This survey of a century of comparative sociology divides its major developments into three phases or generations, separated by abrupt discontinuities. The promise held by comparative sociology from the beginning was the understanding of the diversity of cultures and civilizations and their respective institutions in different regions of the world. I will argue that each generation made significant progress toward realizing the original promise of comparative sociology but was abandoned by the dominant trends in metropolitan social theory. The third generation can appropriate the intellectual heritage of the first two, and thus equip itself for the reception of theoretical concepts that capture distinctive historical experiences of the world regions in the global age. The way is once more open for redeeming the original promise of comparative sociology.

From Durkheim and Weber to Redfield’s comparative analysis of civilizations

In The Rules of Sociological Method (1982 [1895], p.139), Durkheim declared: “Comparative sociology is not a particular branch of sociology; it is sociology itself.” Durkheim’s forgotten dictum shows no intention of disguising the metropolitan patterns of social development as universal; nor does Weber’s final view of different directions of rationalization in the world religions. Weber had no immediate followers and the Weberians of the next generation, notably Norbert Elias and Reinhard Bendix, worked on Europe. Not so with the Durkheimians. In the Elementary Forms of Religious Life (1912), Durkheim used the data on the Australian aborigines, the remotest possible from the metropolitan setting, to construct a general theory of religion, and applied the concept of “collective effervescence” derived from that data to throw light on such metropolitan phenomena as the rise of universities in medieval Europe and the French revolution. Durkheim demonstrated his commitment to comparative sociology in a note on the notion of civilization, written with his nephew, Marcel Mauss, where he developed a concept of civilizations in the plural. This diverged radically from the imperialist notion of civilization in the singular, and incidentally helped Durkheim transcend his own hallmark, reified concept of “society”. Mauss (Reference Mauss, Barthel-Bouchier, Arjomand and Tiryakian2004[1930]) later elaborated this concept of civilizations as social but transnational, trans-societal regimes or super-systems, extending beyond any given society. This was a significant move toward a new paradigm particularly suited to our global age, when, as Albrow (Reference Albrow1996, p. 58) puts it, “the social takes on a meaning outside the frame of reference set by the nation-state”.

Shils (Reference Shils1982, pp. 286-287) bemoaned the fact that the Durkheimians produced almost no work on modern society, and Connell (Reference Connell2007a, p. 6) confirms that, in Année sociologique, their interest in reviewing works on modern society, not to mention industrial society, was marginal. Instead, the research-oriented Durkheimians embarked either on social anthropology or on civilizational analysis, while the mainstream, or rather the politically influential Durkheimians, turned his sociology into a civic ethical philosophy of the Third Republic. The latter trend concerns us only indirectly as the cause of the great institutional weakness of Durkheimian sociology (Karady Reference Karady and Besnard1983, pp. 88-89). Durkheim had not succeeded in creating chairs and departments of sociology, partly because of the devastating First World War, and the Durkheimians dispersed into other disciplines or into academic administration. The research-oriented Durkheimians mostly moved to the disciplines of classics, linguistics and Oriental studies as there were no more than four university chairs in sociology in France. As a result, the Durkheimian circle was in a peculiar situation. As Célestin Bouglé put it in 1927, its center was nowhere, its circumference everywhere (cited in Heilbron Reference Heilbron1985, p. 210). Bouglé’s book on the caste system had appeared in 1908, but he soon moved to the other camp and to the top of the French educational administration. Before being rejected by the researchers led by Marcel Granet in a major split among the Durkheimians (Heilbron Reference Heilbron1985, pp. 212-223), Georges Davy had teamed up with the Egyptologist Alexandre Moret to write a book on ancient Egypt in 1923. Louis Gernet wrote the famous Anthropology of Ancient Greece, followed by Granet’s Chinese Civilization in 1929. What is remarkable about these works is that not a single one of the civilizations covered by the Durkheimians was a part of the French empire. From the École Pratique des Hautes Études,Footnote ¹ Mauss directed Durkheimian research into social anthropology, co-founding the Insitut d’Éthnologie in 1925 and maintaining strong ties with the new Musée de l’Homme in the 1930s. In thus providing a central institutional base for social anthropology in France, Mauss inadvertently established the pattern of research that Hountondji (Reference Hountondji and Odora Hoppers2002) has properly characterized as colonial production of knowledge, with a division of labor between data collection in the French colonies, and analysis and theory-building in Paris. Given the imperialism of the universal (Bourdieu Reference Bourdieu, Fauré and Bishop1992), the slope from commonplaces of the civilizing mission to sociological theory was slippery indeed. Durkheimian comparative sociology slipped and failed to realize its promise.

The Durkheimian tradition of civilizational analysis had one last distinguished French representative: Louis Dumont (1911-1998), a student of Marcel Mauss,Footnote ² who published Homo Hierarchicus (to be discussed below) in 1967. Meanwhile, Durkheim had had a preponderant influence in shaping British social anthropology through A.R. Radcliffe-Brown. Greatly influenced by Radcliffe-Brown’s structural-functionalism as a student in London, M.G. Smith returned to Jamaica in 1952 and was able to develop Durkheimian comparative sociology significantly by basing his generalizations and analytical deductions, as well as contrasts, on the colonial experience of his native region, the Caribbean – a region not explicitly covered by the French Durkheimians. Struggling against Parsons’ allegedly universalistic social system in the 1950s, M.G. Smith put forward the concept of “plural societies”, challenging the assumption that the integration of Caribbean societies in general – and of Granada and Jamaica, in particular – could be based on culturally-shared values (Smith Reference Smith1965; 1990, pp. 6-8). Plural societies consisted of component sections or segments that were primordial collectivities or ethnic groups. These ethnic groups had been differentially incorporated into Caribbean societies under colonial rule. This condition invited comparisons with other plural societies created as a result of conquest and colonialism in other regions of the world. In the following decade, Smith constructed a general theory of “corporations” to put the Caribbean historical experience in comparative perspective, thus generalizing Durkheim’s dichotomy of segmentary (primitive) and differentiated (modern) societies: the differential incorporation of collectivities on the basis of legal inequality created hierarchical societies marked by structural pluralism (as in the Caribbean region). This was in double contrast to the segmental incorporation of collectivities on the basis of formal equivalence, which resulted in societies marked by social pluralism (as in Lebanon and Nigeria), and universalistic or uniform incorporation of collectivities that resulted in (non-plural) societies marked by cultural pluralism (as in the post-segregation United States) (Smith Reference Smith1974; Reference Smith and Ryan1991, pp. 14-15, p. 22).

A second notable metropolitan attempt at the comparative study of civilizations was made when the United States became the dominant world power after World War II. In Europe, Oriental studies had developed in fairly close connection with the British, Dutch, French and belatedly Russian and German empires. The idea of area studies that emerged in the United States was different. It was generated, at the deepest level, “out of an ideological coding of America’s universal mission” and as a response to the intellectual and moral challenge of a new world order (Sartori Reference Sartori1998, p. 56). Area studies were conceived as the marriage between the social sciences and the study of non-Western civilizations. The promise of the original idea of area studies as formulated in the 1940s is well brought out by Wallerstein (Reference Wallerstein and Noam1997). As early as 1943, an internal report of the Committee on World Regions of the Social Science Research Council (SSRC) saw the concentration on regions by area studies as a catalyst to “one of the major and most distant goals of many outstanding social scientists: a weakening of the rigid compartments that separate the disciplines” (cited in Wallerstein Reference Wallerstein and Noam1997, p. 197). Area studies, furthermore, were not meant to serve as the maidservant of American social sciences and help disguise their scientific pretensions as universally valid but rather to provincialize them.

Sociologists were not at all conspicuous among the social scientists who responded to the challenge of the post-World War II reconstruction of the world under the hegemony of the United States, and their intellectual resources for doing so were meager (Crawford Reference Crawford1948). During and after World War I, Edward Ross had elaborated a sociology of the world regions – what he called “worldly sociology” – in particular of Latin America, but it was dated and badly flawed by his racial theory (Ross Reference Ross1915). Calls were made for the study of cultures and civilizations and the placing of national cultures in their respective “culture area” (Callis Reference Callis1947, p. 329), but there was little or no response. This was evident in a symposium held by the American Sociological Society in December 1947, in which the young C. Wright Mills (Reference Wright Mills1948, pp. 272-273) charged the American sociologists with failing to make “the university a permanent third camp in world affairs”, and thus acting “as a vanguard of world citizenship”. Instead, “the comparative sociologist […] moves from provincialism to nationalism”, providing “nationalistic public relations men with materials to use in propagandizing American culture to those in enemy and backward areas” (Wright Mills, pp. 271-272).

The most important project of the period for unifying Oriental studies and the social sciences, by contrast, was Robert Redfield’s social anthropology of civilizations at the University of Chicago. Against the background of a proliferation of postwar Western civilization courses and America’s drift to ideological rigidity with the Cold War, Redfield saw the study of other civilizations as the means of transcending American parochialism while being “only more true to the universalism that underpins its identity”Footnote ³ (Sartori Reference Sartori1998, p. 37). It required the reconciliation of the anthropologist’s and the humanist’s notions of “culture” in an integrated and ultimately comparative study of civilizations. Redfield distinguished between the “societal” and the “cultural structure” of civilizations, and finally developed the idea of their “historic structure” as relations of temporal hierarchy between a Great Tradition and its Little Traditions, embodied in a respective hierarchy of social communities. Redfield found that the methodological problems of studying historic structures of civilizations could not be solved by social anthropologists alone and turned to Orientalists for help. Gustav von Grunebaum joined Redfield’s project, and was the first Orientalist to embark on the first integrated study of Islamic civilization, followed by a number of distinguished Sinologists, including John Fairbanks (Reference Fairbanks1957) and Arthur Wright (Reference Wright1960). Grunebaum’s student, Marshall Hodgson, in The Venture of Islam (1974), published posthumously after his untimely death in 1968, applied the mode of integrated civilizational analysis, reinforced by the idea of Oikumené as developed by Redfield’s colleague, Alfred Kroeber (Arnason Reference Arnason2006). The project’s greatest impact was on the study of Indian civilization, notably in the works of Milton Singer (Reference Singer1955, Reference Singer1972) and Bernard Cohn (Reference Cohn1987 [1955, 1958, 1961a, 1961b]) in Chicago and, as we shall see, of M.N. Srinivas in India.

Meanwhile, the SSRC acted as the most important promoter of area studies in the United States. It is interesting to note that, in the SSRC meetings, Talcott Parsons saw a critical role for area studies in compensating for the limitations of the newly professionalized social sciences, and for cleansing social theory of its provincialism. Through the cooperation necessary for the development of an integrated structure of knowledge, area studies could thus have, in his words, “a profound effect on social science research” (cited in Mitchell Reference Mitchell2003, pp. 8-9). The SSRC set up a Near and Middle East Committee in 1951. The Orientalist H.A.R. Gibb moved from Oxford to Harvard in 1955 and joined it, proposing that the Orientalists’ contribution should be to gather and correlate the findings of separate social studies (Lockman Reference Lockman2004, ch. 4).

Generally speaking, however, things worked the other way, and area specialists and social scientists increasingly went their separate ways for two main reasons. First, the area specialists were greatly outnumbered by social scientists without any area interests who used primarily American data and socio-cultural patterns for pseudo-universal generalizations.Footnote ⁴ Parsons, for instance, moved in the completely opposite direction to the one he had indicated in the late 1940s, proposing in 1954 “a long-term program of scholarly activity which aims at no less than a unification of theory in all fields of behavioral sciences” (cited in Rudolph Reference Rudolph2005, p. 8). Secondly, the Cold War began soon after the launching of area and development studies and imposed an extra-epistemic frame on the whole project. It can be argued that the intellectual terrain of the Cold War era doomed the synthesis attempted by the area studies project because of the division of academic labor required by its Three-Worlds “deep structure” – a structure which inexorably made for the mutual insulation of the “third world of tradition, culture, religion, irrationality [and] underdevelopment”, and the first world of modernity, technology and rationality, a democratic, free and natural society unfettered by religion and ideology” (Pletsch Reference Pletsch1981, p. 574). The result was a failure to realize the promise of comparative sociology a second time.

It can be plausibly argued that what killed this second interdisciplinary phase of comparative sociology was the modernization theory that bore the mark of Parson’s general theory. The “historical turn” in metropolitan social theory in the mid-1970s (Tilly Reference Tilly1975) seemed to signal the fall of the modernization theory to many (Arjomand Reference Arjomand2004c, p. 336). This turned out to be cold comfort to comparative sociologists, however. The “historical and comparative sociology” that has developed in the United States as an alternative to modernization theory since the 1980s, curiously succumbed to the same temptation of regarding Western patterns as paradigmatic.Footnote ⁵ The revisionist Marxists and Weberians who founded it followed Barrington Moore’s, Reinhard Bendix’s and Charles Tilly’s heavy reliance on the Western historical experience for analysis and in forming concepts.Footnote ⁶ Moore (Reference Moore1966), for example, forged his key concepts in order to trace the origins of democracy and dictatorship to the class-coalition in the course of the commercialization of agriculture in the West, and then extended its class analysis to India and Japan. The application of concepts formed out of Western historical experience to other cases is an imposition, as is most evident in Moore’s analysis of inter-War Japanese “fascism” as the “labor-repressive” outcome of Japanese class alignment during the critical transition. The same imposition of metropolitan categories on the periphery is true of Rueschemeyer, Stephens and Stephens (Reference Rueschemeyer, Stephens and Stephens1992), who extend Barrington Moore’s class analysis to highlight the contribution of the working class and its organization to the development of democracy. Despite their extensive coverage of the Caribbean, it is difficult to detect any acknowledgement of M.G. Smith’s periphery-derived idea of plural societies.

Indeed it is more accurate to describe this flourishing school of macro-sociology simply as “historical sociology” (Adams, Clemens and Orloff Reference Adams, Clemensand, Orloff, Adams, Clemensand and Orloff2005). The progress made by this group has been largely driven by methods appropriate to historical sociology – analyses of temporal sequences, path dependency and rational choice in institutional development, and network analysis. The comparative element has, by and large, remained limited to those implicit in case studies. It is unmistakably anemic and at best a secondary feature, expressed in ad hoc explanatory parallels and contrasts adduced in case studies, which Mahoney and Rueschemeyer (Reference Mahoney, Rueschemeyer, Mahoney and Rueschemeyer2003, p. 14) call “contextualized comparison”.Footnote ⁷ Even the so-called “cultural turn” in historical sociology was largely methodological and historiographic, entailing an understanding of patterns of meaning with little or no attention to cultural comparisons (Mahoney and Rueschemeyer Reference Mahoney, Mahoney and Rueschemeyer2003).

To the extent that the periphery was not ignored after the failure of the second generation of comparative sociologists, its experience was fitted into the straightjacket of allegedly universal processes such as modernization and development. These generalized what was taken as the dominant Western pattern into a universal teleology. I have traced the line of theoretical development as viewed from the center as a universal pattern of value-rationalization spreading from the center to the periphery. This development was characterized as a discontinuous process of expansion of the scope of rational judgment driven by periodic shifts in dominant value-ideas in the course of the 20^th century (Arjomand Reference Arjomand2004c). The view from the periphery was different, however. It was from the periphery that challenges to the dominance of metropolitan theory and its claim to universality originated.Footnote ⁸ These challenges sought to rectify both the denial of diversity implicit in the putatively universalistic Western-based categories and the patterns of the modernization theory. This amounted to provincializing metropolitan theory through studies on different regions of the world, seeking to correct what Raewyn Connell (Reference Connell2007a, p. 46) has described as “the erasure of the experience of the majority of humankind from the foundations of social theory”.

I have already examined one significant attempt at theorizing on the basis of the historical experience of a specific world region, namely M.G. Smith’s conceptualization of plural societies distinctive of the Caribbean region. A far more influential theory born in the periphery to challenge the metropolitan economic development theory was the dependency theory that emerged from Latin America. As Roberto Briceño-León emphasizes (2010, 179), it arose from the conviction that “sociologists needed to explain the singularity of Latin America”. Its economic antecedent was the center-periphery theory of Raul Prebisch, who took over the presidency of the UN’s Economic Commission for Latin America (ECLA, Spanish acronym: CEPAL) in 1949. Fernando H. Cardoso and E. Faletto (Reference Cardoso and Faletto1969) sought to put Prebisch’s economic theory in a sociological context, with Cardoso (Reference Cardoso1970) presenting dependency as rooted in “a particular type of articulation between social classes, the productive system and the state, in a particular historical situation”. In other words, Cardoso and the dependentists opposed “the idea of a ‘universal’ methodology” and “believed that the possibility of explaining Latin American reality depended on the determination of its specific problems” (Beigel Reference Beigel and Patel2010, p. 194). If Redfield’s comparative analysis of civilizations was the North American response to the intellectual challenge of post-World War II reconstruction, the dependency theory was the South American response to the same challenge in the era of decolonization.

Although dependency theory had an immediate impact, it just as quickly lost its Latin American specificity. O’Donnell used dependency analysis to explain Latin American authoritarianism, but in doing so it quickly left the comparative sociological field for the generic theories of authoritarianism and transition to democracy.Footnote ⁹ The most drastic loss of Latin American specificity came with Immanuel Wallerstein’s (Reference Wallerstein1976) blanket extension of dependency theory’s analytical terms to the rest of the globe in his world system theory.

The third generation: axial civilizations and multiple modernities

Noting the dissatisfaction with the dichotomous view of tradition and modernity presumed by the modernization theory, S.N. Eisenstadt (Reference Eisenstadt and Eisenstadt1972) turned to the problem of the “continuity and reconstruction of tradition” and even persuaded the American Academy of Arts and Sciences to put it on its agenda, experimenting first with the term “post-traditional societies” (1972) and finally settling for “multiple modernities”Footnote ¹⁰ (Eisenstadt Reference Eisenstadt2000). Tradition and modernity were combined in “new foci of collective national identity” (Eisenstadt Reference Eisenstadt and Eisenstadt1972, p. 7). Collective identities become central to the political and cultural program in the struggle for the appropriation of modernity on the global periphery, thus prompting the reconstruction of a diversity of traditions throughout the world which produces multiple modernities (Eisenstatdt Reference Eisenstadt2000). This reappraisal of tradition also prompted Eisenstadt’s turn to civilizational analysis. Eisenstadt (Reference Eisenstadt1986, pp. 6-7) drew inspiration from Karl Jaspers’s idea of the Axial Age, but attributed the breakthrough to dynamism in Axial Age civilizations to the “chasm between the transcendental and the mundane”. For a while, he remained faithful to Jaspers’ temporal component of the idea of a breakthrough to transcendence in a specific age – that of the Hebrew Prophets, the Buddha and Plato. Eisenstadt later undertook an “historical-sociological reconceptualization of the Axial Age”, offering a typological conception of axiality as a cluster of dynamic characteristics (Arnason Reference Arnason, Arnason, Eisenstadt and Wittrock2005, p. 37). Furthermore, Eisenstadt put forward the idea that the basic premises of the Enlightenment generated the breakthrough to a new and distinct axial civilization, the civilization of modernity. This civilization of modernity is the context of the new dialectic of tradition and modernity that produces “multiple modernities” (Eisenstadt Reference Eisenstadt2003). Arnason (Reference Arnason2003, pp. 304-314) has followed Eisenstadt in focusing on the formation and transformation of tradition in civilizations by highlighting the persistence and temporal integration of civilizations. Civilizations are coherent units through time because they relate to their past by means of continuous interpretation and codification, through their historical memory, and by the canonization of certain texts – in short, through the formation and transformation of tradition.

The number of studies contributed by area specialists and historians who subscribe to the axial civilizations paradigm is truly impressive (Arnason, Eisenstadt and Wittrock Reference Arnason, Arnason, Eisenstadt and Wittrock2005). Furthermore, the question of the distinctiveness of traditions and civilizations and their culturally specific developmental patterns is also addressed by several other attempts at civilizational analysis. Claudio Véliz (Reference Véliz1994), for instance, draws on Vico’s conception of civilization and Isaiah Berlin’s metaphor of foxes and hedgehogs to compare the two variants of Christian, Atlantic civilization. Véliz effectively captures the consistency of civilizational styles through time on the basis of the cultural clusterings produced by elective affinities in English and Spanish America, and presents them as the contrasting and historically unique ideal types of “Gothic foxes” and “Baroque hedgehogs”. Véliz, however, largely ignores the peculiar originality of the pre-Colombian cultures and the persistence of the indigenous past which Octavio Paz (Reference Paz and Kemp1961 [1950], p. 92, p. 144) had emphasized in his celebrated study of Mexico.Footnote ¹¹

Samuel Huntington’s idea of the clash of civilizations (1996) was a major stimulus to the spread of interest in civilizational analysis, provoking as sharp a reaction in social theory (e.g. Melleusch 2000) as it did in political debate. John Rundell and Stephen Mennell (Reference Rundell and Mennell1998) came up with a truly impressive genealogy for the study of cultures and civilizations in the classics of social theory. Quantitative studies in civilizational analysis also began to appear. Yilmaz Esmer (Reference Esmer2002) found the distinctive feature of Islamic civilization in contrast to others to consist in higher religiosity, and greater support for marriage and the family, which make it least supportive of gender equality. Richard Marsh, a China expert noted for his important early book on comparative sociology (Marsh Reference Marsh1967), was persuaded that “the concept of civilization should be taken more seriously as a consequential macro-level factor in the analysis of values” (Marsh Reference Marsh2009, p. 269). He did so by comparing eight contemporary civilizations on the basis of the World Values Surveys, confirmed Esmer’s finding on Islam and also found indirect support for Huntington in that “Islamic civilization stands in sharpest contrast to the West in its religious, family and gender values” (Marsh Reference Marsh2009, p. 300). Brian Min and Andreas Wimmer (Reference Min, Wimmer, Anheier and Isar2007, pp. 70-73), by contrast, found no direct support for Huntington’s thesis; on the contrary, the post-Cold War violence along the “civilizational faultlines” pales before the ethnic civil wars caused by the wave of diffusion of the nation-state form following the end of the European empires in the post- World War II decades, which is replicated after the collapse of the Soviet empire.

In view of the growing momentum of this new wave in comparative sociology, Edward Tiryakian and I suggested that the time was ripe for rethinking civilizational analysis (Arjomand and Tiryakian, 2004 [2001]). Among the older ideas taking the opposite perspective to Huntington’s, we highlighted the seminal idea of inter-civilizational encounters by the late Benjamin Nelson (Reference Nelson1980) who had founded the International Society for the Comparative Study of Civilizations (Nielsen Reference Nielsen2001). In my own rethinking, detecting the old German distinction between civilization (as techno-scientific and general) and culture (as particular) behind Max Weber’s discussion of the types of rationality and the processes of rationalization, I came to the conclusion that he exaggerated the importance of instrumental and formal rationality in world history. The civilizational processes I was familiar with seemed to involve collective striving for a different kind of rationality: they were processes of value-rationalization. I argued that value-rationalization is a process of harmonization of heterogeneous principles of order that is driven by the judgment of meaningful consistency (Sinnzusammenhang) – a consistency that we recognize as “civilizational style”. The challenge was to specify the complex combination of logic, poetic judgment and historical contingency in this architechtonic process of construction of meaning (Arjomand Reference Arjomand, Camic and Joas2004a). I think that various periphery-generated contributions to the role of religion and tradition in the evolution of different patterns of alternative modernities can be understood under this general rubric.

The clear implication of this rethinking was that Elias’s notion of the civilizing process had to be discarded in principle and replaced by culturally specific developmental patterns in different civilizations. This is in no way a denial of the significance of the process Elias had identified or the value of his linking it to state formation and gradual concentration of the legitimate use of violence in European history. There is an interesting parallel with imperial China, where “civilization or wenming could be seen mainly as a process: the spread of virtue from the moral center to barbarians and people with ‘depraved’ customs” (Duara Reference Duara2004, p. 2). Eiko Ikegami (Reference Ikegami1995) also offers a striking Japanese parallel to this long-term developmental pattern, but the pattern varies significantly. Arnason (Reference Arnason2003), Eisenstadt (Reference Eisenstadt2003) and other contributors to civilizational analysis also underline the close connection between culture and power and their entwined symbolism in the dynamics of axial civilizations. The fact remains, however, that Elias’ (Reference Elias and Jephcott1978) idea of the civilizing process presumes a unitary concept of civilization that is Euro-centric and is at sharp variance with the pluralistic conception of civilization. If there are many civilizations, there cannot be only one civilizing process and each civilization must have its distinctive dynamics that, in the long run, produce distinctive (value-)rationalities.

In what follows I will argue for the convergence in the third generation of comparative sociologists between two groups with compatible theoretical orientations: the one led by Eisenstadt, which studies axial civilizations and culturally specific developmental paths to multiple modernities through various adaptations of the core institutions of the civilization of modernity (Eisenstadt Reference Eisenstadt2003), and a broader group that mounted the second wave of challenges to metropolitan theory from the periphery. This latter group’s primary concern is with forming ideas and formulating concepts on the basis of distinctive historical experiences of different world regions.Footnote ¹²

Rethinking civilizational analysis: tradition and religion in civilizational processes

Redfield had offered a model of two distinct civilizational processes, which he called orthogenetic, referring to intra-civilizational approximation of the little to the great tradition, and heterogenetic, referring to innovative trends quite possibly under inter-civilizational influences. His comparative study of civilizations was on a paradigm of the social organization of tradition as continuous communication between local, living Little Traditions and their representatives and the Great Tradition to which they were affiliated (Wolf Reference Wolf1967, p. 460). Modernization would then create “a double structure of tradition”. In Mayan villages, the old structure appeared truncated, broken off and subsisting in folk culture, while the new was continuing and changing (Redfield Reference Redfield1955). Elsewhere, both layers of this double structure were active. His example of the latter case was the Indian civilization, and he cited the works of two of his collaborators: Bernard Cohn, for showing that a caste of “leather-workers have improved their position by adopting customs authorized by the high Sanskritic tradition”, and M.N. Srinivas, for demonstrating that the Coorgs, once largely outside the Indian Great Tradition, had come to consider themselves as Kshatriyas, people of the warrior caste, through their world-renouncing holy men (sannyasis) (Redfield Reference Redfield1955, pp. 17-18).

Cohn had indeed shown that, as a result of literacy and urban experience, “the complex interaction of modernization and traditionalization” in the late-19^th-century Arya Samaj movement had penetrated an Indian village in consonance with the religious life of the low but upwardly mobile Chamar caste (Cohn 1990 [1961], p. 98). Srivinas elaborated his insight about the Coorgs into the justly famous concept of Sanskritization. An intriguing aspect of rapid social change in independent India was that the lower castes were being “Sanskritized”, taking over Sanskritic rituals on occasions of birth, marriage and death and employing Brahmin priests. “Sanskritization refers to a cultural process but it is […] usually a concomitant of the acquisition of political or economic power by a caste. Both are parts of the process of social mobility.” Sanskritization was the opposite of Westernization, but these two processes were ongoing side by side (Srinivas Reference Srinivas1992 [1962], p. 119), and were actually linked together in a dynamic relationship (Srinivas Reference Srinivas1966, Madan Reference Madan1995, pp. 41-43). Westernization is Redfield’s heterogenetic process or Benjamin Nelson’s inter-civilizational encounter. Sanskritization, in contrast, is the distinctive Indian intra-civilizational process required to provincialize Elias’s Eurocentric idea of the civilizing process.

Srinivas has been presented as the father of India’s “nationalist sociology” and a chief architect of its Brahminical or savrana (upper caste) vision of Indian society (Patel Reference Patel and Patel2010, pp. 283-285). He has rightly been criticized for blacking out the non-Brahmin perspective and non-Hindu groups as well as diverse Hindu sectarian traditions. Although Srinivas himself (Reference Srinivas1992, p. 57) later hinted at “Islamisation” as a process set in motion by national integration, alongside Westernization and Sanskritization, his methodological Hinduism, to use T.K. Oommen’s (Reference Oommen2008, p. 76) apt characterization, inevitably tended to exclude well over 40 percent of the Indian population that did not subscribe to Sanskritic Hinduism. As Oommen (Reference Oommen2008, pp. 73-75) pointed out, “Islamization and Tamilisation” are also major current intra-civilizational processes in contemporary India. Needless to say, Islamicization is a major intra-civilizational process throughout the Muslim world, consisting of distinctive historical patterns of intensive penetration of Muslim societies by scriptural Islam and of the extensive spread of Islam along its frontiers. Both intensive and extensive Islamicization were reinforced by and interacted with the post-World War II processes of urbanization, spread of education, national integration, and finally globalization to produce what is called the Islamic resurgence of the past four decades (Arjomand Reference Arjomand2004b).

While Parsons was developing his dichotomous “pattern-variables” and the modernization theory was being constructed on the basis of the invidious contrast between tradition and modernity, a more distinctively peripheral view of tradition was being elaborated in India. D.P. Mukerji (1894-1961), president of the first Indian Sociological Conference (1955), wanted sociology to serve the “task of reconstructing Indian culture through intelligent adaptation and assimilation of new forces in the light of a reinterpreted past” (cited in Madan Reference Madan1995, p. 3). Drawing on Dilthey’s hermeneutic understanding of tradition, he considered it “a condition of rather than an obstacle to modernization”, and elaborated the concept Madan calls “generative tradition” (Madan Reference Madan1995, p. 18, p. 22). He insisted on the historicity of tradition and opposed mythical reconstructions of the past, such as Gandhi’s Rama-rajya (Kingdom of Rama) or Panchayati Raj (village republic). At the same time, though considering himself a Marxist, he opposed the uncritical application of dialectical materialism to India, proposing instead that “the study of Indian traditions […] should precede the socialist interpretation of changes in Indian traditions in terms of economic forces” (cited in Madan Reference Madan1995, p. 16). In other words, he thought the object of Indian sociology should be the study of the distinctive dynamics of the evolving Indian tradition. Mukerji was thus a forerunner of the project of an alternative Indian modernity.

In his celebrated The Intimate Enemy (1983), Ashis Nandy stated that his aim was “to make sense of some of the relevant categories of contemporary knowledge in Indian terms and put them in a competing theory of universalism”. He implicitly subscribed to Mukerji’s conception of tradition and regarded his work as belonging to “the tradition of reinterpretation of tradition to create new traditions” (Nandy Reference Nandy1983, pp. xiii-xiv). This was in fact the way Gandhi had related to the Hindu tradition. In his restructuring of the Hindu tradition, Gandhi had brought marginal, low-status commercial and peasant elements to the fore, thus “making its cultural periphery its center”. It was no accident that his assassin was, like the two men who had tried to kill him earlier, a Hindu nationalist and a Brahmin from Maharashtra, a region of Brahminic dominance, and thus “a representative of the center of the society that Gandhi was trying to turn into the periphery” (Nandy Reference Nandy1980, p. 72, p. 76; also Reference Nandy1995, p. 65). Gandhi, however, appears in a very late stage of Nandy’s story of the impact of colonialism on the Indian culture and personality. The key period in his historical narrative of the psychology of colonialism is the second half of the 19^th century when Western cultural ideas percolated “to the deepest levels of Hindu religious idea”, and “Western cultural theories of political subjugation and economic backwardness” were accepted and internalized. Nandy’s hero is a Brahmin pandit, Iswarchandra Vidyasagar (1820-1891), who maintained his authenticity and sought to reform the Hindu tradition from within on the basis of a theory of cultural progress through resolution of its internal contradictions and “refused to Semiticize Hinduism” (Nandy Reference Nandy1983, pp. 26-27). Vidyasagar, a forerunner of Indian nationalism, was an outstanding exception, however. Everyone else succumbed to the clash of cultures, and the culture of the colonizer was lodged within the personality of the colonized, making the former the latter’s intimate enemy. Vidyasagar’s attempted reform of Hinduism can be said to be an intra-civilizational process, like Srinivas’s Sanskritization, while Nandy’s post-colonial Indian personality is the product of an inter-civilizational encounter forced by imperialism. In both cases, the intra- and inter-civilizational processes are interdependent; each one is stimulated by the other and interacts with it. Be that as it may, with the intimate enemy within, the dispossessed Brahmins who created Hindu nationalism in competition for mass mobilization with Gandhi were far from intent on the restoration of the dislocated center and polymorphous traditional Hinduism. To convert Hindus into a modern nation, they redefined Hinduism into a “religion along Semitic lines”, while transforming its living traditions into a nationalist political ideology, Hindutva, which began its vigorous second life in the closing decades of the 20^th century (Nandy et al. Reference Nandy1995, pp. 57-69).

Max Weber had highlighted the “status honor” attached to “caste” as a distinctive feature of social organization in India. This was effective for differentiating caste as a social group from “class” but as a result caste was put alongside the “estates” of medieval Europe from which it differed significantly. Weber, however, did not center the Indian civilization on caste but on Hinduism, and did so mainly from a somewhat extraneous viewpoint of the obstacles to the development of capitalism. Dumont’s Homo Hierarchicus, the consummation of his project for a sociology of India “at the point of confluence of Sociology and Indology”,Footnote ¹³ presented the hierarchy of castes, which he derived from the binary opposition of purity and pollution underlying the four Vedantic varnas, as the core of the civilizational unity and distinctiveness of India (Madan Reference Madan2001, pp. 478-479). As a work in Durkheimian (comparative) sociology, it was marked by a holism that made it unsuitable as an explanation of social change. Dumont had tried to tamper with the holistic character of caste as a total system by locating change in the politico-economic domain, artha, as relatively autonomous with regard to absolute values pertaining to the domain of dharma. In studying the relationship between these two domains, dominated by the Brahmins and the kings respectively, Heesterman (Reference Heesterman1985), who did not share Dumont’s Durkheimian presumption of complementarity of the two domains, found in them a contradiction at the core of Indian civilization that could explain change (Madan Reference Madan1995, pp. 63-71). All this still leaves out of the picture the lower castes and the untouchables which constitute at least one half of the population and, with them, the issue of a possible resistance to dominant cultural patterns.

Resistance to the dominant or hegemonic cultural pattern would not be missed by Indian Marxist intellectuals. They were at a great disadvantage when highlighting distinctive Indian categories and historical patterns, however, and were indeed in a state of denial. As Srinivas (Reference Srinivas1992, p. 3) puts it, the Indian Marxists “had imposed a taboo on caste” that was not lifted until the end of the 1980s. That Marxian class could be provincialized as “subaltern groups and classes”, however, presumed the unacknowledged fact that the fundamental unit in Indian social structure was caste and not class. Starting from a Gramscian-Marxian point of departure, the founder of subaltern studies, Ranajit Guha (Reference Guha1982, p. 4), stated that the autonomous domain of subaltern politics “was ‘traditional’ only in so far as its roots could be traced back to pre-colonial times, but it was by no means archaic”. Either Mukerji’s or Redfield’s conception of tradition would have obviated the need for such an awkward acknowledgement of the transformation of tradition by denying its presumed opposition to modernity.Footnote ¹⁴ But the word traditional was at least allowed in. The studious avoidance of the term “caste” (or, for that matter, “community”) is remarkable in what follows. Subaltern politics

The suppressed narrative that is crying out is that of caste, community and religion – precisely what later subaltern studies revealed as distinctive features of the colonial legacy, and of post-colonial India!

Dipesh Chakrabarty (Reference Chakrabarti2007, p. xi, p. 15, p. 19) explains how he was intellectually formed “in the 1970s in the passionate scholarship of Indian historians of Marxist persuasion” and had to move from “the analytics of capital and nationalism available to Western Marxism” and “to open up the Marxist narratives of capitalist modernity to issues of historical difference”. Only by doing so could “the experiences of political modernity in non-Western nations” serve the project of “provincializing Europe” (Chakrabarty Reference Chakrabarti2007: 16).Footnote ¹⁵

Partha Chatterjee (Reference Wright1993) pushed forward the provincializing project with regard to Indian nationalism. His study of nationalism also illuminated caste and community, which are presently under discussion. Nationalism, as conceived under the British Empire, became the hegemonic discourse of post-colonial India, as the colonial state was taken over intact by the nationalist elite. The nation’s fragments, overlooked in that discourse, had to be taken into account once the Constitution of 1950 instituted universal suffrage, which, according to Srinivas (Reference Srinivas1992, p. 12, p. 28), set in motion India’s revolution, a revolution that “began quietly and non-violently” but was by 1986 “entering an increasingly violent phase”. Chatterjee brings the nation’s subaltern fragments into the picture, adding to women and peasants the “religious minorities” created by the nationalization of Hinduism, with mounting communitarian violence as one of its consequences. He also breaks the Marxist taboo by offering his synthetic theory of caste. He departs from Dumont further than Heesterman by considering the oppositions reconciled at the level of caste ideology and the Hindu self-consciousness in fact socially unresolved, and generative of constant change in the context of democracy and mass mobilization. The politics of categories, both under colonial administration and postcolonial mobilization, cumulatively restricted the manifold uses of the term jāti, applied to a variety of fuzzy communities, including species, race and nation, for two types of communities made distinct on the basis of caste and religion (Chatterjee Reference Chatterjee1993, pp. 220-223). The key shift and fundamental change, however, is that the axial unifying concept is no longer dharma but “‘nation’ as concretely embodied in the state”. The result is

As a matter of fact, caste identities have become much more assertive in the public sphere and politics, and caste competition and conflict have intensified. Defying the segmentary logic of the traditional caste system, highly effective, caste-based electoral coalitions have consolidated themselves as vote-banks (Srinivas Reference Srinivas1992, pp. 6-24, p. 41). On the other hand, the kind of claims the castes make upon the modern state are rooted in the traditional Indian pattern of the hierarchy of status honor. A major concern of democratic politics in India is the redistribution of dignity. Democratization is focused not on individual but on group rights, especially the “recognition of equality among self-recognizing caste groups”. A strident new language of group rights has emerged, resulting, incidentally, in the marked vernacularization of Indian politics since the 1970s, and the most radical demands are “for group rights rather than income equality between individuals […] The largest number of the Indian poor themselves seem to be more intent on removing degradation rather than poverty” (Kaviraj Reference Kaviraj, Skinner and Stråth2003, p. 356, p. 370).

Redfield’s notion of tradition as possessing a temporal structure and evolving through time has been independently discovered and drawn upon in the search for an alternative modernity outside of India as well. The contemporary rise of Islamic fundamentalism can thus be seen as a new stage in the evolution of the Islamic tradition, as it is by Talal Asad (Reference Asad2003, pp. 222-248). Asad conceives tradition independently of the invidious contrast with modernity, drawing on Koselleck’s idea that it has a temporal structure centered on the present, and sees the Shari‵a as “a traditional discipline”. “Islamists”,Footnote ¹⁶ taking for granted and working through the nation-state, “relate themselves to the classical theological tradition by translating it into their contemporary political predicament” (Asad Reference Asad2003, p. 198). The proponents of Islam as a public religion in the 1960s and 1970s drew on marginal elements and figures in the Islamic tradition to elaborate a model of “the Islamic state” as the basis of Islamic political ideologies.

Multiple modernities in old and new nation-states and diversity in their patterns of secularization

Max Weber compellingly argued for the world-historical significance of the modern European state in terms of its superiority in rational administration and law over earlier forms of political organization, and its profound penetration of and impact on society. It was not the European state, however, but its idea – the ideal-type of the modern state of a sovereign nation –, that spread throughout the world and created the single most powerful collective agency in the non-Western world. This is true even in the increasingly frequent cases of state failure, as in Sub-Saharan Africa and Afghanistan, where the social transformation set in motion by the failed states has been profound. The nation-state is thus the fundamental element in the political structure of what Eisenstadt calls the civilization of modernity. Yet its organization and institutional components have been selectively adopted and undergone marked transformations in different regions of the world because of different conditions, including diverse expressions of the will to alternate modernity. Diversity is thus a far more striking feature of global state formation than uniformity.

The independent states of Latin America were born in the early 19th century with the disintegration of the Spanish empire. In the following decades, we witnessed the incipient transformation of imperial and traditional monarchies into modern states in the Middle East, Japan and China. The breakup of the Hapsburg and Ottoman empires after World War I coupled statehood with nationhood in the international system. The era of the nation-state was extended to the rest of Asia and Africa after the World War II. Africa had been colonized in the latter part of the 19th and early 20th centuries mainly by imperial conquest, but the Indian and Indonesian empires had seen the much slower creation of trading companies, namely the English and Dutch East India Companies. The universalization of the nation-state went hand in hand with decolonization in Asia and Africa. The sharpest contrast among the post-colonial states lies between the general failure to build national states in sub-Saharan Africa and its success in India and Indonesia, the two most populous new nation-states.

The old Republics of Spanish America and the new states of Africa share a common characteristic: they gained independence as administrative regions of disintegrating empires. Nation-building and nationalism were much more the consequences of independence than its causes. In Latin America the process of national integration was gradual and took many decades, with negligible impact from the international system of sovereign states. This mode of national integration also left a permanent imprint on the character of Luso-Hispanic states of Latin America. Caudilismo and Caciquismo are systems interlocking local and national systems, and the delivery of votes in exchange for patronage. These, however, were unstable systems of personal power (Linz Reference Linz2000, pp. 155-157). In his analysis of the development of the Mexican state since the revolution of 1911, Alan Knight (Reference Knight, Centeno and López-Alves2001) shows how these systems of local power and brokerage were integrated by the 1940s, after decades of revolutionary mass mobilization, into a stable political state system dominated by the Revolutionary Institutional Party, creating a political regime and state-party system, distinct from both the totalitarian and the liberal democratic state, that lasted to the end of the 20th century.

British colonial rule in India after 1858 constituted a major rupture with pre-modern Indian polities. To highlight this rupture, Kaviraj characterizes the Hindu and Muslim kingdoms that preceded it, including the Mughal empire, as belonging to the category of “subsidiary”, as distinct from “sovereign”, state because they did not in principle attempt to change the society they ruled and respected its autonomy.Footnote ¹⁷ The British colonial state was like the sovereign states of Europe but its colonial character prevented it from developing all aspects of a sovereign state (Kaviraj Reference Kaviraj2005a, p. 275). With independence and the transfer of sovereignty from the British monarch to the people of India, this intermediate form of state became a society-centered, sovereign democratic state. After independence, the three pillars of the colonial state – the bureaucracy, the police and education – were taken over intact. It should be added that the British Colonial Development and Welfare Acts of 1929 and 1940 had made development “the highest stage of colonialism” and thus the prerequisite for independence (Jackson Reference Jackson1990, p. 93). From 1937 until his imprisonment in 1940, Nehru had chaired the important Planning Committee of Congress. The idea that thus took shape in his mind, and that he was able to realize after 1947 (with little opposition due to the death of Gandhi and Patel shortly thereafter), was that of a developmental state. And the Nehruvian developmental state was society-centered and interventionist (Kaviraj Reference Kaviraj, Skinner and Stråth2003, p. 346).

“Indian society” Kaviraj (Reference Kaviraj2005a, p. 274) tells us, “could not draw upon the existing body of conceptual and theoretical resources to make sense of […] the new institutional and practical forms of political power”. Nationalist thinking about a distinctively Indian sociopolitical order was based not on the historical experience of pre-colonial India, nor on Indian statecraft, arthashastra. It was rather the product of romantic free imagination. Nationalist ideas, from Bhudev Mukhopadhyay’s idea (1892) of an “inner organization”, whose “normative principles derived from a collectively accepted and intelligible normative order of dharma”, to Gandhi’s “government of the self and government of society”,Footnote ¹⁸ had little impact on the Indian constitution crafted by Nehru and the untouchable leader Ambedkar, except for a concession to the memory of Mahatma Gandhi’s “Panchayati Raj” with respect to local government by village councilsFootnote ¹⁹ (deSouza Reference de Souza and Bhargava2008).

The Nehruvian developmental state served as an effective instrument of social change. It led the industrialization of India and supervised the rapid growth of a modern middle class. Its constitution abolished untouchability and directed the state to remove the inequities of the caste system. The Nehruvian state thus “appealed powerfully to subaltern imagination of politics” (Kaviraj Reference Kaviraj2005a, p. 291). Its affirmative action took the form of a system of reservations in legislatures, civil service and educational institutions. Its colonial origins notwithstanding, the state was “transformed into a central moral force, producing an immense enchantment in India’s intellectual life” (Kaviraj Reference Kaviraj2005a, p. 263); and it was consequently sustained by considerable popular legitimacy.

State formation in Africa followed a very different trajectory from the ones discussed so far. The first Article of the UN Charter on “self-determination of peoples” laid the foundation for the establishment of what Jackson (Reference Jackson1990) has called “quasi-states” or “negative sovereign regimes”, where the new states could claim territorial sovereignty but without administrative capacity or broad, if not consensual, legitimacy. After independence for the African colonies of Britain, France, Portugal and Belgium, the juridical sovereign state replaced a bifurcated colonial state with a minute though growing proportion of population enjoying the status of citizens under civil law in the cities and the bulk of rural population constituting the subjects, organized under “indirect rule” into ethnic communities defined by their customary law administered by tribal and local chiefs (Mamdani Reference Mamdani1996). Nation-building was not a cause but a consequence of independence, as in Spanish America but, in contrast to the latter case, it was often a failure. Consequently, ethnic civil war and genocide, as well as strident kleptocracy and state failure, are prominent features of the contemporary world order in Africa, shared by other countries where the presumption of nationhood for tribally or communally-organized colonies as imperial administrative units was similarly erroneous, and where the idea of the state had not grown or generated internal legitimacy. The endemic spread of civil wars in post-colonial Africa predominates in the global pattern of war and violence, and is only replicated, though not matched, by those following the collapse of the Soviet empire (Wimmer Reference Wimmer1997, Min and Wimmer Reference Min, Wimmer, Anheier and Isar2007, pp. 67-68, pp. 73-75).

Senegal is one of the few exceptions. Its post-colonial state has been sustained for half a century without a civil war. However, the developmental state conceived according to the 1945 blue-print for sovereign states and set up after independence in 1960, gradually succumbed to the corruption and clientelism typical of Africa, only to be debilitated externally in the 1990s as a result of the internationally imposed structural adjustments that stemmed from the discarding of the idea of the developmental state in favor of the market economy (Diop 2004). But a theoretically more interesting feature of Senegalese historical experience is the failure to generate a civilizational process comparable to those producing social transformation in India and Iran. Senegal’s independence leader and first president, Léopold Senghor, was also a major African intellectual and architect of the philosophy of negritude. Senghor’s idea of negritude, however, failed to act as the engine of a culturally distinct pattern of social transformation. It failed to find effective social bearers and generate an intra-civilizational process comparable to Islamicization or Sanskritization. In fact, the Muslim maraboutic clergy of the Sufi congregations in rural Senegal – the social stratum comparable to the Shi‵ite clerical estate in Iran or the Brahmins in India as bearers of religious tradition – acted as political brokers of the new political elite in the countryside, perpetuating the clientelism and corruption that pervaded Senghor’s developmental state by the late 1970s (Diop Reference Diop and Diop1993, pp. 4-7).

Diversity also marks the range of variation in the modern state’s relation to religion and society, which is left out in the typologies of political regimes, notably totalitarianism, authoritarianism and democracy (Linz Reference Linz2000 [1975]). The typological approach is increasingly focused on the transition from authoritarianism to democracy (Mahoney Reference Mahoney, Mahoney and Rueschemeyer2003), and it leaves out the differences in the political culture that sustains the modern state and gives it legitimacy in different parts of the world. In historical record, however, the placing of religion in the constitutional order presented a vexing problem to non-Western modernizers when they were considering different variants of the European constitutional state for adoption.

The cultural aspect of secularization, culminating in secularity in our time, defined by Charles Taylor in his celebrated A Secular Age (2007) as the conditions for the experience of and search for the spiritual in an immanent frame, is a distinctive Western developmental pattern spread in the global age.Footnote ²⁰ As a culturally specific developmental pattern, it should be compared to contrary intra-civilizational processes, such as Sanskritization and Islamicization discussed earlier. Long before the Secular Society of London, founded in 1846, had “secularism” written on its banner to celebrate the coming of Taylor’s new age, however, “secularization (saecularisatio)” had been a legal term meaning alienation of church property and a person’s change of status from monastic to civilian, and it had entered the language of constitutional law concerning church-state relations (Brunner, Conze and Koselleck Reference Brunner, Conze, Koselleck, Brunner, Conze and Koselleck1984).

Matthias Koenig (Reference Koenig2007) has highlighted the institutional varieties of secularism in the West, demonstrating the dependence of the articulation of the secular and the political on state formation. But further provincializing of the already battered notion of the secular state is needed for a more adequate comparative understanding of the impact of state-building on religion. Modernization of the state has had a strong but varied impact on redrawing the boundaries between religion and politics in the Muslim world and India, and consequently on their respective patterns of popular resistance and revolution.

The word “secular” had no obvious equivalent in Arabic, Persian, Turkish or Hindi; nor was its idea familiar in the cultures based on those languages. The Iranians never discussed the idea in their constitutional revolution, and the Fundamental Laws of 1906-1907 entrenched Twelver Shi‵ism as Iran’s official religion and required all legislation to conform to the Shari‵a (Islamic law), making it subject to the approval of a committee of five religious jurists. The Turkish Republic went to the other extreme and chose the French doctrine of secularism or laïcisme, adopting the word and the concept in its constitutional amendment of 1928. Egypt, in contrast, devised its own variant of secularization in the course of state-building, leaving the Arabic neologism for secularism, ‵almāniyya, to be coined later and after the fact.

Talal Asad (Reference Asad2003) discussed the decisive connection in Egypt between state-building and the late-19th century Islamic reform movement of Muhammad ‵Abduh (who died in 1905), Chief Mufti of Egypt. With ‵Abduh’s legal reform project, carried out in the decade following his death, the Shari‵a came to be equated with legal rules for marriage, divorce and inheritance – the substantive fields historically under its exclusive jurisdiction. The result was the reconfiguration of law, ethics and religious authority, “not simply abridgement but […] a re-articulation of the concepts of law and morality” (Asad Reference Asad2003, p. 248). Law became the domain of the state and religious courts incorporated into its judicial hierarchy until they were abolished by Nasser in 1955, while morality was relegated to the private, religious sphere.

By that time, however, a notion had been born that grew monstrously in the following decades of Islamic resurgence: “the Islamic state”, defined as the counterpart to the secular state by its function of execution of the Shari‵a (misconceived as positive law of the state). Categories of “politics” and “religion” come to implicate each other profoundly in the contemporary resurgence of political Islam because “Islamism’s preoccupation with state power is the result […] of the modern nation-state’s enforced claim to constitute legitimate social identities and arenas” (Asad Reference Asad2003, p. 200). As the Indonesian thinker, Nurcholish Madjid (who died in 2005), already recognized in 1972, the idea of the Islamic state was doubly apologetic: it was an apology in relation to modern ideologies which “gave rise to an ideological-political understanding of Islam”, and it was an apology in relation to modern constitutional law in order to demonstrate that the Shari‵a amounted to “laws and regulations that are superior to other laws” (Madjid Reference Madjid and Kurzman1998, pp. 293-294). The concept of the “Islamic state”, Madjid (Reference Madjid and Kurzman1998, p. 294) concluded, “is a distortion of the [properly] proportioned relationship between state and religion”. More recently, Abdollahi An-Na‵im (Reference An-Na‵im2008, p. 3) has shown the idea of “the Islamic state” to be a product of post-colonial discourse based on “European notions of the state and positive law”. This distorted idea nevertheless had a deep impact on the reconstruction of modernized states, resulting in the creation of the first successful modern theocratic state in the Islamic Republic of Iran, and of the less successful attempts to set up Islamic states in the Sudan and Afghanistan, not to mention the Islamicization of the state in Pakistan and quite a few other Muslim countries.

The revolutionary transformation of Shi‵ite Islam in the context of this cultural impact of the modern state emerged with the Islamic revolution of 1979 in Iran. Putting that revolution in comparative perspective, I concluded that

The legitimacy of the Ottoman and Iranian states was deeply rooted in history and in the centuries-long process of defensive modernization against Western imperialist encroachments since the 19^th century (Rustow Reference Rustow, Holt, Lambton and Lewis1970). With the Islamic revolution of 1979, it was the monarchy that was delegitimized and the Shah who was overthrown, but the modern state lost little or none of its legitimacy. The Shah’s state was captured by a Grand Ayatollah leading a Shi‵ite hierocracy weakened but not destroyed by its secularizing policies. The dispossessed militant clerics under Khomeini’s charismatic leadership thus acted as the bearers of the process of Islamicization in Iran. Khomeini’s Mandate of the (Religious) Jurist (velāyat-e faqih) was constitutionalized and subjugated the state to clerical authority; and it set in motion a process of Islamicization of the Iranian state and its legal frame. In this process, the transformation of Shi‵ite Islam has been as profound as the Islamicization and clericalist Shi‵itization of the Iranian state. The result is a distinctive kind of modern state, the first theocratic Islamic republic in history. Under the Supreme Jurist, President Ahmadinejad currently presides over the centralized bureaucratic state created by the Pahlavi Shahs but relying on the Revolutionary Guards whose organization has been rationalized and bureaucratized in place of the Shah’s army. During Iran’s variant of “privatization”, the Revolutionary Guards have been given an economic empire that predominates over the revolutionary foundations created out of confiscated property in a corporatist economy (Arjomand Reference Arjomand2009, ch.2).

Walter Skya (Reference Skya2009, p. 297, p. 316) has found my analysis of the distinctive feature of the Islamic revolution in Iran persuasive, and concluded his comprehensive study of what Barrington Moore had reductively characterized as Japan’s labor-repressive fascism with the suggestion that “the religious revolt against the secularized Western world started not with the Iranian revolution based on the mass-based radicalization of the Islamic religion, but with the radical Shintō ultranationalist movement and its war against the Western world”. Skya traces the emergence of a distinctly Japanese authoritarian theory of the family-state around the turn of the 20^th century and its transformation, in contestation with liberal democracy and socialism in the era of mass politics, into a totalitarian ideology of Shintō utranationalism. This ideology became the official ideology of the Japanese nation-state in 1937 and guided its entry into World War II as a holy war in the name of the emperor. It is interesting to note that this ideological development was mainly the work of constitutional lawyers who were trained in Germany but came to consider secularized Western civilization as Japan’s mortal enemy. Step by step, they interpreted the 1889 Constitution of the Empire of Japan away from its German model, integrating it into Shintō cosmology and the “Way of the Gods as Such”. According to Kaheki Katsuhiko, the last in this line of constitutional interpreters, the “great life” was “the emperor and the masses united as one heart, same body. The state […] was composed of the emperor and the masses (okuchō) who had abandoned their individual selves to serve the emperor” (cited in Skya Reference Skya2009, p. 198). Japan’s Asian empire was justified because the Japanese were the race that descended from the gods and the emperor was the reincarnation of the sun god, Amatersu Ōmikami.

North Korea, the most interesting post-colonial state of the Japanese Asian empire, adopted Japanese racist ultra-nationalism as its own nationalist ideology under the guise of Marxist-Leninism. Some of the main architects of North Korean racist ideology as the nationalism of “the cleanest race” had actually worked in the Japanese propaganda offices of the colonial period (1910-1945), and gave Japanese racism a heavy victimization inflection as befitted a liberated colony. Korea’s history was the history of a child-race sequentially abused by the Chinese, the Japanese and the Americans. Both Dear Leaders were accordingly passive-aggressive mother figures: Kim Il Sung tucked children into bed, and his son and successor, Kim Jong Il, is the “Mother General” at whose “breast” lies the nation. The fact that North Korea has now survived the collapse of communism by two decades can in part be plausibly attributed to this distinctively North Korean adaptation of the racist ideology of Japanese imperialism (Myers Reference Myers2010).

As Rajeev Bhargava (Reference Bhargava2008, p. 2) points out, “for all the talk of pluralism and multiculturalism in Western political theory, there is barely a mention of how these issues are tackled […] in India”.Footnote ²¹ The embeddedness of Hinduism in ritual practice rather than belief and the diversity of religious communities in India, their hierarchical nature and oppressiveness of individualism presented difficult problems to India state-builders. In response the independent developmental state committed to the promotion of equality, on the one hand, and to the contradictory amalgam of valorization of communities due to the mass mobilization for independence and resulting from subsequent democratization, on the other. Makers of the Indian Constitution of 1950 recognized the rights of religious communities and established a distinctive pattern that recognized religious pluralism. A developmental state committed to social reform and affirmative action could not build any wall of separation between itself and religion. “In accepting community-based rights for religious minorities and endorsing state intervention in religion”, the Indian Constitution “developed its own modern variant” of secular principles (Bhargava Reference Bhargava and Srinivasan2007, p. 36). The Indian state accordingly opted for a policy of “principled distance”. In theory, this means strict neutrality in helping or hindering all religions to an equal degree (Bhargava Reference Bhargava and Srinivasan2007, pp. 38-40). In practice, it came to mean a good deal of state support for religious practices and institutions by the Indian secular state.

The tension between the contradictory values of communal rights, social equality and religious freedom in the Indian constitutional settlement has thus built political compromise into India’s secular principles. Indian secularism has consequently met with sharp criticism from liberals and conservatives alike. Mehta (Reference Mehta and Hasan2005) highlights the practical impossibility of principled distance and parity or “Congress secularism”. By reinforcing communal identity, it has failed to protect the religious freedom of the individual and to extend its reformist interventionism through legislation and judicial review of the Supreme Court from Hinduism of the majority to the Islam of the largest minority. T.N. Madan (Reference Madan1997) has rejected secularism as an allegedly generic and universal principle as well as Indian secularism as its specific variant, arguing against the transplantation of secularism from its Western Christian home to the vastly different religious sphere of India. As shown by the contradictions among different Articles of the Constitution, “a communally divided society and a secular state could be mutually contradictory”Footnote ²² (Madan Reference Madan1997, p. 249). Furthermore, “the emergence of state-sponsored religious pluralism, summed up in the slogan sarva dharma samabhava (equal respect for all religions)”, enshrined in the 1950 Constitution, “does not go very far in strengthening inter-religious understanding and appreciation” (Madan Reference Madan, Malek and Reifeld2005, p. 72). Nandy is more vehement in his rejection of secularism than Madan and finds it contradictory to democracy in India. He proposes closing the debate on secularism in order to clear up “the rubble of dead categories occupying public space”Footnote ²³ (Nandy Reference Nandy, Needham and Rajan2007, p. 108).

Varieties of nationalism and their civilizational background

Unlike the state, for which Weber had given us the definition, the conjoint “nation” in the conception of nation-state was notoriously under-theorized in social theory, and relegated to particular histories and “area studies” by the modernization theorists (Chatterjee Reference Chatterjee1993, pp. 3-4). Elie Kedourie (Reference Kedourie1966) considered nationalism an invention of stateless German intellectuals such as Fichte in early the 19^th century. This German genealogy of nationalism was put forward as universally relevant as it was allegedly imitated by the new nation-states of the post-colonial world. Gellner (Reference Gellner1983) freed the idea from its German genealogy and put forward a general concept of nationalism within the frame of the modernization theory. Hobsbawm (Reference Hobsbawm1992, pp. 9-11, p 23) corrected what he saw as Gellner’s bias for viewing modernization from above by historicizing the impact of “the principle of nationality” in the context of European mass politics from 1830 to 1880. Hobsbawn’s historical sense made him recognize that “the extension of ‘nationalism’ beyond its region of origin moves it beyond the range of the original analysis of the phenomenon” (Hobsbawm, p. 160). Had he considered the phenomenon after the declaration in 1918 of Woodrow Wilson’s Fourteen Points, including national self-determination, he may have been struck by the civilizational perspective brought to light by Prasenjit Duara (Reference Duara2001). The triumph of nationalism and the world system of nation-states in the inter-War period, in fact, gave rise to a new conception of civilization, or more precisely, was decisive for the shift from the conception of civilization in the singular to that in the plural. Japan and other nation-states of Asia, and the Indians who aspired to independent nationhood, needed this new conception of an alternative Asian civilization as the transcendent basis for their nationalisms. The architects of modern Iranian nationalism under Reza Shah, for instance, invited the Indian poet and Nobel laureate, Rabindranath Tagore, for an elaborate state visit in 1932 to celebrate their common Asian civilization (Marashi Reference Marashi2010).

Gellner and Hobsbawm had little impact as compared to Benedict Anderson’s Imagined Communities (1983). Anderson presented the nation as a universal and uniformly-imagined community born of the impact of “print capitalism”.Footnote ²⁴ The conception of the “nation” was a good deal more ambiguous than Anderson allows. Secularism as the distinctive feature of Anderson’s modularly imagined nationalism is particularly problematic. In Spanish America, “national consciousness emerges as an offshoot of the religious expansionism” of the Spaniards as “old Christians and the chosen people” (Lomnitz Reference Lomnitz, Centeno and López-Alves2001, p. 339). The priests who led the movement for national independence in Mexico, Morelos and Hidalgo, accused the Spaniards of betraying their true Christian mission, with the former proclaiming, “Know that when kings go missing, Sovereignty resides only in the Nation” (Lomnitz Reference Lomnitz, Centeno and López-Alves2001, p. 348). The Creoles, whom Anderson takes as the bearers of nationalism, to use Max Weber’s term, must share that honor with the Bourbon reformers who inadvertently set the nationalist movement in motion (Lomnitz Reference Lomnitz, Centeno and López-Alves2001, pp. 346-348). Furthermore, the presumption of horizontal and egalitarian fraternity in modularly imagined nationalism is not borne out in Spanish America with its corporatist, communal and hierarchical institutions of the empire.

In Spanish America, empire was intimately connected with the birth of nationalism. The same is true in the case of India, albeit in terms of a completely different empire with very different characteristics. It was with reference to India that Anderson’s account of the allegedly uniform making of nationalism was most vehemently challenged by Chatterjee’s (Reference Chatterjee1993) question: Whose imagined community makes for which nationalism? The bearers or carriers of nationalism in India were a new social class formed under British imperial rule in the latter part of the 19^th century: the middle class babus, most notably the Bengali bilingual intelligentsia calling itself the bhadralok (respectable folk).Footnote ²⁵ The peculiar social position of this group left an indelible mark on Indian nationalism. These architects of Indian nationalism did not swallow any “modular” nation imagined for them by Fichte or anyone else in Europe; on the contrary, they conceived it as the difference marking them off from the colonial power. This nationalism was inserted into the new public space of the Raj as a new form of political mobilization in order to overcome their subordination as the colonized middle class. The nation was already sovereign for them, even though the state was in the hands of the colonial power (Chatterjee Reference Chatterjee1993, pp. 5-6, p. 74).

If secularism as a presumed feature of modularly imagined nationalism did not fit the earlier Spanish-American nationalism, it was equally at odds with the preeminence of Oriental religion in Indian nationalism. Chaterjee refuses to take the claim of nationalism to be a political movement at face value. It was primarily a cultural project, and Chaterjee follows Nandy in considering the claim to religion as a distinctively Oriental and essential component of the Indian nationalist imagination. He accordingly gives considerable attention to the middle class movements to reform Hinduism. Nationalism was born of colonial difference. It was an attempt to appropriate language, religion and family life as the “spiritual” or “inner” aspect of culture, and “was of course premised upon a difference between the culture of the colonizer and the colonized” (Chatterjee Reference Chatterjee1993, p. 26). The nationalists thus constructed “a new sphere of the private in a domain marked by cultural difference: the domain of the “national” as defined as one that was different from the “Western”.Footnote ²⁶ As a result of this colonial legacy, “the search for a postcolonial modernity has been tied, from its very birth, with its struggle against modernity” (Chatterjee Reference Chatterjee1993, p. 75).