Joanna Russ writes, in her F&SF column of January 1975 (reprinted in The Country You Have Never Seen):
A reviewer's hardest task is to define standards. "Good" can mean almost anything: what the British call "a good read," "for those who like it, this is what they'll like," "it won't poison you," "good enough for minor entertainment," "mildly pleasant," "intelligent, thoughtful, and interesting," "charming!" and just plain "good"--excluding the range of better, from fine to splendid to superb to great. Reviewers also tend to adopt a paradoxical sliding scale in measuring a book's quality, i.e., the more ambitious a book, the more it's likely to fail; yet the competent, low-level "success" can be less valuable and interesting than the flawed, fascinating, incomplete "failure." For example, in July 1973 I reviewed James Gunn's The Listeners (which belongs emphatically in category two, above) and managed to make it sound worse than Norman Spinrad's The Iron Dream, a considerably lower-level (although fun and interesting) category one. Novels don't only provide different kinds of pleasures; they involve a reader more or less profoundly. Listeners was "bad" because parts of it were so wonderfully good. Dream was "good" partly because it demanded so little of the reader-- some of this by the author's deliberate choice, which only adds to the complexity of the whole business.
None of this months' hardcover novels* lives up to its author's own best work and in that sense they are not good books. They're certainly not in the "good-by-any-standards" class. Yet none of them is in the droopy-eyeball or loathsome class, either, and all have some excellences. The reviewer's business (as so many reviewers have said) is distinguishing between various levels of failure, keeping in mind that by "good" here I mean very high standards indeed.
I'm variously a reader of reviews, a writer of reviews, a writer's whose work receives reviews, and an editor/publisher who hopes to see the work she publishes reviewed. I have long read reviews for pleasure-- and not as a guide to what I ought or ought not to be reading. I think I first began reading reviews as a graduate student, most often in the form of review essays (which I also, of course, was required to write). Review essays encompass more than a simple review does, exploring the subject matter of several books in a thoughtful, critical way. This probably has something to do with my interest--as a reader--in reviews these days. When I read a review for pleasure (as I still often do), it's in the hope of being given something interesting to think about, or even the occasional gem of insight. This is certainly why I'm reading Russ's old reviews. Reviewers who offer intellectually or emotionally threadbare descriptions or assessments of the books they're reviewing don't interest me; if after several encounters I see that their reviews consistently fail to give me anything worth thinking about, I stop reading their reviews. I also stop reading their reviews for pleasure if I see that a reviewer is intractably ignorant or stupid-- that they have a penchant for misreading or under-reading or for continually praising the banal. I'm a bit like Mr. Darcy, I'm afraid, in that my good opinion of a reviewer once lost, is unlikely to be regained. I'm a busy woman, you know?
As an editor/publisher, I'm largely unconcerned with whether or not a review is pleasurably insightful or even intelligent. It's still a mystery to me what actually sells books. A reviewer's writing intelligently about a book will likely make its author happy, but it probably won't make that much difference to sales. Because I read every review of Aqueduct's books (that I know of), I'm constantly made aware of how few reviewers are careful readers. More to the point, though, is that many of those reviewers who have passed beyond the naive-reader stage are so caught up in their own particular set of dogmas that their readings tend to be inflexibly rigid and skewed to conform to their dogmas such as they are. This doesn't concern me as a publisher, but I can't help reflecting on it as a reader and writer of reviews myself.
Not surprisingly, I try to write the kind of reviews I enjoy reading. It can be hard to do it, though, when a book is so poorly written that I can barely bring myself to keep reading it. (I've twice forced myself to finish really bad books and reviewed them, but the experiences were so painful that I have a deep aversion to doing it ever again.) Russ has no hesitation in characterizing books as "bad." But aware as I am of what inadequate readers so many reviewers are, I've been reluctant to deliver such bald judgments myself. [Though so as not to mislead anyone, I will cop to once having delivered an unequivocally scathing judgment of a book that really should not have been published.] And so I have to wonder just how useful terms like "good" and "bad" can be in reviews today. Russ, unlike most reviewers, offers insight into the books she reviews; so I'm tempted to think she knows what she's talking about. Even so, for me at least it is more her discussion of the books rather than her verdict on a given book's quality that "defines" her "standards." The "whole business" is indeed complex.
*The books under review were by Robert Silverberg, James Gunn, John Brunner, Philp K. Dick, and Howard Waldrop & Jake Saunders.