The Never-ending Story of Modernization Theory

Wolfgang Knöbl. Handbook of Historical Sociology. Editor: Gerard Delanty & Engin F Isin. Sage Publication. 2003.

Although modernization theory has been one of the most influential theoretical paradigms in the social sciences since the 1950s, there is no canonical text expressing all hypotheses of the theory, and no author has really dominated and structured the whole debate. It rather seems to be that modernization theory often was not much more than a bundle of hidden, but decisive, assumptions in the minds of social scientists who tried to link empirical research to various large-scale historical and social processes diffusely called ‘modernization.’ This makes it extremely difficult to talk about ‘modernization theory proper’ as there are no true ‘believers,’ no aggressive ‘renegades’—the missing canonization of the theory obviously doesn’t allow such clear-cut categories. Nevertheless, there is no doubt that it is possible to circumscribe the contours of the theory by pointing to ground-breaking works published between the mid-1950s and mid-1960s, mostly by American authors—for example, Daniel Lerner’s The Passing of Traditional Society (1965 [1958]), Seymour Martin Lipset’s Political Man (1988 [1959]), Neil J. Smelser’s Social Change in the Industrial Revolution (1960 [1959]), Walt Rostow’s The Stages of Economic Growth (1971 [1960]), David McClelland’s The Achieving Society (1961) or Gabriel Almond and Sidney Verba’s The Civic Culture: Political Attitudes and Democracy in Five Nations (1989 [1963]). As a non-Marxist, macro-sociological and often interdisciplinary theory of social change, modernization theory tried to conceptualize either historically or typologically the development of societies, focusing in the beginning mostly on the relationship between culture and economic progress, but increasingly also on that between culture and political development and between economic growth and democracy. As has been shown by various interpreters of the history of this approach (see Alexander, 1994: 168ff; Harrison, 1988: 30ff; Huntington, 1971: 288-90), modernization theorists assumed that:

  • Modernization is a global and irreversible process, which began with the Industrial Revolution in the middle of the eighteenth century (or even earlier) in Europe, but which nowadays, that is, since the end of the Second World War, concerns societies all over the world;
  • Modernization is a historical process leading from traditional to modern societies, thus implying a sharp antithesis between tradition and modernity.
  • In traditional societies and countries of the so-called ‘Third World’ there is a dominance of personal attitudes, values and role structures which can be characterized by terms like ‘ascription,’ ‘particularism’ and ‘functional diffuseness,’ and which are to be interpreted as powerful barriers for economic and political development;
  • In modern societies of the Euro-American civilization there is a predominance of secular, individualistic and scientific values and corresponding role clusters;
  • Modernization is a more or less endogenously driven process to be localized within societies which should be regarded as coherent wholes and—if possible analysed with the theoretical instruments of structural functionalism; and
  • Social change towards modernity in different societies will take place in a rather uniform and linear way.

These were the guiding premises of research on developing countries leading to the further assumption that the Western economic and political model soon ‘reappears in virtually all modernizing societies on all continents of the world, regardless of variations in race, color, creed’ (Lerner, 1965 [1958]: 46). But modernization theory seemed not only well suited to analysing contemporary social change; the theory could obviously also bring forward historical explanations for the rise of the West in the nineteenth century (Rostow, 1971 [1960]) as well as detailed prognoses concerning the future structure of traditional, developing and already quite advanced countries in which processes like ever-increasing economic growth, a continuing structural differentiation and a further weakening of traditional values were to be expected. Thus, modernization theory could not only be applied to the rather narrow field of the sociology of developing countries; the theory very soon had a far larger claim, namely that it is really a global theory of social change comparable to Marxism.

It is well known that modernization theory as characterized had its good and bad times. It thrived from the beginning of the 1950s to almost the end of the 1960s; it was ‘dead’—as Marxists like Immanuel Wallerstein (1979: 132-37) asserted, not without reason—from the beginning of the 1970s to the middle of the 1980s; and there has been another heyday since the end of the 1980s. Because of these ups and downs, it is necessary to historicize modernization theory, to ask why and how it suddenly appeared in the beginning of the 1950s, how the theory was related to the classical works of sociology, and how the theory began to change internally during the late 1950s and 1960s. Answering these questions is particularly important as the new modernization theory emerging in the last two decades of the twentieth century claims to have solved all serious theoretical problems of the older version of the paradigm. This, however, is open to considerable doubt, as the structure of the new theory—as will be shown at the end of this chapter—is not very different to the structure of the old one.

The Origins of Modernization Theory

The emergence of modernization theory was certainly related to political events: to the Cold War in general and Harry S. Truman’s so-called ‘Point Four’ programme in particular. Truman’s Second Term Inaugural Address from January 1949 focused almost exclusively on foreign affairs and—this was Point Four of his speech—gave the American public a vision of how underdeveloped regions in the world could prosper with the help of American technology and know-how to counter the attractions of communist ideology. Soon new government agencies were founded for sending hundreds of technicians into these regions to increase agricultural production and to build up adequate health and educational systems. As it turned out, however, these projects often encountered huge difficulties because of unfamiliar social structures and cultural patterns. It was obvious that most of them needed support by social scientists, who, as historians, political scientists, anthropologists and sociologists, should have had expertise concerning these structures and patterns. Conferences were organized to pool existing knowledge and to spread information among the experts of different disciplines. One of the results of such a meeting in Chicago in the summer of 1951 was a quite influential reader edited by Bert Hoselitz (1963 [1952]), in which an essay by Marion J. Levy really tried to come to grips theoretically with the phenomenon of macro-sociological change by using and historicizing Talcott Parsons’s ‘pattern variables.’

Levy argued that industrial societies are characterized by rational, universalistic and functionally specific value orientations and role structures; non-industrial societies, in contrast, are characterized by non-rational, particularistic and functionally diffuse values and roles. Economic growth—so he claimed will sooner or later completely change non-industrialized societies and thus bring forward the same cultural and social patterns as known in the industrialized West. Levy didn’t deny that there might be difficulties and even ruptures in this process; but they won’t last very long as—this was Levy’s argument—there is a functional interdependency between societal spheres and subsystems: wherever modern value orientations begin to dominate, they will have consequences in other fields: ‘The allocation of goods and services is only analytically separable from the allocation of power and responsibility. Highly universalistic relations in the economic aspects of action are functionally incompatible with highly particularistic ones in the political (i.e., allocation of power and responsibility) aspects of action’ (Levy, 1963 [1952]: 123). Thus, Levy formulated an extremely elegant theoretical position the other participants of the conference obviously could agree with, even if some of them were sceptical whether the transformation process would proceed as fast as Levy assumed. It was this elegance and especially the novelty of the theory which made it possible for Levy’s ‘pattern variables,’ or at least the very idea behind them, quickly to begin to serve as strong background assumptions for macro-sociological research on social change in the following ten or fifteen years, research done by psychologists, historians, political scientists, economists and—of course—sociologists. And a new theory—in various respects—it certainly was, even if the term ‘modernization’ wasn’t used very often in the beginning of the 1950s, and the expression ‘modernization theory’ did not become familiar until the 1960s.

The theory was so new and attractive because of the lack of a strong tradition of macro-sociological research within American sociology, anthropology and political science from the 1920s to the end of the 1940s. Although at least some sociologists in the second decade of the twentieth century analysed large-scale social change—see, for example, Thomas and Znaniecki’s brilliant description of economic and cultural processes in Poland and the United States in their The Polish Peasant in Europe and America (1974 [1918-20]. see esp. Vol. 1, 156ff.), this kind of research and thinking became very rare later on within a sociological discipline specializing more and more in meso- and above all micro-sociological processes. The same was true for social anthropology, in which the strong position of Boasian culturalism and relativism and a predominant focus on quite static ‘tribal’ societies constituted a largely unfavourable climate for theorizing social change. There were exceptions, of course, most notably Robert Redfield who specialized in Mexico, used—influenced by his father-in-law, Robert Park—the rural-urban dichotomy in conceptualizing social change, and thereby even spoke of ‘modernization’ (1968 [1930]: 4). But Redfield was not typical, and neither was the Austrian immigrant Hans Kohn, who—as a political scientist and historian of nationalism also introduced the term of ‘modernization’ in conceptualizing social change in the ‘Orient’ (1937: 262).

Thus Marion J. Levy’s conceptualization of social change, and especially the research which was built on these theoretical ideas, could be considered highly original and new. And all this was so attractive because the new paradigm seemed to incorporate the most valuable theoretical insights of the sociological classics (Joas, 2003 [1996]: 44). At least at first sight Levy’s historicization of the ‘pattern variables’ was able to encompass Max Weber’s ideas concerning the process of occidental rationalization, to clarify Ferdinand Tönnies’s crude dichotomy between Gemeinschaft and Gesellschaft or ‘mile Durkheim’s’ distinction between ‘mechanical’ and ‘organic solidarity.’ A closer look at this subject matter, however, could have raised the question whether Levy’s theoretical position and especially his rather linear view of historical progress was at all compatible with the oftentimes rather sceptical view of the classics on historical processes. Was, for example, Weber’s idea of rationalization, with his emphasis on the clash of value spheres and the concomitant occurring tensions, really compatible with Levy’s rather simple thesis of a switch from non-rational, functional diffuse and particularistic values to modern ones? Did Tönnies really think that there would be a complete loss of Gemeinschaft as could be expected if one translated his terms into Levy’s ‘pattern variables’? And a last ‘strange’ relationship should also not be forgotten. For his conceptualization of social change, Levy used Parsonian theoretical tools—the ‘pattern variables.’ But was Levy’s project really a fruitful concretization of Parsons’s often highly abstract project? It seems rather doubtful that Parsons wholeheartedly supported Levy’s use of the pattern variables for theorizing macro-sociological change. Parsons had a much more complex understanding of modernity than had Levy and therefore could not have been very pleased by the latter’s rather simple historicization of the pattern variables. Parsons never believed—as Levy seemed to—that modernity will bring forward a complete dissolution of all particularistic values, of all ascriptive ties and functionally diffuse role structures. The idea of a partial modernization, a modernization process which mixes modern and traditional elements, was not an alien element in Parsons’s thinking, even if this idea became prominent only later on by internal critics of modernization theory (see Rüschemeyer, 1970). This was one of the reasons why Parsons—nearly at the same time as Levy began to historicize the ‘pattern variables’—claimed in The Social System (1951: 534) that he has no general or systematic theory of social change. In regard to the analysis of cultural processes, Parsons argues that even Max Weber’s theory of rationalization doesn’t tell us very much about a clear direction of these processes as ‘systems of expressive symbols and systems of value orientations’ (1951: 498), cannot be rationalized like ‘belief systems,’ so that there is always the chance of plural and thus unpredictable paths of development (500). This means that, according to Parsons, right now sociology has no chance to make reliable predictions on the overall direction of social and historical change, as this change is always dependent on more or less contingent constellations of various actors.

All these insights were in contrast to the kind of linear historical thinking so prominent in Levy’s historicization of the pattern variables, which shaped the very roots of modernization theory. And this neglect of Parsonian insights, and the similar neglect of the complexity of Weber’s and Tönnies’s thought for example, haunted modernization theory from the beginning and was the decisive reason why this theory began a very fast process of internal revision which led to the dissolution of the theory at the end of the 1960s. The talk of ‘modernization theory’ as if it is one single and stable theory is therefore as misleading as the claim that there were quite different versions of modernization theory. The last statement is certainly true, but it is more than that. The emergence of different versions was in most cases not dependent on certain idiosyncratic tastes of particular researchers, it was more the result of a learning process within modernization theory, even if this learning didn’t lead to a happy end. Modernization theory had an extremely interesting early history, and this history has to be understood if one wants to judge the ‘death’ of the theory and its rebirth in the late 1980s and early 1990s (see Knöbl, 2001).

A Short History of Modernization Theory in the 1950S and 1960S: Problem-Solving Without a Happy End

The first major empirical studies within the modernization paradigm tried to verify Levy’s theoretical ideas concerning social change. As it turned out very quickly, demonstrating the alleged dichotomy between preindustrial/traditional societies and modern ones was much easier than pointing to concrete processes which led to the coming of modern society: the search for typical structures of modern societies, for psychological characteristics of modern man/woman, seemed easier than looking for the causes of industrialization. This can be seen in the work of, for example, Daniel Lerner and David McClelland.

Lerner was one of the first authors who made the term ‘modernization’ central in analysing non-industrialized societies. However, in contrast to the title and subtitle of his famous book The Passing of Traditional Society: Modernizing the Middle East (1965 [1958]), his major arguments didn’t focus on dynamic processes. He was more concerned with demonstrating the fundamental cultural and psychological differences between members of traditional and modern societies, arguing that one of the preconditions for modernization is a kind of psychological mobility of the population that he called ‘empathy.’ Empathy he defined as the ability of people to act and think along abstract criteria so that they are able to leave behind the somewhat narrow personal and familial horizon so typical of traditional societies, to think of themselves as active members of a modern and mobile society.

Traditional society is nonparticipant—it deploys people by kinship into communities isolated from each other and from a center; without an urban-rural division of labor, it develops few needs requiring economic independence; lacking the bonds of interdependence, people’s horizons are limited by locale and their decisions involve only other known people in known situations. Hence, there is no need for a transpersonal common doctrine formulated in terms of shared secondary symbols—a national ‘ideology’ which enables persons unknown to each other to engage in political controversy or achieve ‘consensus’ by comparing their opinions. (Lerner, 1965 [1958]: 50)

Working extensively on the psychological shape of modern (and traditional) individuals, Lerner’s answer to the question how this psychological mobility comes into being, or will be increased, was rather short and simple. He just pointed to the influence of the mass media, which would distribute new and modern role models and which would spread this ‘empathy.’ However, this hint to the media, so often heard within modernization theory even in later times, was not very convincing because it naïvely assumed simple and one-directional effects of the media, which supposedly easily change enduring attitudes, capabilities and predispositions. That assumption was scarcely in accordance with the insights of contemporary media experts (not even those sympathetic to modernization theory), who argued that, although the media indeed might shape tastes, images or foci of attention, a change of deep attitudes is a much more complex process depending on stable relationships of interaction, on the influence of peer groups, on influential figures within a community, and so on (Sola Pool, 1963). The other problem with Lerner’s reference to the media was that he didn’t discuss the causes of the increasing use of mass media; thus one could doubt whether the spread of the media really is a valid variable as it might be too dependent on other variables, for example economic factors. If that doubt is justified, then Lerner’s explanatory model looks rather tautological as he seems to explain economic growth and industrialization by alluding in fact to economic growth.

David McClelland’s theoretical ideas in The Achieving Society (1961) were not too different from those in Lerner’s model. Psychologist McClelland presented a huge amount of comparative and historical data trying to demonstrate that the will for ‘achievement’ is one of the most important character traits of modern man/woman and the decisive precondition of economic activity and thus economic growth. But he was quite vague about the causes for the development of such psychological traits. Certainly, he argued that education will increase an achieving attitude. But he never really asked who in underdeveloped regions in fact will push forward educational measures and what will happen if there are no major actors interested in such kind of reforms. And—even more important—he never analysed the concrete effects of the supposedly modern values and attitudes if they really emerge within a formerly traditional society. What are the societal consequences of those values if it is also true that attitudes and values are often difficult to translate into new political and social institutions? The existence of modern values might indeed be a necessary precondition for modern institutions, but it is certainly not a sufficient one as values have to be institutionalized by concrete actors.

Thus the theoretical debate already at the very beginning of the history of modernization theory clearly showed that it is not enough to establish typologies between modern and traditional society, modern and traditional man/woman. To paint a dynamic picture of modernizing processes and—above all—to explain them, it was necessary to focus much more intensively on concrete actors than Lerner and McClelland were willing to do in their research designs. ‘Who wants modernization?’—that became the decisive question for those who saw the deficit of a purely typological approach.

Focusing on certain groups within a population should, of course, allow explanations why some countries modernized faster than others. But such a research design was far more complex than the one Levy, Lerner and McClelland had used because researchers now had to identify concrete constellations of actors which make possible or might accelerate the transition between traditional and modern societies. Who were the decisive actors? Strangely enough, most modernization theorists more or less neglected the majority of the population in underdeveloped or developing countries: studies of the rural masses and/or peasants were as rare as thorough investigations of the very different forms of land-ownership all over the world. It seemed as if modernization theorists didn’t regard these rural classes or groups (and questions related to them) as very important for the transformation process: as it was supposed that urbanization will quickly and completely change the demographic and professional structure of a population, as, thus, rural people were a group more or less doomed to disappear, most researchers didn’t problematize the neglect of these research questions (for an exception, see Stinchcombe, 1966 [1961]: 496).

If it is not the rural population who will be the decisive group for the modernization process, who else remains to be focused on? Modernization theorists usually analysed two groups, elites and the middle classes.

Already in the mid-1950s, the economic historian Walt Rostow had argued that economic progress in the past has always been dependent on a certain constellation of actors which allows the emergence of an elite group willing to invest huge amounts of capital:

What appears to be required for the emergence of such ‘lites’ is not merely an appropriate value system but two further conditions: first, the new ‘lite must feel itself denied the conventional routes to prestige and power by the traditional less acquisitive society of which it is a part’; second, the traditional society must be sufficiently flexible (or weak) to permit its members to seek material advance (or political power) as a route upwards alternative to conformity. (1966 [1956]: 249)

Thus, Rostow argued that industrialization is not an automatic process but one dependent on constellations of powerful groups within a society (see the similar argument in Hoselitz, 1966 [1957]). Although this was quite convincing, it didn’t solve the most important problems: even if one is able to find certain actors in a given traditional or transitional society with a strong will to modernize, it was still not clear which kind of actions these actors will pursue. Will they push forward a modernization process similar to the one in the democratic West or one along the lines of the Soviet model? This was a scientifically and of course politically decisive question, and it also means that Rostow’s allusion to constellations of actors was still not specific enough. Therefore it was reasonable to focus more strongly than before onpolitical elites, on intellectuals, as these are the groups who usually shape the political outlook of modernization processes. However, as it turned out, the actions of political elites in many underdeveloped countries, and especially those of intellectuals, were also difficult to predict; their behaviour was often quite erratic, so that it was impossible to answer definitively those questions which were left open by Rostow. The preconditions for a capitalist or socialist industrialization were still not clear.

One way out of this dilemma was a shift of attention to other groups the behaviour of which might be more predictable. This was one of the steps made by Edward A. Shils, who argued that the centre of a society has to be made up by actors and opinion leaders who might counter the irrational impulses of the masses and those of (oftentimes left-wing) ideologists. He argued that especially those groups with explicitly rational value standards a professional middle class, that is, scientists, engineers, and so on—might be able to push forward a persistent process of modernization as their rational values might—in the long run—diffuse throughout the whole population, thus guaranteeing a smooth transition to an institutional model shaped along Western lines:

Indirectly, the building up of the professional community or of professional sub-communities, will contribute to the civility by creating alternative modern objects of attachment, and will thus provide an alternative to hyper-politicization, which the demagogue practices and which he wrongly regards as the ideal. It will contribute to the civil culture through providing a field of sober, realistic, and responsible judgement. (Shils, 1963: 76-7])

The problem even within this kind of argument, however, was that the role of the middle classes within developing countries was always a precarious one as the growth of these classes was dependent on economic growth. But what happens if, as actually happened between 1950 and 1966 (Packenham, 1973: 9), the annual growth rate declined and the developing countries thus fell further behind the West? In this case, the debate about the role of the middle classes was in fact an academic one, because there were in most countries not enough members of a middle class to be regarded as a critical mass for modernization processes along the Western path.

All in all, the debate about constellations of actors—elites or (professional) middle classes turned out to be not very fruitful. In the end even the question arose whether modernization—against all assumptions—might possibly not be a uniform process at all, but one with very different political results. Was it possible to think about different modernities?

This was an increasingly important question in the 1960s. It was Joseph LaPalombara and again Edward A. Shils who most clearly saw the consequences for modernization theory if this question received a positive answer. Although Shils was still optimistic with regard to the establishment of more or less democratic regimes in non-Western regions of the world, he had to admit that these democratic structures would be different from Western ones: Shils called them ‘tutelary democracies,’ a specific form of government with strong executive institutions and with a really dominant position accorded to political elites, a form of government ‘which recommends itself to the elites of the new states. It does so because it is more authoritative than political democracy, and also because the institutions of public opinion and the civil order do not seem qualified to carry the burden which political democracy would impose on them’ (Shils, 1966 [1962]: 466). This kind of argument touched very dangerously one of the central hypotheses of modernization theory, a hypothesis already formulated by Marion J. Levy with his insistence on the functional interdependence of economic growth and democratization. Shils, as has been shown, talked about new forms of democracies, but he was still optimistic that at least in the long run this functional interdependence would show its relevance. But there were others, like Joseph LaPalombara, who seemed willing to abandon this assumption, who didn’t believe in this interdependence any longer. LaPalombara (1963) made quite clear that in the theoretical debate on modernization it has become impossible to find a consensual definition of political development: everybody seems to define this term according to idiosyncratic tastes and there is always—so LaPalombara argued—the danger that the Western or US institutional system will be interpreted as some sort of a telos of history.

‘It is precisely at this point that it becomes apparent that comparative analysis is not facilitated by a definition of political ‘modernity’ that is culture-bound and narrowly restrictive through its assumption of the evolutionary inevitability of the Anglo-American model. Only a rigid Wilsonian faith in the inevitability of democracy would justify a retention of a parochial and deterministic definition in the face of the historical and contemporary evidence that surrounds us. (1963: 38)’

Thus LaPalombara pleaded for more open models of political change, arguing that the New States will develop quite new mixtures of institutions unknown to the West. This, of course, implied that he had foregone the idea of a functional interdependence of economic and political processes. Other authors within modernization theory didn’t make such radical conclusions. But it could nevertheless be seen that even they—Shmuel Eisenstadt (1963: 96) or Lucian Pye (1966 [1965]: 90), for example—defined the concept of democracy in more and more abstract terms, thus indicating that economic progress will notnecessarily lead to Western-style political systems.

It became increasingly clear that central hypotheses, assumptions and concepts of early modernization theory à la Levy were mistaken, or diffuse—and that was even true with regard to the term ‘tradition.’ And if the interpretation of ‘tradition’ became increasingly difficult, what about the important dichotomy between ‘tradition’ and modernity? What, then, is ‘tradition’ if it is not the counterpart of ‘modernity?’

These questions, too, began to be asked, even if implicitly, at quite an early phase of the development of modernization theory. This could be seen in the work of Robert Bellah on Japan as he argued that Japanese modernization was based on strong particularistic ties of all societal elites to the family of the emperor: ‘The continuity of the imperial line and of the national religion served to symbolize an almost ‘primitive’ particularism. The high evaluation of military achievements and the fulfillment of one’s lords commands became generalized beyond the warrior class into a high level valuation of performance in all spheres’ (Bellah, 1985 [1957]: 183). Thus a clear division of the pattern variables as suggested by Marion Levy was definitely missing in Japan: particularistic values were and are not just characteristic features of traditional societies, so that the assumption of modernization as a unidirectional, linear, process was wrong. Modernization—as Bellah (1958: 5) argued-doesn’t lead to a clear dominance of rational and secular values.

Bellah didn’t elaborate on the implications of such a statement for modernization theory, but other authors did: Bert Hoselitz, for example, began to realize that the term ‘tradition,’ or even Max Weber’s term of ‘traditional action,’ was heavily under-theorized, often encompassing various and highly different forms of actions ‘ranging from the purely automatic,’ often not meaningfully oriented, behavior to a highly self-conscious behavior whose underlying principles are reflected upon and often highly ‘rationalized…’ (1961: 85). Hoselitz tried to clarify this and differentiated between the terms ‘habit,’ ‘usage,’ ‘norm’ and ‘ideology.’ This was certainly helpful, but provoked the problem that the former distinction between tradition and modernity began to disappear as habits, norms, and so on, certainly are not alien elements within modernity. Hoselitz was aware of this, admitting that traditions might accompany modernization for quite a long time. And he was certainly not the only one who felt uneasiness regarding this simple dichotomy between tradition and modernity as suggested, for example, by Levy: especially researchers interested in the political consequences of modernization realized that there is no smooth path towards a rational society within underdeveloped regions. David Apter (1963) increasingly theorized the emergence of political religions like Marxism, which often served as modernizing ideologies and in which rational elements were hopelessly interwoven with irrational ones. And Clifford Geertz even radicalized such a position: whereas Apter still believed that at some time the rational elements within modernizing ideologies might overcome the irrational ones in the process of modernization, Geertz was more pessimistic as he could demonstrate that it was only by modernization that non-rational belief-systems like nationalism came into being. Nation-building in New States doesn’t destroy ethnocentric feelings, it creates them or at least modernizes them (Geertz, 1963: 154). Primordial and civic feeling are not in an evolutionary opposition towards each other: ethnic differences are—as Geertz argued—not traditional remnants of past periods, but they will precariously interact with other loyalties and ties—not necessarily getting weaker.

If it turned out that the central dichotomy between tradition and modernity was not a very meaningful one, the question arose as to how to proceed without abandoning the idea of modernization. One of the solutions was pushed forward within the circle of theorists around Talcott Parsons. It was Neil Smelser who very early saw all the difficulties inherent in the term ‘tradition’ and thus began his famous book on Social Change in the Industrial Revolution (1960 [1959]) with the following two sentences: ‘When comparing’ a society with its past or with another society, we often employ a dichotomy such as ‘advanced vs. backward,’ ‘developed vs. underdeveloped,’ ‘civilized vs. uncivilized,’ or ‘complex vs. simple.’ Sometimes these words yield too little information, because they claim simply that one society is superior to another’ (1960 [1959]: 1). The concept of ‘differentiation’—so Smelser suggested—is a better tool to deal with processes of modernization as it avoids all value judgements and especially this precarious dichotomy between tradition and modernity: differentiation, as ‘the evolution from a multi-functional role structure to several more specialized structures’ (Smelser, 1964: 271), is a concept which allows us to think of smooth transitions and gradual differences, in contrast to the aforementioned dichotomy of early modernization theory.

Smelser’s concept of differentiation was important for functionalist theorizing in so far as it allowed the countering of criticisms brought forward by conflict theorists like Lewis Coser arguing that functionalism is unable to analyse social change and prefers to deal with static structures. By working with the model of differentiation, this reproach seemed no longer to have any substance and even more important—using this model also seemed to allow the incorporation of conflicts into functionalist thought: for Smelser, conflicts and social movements were decisive factors which would disturb old societal structures and allow the emergence of situations in which new social structures, structures of a higher level of differentiation, could be built.

However, there was one major problem with this concept: most of the early versions of modernization theory at least tried to find causal mechanisms which would lead to the take-off of a society; they at least tried although without much success—to identify constellations of actors which allow a society to change its old structures. Differentiation theorists no longer did so—as if they had realized the futility of such attempts. Authors working with the differentiation concept almost automatically abandoned the search for causal statements with regard to ‘differentiation’ as ‘differentiation’ is in fact a post-hoc description of results of social processes, and not an explanation. Smelser saw the problems with such a theoretical move: he sometimes calls differentiation theory an explanation, but sets ‘explanation’ in inverted commas, or he defines ‘structural differentiation’ as a ‘scheme… certainly… not meant to encompass all other possible explanations’ (1960 [1959]: 384). Thus, in the end, the differentiation concept as even Smelser seemed to have felt—didn’t help very much in concrete empirical analysis of macro-social change: the concept just tells us pretty vaguely that—in the long run and by whatever causes—the disintegration of systems is followed by systems characterized by a higher level of differentiation.

A still more radical separation between a causal explanation and a pure description of the results of social processes can be seen in the evolutionary turn of modernization theory which was begun by Parsons (1964) and Bellah (1964) around the mid-1960s. Although these interpretations of the development of human societies brought forward many interesting insights and caused many sociologists to take history more seriously than before (see the chapter by Holmwood and O’Malley in the present volume), this theoretical move meant in the end an almost exclusive analytical focus on the logic of social developments and a parallel neglect of real historical and social processes driven forward by certain groups and actors. One could argue that the theoretical debate within the modernization paradigm on the relationship between agency and structure was brought to an end by a somewhat arbitrary decision: as it was not possible to identify the carriers of modernization processes, influential theorists began to disband the action theoretical ground of the theory altogether and instead focused their research on somewhat static descriptions of sequences of social structures. The results were often highly abstract theoretical and historical schemes not very helpful for those who were really interested in the causes and preconditions of social change, in the causes and precondition of a process called ‘modernization.’ Thus, the original intent of early modernization theorists to offer and further develop an empirically fruitful theory of social change was certainly not fulfilled by those who stepped too deep into the field of evolutionary or evolutionist theorizing.

Regarding all these theoretical shifts within modernization theory in the 1950s and the first half of the 1960s, it became obvious that the deficits and aporetical problems of the original formulation of the theory as originally formulated by Marion Levy could not be eliminated. Although the theoretical debate undoubtedly showed some progress, many of the foundations of the theory turned out to be flawed or at least highly implausible. And this was recognized even by those who were at the beginning of the theoretical debate quite sympathetic towards the modernization paradigm.

Contrary to the interpretation of Jeffrey Alexander, who claimed that modernization theory died ‘sometime in the later 1960s… because the emerging younger generation of intellectuals could not believe it was true’ (1994: 175), and contrary to his allegation that it was a different ideological (Marxist) Zeitgeist which brought modernization theory into disrepute, it was the internal debate within modernization theory or at its margins which led to massive criticisms and to its ‘death’ in the late 1960s. It was especially since 1967 that more and more theorists and researchers argued and demonstrated that tradition is not at all the opposite of modernity (Gusfield, 1966-7; Levine, 1968; Singer, 1971), that the idea of a stable, homogeneous, inflexible ‘traditional’ culture is wrong (Lauer, 1971; Whitaker, 1967), that modernization means completely different things to different countries as the variable temporal starting points of national modernization processes have to be seen in an international context (Bendix, 1967; Huntington, 1971; Rudolph and Rudolph, 1967), and that the functionalist assumption of an ‘eurythmic’ change of societal subsystems is a myth which nourishes, for example, the mistaken belief that a market society will automatically bring forward a democratic parliamentary system (Rüschemeyer, 1970; Whitaker, 1967). At that time the theory was really shattered and—to stress this again—not only because Marxism in its different guises became more and more attractive to younger social scientists. The theory was no longer convincing, and it seemed even difficult to imagine a future in which modernization theory could again play a dominant role within macro-sociological theorizing. But in the mid-1980s such a new future for modernization theory emerged.

Old Wine in New Bottles: The Revival of Modernization Theory Since the 1980s

The reason for such a revival had above all to do with the economic rise of the so-called Asian tiger-states’ in the late 1970s and early 1980s and the slow decline of the Soviet Empire. These political processes seemed to confirm a posteriori modernization theory’s original assumption that—in contrast to the claims of dependencia theory or world-systems theory there are indeed chances of sustainable development for non-Western nations, and that the very stability of Western modernity with its particular institutional system seemed really to be some sort of a telos of history: the fall of the Soviet Union could be interpreted as an indication that there is one, and only one, route into modernity, that the Soviet model failed because of its insufficient structural differentiation—a point always stressed by modernization theorists. Modernization theory—so Edward Tiryakian argued in 1991 was verified by history itself; the theory could be revived and start a new life. And indeed, American authors like Jeffrey Alexander and Paul Colomy (1990) tried to reintroduce into sociological theorizing a modified concept of differentiation which takes into consideration the actions of groups and social movements and their oftentimes contingent results; and Europeans like Ulrich Beck and Anthony Giddens (see Beck/et al., 1994) were heading towards a new description of modernization processes they called ‘reflexive modernization,’ thus indicating that the process of modernization is not a simple story of progress but one in which many risks can be found.

The question, however, was whether these new modernization theories really were much better than the old paradigm, whether the theoretical debate in the 1980s and 1990s really got much further ahead than the one in the 1960s. There is some doubt about that. As Alexander (1996), in criticism of the writings of Beck and Giddens, has rightly remarked: the old and highly problematic dichotomy between tradition and modernity is again lurking behind their new approach, which makes the theory as vulnerable as the one Levy had proposed in 1952. Furthermore Alexander and Colomy were themselves critically asked whether their proposal for a new concept of differentiation really amounts to a new theory, as they admit that differentiation is so much dependent on actors, constellations and contingent actions that the explanatory value of such a theory must necessarily be quite weak (Joas, 1996 [1992]: 230). Thus, if one summarizes the history of modernization theory in the 1950s and 1960s, one quickly realizes that most of the arguments now being pushed forward to formulate a new concept of differentiation and to build a new theory of modernization have already been tried in the past. The chances that a solid and convincing new theory of modernization will arise are therefore quite small. Nevertheless, modernization theory is still alive, and one might even dare to predict that the theory will live on for quite a long time. Why is that the case? Why will the theory not ‘die’ despite all its weaknesses and failures?

Dean C. Tipps answered this question as early as 1973. He argued that the exact meaning of the term ‘modernization’ is unclear and highly contested, which paradoxically is one of the reasons why modernization theories will proliferate endlessly despite all criticisms: ‘Every theory’ of modernization attacked and destroyed will only raise two in its place’ (Tipps, 1973: 217). That endurance of the theory is based on the fact that the term ‘modernization,’ as well as the related concept of ‘modernity,’ has a strange kind of (normative) attraction for all those—politicians or intellectuals—debating the contours of contemporary and future societies. Even the mostly philosophically inspired talk about postmodernity (for an overview, see Yack, 1997) hasn’t changed the situation, since modernity is the nation’s final goal and justification. Histories in which nations measure themselves against one another and vie for advantage fall easily into narrative patterns that imply linearity and convergence’ (Cullather, 2000: 646).

So what is the conclusion of all this? Although modernization theory will not disappear in the social sciences in the very near future, one has to realize that there is no stable, empirically grounded theory, no theory at least with strong explanatory claims. All there is is some sort of modernization discourse, some vague ideas about possible developmental paths of contemporary societies. These ideas—Göran Therborn (2000: 69) argues have an inspiring potential for those interested in macro-sociological change if approached sceptically. But they should certainly not be taken as theoretical premises for practitioners of historical sociology.