university chancellorships. Corporations sole and corporate groups share the following characteristics, all of which are necessary for “perfect” or full corporate status: identity, presumed perpetuity, closure and membership, autonomy within a given sphere, exclusive common affairs, set procedures, and organization.
The first four of these qualities are formal and primarily external in their reference; they define the unit in relation to its context. The last four conditions are processual and functional, and primarily internal in their reference. The main differences between corporations sole and corporate groups are structural, though developmental differences are also important. Corporate groups are pluralities to which an unchanging unity is ascribed; viewed externally, each forms “one person,” as Fortes characterized the Ashanti matrilineages. This external indivisibility of the corporate group is not merely a jural postulate. It inevitably presumes and involves governmental processes within the group. In contrast with a corporate group, an office is a unique status having only one incumbent at any given time. Nonetheless, successive holders of a common office are often conceived of and addressed as a group. The present incumbent is merely one link in a chain of indefinite extent, the temporary custodian of all the properties, powers, and privileges which constitute the office.
As such, incumbents may legitimately seek to aggrandize their offices at the expense of similar units or of the publics to which these offices relate; but they are not personally authorized to alienate or reduce the rights and powers of the status temporarily entrusted to them. The distinction between the capital of an enterprise and the personalty of its owners is similar to the distinction between the office and its incumbent. It is this distinction that enables us to distinguish ffices from other personal statuses most easily. It is very possible that in social evolution the corporate group preceded the corporation sole. However, once authority is adequately centralized, offices tend to become dominant; and then we often find that offices are instituted in advance of the publics they will regulate or represent, as, for example, when autocrats order the establishment of new towns, settlements, or colonies under officials designated to set up and administer them.
There are many instances in which corporate groups and offices emerge and develop in harmony and congruence, and both may often lapse at once as, for example, when a given public is conquered and assimilated. These developmental relations are merely one aspect of the very variable but fundamental relation between offices and corporate groups. Despite Weber, there are a wide range of corporate groups which lack stable leaders, 4 Meyer Fortes, “Kinship and Marriage among the Ashanti,” in African Sys- tems of Kinship and Marriage, eds. A. R.
Radcliffe-Brown and Daryll Forde (London: Oxford University Press, 1950), pp. 254-61. 118 / A STRUCTURAL APPROACH TO COMPARATIVE POLITICS much less official heads. Others may have senior members whose authority is at best advisory and representative; yet others have a definite council or an official head, or both. In many cases, we have to deal with a public constituted by a number of coordinate corporate groups of similar type. The senior members of these groups may form a collegial body to administer the common affairs of the public, with variable powers.
Ibo and Indian village communities illustrate this well. In such contexts, where superordinate offices emerge, they often have a primarily sacred symbolic quality, as do the divine kingships of the Ngonde and Shilluk, but lack effective secular control. Between this extreme and an absolute despotism, there are a number of differing arrangements which only a comparative structural analysis may reduce to a single general order. Different writers stress different features of corporate organization, and sometimes employ these to “explain” these social forms.
Weber, who recognizes the central role of corporate groups in political systems, fails to distinguish them adequately from offices (or “administrative organs,” as he calls them). 5 For Weber, corporate groups are defined by coordinated action under leaders who exercise de facto powers of command over them. The inadequacy of this view is patent when Barth employs it as the basis for denying to lineages and certain other units the corporate status they normally have, while reserving the term corporate for factions of a heterogeneous and contingent character. Maine, on the other hand, stresses the perpetuity of the corporation and its inalienable bundle of rights and obligations, the estate with which it is indentified. 7 For Gierke,s Durkheim,9 and Davis,10 corporate groups are identified by their common will, collective conscienc~, and group personality. For Goody, only named groups holding material property in common are corporate. 1! These definitions all suffer from overemphasis on some elements, and corresponding inattention to others. The common action characteristic of corporate groups rarely embraces the application of violence which both Weber and Barth seem to stress.
Mass violence often proceeds independ5 Max Weber, The Theory of Social and Economic Organization, trans. A. R. Henderson and Talcott Parsons (London: Wm. Hodge & Co. , 1947), pp. 133-37, 302-5. 6 Fredrik Barth, Political Leadership among Swat Pathans. Monographs in Social Anthropology, London School of Economics, No. 19 (London: University of London Press, 1959). 7 H. S. Maine, Ancient Law (London: Routledge & Kegan Paul, Ltd. , 1904), p. 155. S Otto Gierke, Natural Law and the Theory of Society, 1500 to 1800, trans. Ernest Barker (Boston: Beacon Press, 1957). Emile Durkheim, The Division of Labour in Society, trans. George E. Simpson (New York: Free Press of Glencoe, Inc. , 1933). 10 John P. Davis, Corporations (New York: Capricorn Books, 1961), p. 34. 11 Jack Goody, “The Classification of Double Descent Systems,” Current Anthropology; II, No. 1 (1961), 5, 22-3. A STRUCTURAL APPROACH TO COMPARATIVE POLITICS / 119 ently of corporate groups. Corporate action is typically action to regulate corporate affairs-that is, to exercise and protect corporate rights, to enforce corporate obligations, and to allocate corporate responsibilities and privileges.
When a group holds a common estate, this tenure and its exercise inevitably involve corporate action, as does any ritual in which the members or representatives of the group engage as a unit. Even the maintenance of the group’s identity and closure entails modes of corporate action, the complexity and implications of which vary with the situation. It is thus quite fallacious to identify corporate action solely with coordinated physical movements. A chorus is not a corporate group.
The presumed perpetuity, boundedness, determinate membership, and identity of a corporation, all more or less clearly entail one another, as do its requisite features of autonomy, organization, procedure, and common affairs. It is largely because of this interdependence and circularity among their elements that corporations die so hard; but by the same token, none of these elements alone can constitute or maintain a corporation. An office persists as a unit even if it is not occupied, providing that the corpus of rights, responsibilities, and powers which constitute it still persists.
To modify or eliminate the office, it is necessary to modify. or eliminate its content. Among ! Kung bushmen, bands persist as corporate groups even when they have no members or heads12 ; these bands are units holding an inalienable estate of water holes, veldkos areas, etc. , and constitute the fixed points of ! Kung geography and society. The Bushman’s world being constituted by corporate bands, the reconstitution of these bands is unavoidable, whenever their dissolution makes this necessary.
As units which are each defined by an exclusive universitas juris, corporations provide the frameworks of law and authoritative regulation for the societies that they constitute. The corporate estate includes rights in the persons of its members as well as in material or incorporeal goods. In simpler societies, the bulk of substantive law consists in these systems of corporate right and obligation, and includes the conditions and correlates of membership in corporate groups of differing type. In such societies, adjectival law consists in the usual modes of corporate procedure. To a much greater extent than is commonly ealized, this is also the case with modern societies. The persistence, internal autonomy, and structural uniformity of the corporations which constitute the society ensure corresponding uniformity in its jural rules and their regular application over space and time. As modal units of social process and structure, corporations provide the framework in which the jural aspects of social relations are defined and enforced. Tribunals are merely functionally specific corporations charged with handling issues of certain kinds. Neither tribunals nor “the systematic ap12 Lorna Marshall, “! Kung Bushmen Bands,” A/rica, XXX (1960), 325- 5). 120 / A STRUCTURAL APPROACH TO COMPARATIVE POLITICS plication of the force of politically organized society”13 are necessary or sufficient for the establishment of law. The law of a primitive society consists in its traditional procedures and modes of corporate action, and is implicit in the traditional rights, obligations, and conditions of corporate membership. In such societies, units which hold the same type of corporate estate are structurally homologous, and are generally articulated in such a way that each depends on the tacit recognition or active support of its fellows to maintain and enjoy its estate.
Thus, in these simpler systems, social order consists in the regulation of relations between the constitutive corporations as well as within them. In societies which lack central political organs, societal boundaries coincide with the maximum range of an identical corporate constitution, on the articulation of which the social order depends. Though the component corporations are all discrete, they are also interdependent. But they may be linked together in a number of different ways, with consequent differences in their social systems.
In some cases, functionally distinct corporations may be classified together in purely formal categories, such as moieties, clans, or castes. The Kagoro of northern Nigeria illustrate this. 14 In other cases, corporations which are formally and functionally distinct may form a wider public having certain common interests and affairs. The LoDagaba of northern Ghana and Upper Volta are an example. 15 In still other cases, corporations are linked individually to one another in a complex series of alliances and associations, with overlapping margins in such a way that they all are related, directly or indirectly, in the same network.
Fortes has given us a very detailed analysis of such a system among the Tallensi. 16 However they are articulated in societies which lack central institutions, it is the extensive replication of these corporate forms which defines the unit as a separate system. Institutional uniformities, which include similarities of organization, ideology, and procedure, are quite sufficient to give these acephalous societies systemic unity, even where, as among the Kachins of Burma, competing institutional forms divide the allegiance of their members. 7 To say that corporations provide the frameworks of primitive law, and that the tribunals of modem societies are also corporate forms, is simply to say that corporations are the central agencies for the regulation of public affairs, being themselves each a separate public or organ, administering certain affairs, and together constituting wider publics or associations of publics 13 Roscoe Pound, Readings on the History and System 0/ the Common Law, 2nd ed. (Boston: Dunster House Bookshop, 1913), p. 4. 14 M. G.
Smith, “Kagoro Political Development,” Human Organization, XIX, No. 3 (1960), 37-49. 15 Jack Goody, “Fields of Social Control among the LoDagaba,” Journal of the Royal Anthropological Institute, LXXXVII, Part I (1957),75-104. 16 Meyer Fortes, The Dynamics 0/ Clanship among the Tallensi (London: Oxford University Press, 1945). 17 E. R. Leach, Political Systems of Highland Burma (London: G. Bell & Sons, Ltd. , 1954). A STRUCTURAL APPROACH TO COMPARATIVE POLITICS / 121 for others. By the same token, they are the sources or frameworks of disorder.
In some acephalous societies, disorder seems more or less perennial, and consists mainly in strife within and between corporations. Centralization, despite its merits, does not really exclude disorder. In concentrating authority, it simultaneously concentrates the vulnerability of the system. Accordingly, in centralized societies, serious conflicts revolve around the central regulative structures, as, for instance, in secessionist or revolutionary struggles, dynastic or religious wars, and “rituals of rebellion. “18 Such conflicts with or for central power normally affect the entire social body.
In acephalous societies, on the other hand, conflicts over the regime may proceed in one region without implicating the others. 19 In both the centralized and decentralized systems, the sources and objects of conflict are generally corporate. Careful study of Barth’s account of the Swat Pathans shows that this is true for them also, although the aggregates directly contraposed are factions and blocs. 20 Societal differences in the scale, type, and degree of order and coordination, or in the frequency, occasions, and forms of social conflict are important data and problems for political science.
To analyze them adequately, one must use a comparative structural approach. Briefly, recent work suggests that the quality and modes of order in any social system reflect its corporate constitution-that is, the variety of corporate types which constitute it, their distinctive bases and properties, and the way in which they are related to one another. The variability of political systems which derives from this condition is far more complex and interesting than the traditional dichotomy of centralized and noncentralized systems would suggest.
I have already indicated some important typological differences within the category of acephalous societies; equally significant differences within the centralized category are familiar to all. This traditional dichotomy assumes that centralization has a relatively clear meaning, from which a single, inclusive scale may be directly derived. This assumption subsumes a range of problems which require careful study; but in any event, centralization is merely one aspect of political organization, and not necessarily the most revealing.
Given variability in the relations between corporations sole and corporate groups, and in their bases and forms, it seems more useful to distinguish systems according to their structural simplicity or complexity, by reference to· the variety of corporate units of differing forms, bases, and functions which they contain, and the principles which serve to articulate them. Patently, such differences in composition imply differences in the relational networks in which these corporations articulate. Such ifferences in structural composition simultaneously describe the variety of political forms 18 Max Gluckman, Rituals of Rebellion in South East Africa (Manchester: Manchester University Press, 1954); “Introduction” to Gluckman, Order and Rebellion in Tribal Africa (London: Cohen & West, 1963). 19 Leach, Political Systems 0/ Highland Burma. 20 Barth, Political Leadership among Swat Pathans. 122 / A STRUCTURAL APPROACH TO COMPARATIVE POLITICS and processes, and explain differences in the scale, order, and coordination of polities.
This is so because corporate organization provides the framework, content, and procedures for the regulation of public affairs. For this reason, the analysis of corporate structure should be the first task in the case study of a political system and in comparative work. For many political scientists, the concept of sovereignty is essential as the foundation of governmental order and autonomy. In my view, this notion is best dispensed with. It is a hindrance rather than a help to analysis, an unhappy solution of a very real problem which has been poorly formulated. In a system of sovereign states, no state is sovereign.
As etymology shows, the idea of sovereignty derives from the historically antecedent condition of personal dominion such as kingship, and simply generalizes the essential features of this form as an ideology appropriate to legitimate and guide other forms of centralization. The real problem with which the notion of sovereignty deals is the relation between autonomy and coordination. As the fundamental myth of the modern nation-state, the concept is undoubtedly important in the study of these states; its historical or analytical usefulness is otherwise very doubtful.
It seems best to formulate the problems of simultaneous coordination and autonomy in neutral terms. As units administering exclusive common affairs, corporations presuppose well-defined spheres and levels of autonomy, which are generally no more nor less than the affairs of these units require for their adequate regulation. Where a corporation fully subsumes all the juridical rights of its members so that their corporate identification is exclusive and lifelong, the tendencies toward autarchy are generally greatest, the stress on internal autonomy most pronounced, and relations between corporations most brittle.
This seems to be the case with certain types of segmentary lineage systems, such as the Tallensi. Yet even in these conditions, and perhaps to cope with them, we usually find institutional bonds of various types such as ritual cooperation, local community, intermarriage, clanship, and kinship which serve to bind the autarchic individual units into a series of wider publics, or a set of dyadic or triadic associations, the members of which belong to several such publics simultaneously.
Weber’s classification of corporate groups as heteronomous or autonomous, heterocephalous or autocephalous, touches only those aspects of this problem in which he was directly interested. 21 We need also to analyze and compare differing levels, types, and degrees of autonomy and dependence in differing social spheres and situations. From comparative studies of these problems, we may hope to derive precise hypotheses about the conditions and limits of corporate autonomy and articulation in systems of differing composition and span. These hypotheses should also illuminate the conditions and limits of social disorder.
Besides the “perfect” or fully-fledged corporations, offices and corpo21 Weber, Theory 0/ Social and Economic Organization, pp. 135-36. A STRUCTURAL APPROACH TO COMPARATIVE POLITICS / 123 rate groups, there are “imperfect” quasicorporations with must also be studied explicitly. The two main forms here are the corporate category and the commission. A corporate category is a clearly bounded, identifiable, and permanent aggregate which differs from the corporate group in lacking exclusive common affairs, autonomy, procedures adequate for their regulation, and the internal organization which constitutes the group.
Viewed externally, acephalous societies may be regarded as corporate categories in their geographical contexts, since each lacks a single inclusive frame of organization. But they are categories of a rather special type, since, as we have seen, their institutional uniformity provides an effective basis for functional unity. In medieval Europe, serfs formed a corporate category even though on particular manors they may have formed corporate groups.
Among the Turkana22 and Karimojong23 of East Africa, age-sets are corporate categories since they lack internal organization, exclusive affairs, distinctive procedures, and autonomy. Among the nearby Kipsigi24 and Nandi25 clans are categorical units. These clans have names and identifying symbols, a determinate membership recruited by agnatic descent, certain ritual and social prohibitions of which exogamy is most important, and continuity over time; but they lack internal organization, common affairs, procedures and autonomy to regulate them.
Though they provide a set of categories into which all members of these societies are distributed, they never function as social groups. Not far to the south, in Ruanda, the subject Hutu caste formed a corporate category not so long ago. 26 This “caste” had a fixed membership, closure, easy identification, and formed a permanent structural unit in the Tutsi state. Rutu were excluded from the political process, as a category and almost to a man. They lacked any inclusive internal organization, exclusive affairs, autonomy, or procedures to regulate them.
Under their Tutsi masters, they held the status of serfs; but when universal suffrage was recently introduced, Rutu enrolled in political parties such as the Parmehutu Aprosoma which succeeded in throwing off the Tutsi yoke and expelling the monarchy. 27 In order to become corporate groups, corporate categories need to develop an effective representative organization, such for instance as may now be emerging among American Negroes. In the American case, this corporate category is seeking to organize itself in order to remove the disprivileges which define it as a category.
Some corporate 22 Philip Gulliver, “The Turkana Age Organization,” American Anthropologist, LX (1958), 900-922. 23 Neville Dyson-Hudson, to author, 1963. 24 J. G. Peristiany, The Social Institutions of the Kipsigis (London: Routledge & Kegan Paul, Ltd. , 1939). 25 G. W. B. Huntingford, The Nandi of Kenya (London: Routledge & Kegan Paul, Ltd. , 1953). 26 J. J. Maquet, The Premise of Inequality in Ruanda (London: Oxford University Press, 1960). 27 Marcel d’Hertefelt, “Les Elections Communales et Ie Consensus Politique au Rwanda,” Zaire, XIV, Nos. -6 (1960), 403-38. 124 / A STRUCTURAL APPROACH TO COMPARATIVE POLITICS categories are thus merely formal units lacking common functions; others are defined by common disabilities and burdens, though lacking common affairs. Under Islam, the dhimmi formed such a category; in India, so do the individual castes. The disabilities and prohibitions which define categories are not always directly political; they include exogamy and ritual taboos. Commissions differ from offices along lines which recall the differences between corporate categories and corporate groups.
Like categories, commissions fall into two main classes: one class includes ad hoc and normally discontinuous capacities of a vaguely defined character, having diffuse or specific objects. The other class includes continuing series of indefinite number, the units of which are all defined in such general terms as to appear structurally and functionally equivalent and interchangeable. Familiar examples of the latter class are military commissions, magistracies, professorships, and priesthoods; but the sheiks and sa’ids of Islam belong here also.
Examples of the first class, in which the powers exercised are unique but discontinuous and ill-defined, include parliamentary commissions of enquiry or other ad hoc commissions, and plenipotentiaries commissioned to negotiate special arrangements. In some societies, such as the Eskimo, Bushman, and Nuer, individuals having certain gifts may exercise informal commissions which derive support and authority from public opinion. The Nuer “bull,” prophet, and leopard-skin priests are examples. 28 Among the Eskimos, the shaman and the fearless hunter-warrior have similar positions. 9 The persistence of these commissions, despite turnover of personnel and their discontinuous action, is perhaps the best evidence of their importance in these social systems. For their immediate publics, such commissions personalize social values of high relevance and provide agencies for ad hoc regulation and guidance of action. In these humble forms, we may perceive the seeds of modern bureaucracy. Commissions are especially important as regulatory agencies in social movements under charismatic leaders, and during periods of popular unrest.
The charismatic leadership is itself merely the supreme directing commission. As occasion requires, the charismatic leader creates new commissions by delegating authority and power to chosen individuals for special tasks. The careers of Gandhi, Mohammed, Hitler, and Shehu Usumanu dan Fodio in Hausaland illustrate this pattern well. So does the organization and development of the various Melanesian “cargo cults. “30 But if the commission is to be institutionalized as a unit of permanent administration, its arbitrary 28 E. E. Evans-Pritchard, The Nuer (London: Oxford University Press, 940). 29 Kaj Birket-Smith, The Eskimo (London: Meuthuen & Co. , Ltd. , 1960); V. Stefansson, My Life with the Eskimo (New York: The Crowell-Collier Publishing Co. , 1962). 80 Peter Worsley, The Trumpet Shall Sound (London: McGibbon & Kee, 1957). A STRUCTURAL APPROACH TO COMPARATIVB POLma / 125 character must be replaced by set rules, procedures, and spheres of action; this institutionalization converts the commission into an office in the same way that its organization converts the corporate category into a corporate group.
Moreover, in the processes by which corporate categories organize themselves as groups, charismatic leadership and its attached commissions are the critical agencies. The current movement for civil rights among American Negroes illustrates this neatly. Any given public may include offices, commissions, corporate categories, and corporate groups of differing bases and type. In studying governmental systems, we must therefore begin by identifying publics and analyzing their internal constitution as well as their external relationships in these terms.
It is entirely a matter of convenience whether we choose to begin with the smallest units and work outwards to the limits of their relational systems, or to proceed in the opposite direction. Given equal thoroughness, the results should be the same in both cases. Any governmental unit is corporate, and any public may include, wholly or in part, a number of such corporations. These units and their interrelations together define the internal order and constitution of the public and its network of external relations.
Both in the analysis of particular systems and in comparative work, we should therefore begin by determining the corporate composition of the public under study, by distinguishing its corporate groups, offices, commissions, and categories, and by defining their several properties and features. As already mentioned, we may find, in some acephalous societies, a series of linked publics with intercalary corporations and overlapping margins. We may also find that a single corporate form, such as the Mende Para or the Roman Catholic Church, cuts across a number of quite distinct and mutually independent publics.
An alternative mode of integration depends on the simultaneous membership of individuals in several distinct corporations of differing constitution, interest and kind. Thus, an adult Yako81 simultaneously belongs to a patrilineage, a matrilineage, an age-set in his ward, the ward (which is a distinct corporate group), one or more functionally specific corporate associations at the ward or village level, and the village, which is the widest public. Such patterns of overlapping and dispersed membership may characterize both individuals and corporations equally.
The corporations will then participate in several discrete publics, each with its exclusive affairs, autonomy, membership, and procedures, just as the individual participates in several corporations. It is this dispersed, multiple membership which is basic to societal unity, whether or not government is centralized. Even though the inclusive public with a centralized authority system is a corporate group, and a culturally distinct population 81Daryll Forde, Yako Studies (London: Oxford University Press, 1964); Kenneth Little, The -Mende of Sierra Leone (London: Routledge & Kegan Paul, Ltd. 1951). 126 / A STRUCTURAL APPROACH TO COMPARATIVE POLITICS without this remains a corporate category, functionally both aggregates derive their underlying unities from the same mechanism of crosscutting memberships, loyalties, and cleavages. In the structural study of a given political system, we must therefore define its corporate constitution, determine the principles on which these corporate forms are based, and see how they articulate with one another.
In comparative study, we seek to determine what differences or uniformities of political process, content, and function correspond with observable differences or uniformities of corporate composition and articulation. For this purpose, we must isolate the structural principles on which the various types of corporations are based in order to determine their requisites and implications, and to assess their congruence or discongruence. To indicate my meaning, it is sufficient to list the various principles on which corporate groups and categories may be based.
These include sex, age, locality, ethnicity, descent, common property interests, ritual and belief, occupation, and “voluntary” association for diffuse or specific pursuits. Ethnographic data show that we shall rarely find corporate groups which are based exclusively on one of these principles. As a rule, their foundations combine two, three, or more principles, with corresponding complexity and stability in their organization. Thus, lineages are recruited and defined by descent, common property interests, and generally co-residence.
Besides equivalence in age, age-sets presume sameness of sex and, for effective incorporation, local co-residence. Guilds typically stressed occupation and locality; but they were also united by property interests in common market facilities. In India, caste is incorporated on the principles of descent, ritual, and occupation. Clearly, differing combinations of these basic structural principles will give rise to corporations of differing type, complexity, and capacity; and these differences will also affect the content, functions, forms, and contextual relations of the units which incorporate them.
It follows that differing combinations of these differing corporate forms underlie the observable differences of order and process in political organization. This is the broad hypothesis to which the comparative- structural study of political systems leads. It is eminently suited to verification or disproof. By the same token, uniformities in corporate composition and organization between, as well as within, societies should entail virtual identities of political process, content, and form.
When, to the various possible forms of corporate group differentiated by the combination of structural principles on which they are based and by the relations to their corporate contexts which these entail, we add the other alternatives of office, commission, and category, themselves variable with respect to the principles which constitute them, we simultaneously itemize the principal elements which give rise to the variety of political forms, and the principles and methods by which we can reasonably hope
A STRUCTURAL APPROACH TO COMPARATIVE POLITICS / 127 to reduce them to a single general order. Since corporations are essential regulatory units of variable character, their different combinations encompass the entire range of variability of political systems on the functional, processual, and substantive, as well as on the structural levels. Within this structural framework, we may also examine the nature of the regulatory process, its constituents, modes, and objectives.
The basic elements of regulation are authority and power. Though always interdependent and often combined, they should not be confused. As a regulatory capacity, authority is legitimated and identified by the rules, traditions, and precedents which embody it and which govern its exercise and objects. Power is also regulatory, but is neither fully prescribed nor governed by norms and rules. Whereas authority presumes and expresses normative consensus, power is most evident in conflict and contraposition where dissensus obtains.
In systems of public regulation, these conditions of consent and dissent inevitably concur, although they vary in their forms, objects, and proportions. Such systems accordingly depend on the simultaneous exercise and interrelation of the power and authority with which they are identified. Structural analysis enables us to identify the various contexts in which these values and capacities appear, the forms they may take, the objectives they may pursue, and their typical relations with one another within as well as between corporate units.
In a structurally homogeneous system based on replication of a single corporate form, the mode of corporate organization will canalize the authority structure and the issues of conflict. It will simultaneously determine the forms of congruence or incongruence between the separate corporate groups. In a structurally heterogeneous system having a variety of corporate forms, we shall also have to look for congruence or incongruence among corporations of differing types, and for interdependence or competition at the various structural levels.
Any corporate group embodies a set of structures and procedures which enjoy authority. By definition, all corporations sole are such units. Within, around, and between corporations we shall expect to find recurrent disagreements over alternative courses of action, the interpretation and application of relevant rules, the allocation of positions, privileges and obligations, etc. These issues recurrently develop within the framework of corporate interests, and are settled by direct or indirect exercise of authority and power.
Few serious students now attempt to reduce political systems to the modality of power alone; but many, under Weber’s influence, seek to analyze governments solely in terms of authority. Both alternatives are misleading. Our analysis simultaneously stresses the difference and the interdependence of authority and power. The greater the structural simplicity of a given system, that is, its dependence on replication of a single corporate form, such as the Bushman band or Tallensi lineage, the greater its decen- 28 / A STRUCTURAL APPROACH TO COMPARATIVE POLITICS tralization and the narrower the range in which authority and power may apply. The greater the heterogeneity of corporate types in a given system, the greater the number of levels on which authority and power are simultaneously requisite and manifest, and the more critical their congruence for the integration of the system as a whole.