Concepts and Generalizations in Comparative Sociological Studies

Reinhard Bendix

American Sociological Review, Vol. 28, No. 4. (Aug., 1963), pp. 532-539.

Stable URL: http://links.jstor.org/sici?sici=0003-1224%28196308%2928%3A4%3C532%3ACAGICS%3E2.0.CO%3B2-T

American Sociological Review is currently published by American Sociological Association.

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

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

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

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

http://www.jstor.org Thu Feb 28 08:30:58 2008 CONCEPTS AND GENERALIZATIOKS IN COhfPARATIVE SOCIOLOGICAL STUDIES * REINHARDBENDJX , Berkeley

Comparative sociological studies attempt to develop concepts and generalizations at a level between "pure theory" and descriptive area studies. They help to elucidate the time-and- space limitations of sociological concepts that have less than universal applicability and to uncover the generalizations hidden in many "composite concepts." Positively, they can help us develop typologies of social actions and structures and assess their characteristic range of variation. These contributions are exemplified on the basis of classic comparative studies and of the author's studies of the "political conznzunity."

IKE the concepts of other disciplines, Other concepts can be related logically to sociological concepts should be uni- 'Lascription": emphasis on personal attri- L versally applicable. The concept "di- butes is incompatible with emphasis on uni- vision of labor," for instance, refers to the versal standards, a neutral or impersonal fact that the labor performed in a collec- attitude, and equalitarianism. Empirically, tivity is specialized; the concept is universal however, the meaning both of these terms because we know of no collectivity without and of their interrelations is in doubt. To such specialization. Where reference is made be analytically useful, universal concepts re- to a principle of the division of labor over quire specifications which will help us bridge time-irrespective of the particular indi- the gap between concept and empirical evi- viduals performing the labor and of the way dence. Emphasis on the ascriptive criterion labor is subdivided (whether by sex, age, of birth may refer equally well to a person skill or whatever)-we arrive at one mean- with an ancient aristocratic lineage as to a ing of the term "social organization." We person whose family acquired its title by pur- know of no society that lacks such a princi- chase. In one cultural setting emphasis on ple; furthermore, we can compare and con- beauty may outweigh emphasis on high trast the social organization of two societies birth; in another it may not. Such differ- by showing how their division of labor ences, rather than the predominance of as- differs. criptive criteria as such, provide the clues It is possible to remain at this level of for a sociological analysis of diverse social universal concepts. A whole series of mu- structures. tually related concepts can be elaborated These considerations point to a persistent deductively in an effort to construct a frame- problem in . Concepts and theories work of concepts applicable to all societies. are difficult to relate to empirical findings Such efforts in "pure theory," however, on the one hand, while much empirical re- should be subjected to periodic checks of search is devoid of theoretical significance the analytic utility of the concepts. For on the other. Many sociologists deplore this example, the concept "ascription" apparently hiatus, but the difficulties persist and tend refers unambiguously to a principle of as- to reinforce the claims of "pure theory" and signment to roles based on the attributes "pure methodology," respectively. Compara- rather than the performances of a person. tive sociological studies represent an attempt to develop concepts and generalizations at a * Revision of a paper presented to the Fifth level between what is true of all societies and World Congress of Sociology, Washington, Septem- ber, 1962, and published in its original form in what is true of one society at one point in Transactions of the Fifth World Congress of Soci- time and space. In fact, many sociological ology, Louvain: International Sociological Asso- concepts impIy such an "intermediate level" ciation, 1963. I am indebted to my colleagues Wit- of analysis, though frequently they are used liam Petersen and for their criticisms of the earlier draft. as if they applied universally. COMPARATIVE SOCIOLOGICAL STUDIES 533

CONCEPTS OF LIMITED APPLICABILITY tangled conceptually, and once untangled each of these dimensions would be universal. Concepts of socio-historical configurations Perhaps so, but it has yet to be demonstrated are a case in point. Stratification is present that this gain in universality is not obtained in all societies, but stratification by "class" at the expense of analytic utility. So far it is present only in some. Classes depend upon appears to me that it is necessary to use the voluntary coalescence of interests among "composite concepts" if one wishes to apply individuals and thus they differ from strati- concepts rather than elaborate them de- fication by "estate," or court-rank, or clan- ductively. Hence I see the utility of logical affiliation. Again, the exercise of authority decomposition more in the clarification of requires subordinate agents everywhere, but concepts than as a ground for abandoning their organization in a "bureaucracy" is a them. more specific phenomenon. Bureaucracy in Usually, we meet these difficulties by con- the sense of 's concept of govern- structing a contrast-conception. "Bureauc- mental organization under the rule of law racy" is hardly a usable concept as long as applies principally to the countries of it stands alone. It gains clarity when we con- Northwestern Europe from the nineteenth trast it with the "patrimonial" form of gov- century onward. However, several elements ernment, as Max Weber has done, because of bureaucratic organization can be found in this way we learn of a non-bureaucratic centuries earlier, as T. F. Tout has docu- type of government administration that has mented in his five-volume work, deceptively a century-long development of its own. entitled Chapters in the Administrative His- Again, stratification by "class" is a better tory of England. Also, elements from the analytic tool when contrasted with alterna- "bureaucratic" complex have been adopted tive types of stratification. While such paired in many countries throughout history-with concepts are never wholly satisfactory, they varying success to be sure. do enable us to delimit the space-and-time Such historical delimitations of the ap- dimension of a given concept to some e~tent.~ plicability of a concept are clear only in Comparative analysis reveals also that principle; they are very vague in practice. many concepts are generalizations in dis- Though it is possible to date the inception guise. Urbanism is a case in point, as are of "bureaucracy" in England from the other concepts of complex structures, such Northcote-Trevelyan Reforms of 186 1, even as industrial society, bureaucracy, democ- so marked an institutional innovation is no racy, feudalism, caste society, etc., together more than a "high-water mark" of changes with related "developmental" terms, such in English administration whose century- as urbanization, industrialization, and so old continuity can be documented easily. on. These concepts define social structures The delimitation in time and space of other with regard to their several distinguishing sociological concepts presents even greater characteristics. If we are to refer to social difficulties, since most such concepts are not structures, we must define a cluster of at- reflected in legal or administrative docu- tributes that distinguishes one structure ment~.~Thus it is both difficult to identify from an~ther.~It is a fiction to suppose that the space-and-time dimensions of certain

sociological concepts and difficult to deny 2 1prefer to avoid the term "ideal type" since it that they possess an historically limited ap- requires too many explanations of its meaning to plicability. It may be argued that this is be useful. Cf. the earlier discussion of "paired con- cepts" in Reinhard Bendix and Bennett Berger, all the more reason for treating such con- "Images of Society and Problems of Concept-For- cepts as unanalyzed composites of several mation in Sociology," in Llewellyn Gross, (ed.) analytic dimensions which need to be un- Cont~ibz~tiorzsto Sociological Theory, Evanston: Row, Peterson & Co.. 1959, pp. 92-118. 3 The "pattern variables" as formulated by Tal- 1 No doubt it is due in part to these difficulties cott Parsons are an aid in this respect, provided of chronologicaI delimitation that sociologists and their application to spzcific structures is spelled out. historians do not get along well intellectually; yet "Universa!i5m," for instance, nlay be an attribute they need each other, for historians use sociologi- of urbanism and of bureaucracy, but it is necessary cal concepts and sociologists ought to derive part to explicate how this characteristic, which distin- of their evidence from history. guishes urban from folk and bureaucratic from 534 AR1II;RICAN SOCIOLOGICAL KEVlEW these attributes generally occur together; phenomena from the earlier history of West- after all, the conventional definition, say, of ern cities also and hence it is tempting to urbanism puts into abstract terms what we predict that with sufficient economic devel- have learned about some Western cities as opment and urbanization this tie to the distinguished from non-urban types of settle- country will be broken in India as well. ment. Hence, comparative sociological But what began as a definition has subtly studies are needed to delimit the applica- turned into a prediction based on a general- bility of those attributes; here we are back ization about "urbanism," though this pre- to the space-and-time dimension of socio- diction is hazardous. For in the West the logical concepts. Even more important, such religious consecration of family ties had been studies would enable us to examine critically broken long before modern urbanization the implicit and, in my judgment, unjustified occurred. If by individualism we understand generalization, according to which the sev- this destruction of the fetters of kinship, eral attributes of "urbanism" tend to occur then individualism was a precondition as and vary together. much as a consequence of that urbanization. Recent observations in India suggest that And as urbanization in India occurs in the the generalizations and expectations we as- absence of comparable preconditions we sociate with the term "urbanization" may be must expect that it will take unfamiliar excessively culture-bound. In India kinship forms. Furthermore, it is difficult to antici- ties between urban and rural residents re- pate these forms. In the UKESCO Seminar main strong. Examples: In a recent flood Report on this subject we read the following: disaster in Poona about one-third of the Although the great cities of Asia have large people made homeless (some 30,000) simply size, high density and heterogeneous popula- rejoined their families in the villages; in tions these characteristics (which according Bombay, textile workers on strike go back to Wirth essentially give rise to the urban to their villages for the duration of the way of life), have not produced the basic changes in interpersonal relations, the nature strike; in many cities husbands go to work of human beings and the social institutions, by themselves, leaving their wives and chil- as in the Western context. Despite these dren in the countryside. A recent survey of relatively high densities, life has not neces- the ((urban social situation in India" con- sarily become largely secularized, great dif- cludes that rural-urban differences with re- ferentiation of function has not taken place and the way of life has not changed markedly gard to such key factors as caste and joint for many of the indigenous population gr0ups.j family have not in fact developed as ex- pe~ted.~We know about somewhat similar Although the report goes on to deny "in- creased sophistication, rationality of be- patrimonial administration, becomes manifest in havior, cosmopolitanism in outlook, or in- each, and how these manifestations may be linked. novation and social change," with regard Moreover, sociologists interested in social change have often used the logical compatibility among to these cities, these qualities of "urbanism" such pattern-variables as universalism, achieve- surely exist. But they may be more suffused ment, affective neutrality, self-orientation and with elements from the traditional culture others as if this indicated an empirical coherence than in Western cities, and at the same time among these several elements. Thus, they harken back to W. G. Sumner's "strain of consistency [of certain sections of the urban elite may live the folkways] with each other, because they all an- - -- swer their several purposes with less friction and some evidence concerning the compatibility be- antagonism when they cooperate and support each tween Western urbanism and a "modified extended other." See W. G. Sumner, Folkways, Boston: family," as discussed by Eugene Litwak, "Occupa- Ginn & Co., 1940, pp. 5-6. One would suppose that tional Mobility and Extended Family Cohesion," Sumner was too much of an evolutionist to be and "Geographic Mobility and Extended Family aware of the Hegelian legacy in this approach, and Cohesion," Afnerican Sociological Review, 25 (Feb- given the decreasing interest in the history of ideas ruary and June, 1960), pp. 9-21, 385-94. See also among sociologists a number of them may no Sidney M. Greenfield, "Industrialization and the longer be aware of the evolutionary theory implicit Family in Sociological Theory," American Joulnal in their use of "pattern variables." of Sociology, 67 (November, 1961), pp. 312-322. 4N. V. Sovani, l6The Urban Social Situation in Urba~zization in Asia and the Far East, Pro- India," Artha Vijnana, I11 (June-September, 1961), ceedings of thc Joint UN/UNESCO Seminar, Esng- pp. 85-105, 195-222. The expectation to which kok, August 8-18, 1956, Tensions and Technology Sovani refers should be qualified in the light of Series; Calcutta: 1957, p. 87. COMPARATlVE SOCIOEOGlCAL STUDIES 535 at a greater social and cultural distance from of unresolved theoretical questions. Com- the common people in Indian than in Euro- parative sociological studies, I believe, can pean or American cities. In addition, as an make their own specific contribution to their outsider examines the statistics of unem- resolution. ployment, housing conditions, and popula- tion with sympathy for the human condition, CONTRIBUTIONS OF THE COXIPARATIW he naturally wonders whether the resilience APPROACH of kinship- and caste-ties under urban condi- tions represents the one remaining social Among these contributions at least three security for the individual in fiercecompe- may be distinguished. First, comparative tition for the scarce opportunities available. studies illuminate the meaning of sociological 1f these impressions are near the mark, universals by exhibiting the range of "so- then every increase in population, every lutions" that men have found for a given further crowding of the cities will militate problem in different societies. Second, since against that individualism most needed to "any sociological concepts are curb population and make rndian cities [cur- terms, such studies provide an important banized" in our sense of that word. T~ dis- check on the generalizations implicit in these miss all this as a transitory phenomenon that terms. Third, insofar as our concepts are of will give way to more familiar features limited applicability, such studies also en- city life presupposes what we need to ex- zble us to characterize these limits and hence amine, namely that the cluster of attributes the re- constituting ourbanismu represents a valid ferents of contrasted social structures. generalization of a pattern of interrelated 1. Com~arisolls between "related" Phe- changes. ~ ~ it is ~probable h that~ nomena~ in , different societies are made possi- the cities of India have structural anteced- ble by referring them to some Soci~l~gical ents of their own that will eventually blend universal. "Sociological universals" is another with the familiar physical attributes of cities phrase for the prob1ematics of the social under the impact of modern industry-and condition; they are not generalizations in in this way will create a distinctive type of the ordinary sense. A very detailed, deduc- urbanization. tive elaboration of these "universalsn-- ~f we admit this possibility, then we face such as that of -deals with difficult questions of nomellclature and social these problematics as if were logical theory. we would have to conclude that attributes of all societies conceived as "sys- terms sociologists have adopted or tems." Comparative sociological studies, on from ordinary speech (e,g., city, village, in- the other hand, are less deductive and am- dustry, bureaucracy, etc.) are not readily bitious. they take a single issue applicable in their usual connotations. Since that is be found in mally (conceivably such terms have more connotations than we in all) societies and seek to iiluminate it are aware of in ordinary usage, it is indis- by showing how different societies have pensable for scholarly purposes to make dealt with the same issue. IVhen Max Weber these connotations explicit. As a result, when ,Hence, Parsons, that are true of we use the term "urbanism" with reference all societies take a form such as the folIoc17ing. to India we should not also apply connota- "From the point of view of functioning of the so- tions of the term that are inappropriate. cid system, it is not the needs of all the participant actors which must he met, nor all the needs of any Some theorists cite these diffi- one, but only a sufficient proportion for a sufficient culties as their reason for discarding ordinary fraction of the population." See The Social System, terms altogether and substituting for them G!encoe: The Free Press, 1951, p. 28. Such a state- a new language. ~~t that approach raises ment cannot be falsified or verified, and its remote- ness from the evidence is suggested by the fact even greater it is remote from that the history of all societies records controversies ordinary experience and it interferes with over the meaning of phrases like LLsufficientpropor- effective communication since it makes ref- tion" and "sufficient fraction." Cf. here the critical erences to that experiellce unnecessarily Ob- comments of T.H. htarshall concerning the use of the term "social system" in "The Welfare State: .!I scure. These terminological Problems do not Sociological Interpretation," ~ r ~ pi^^^^^^h i ~ ~ ~ exist in isolation. They are often a symptom de Sociologie, 11 (1961), pp. 285-86. AMERICAN SOCIOLOGICAL REVIEW wrote on the secular causes and consequences underscore the inescapable artificiality of of religious doctrines, he identified one such conceptual distinctions and the consequent issue. We may call it the inner-worldly in- need to move back and forth between the centives implicit in religions; this issue is empirical evidence and the benchmark- examined in the Western religions (culmi- concepts which Max TVeber called "ideal nating in Puritanism), which are contrasted types." with the inner-worldly incentives implicit 2. Many sociological concepts are com- in other religions, such as Hinduism or Bud- posite terms formdating a limited body of dhism. When, in his Ancient City, Fustel de the evidence. If we use such terms without Coulanges wrote of the steps by which a regard to this limitation, we make unwar- consecrated deity of the community gradu- ranted generalizations-however inadvert- ally prevails over the worship of separate ent. Here comparative studies help us to deities of family and tribe, he identified be on guard. For example, if city life as another issue. We may call it the social- we know it goes with secularism, are there or in this case the religious-preconditions other types of city life which go with the of civic unity, which Coulanges examines maintenance of religious beliefs? More over time and in a comparison between generally, if X goes with Y, can we also Greek and Roman society. Or, to take a find evidence that X can go with non-Y? modern work, when Hannah Arendt dis- The second type of evidence would not in- cusses anti-Semitism in Europe and race- validate the first. But by considering both relations in South Africa, she identifies an- we protect ourselves against spurious gen- other issue; we may call it the moral crisis eralizations. The gain is not only negative, of discrimination. Both those who discrimi- however. nate and those who are discriminated against, In Contributions to Indian Sociology lo lose, or are made to lose, their humanity, the question has been raised, for example, either because they claim and exploit as whether the term "village" is applicable to virtues what are accidents of birth or because Indian society, because in all too many in- they lose the standards of one community stances the minimum degree of cohesion without quite acquiring the standards of commonly associated with this term, is ab- another? sent. Such a question should not remain on In these and similar studies a recurrent the conceptual level, however. People's ori- issue of the human condition is identified entations toward kin and caste always com- in order to examine empirically how men in pete to some extent with the demands of different societies have encountered that is- the village as a community which more often sue. If the emphasis is to be on men acting than not are articulated by political author- in societies, these studies will have to give ity. Accordingly, two relatively antagonistic full weight not only to the conditioning of principles of "community" are at work here. these actions but in principle also to the Research into the prevalence of one or fact that men have acted in face of the agonizing dilemmas that confront them. To 8 Perhaps the simplest statement of this issue is maintain this balanced approach, compara- contained in Max LXTeber, "Agrargeschichte des tive studies should not only highlight the Altertums," Gesaiilrnelte Aufsiitze eur Soeialzlnd contrasts existing between different human TViutschaftsgeschiclzte, Tiibingen: J. C. B. Mohr, situations and social structures, but also 1924, p. 280, where an enumeration of different va- rieties of craftsmen is followed by the assertion that it is necessary. nevertheless, to make do with one 7 I note in passing, but with emphasis that all concept of the "ancient Greek craftsman" to refer these are moral issues and that it greatly curtails to all of them. the "sociological imagination" in my judgment if Q See, for example, Ililton Singer, "The Great this moral dimension is neglected. Some sociologists Tradition in a Metropolitan Center: Madras," manage to write even about values or power, as if Journal of Anzerican Folklore, 71 (July-September, values existed outside a moral framework or power 1958), pp. 347-388. involved simply a distinction between a few "bad loCf. the discussions of the editors, Louis Du- guys7' and the masses of the people whose depriva- mont and D. Pocock, and of F. G. Bailey in Con- tions are a synonym of their virtue. The intellectua1 tributions to Indian Sociology, I (April, 193'7), pp. challenge of sociological concepts can only gain if 26-27 and pa\sim; 111 (July 195Y), pp. 88-101; the moral issues inherent in them are laid bare. and IV (April, 1960), pp. 82-89. COMPARATIVE SOCIOLOGICAL STUDIES 531 another of these principles might enable us tion of a cluster of interrelated attributes is to formulate a typology of villages which not sufiicient for this purpose. would reveal the special features of Indian Max Weber's analysis of legal domination village^.^^ exemplifies a concept of "structure" in which An example from Max Weber's sociology the variations typical of the rule of law are of religion shows the analytic usefulness of incorporated. A belief in legality means comparative studies for the construction of first and foremost that certain formal pro- such typologies. Weber points out that in cedures must be obeyed if the enactment ancient times religious prophecy involved and execution of a law or regulation is to oracles concerning future political and mili- be considered legal. But while legal rule- tary events. Three types of prophecy may making tends to eliminate the idiosyncracies be distinguished. In the bureaucratic king- of personal rule, it also militates against the doms of Egypt and Rome emotional prophecy exercise of judgment in the individual case of the Biblical type did not appear because --in the interest of developing a consistent the religious police suppressed prophets as body of rules that are the same for everyone. dangerous demagogues. In Israel such proph- Yet attention to rules in the interest of equity ecy had a long tradition, on the other hand, may engender an interest in rule-making for because it was supported by families of pious its own sake-just as too much regard for notables and the monarchy was not strong equity in the individual case can jeopardize enough to suppress it. Ancient Greece rep- the integrity of the rule-making process. represents a type of prophecy "midway" Hence, the rule of law endures as long as between that of Egypt and Israel: only the piecemeal solutions for these conflictin,u im-' famous oracle at Delphi was permitted to peratives are found and neither the concern prophecy, but the ecstatic states of the with equity nor with the formal attributes of priestess, Pythia, were considered portents rule-making is allowed to predominate. In which had to be controlled and interpreted this way a social structure is understood not by the priesthood. The three cases are dis- as a natural system with defined limits and tinguishable in terms of the degree of po- invariant laws governing an equilibrating litical or religious control exercised over the process, but rather as a system of historical prophets. These are primarily political dis- dimensions which we examine in terms of the tinctions differing from the religious distinc- piecemeal solutions men have found for the tion between the ethical prophecy of Judaism characteristic problems of that structure.14 and the exemplary prophecy of India. Both Where analysis emphasizes the chronology the political and the religious dimensions and individual sequence of such solutions, provided Weber with an analytic tool that it belongs to the historian; where it em- enabled him to study prophecy compara- phasizes the pattern of these solutions, it tively, as a clue to the distinctive social belongs to the sociologist. Comparative so- structures of ancient societies.12 ciological studies are especially suited to 3. Social structures have a space-and-time elucidate such patterns because they tend dimension, as discussed earlier. To formulate to increase the "visibility" of oae structure concepts appropriate to such structures it by contrasting it with another. In this way, is necessary to allow for the variations which they may hclp us identify the issues con- are compatible with-or even characteristic of-each type of structure.13 The enumera- ciety are related to each other, then social. struc- tures are distinguished from each nther by their different "principles." Wolvever, such "pri~iciples" 11 See the striking characterization of these dis- may be more or less fully developed, tlic relations tinguishing features of the Indian village in B. B. between different "paris" may hc strcng or we:ik. Misra, The Indian Middle Class, London: Oxford Or, concretely, absolutist monarchy exists whether University Press, 1961, pp. 310-312. the ruler is effective or not, denlocratlc govcmment 12Max Ureber, Ancient Judaism, Glencoe: The exists whether it is sustained by a vkihlc t\vo-parly Free Press, 1952, pp. 270-71, 281, 287-88, 290-92, structure or jeopardized by a rilliltiplicit-ji of 295, and my summary statement in Max Webcr, parties. An Intellectual Portrait, Garden City: Anchor l4Cf. the lucid stale~neilt of tlik difference be.- Books, 1962, pp. 247118. tween natural and historical systeiiis in F:. E. Evans- 1"f by "structurc" we mean the principle ill ac- Pritchard, Social Antlzropology, London: Cohen & cordance with \vliich the different "parts" of a so- West, 1960, pp. 56-62. 538 AMERICAN SOCIO BGICAL REVIEW fronting men in their attempts to develop following inquiries may be distinguished: their country along the lines of one pattern 1. If the subordination of private interest or another. And by exposing concepts and to public authority is a characteristic feature generalizations to a wider range of evidence of political community, then it should be than is sometimes customary, comparative possible to distinguish between types of po- sociological studies are likely to impart a litical communities by types of subordina- salutary degree of nominalism to the terms tion. It is important, for instance, to formu- we use. late the type of subordination characteristic of the "medieval political community," to COMPARATIVE STUDIES OF THE contrast this with the subordination charac- "POLITICAL COMMUNITY" teristic of the Western nation-state and, if possible to formulate the "crisis of transi- Since the foregoing comments are mainly tion," in which the "medieval" subordina- cautionary and descriptive, it is well to sup- tion gives way and is superseded by one plement them briefly by illustrating the type typical of the nation-state.17 of comparative approach here envisaged. As 2. If two social structures differ in the in the classic studies cited earlier, the starting type of subordination characteristic of them, point is the identification of a universal, then they will differ also in the type of pro- in this case the friction between private in- test they provoke among the subordinated. terest and public authority. Where each in- Accordingly, the millenarian movements, so- dividual or group takes the law into their cial banditry, and populist legitimism char- hands until checked by the momentarily acteristic of medieval political life may be superior force of an opponent, anarchy reigns contrasted with the quest for national cjti- and a political community does not exist. zenship characteristic of protest-movements Some subordination of private interest to in some countries of Western Europe during public authority is the sine qua non of such the nineteenth century.ls a community. While governments vary 3. Though social structures may be dis- greatly with regard to the subordination tinguished one from the other, no structure they demand and the rights they acknowl- is static. A further task is to analyze the edge, the term "political community" may transformation a given structure undergoe~ be applied wherever the relations between without losing its distinguishing character- rulers and ruled involve shared understand- istics. In the present case this may be at- ings concerning this "exchange." These un- tempted by comparing the manner in which derstandings concern the legitimacy of public kindred Western European societies have authority, its organization and demands extended national citizenship to those see upon the individual; they exist side by side ments of the population which previously had with the fact that individuals find their been excluded from the rights of citizenship. private interests enhanced through coopera- In the early type of nation-state these rights tion with others. Accordingly, men are en- were available only to social notables, gaged in the pursuit of "ideal and material whereas subsequently these rights became interests'' leading to social relationships an attribute of all adults as citizens of the based on a coalescence of these interests, nation-state.19 and they are engaged in actions "governed by the conception that a legitimate order 16 These illustrations are takrn from 111y or~~nclll- exists." l5 This formulation points to the rent work. My hope is that other scholars interested universal problem of reconciling private con- in comparative studies will accept my character- ization of this level of analysis. cerns and the actions that sustain public 17 For an attempt to formulate these three models authority. with the aid of Tocqueville's analysis see the au- In a comparative study of changing social thor's "Social Stratification and the Political Com- structures since the French revolution the munity," A~chives Europden~zes de Sociologie, I (1960), pp. 3-32. 1s Cf. the discussion of this contrast in my essay, 16 For a fuller statement of this distinction in the "The Lower Classes and the 'Democratic Revnlu- work of Max Weber and an analysis of its intel- tion,' " Indzcstrial Relatioi~s,I (October 1961), pp lectual derivation see Bendix, Max Weber, pp. 476- 91-116. 77 and passim. 1<4 preliniinary formuI~lionof this approach is ON CHURCH AND SECT The three kinds of studies here suggested to these characteristics it should be possible are capable of extension in many directions. to formulate models of the pre-modern social For example, the distinction between the structure, of the transition which followed, medieval political community, the modern and of the modern social structure which nation-state and the crisis of transition is has developed to date.21 applicable principally to the countries of These are only a few positive illustra- Western Europe, and one should explore the tions of comparative sociological studies limits of this applicability. But one may aiming at propositions that are true of more also apply an analogous approach to other than one but less than all societies. This areas of the world which differ from the essay will have served its purpose if it di- Western European pattern, to be sure, but rects attention to a type of inquiry which- which nonetheless possess common structural at the macro-sociological level-seeks to hold characteristics of their own.20 With regard a balance between grand theory and the de- contained in R. Bendix and Stein Rokkan, "The scriptive accounts of area-studies. Extension of Kational Citizenship to the Lower Classes: A Comparative Perspective," Paper sub- different. I have made such an attempt in a com- mitted to the Fifth World Congress of Sociology, parison of German and Japanese modernization. Washington 1962. See Reinhard Bendix, "Pre-conditions of Develop- 20Examples are the Latin American countries ment: A Comparison of arid Japan," which have in common the Spanish colonial herit- Conference on Modern Japan, Bermuda, 1962. age, European frontier-settlements like the United "In an effort to articulate the distinguishing fea- States, Canada, Australia and New Zealand which tures of Western European societies, I have at- have the British legacy in common, and others. tempted to formulate such models for Russia from Such groupings are not always that simple and her autocratic rule in the eighteenth and nineteenth there are countries, like Japan, which probably are centuries to her totalitarian structure of the post- in a category of their own. Such historical cluster- revolutionary period. See my Work and Authority ing of social structures may then be analyzed with in Industry, New York: John Wiley, 1956, Chap- the aid of sociological universals; but I confess to ters 3 and 6 and "The Cultural and Political Set- considerable scepticism concerning the use of such ting of Economic Rationality in Western and East- universals without regard to such clusters, or in ern Europe," in Gregory Grossman (ed.), Value the absence of an attempt to spell out in what re- and Plan, Berkeley: University of California Press, spects two or more social structures are alike or 1960, pp. 245-70.

ON CHURCH AND SECT 4:

University of Oregon

The church-sect typology as developed by Troeltsch and modified by others is of limited use in classifying religious groups because it is applicable only to a specific historical context and it encompasses a variety of elements which tend to vary irzdependently. A single-variable definition of broad applicability is proposed. Grounded in Weber's systewzatic typology of religion, the new definition should facilitate the co+nparative study of most groups in the Jewish, Christian and Islamic traditions. Some guidelines for using the typology are sug- gested. These are illustrated by applying the typology to the major religious groups of the .

INCE Ernst Troeltsch formulated the larly useful device for the analysis of the church-sect typology more than half a characteristics of organized Christian groups S century ago it has come to be regarded in relation to their environment. by most sociologists of religion as a singu- Yet the typology as developed by Troeltsch has been subjected to a great deal of criti- * Revised version of a paper presented to the annual meeting of the Society for the Scientific cism. During the past generation many stu- Study of Religion, New York, October, 1962. dents have reworked it in various ways to http://www.jstor.org

LINKED CITATIONS - Page 1 of 1 -

You have printed the following article: Concepts and Generalizations in Comparative Sociological Studies Reinhard Bendix American Sociological Review, Vol. 28, No. 4. (Aug., 1963), pp. 532-539. Stable URL: http://links.jstor.org/sici?sici=0003-1224%28196308%2928%3A4%3C532%3ACAGICS%3E2.0.CO%3B2-T

This article references the following linked citations. If you are trying to access articles from an off-campus location, you may be required to first logon via your library web site to access JSTOR. Please visit your library's website or contact a librarian to learn about options for remote access to JSTOR.

[Footnotes]

4 Occupational Mobility and Extended Family Cohension Eugene Litwak American Sociological Review, Vol. 25, No. 1. (Feb., 1960), pp. 9-21. Stable URL: http://links.jstor.org/sici?sici=0003-1224%28196002%2925%3A1%3C9%3AOMAEFC%3E2.0.CO%3B2-A

4 Industrialization and the Family in Sociological Theory Sidney M. Greenfield The American Journal of Sociology, Vol. 67, No. 3. (Nov., 1961), pp. 312-322. Stable URL: http://links.jstor.org/sici?sici=0002-9602%28196111%2967%3A3%3C312%3AIATFIS%3E2.0.CO%3B2-K

9 The Great Tradition in a Metropolitan Center: Madras Milton Singer The Journal of American Folklore, Vol. 71, No. 281, Traditional India: Structure and Change. (Jul. - Sep., 1958), pp. 347-388. Stable URL: http://links.jstor.org/sici?sici=0021-8715%28195807%2F09%2971%3A281%3C347%3ATGTIAM%3E2.0.CO%3B2-B

NOTE: The reference numbering from the original has been maintained in this citation list.