At the Anxious Bench, Thomas Kidd has a post titled “The Art of the Book Review.” Kidd argues:

When reviewing a book, be as charitable as possible. This is a “Golden Rule” issue. Once you’ve experienced the years of struggle behind producing a book, you should be cautious about casually dismissing the value of another author’s work. Even when I have major problems with a book, I try to praise it as much as I can, and include a few lines that a press could likely use to promote the book. …

It is certainly in-bounds to disagree with an author’s interpretation or ideological assumptions, of course. But do this in a polite way, as a means of discussion rather than denunciation.

I’d add that there are two kinds of book reviews: reviews that appear in professional, “serious” journals, and reviews that appear everywhere else. These reviews have entirely different purposes. When I get my copy of the American Historical Review or Church History, I browse the reviews for anything that is immediately interesting or relevant. But most of the time I come across reviews in journals through a search in JSTOR or some other database. In such cases I’m looking for reviews to find out the scholarly consensus about the worth of the book that has been established over time. And make no mistake, it takes time—years—for books to be reviewed in scholarly journals. I want those reviews to be the plainest possible explanation of what is good and bad about a book, perhaps even erring on the side of letting readers know where they could be mislead.

But plenty of reviews are published on blogs like Religion in American History, The Anxious Bench, and the S-USIH blog, as well as in general interest publications like magazines and newspapers. These reviews can be of the highest quality—for example, take Paul Putz’s recent review of Matthew Bowman’s The Urban Pulpit. But the purpose of these reviews is different. To borrow a phrase from Dan Cohen, their purpose is “catching the good.” When I read a review at one of these sites, I want the authors to alert me to the best books—books they’ve found useful for their research or teaching, or books that are a model of the historian’s craft. I can’t think of a reason why someone would publish a highly critical review in venues like those, especially since you get to pick what you review. If it’s a bad book, don’t bother reviewing it.

If you agree to write a review for a scholarly journal, then be as critical as you must, keeping in mind what Kidd writes about charity and generosity. But if you are writing a review for any other venue, find a book that’s good, and tell people about it.