Historical
The ancestors of the Romani-speaking peoples left India some one thousand years ago, moving along trade routes trodden over the centuries by countless other migratory nations. Some two or three hundred years later, contemporary documents attest their arrival in eastern Europe; before the end of the fifteenth century their presence is recorded in the British Isles.1 They brought with them a language whose Indian construction was in the eighteenth century to betray their history to the learned world; but with the Indian base came loan words from every country on their path; and hybridisation and creolisation with other languages has fragmented the Romani language into hundreds of dialects. Today, like the Jews, they live throughout the world, sometimes intermarrying, sometimes not, disunited politically, heterogeneous culturally, and with the most diverse aspirations.
Values and methods; the collection of data
Although I have probably spent more time in ‘participant observation’ than in reading about Gypsies, I have had the benefit of a vast literature on the subject, including several anthropologically respec table participant-observation studies, and even one or two attempts, on non-British Gypsies, to use questionnaire surveys. The differences between the conclusions of these studies serve as a valuable reminder of the heterogeneity of Gypsy culture. The great fault of the literature on Gypsies, both official and academic, is overgeneralisation; observers have too easily been led to believe that the practices of particular groups are universal, with the concomitant suggestion that any group not having such a practice are not ‘true’ Gypsies.
I felt, therefore, that the great need was not for another detailed study of some small group of South Essex Gypsies, with whom I did my most intensive field-work, but for synthetic studies, drawing together all the data on England and Wales as a whole, and comparing them, where appropriate, with the situation in other countries of Gypsies or other cultural groups. Moreover, since time has not lessened the fury of controversy over many of the matters with which I am dealing, I have preferred to rest my argument as much as possible on documentation rather than merely on personal observations.
But although the 100,000 words of my field-log are quoted comparatively rarely here, my field-work was still of paramount importance in my research; for it gave me critical insight into the literature. I do believe, with Goffman, that
any group of persons … develop a life of their own that becomes meaningful, reasonable, and normal, once you get close to it, and that a good way to learn about any of these worlds is to submit oneself in the company of the members to the daily round of petty contingencies to which they are subject. (Goffman, 1961, Preface)
Equally, I believe, however, that this approach cannot stand on its own; the analytical confusion of one of the classic early Chicago participant-observation studies, Nels Anderson’s The Hobo (1923), can stand as a warning of the dangers of shortsightedness in the participant-observer.1
part one Theoretical perspectives
The complaint that there is no general theory of the causal processes of race relations, or any stock of widely confirmed empirical generalisations, is a commonplace. Equally constant are the complaints about the number of books, all with their own typologies and other theoretical devices personal to the authors, which have to be considered before one can gain a general picture, even at an elementary level, of how race relations theorising has been done. Not that many people consciously set themselves to survey the whole vast literature, but those who do resemble each other in this if little else. The latest complaint to come to my notice is that of Herbert H. Hyman (1969) in Katz and Gurin’s collection of essays on possible research programmes, who comments: ‘Alas one becomes weary at the mere thought of reading so much, let alone understanding i t …. Confusion falls upon weariness as one begins to read the empirical studies.’ Over a decade earlier, with a comic disdain that would probably prove unacceptable in scholarly writing today, Professor M. B. King (1956) wrote similarly of a pile of textbooks which had been sent to him unsolicited:
Nowhere in the ‘minority’ literature can one find a systematic and comprehensive set of sociological generalisations. This lack is not so culpable, for such a set does not now exist. What surprises me is that many authors do not seem to miss it.
Leaving aside concern for sects, homosexuals and cripples, King divided the books into three categories. The first concentrated on racial, ethnic, inter-group, inter-cultural, or even human relations. The second consisted of lists of groups and their contribution to the ‘American way of life’. The third type ‘explained everything in terms of psychology’, ignoring the possibility that real conflicts of interest and culture-clashes as well as ‘prejudice’, might have to be resolved; it is regrettable that since the experimental tradition in the study of the social psychology of prejudice has produced more concrete results than the sociological work on the situations in which prejudice is found, it is still the case that many social psychologists tend towards a psychological reductionism. A recent book by P. I. Rose (1968), on the teaching of race relations and related subjects in America, documented at great length an even greater variety of title, content, emphasis and conclusions of teaching and research in the field. British students are even worse off: in addition to reading the American literature, they also have to read an equally diverse English literature. I have not space here to survey all of this literature, although I shall comment on some of the surveys that have been made; my interest will be to speculate about the theoretical confusion itself. I wish to examine why people are continually proposing and using such differing methods, which nearly always fail to yield generalisations which can in any sense be divorced from particular...