The Molokan sect, one of the strongest and most influential in prerevolutionary Russia, is today moving towards extinction. This process, however, has not gone equally far in different geographical areas of the Soviet Union. This paper puts forward an assessment of the extent of the sect's decline and attempts to explain both the general decline and its differential rate in different communities. It analyses the sect's development over a long period of time and under radically different socio-political conditions and it tries to systematize the Molokans' changing response to their different social environments, locating the sect in the typology devised by Bryan Wilson. It concludes that the severe decline in membership during the Soviet period is due to the fact that the Molokan sect is no longer an organization through which political dissent can be expressed but has fully adjusted to and affirms the values of Communist society. In the paper it is pointed out that although the sect can be typified in Wilson's terms during its early period, its changed response in Soviet society is not anticipated in Wilson's scheme. Existence in the exceptional social environment of socialist society has called forth an exceptional sectarian response. To illuminate this process we will consider both the Molokans' historical development and their religious belief, as well as their social complexion today.

The purpose of this paper is to argue that reciprocity may be of two types: prestations of things and prestations of actions or, expressed in another way, gift exchanges and cooperation. Evidence from two societies will be used to show that each of these two types of reciprocity is appropriate to a different type of society or social situation.

In recent years historians and sociologists alike have given increasing attention to the relationship between education and social structure in France. Louis Althusser has argued that the French educational system was designed to preserve elites. Michel Crozier concluded that the system prevented a too rapid mobility that would upset the status-quo and that ‘graduates [of grandes é'coles] have a de facto monopoly on the top brackets of the Civil Service, the universities, and medicine, and decisive advantages in entering most professions and many industrial organizations’.

The study of festivals has only recently caught the attention of historians, who have traditionally shown themselves more concerned with the labours and afflictions of men than with their pleasures and diversions. If from this time onward, festivals become an object of historical enquiry, it will be perhaps because industrial society no longer has festivals—or at least that it has the festival only as pageant in which the passive community of onlookers has been substituted for the active community of celebrants—and because our interest in the festival increases to the extent that we lose it. However, we should also mention the historian's debt to the double stimuli of folklore and of ethnology. From folklorists and ethnologists historians have learned to concern themselves with the armature given to human experience by ritualization, even if anonymous, even if destitute of an explicit system of regulation or of a conscious cohesiveness.

In a recent essay in CSSH (16:1) Oriol Pi-Sunyer has joined in debate with our article ‘Modernization and Development…’ (CSSH 14:3). The thrust of his critique is that our model of social change in the Western Mediterranean leads us inexorably to lump Sicily and Catalonia together with respect to general social and economic developments since World War II. He argues in terms of modernization theory that Catalans long ago developed industry and have always shown a propensity to organize in corporate groups whenever the central government gives them the opportunity to do so. In short, our model is lacking in ‘time depth and political sensitivity’ (pp. 128–9) since it puts the Catalans in the same class as the Sicilians. We wish here to demonstrate that (1) in our article we discussed the differences between Sicily and Catalonia in much the same way that Pi-Sunyer did in his critique; (2) that, in fact, Sicily and Catalonia have undergone similar processes of social change since World War II, in ways that threaten the historical continuity of these societies, both formed over centuries; and (3) that Pi-Sunyer's misreading of our materials is due to his unfortunate confusion of modernization with development, which is precisely what we tried to get away from in our essay.

I suggest that the reader sufficiently interested in our debate on Catalonia consult the two original essays; in this reply to a reply, I will cover only so much of the old ground as is necessary for a rejoinder. ‘From Autonomous Development to Dependent Modernization …’ badly misrepresents some of my observations, essentially denies some of the statements made by Hansen and the Schneiders in their initial essay, injects new elements into the discussion, and is replete with ad hominem arguments. I will do my best to set the record straight.

In a recent essay in this journal Dean C. Tipps subjected the tradition/ modernity concept to a thorough ideological and empirical critique and concluded that the time has come to get rid of it. Although his ideological and empirical analyses were convincing, the methodological remarks at the end of the essay were incomplete. If the contention that we should no longer attempt to revise the tradition/modernity concept is to stick, compelling methodological grounds must be given since ideological bias and empirical inadequacy can in principle be corrected. I want to show that in so far as the tradition/modernity concept is an ideal type, efforts to make it better ‘fit’ the spread of empirical data are in principle wrong. Moreover, efforts to transform the concept into an empirical ranking device or a classificatory system are open to equally serious objections. In order to do this I will first clarify the nature and function of ideal types, empirical types and classification systems.