Lesnotions de frustration relative (relative deprivation) et de groupe de référence constituent a elles deux un ensemble conceptuel dont le succès est dû sans doute à ce qu'il permet, à l'aide de propositions de bon sens, de rendre compte d'observations paradoxales. Comme le note Runciman, les deux notions sont issues d'un truisme familier: à savoir que «les attitudes, aspirations et frustrations, dependent largement du cadre de référence dans lequel elles sont conçues » (“ that people's attitudes, aspirations and grievances largely depend on the frame of reference within which they are conceived”) (I). L'observation montre en effet qu'il est dans la plupart des cas impossible de comprendre pourquoi un individu A éprouve par exemple de l'envie à l'égard de B mais non de C si on ne connaît pas le « cadre de référence » (“ the frame of reference”) de A: comment expliquer que A envie la 504 de son voisin de droite mais non la Jaguar de sonvoisin de gauche, si on ne sait pas que A peut espérer se hausser au niveau de son voisin de droite mais non a celui de son voisin de gauche ? Utilisons l'excellente définition que propose Runciman de la notion de frustration relative: éfinition stricte est difficile. Mais on peut grossièrement dire que A est relativement frustré de x si (I) il n'a pas x, (II) il voit une personne ou plusieurs autres personnes, incluant eventuellement lui-même dans le passé ou dansl'avenir, comme ayant x (quece soit, ou doive ou non être en fait le cas), (III) il désire x, et (iv) perçoit comme plausible l'éventualité d'en disposer. Dire qu'il possède x peut naturellement vouloir dire qu'il est exempt ou indemne de y »(2).

LaComposition sociale du corps des officiers des armées européennes a longtemps constitué un thème classique de recherche et de réflexion pour les historiens et les sociologues intéressés par l'étude des institutions militaires. Les œuvres d'Alexis de Tocqueville ou de Gaetano Mosca, les travaux de C. Wright Mills, d'Alfred Vagts ou de Stanislas Andreski, pour ne nommer qu'eux, en fournissent la preuve. La recherche d'une explication à l'exceptionnel apolitisme de l'élite militaire européenne au cours du xixe siècle et du début du xxe est sans aucun doute à l'origine d'un tel intérêt. Ainsi, l'hypothese la plus communément avancée stipulait que l'origine sociale des officiers était probablement l'une des variables déterminantes de leur comportement politique. Par exemple, en soulignant dans l'Angleterre du xixe siècle l'intime corrélation entre la similarité du recrutement social dans la classe politique et la classe militaire d'une part et la réticence de cette dernière à s'immiscer dans les affaires politiques de l'Éitat d'autre part, Gaetano Mosca admettait implicitement l'existence d'une relation causale étroite entre les deux paramètres (1).

Wirkönnen heute in vielen Entwicklungsländern eine paradoxe Situation beobachten (i). Wir finden dort große ökonomische Ungleichheiten vor, Ungleichheiten, die größer sind als bei uns. In einer solchen Situation sollte man erwarten, daß sich politische Organisationen auf einer Klassenbasis, d.h. auf horizontaler Ebene herausbilden. Statt dessen treffen wir aber häufig vertikale Beziehungen an, die verschiedene Klassen durchschneiden. In einem Parteiensystem stehen sich dann z.B. nicht zwei Parteien gegenüber, die verschiedene ökonomische Interessen vertreten, sondern beide Parteien weisen eine ähnliche Mitglieder- und Wählerstruktur auf. Wir können diese Parteisysteme besser verstehen, wenn wir die zugrundeliegenden Klientelstrukturen aufdecken.

TheMouride brotherhood has by now become the subject of an abundant literature, sociological and other, since its foundation in the 1880s (notably Paul Marty 1913, L. Nekkach 1952, V. Monteil 1962, Shaikh Tidiane Sy 1969, J. Copans et al. 1972, F. Dumont 1975, Donal Cruise O'Brien 1971, 1975) (1). Looking over this literature today, one is impressed by the remarkable resilience and versatility of this holy organization: charismatic community and vehicle of a variant of ‘sacred nationalism’ (1886-), instrument of a massive land settlement (1912-), and of electoral brokerage in national party politics (1951–1966). This is only to mention some of the principal functions which the observers have discerned over the years, leaving aside the theological-didactic mission which the brotherhood's members (both leaders and followers) tend above all to emphasize. Yet for this writer, on the basis of field research in 1966–67, the Mouride success story seemed to be near its close (Marty more than half a century before had reached a similar verdict): but now, after renewed fieldwork in 1975 (2), it would appear that the brotherhood (with a new, younger and more vigorous leader) has made a new and unforeseen departure which invalidates recent premonitions of its imminent demise. Writing in 1970, I had been careful in concluding a book on the subject to cover every apparently foreseeable future trend but had not foreseen the major subsequent development, that the brotherhood would emerge as a (curious and perhaps slightly ambivalent) form of peasants' trade union. Experience showed that the great holy men (shuyukh, marabouts, saints) could be successful estate managers, political brokers, even capitalists, but union bosses!—that last can be explained now, but to me at least the causes have become obvious only after the event.

TheComplex ranking systems of South Asia and Polynesia offer a special challenge and a special trap to the anthropologist or historian attempting to explain their development. He will inevitably trace the political and economic development of the state and link those to the appearance of new types of power groups and new structures of stratification. The danger lies in identifying and correlating directly the actors' concept of stratification, the distinctions and associations made in the emic system with interest groups identified by analyses of the political or economic structures. This may be done consciously or by implication by translating actors' category terms by words such as ‘nobility’, ‘ruling class’, ‘bourgeoisie’. This is likely to lead to gross ethnocentric assumptions and obscure the specific nature of the ethnographic example. The danger of such an approach has been demonstrated well by Dumont and Pocock for the most complex and the most famous of such hierarchical systems. They point out how the Indian caste system does not fit either in structural form or in ideological content models borrowed from feudal or class systems. Indeed such comparisons obscure the essentially religious aspect of Hindu ideology contained in the scheme (Dumont and Pocock 1958, Dumont 1966). Similar criticisms would also be pertinent to the way the hierarchical systems of Central Madagascar have been analyzed and for very similar reasons.

Urban ethnographers can explore spaces forgotten or given up for lost, replacing illusions with maps of social reality. The city is filled with milieux that remain scarcely known to the inhabitants of other spaces. Paradoxically, as the rationalization of urban life continues, boundaries enclose some lives more tightly, isolating and making them more vulnerable. By re-discovering the lives of people in those spaces, by replacing stereotypes about them with descriptions that convey their vitality, dignity and humanity, ethnographers may restore some lost relationships in urban milieux, and in a modest way reduce the isolation.

Britain can avoid the fate of a cold Haiti only if it succeeds in retaining (or rather regaining) its place in the forefront of inventiveness and technical and organizational skills. We must remember (what every business man knows) that even a small initial lag in relation to the competitors can set a snowballing process which can end in total bankruptcy. Faced with the need to make a great effort and tighten the belts, the country can remain a liberal democracy only if it has a capable and public-spirited elite imbued with the ethics of work. To supply this elite—to maintain the standards where they are high and to raise them where they are low—ought to be the main task of the universities. This means that—apart from the moral education of teaching the young to work conscientiously—everything should be subordinated to the function of advancing and disseminating knowledge. Although this is plain common sense, one can hardly find a reference to it in the pronouncements of the educationalists, sociologists, journalists and politicians who seem to view the universities as agencies whose main functions are to help the handicapped, transfer the largest possible number of sons and daughters of useful workers into parasitic bureaucracy, or to conceal unemployment. Some of the academics on the other hand regard them chiefly as dining clubs, many of the students as discotheques, while some of the administrators treat them as if their purpose was to open a field for an endless proliferation of paper shufflers.