1 In Houses of God, Peter Williams provides a regional and historical survey of American religious architecture. He divides the United States into seven regions, what he calls "cultural hearths." One chapter is devoted to each of the following: New England; the Mid-Atlantic states; the South; the Old Northwest; the Great Plains and Mountains; the Spanish Borderlands; and the Pacific Rim. Each, presumably, can be defined by at least one distinctive style (or a distinctive stylistic variant) originating in at least one distinctive religious or ethnic tradition.

2 The organization of this book has one major advantage: geographical comprehensiveness. Some authors would stop there. But Williams adds another kind of comprehensiveness. He is very careful to note the "diversity" and "pluralism" of American religious and art history. Unlike many other books of this kind, his gives Jews, Hindus, Buddhists, and other non-Christians their due. (It could be argued that his title is not inclusive, though, because neither synagogues nor most Protestant churches can be described, at least technically, in connection with the domus dei.) But the organization of this book has one major disadvantage, too. That requires some explanation.

3 As Williams himself points out, some regions have been much less distinctive and influential than others. With a few exceptions (due to the historical presence of the Spanish in Arizona and New Mexico, the Russians in Alaska, the Mormons in Utah, and, more recently, the Buddhists in Hawaii and California) western religious architecture was simply brought there by eastern migrants and modified slightly according to local needs and resources. Although these migrants came from homogenous communities, they settled in what became very heterogeneous ones. The religious architecture of these could be described, therefore, as "some of this and some of that." As a result, most sections of this book dealing with western regions are somewhat less satisfying than those dealing with eastern regions. (By far the best chapters are those on New England and the South.) The organizational principle simply makes less sense when applied to the west, in other words, than when applied to the east. Williams might have used some other organizing principle for his book. He might have used five or six types of religious architecture, for example, or five or six types of religion, for that matter.

4 Williams is at his best when dealing with specifics. Most fascinating of all is his discussion of the New England meetinghouse and its transformation, as an "iconic" building type, to suit the needs of people far removed in time, place, and even spirit from those who first built them.

5 There are a few errors. The editor should have found some of these. On page 210, for instance, we read about "climactic" circumstances (instead of climatic ones). On page 159 Williams discusses some churches that still exist. But then, in the very next line, he mentions that one of them was demolished in 1891. On page 176 we read that "Jewish residential patterns ... followed those of other cities, with a massive outflux into the northwestern suburbs [of Detroit] culminating with the urban riots of 1967." The (no doubt unintended) implication is that Jewish migration patterns led directly to or even caused the riots (in which Jews themselves did not participate). The author himself could have avoided other errors. On page 174 he writes that Cincinnati's Hebrew Union College was "the first rabbinical seminary in the United States." Actually, it was the first Reform rabbinical seminary; the Orthodox had been training their own rabbis at smaller and more informal yeshivot for generations.

6 According to the blurb, this book is "beautifully illustrated." Would that it were so. Even though some illustrations are indeed by famous photographers (such as Walker Evans and Dorothea Lange), they are not beautifully reproduced; some of them are too murky, in fact, to be of much use. But that can hardly be blamed on Williams. In any case, he has provided lengthy bibliographies for every chapter, an index of personal names, and — even more useful for art historians — an index of structures. I strongly recommend this book for college and public libraries.