The subject of Sally Engle Merry's provocative exercise in "urban anthropology" is not so much crime as fear. The distinction, as Merry defines it, is central: crime is an act of violation against person or property; fear is the apprehension that such violation will take place.

It sounds obvious, but it really isn't. Crime and fear -- "danger" is Merry's term for the latter -- do not necessarily go hand in hand. It is possible for a community to be swept by fear yet to experience relatively little actual crime. It is equally possible for a community to be swept by crime, yet to accept that as normal and therefore to experience relatively little actual fear.

The implication of Merry's analysis, though she does not come right out and say as much, is that a "crime wave" such as the country is now supposedly undergoing may be in substantial measure a figment of our imagination -- of our fears. She writes:

"This book proposes the thesis that crime serves as an idiom for expressing and legitimating the fear of the strange and the unknown. Such fears often focus on populations that are racially, culturally and economically distinct. Members of a dominant group may denounce a subordinate group for its criminality rather than denounce it for the real threat it poses to the perpetuation of the existing social order and continued elite dominance of that order. Concern about crime thus justifies and reinforces hostility that stems from class conflict and racial and ethnic differences."

The laboratory in which Merry reached these conclusions is "Dover Square," her name for "a small housing project in an area undergoing extensive urban renewal"; the project, which she studied in 1974 and '76, is in a northeastern city that appears to be Boston. At the time of her study, the project's population was 42 percent Chinese, 27 percent black, 12 percent white and 6 percent Hispanic.

Obviously Dover Square is not a perfect model of an American community; the large Chinese population skews the demographics dramatically. But as a microcosm for the study of racial and ethnic tensions as manifested in a sense of "danger," it is more than sufficient.

Like so many American urban neighborhoods, Dover Square is a community of strangers -- people who are largely unknown to each other and therefore likely to fear each other. Some of these strangers are people who merely pass through. But most of them actually live there:

"This is the person who has lived next door or down the street as a neighbor for five or 10 years, but who is still not known as an individual with a name or personal history. These people share public spaces, walkways, stores and the laundromat with one another, yet remain anonymous. They continue to view one another only as members of social categories."

The Chinese fear the blacks and whites, believing them to be loud, violent and domineering. The blacks and whites fear the Chinese, believing them to be clannish, secretive, parsimonious and skilled in the martial arts. But these fears on all sides are grounded almost solely on rumor. Only the blacks and whites, who in Dover Square get along reasonably well, have sufficient actual acquaintance with each other to have a sense of individual personalities and behavior patterns; otherwise, the "danger" that members of the project feel is based on stereotypes and fantasies about "the dangerous foreigner."

Those fantasies and fears diminish greatly, Merry finds, when individual identities emerge. The Chinese and some whites are terrified by the "street youths," most of whom are black. Yet a young black woman told Merry: "To you, the boys are all dangerous because you can't tell them apart, but to me, they are George, Johnny, and Jamesy, and I know who to look out for and who will not bother me."

Bingo. Cut away all the anthropological jargon, and there you have it: What is known is understood and therefore less likely to be feared. It's true, of course, that a large percentage of violent crime (murder, rape, assault) is committed by and against people who know each other. But such crime is not Merry's subject; she is discussing "random, vicious, unwarranted attack by a stranger who belongs to a hostile group" -- what we mean when we say "crime in the streets."

Historically, Merry finds, "danger is an issue only in large, industrial, heterogeneous cities during periods of rapid growth, social change, and deepening class cleavages." What do you find under much of the rhetoric about the "crime wave"? You find white America's "fear of unfamiliar cultural groups, of people who are not committed to the established urbanites' life style or social order, who express their antagonism in part through crime" -- fear, that is, of blacks and Hispanics.

All of this would sound like the sentimental liberalism of the '60s except that it quite strongly appears to be true. The current apprehension over urban crime clearly has its roots in the early '60s, when nonviolent civil rights protests, and then the urban riots, raised white anxieties about interracial conflict. Whites, most of whom have little real experience with blacks (the reverse, of course, also being true), tend to fear crime by blacks in precisely the vague and stereotypical terms that Merry describes.

The real value of "Urban Danger" is that it requires the reader to think past the stereotypes, to see crime, and fear, in more objective and dispassionate terms. In no way does Sally Engle Merry minimize the actuality of crime or the individual and collective grief it exacts. But she makes a convincing case that there can be a wide gulf between what we fear and what actually takes place -- a gulf that often has a lamentable effect on the rhetoric by which we obscure and evade the issues.