MANY OF US are fond of referring to "American religion," as if that were a thing to be described. But anyone who has spent much time on the ground (or in churches) knows that there are lots of ways in which that term has to be modified, and one of the most significant modifiers is regional. I have lived in six of the eight regions identified in the Religion by Region series, and who I am religiously, and who people think I am, has varied enormously according to where I was living.

As a kid in Missouri, I was a Baptist, and that identity said not only that I belonged to an important church but that I was on the right side of the great eternal divide, ready to defend my salvation against the other contenders around me. When my family moved to Arizona, I joined the tail end of the white evangelizers who hoped to bring faith and education to the Native American and Mexican laborer population that surrounded us. In southern California, our next stop, people thought we were from Texas and just figured we were one more in a trail of exotic breeds that seemed to flourish on the Pacific shores. What they didn't know was that Southern Baptists and other evangelicals were becoming an institutional force to be reckoned with.

More than a decade later, when I was a pastor's wife on the other coast, the people who heard "southern" and "Baptist" seemed to assume that meant "snake handler" and closed the door as quickly as possible. Then, when we actually moved to the South, we experienced the cognitive dissonance of being assumed to be part of the irresistible evangelical mainstream while practicing a form of Baptist life that eventually got our church kicked out of the denomination.

Now I live in New England, and I'm an American Baptist--part of the "mainline," but we're anything but mainstream and powerful. People are polite about our religious identity, but no one assumes that our church will make the news. The pages of our local paper are largely reserved tar the doings of the Roman Catholic Church.

Who makes the news--and why and how--is part of the story this series seeks to tell. In part, it is an attempt to educate the public, including reporters, on the unique religious history and ecology of America's regions. With funding from the Lilly Endowment, the Leonard E. Greenberg Center for the Study of Religion in Public Life at Trinity College in Hartford undertook the massive task of assembling scholars and data and producing eight books, each containing both demographic overviews and focused essays on the features that distinguish each region's religious life. Mark Silk and Andrew Walsh, the center's director and associate director, persuaded an impressive array of historians, sociologists and religion scholars to contribute their formidable insight.

As with any edited collection, some chapters are better than others, but on balance, there is plenty here to reward readers who want a closer look at the role of religion in the particular places that make up this extremely varied nation. Each volume includes introductions and conclusions that draw out common regional themes. There is also a general overview chapter that presents the necessary numbers--from ethnic and educational patterns to just how many Baptists or Catholics or, for that matter, Sikhs and Muslims there are in the U.S. The remaining chapters address the distinct practices and histories of the dominant groups and include a variety of essays that take up such topics as the roles of southern religious women, patterns of new religious immigration, and even the religious urban geography of Chicago. Taken together, they constitute an encyclopedic introduction to the myriad stories that make up "American religion."

THE QUALITY of this team of authors and editors was essential, because the statistical data from which they had to work simply cannot tell the whole story. The absence of a national religious census in the U.S. means that we are always confined to filling in numerical gaps with educated guesswork. Each of the sources used here provides a wealth of information, but each has significant limits as well--and the three do not overlap perfectly. The North American Religion Atlas (NARA) depends on the 2000 membership numbers compiled by the Glenmary Research Center (a decennial project undertaken since 1950). For many groups, these numbers provide a fairly accurate county-by-county picture, but to the extent that a group provided sketchy data (lots of round numbers always raise questions) or no data at all, these "adherent" numbers fail to reflect the actual religious composition of a given area.

For instance, Samuel Hill argues convincingly that many of the religions that thrive in Appalachia are precisely the sort that don't look kindly on anyone who wants to count them. The high number of presumed "nonadherents" in those counties distorts the picture.