All of literature is a conversation. Writers create worlds and invite us to imagine how other people live. They bring us news, history, argument, and ask us to argue back. They impel us to feel, to notice, to observe as they change their minds, or endeavor to change ours. The critic has always been a crucial part of this tradition.
“Art,” Henry James wrote, “lives upon discussion.” And as long as there have been critics, there has been debate over the role they should play in this discussion. What sort of expertise should they bring? When they pass judgment, what form should it take? When it comes to books, what do critics owe the reader? What do they owe the writer? What do they owe the culture?
In this spirit, I have been asked to consider a question: What makes for a good book review?
“Good” is of course a matter of taste. Seen another way, it’s a matter of best practices — ethical, procedural, intellectual. “Good” being subjective, you may quarrel with what I say. I would argue this, too, is good. Quarrel with me, and I’ve already fulfilled one job of the critic: to prompt you, the reader, to think.
Were I forced to provide a short answer to the above question, it would look something like this: A good review introduces a book and attempts a rigorous appraisal, while demonstrating fairness, intelligence, clarity, discernment, and style.
That may sound straightforward, but it could use some interpretation. So let’s start where I always try to begin: with fairness.
Fairness comes into play early, with the very commissioning of a review. When I query writers, I provide enough information about a book to entice, but not to create bias; the reviewer shouldn’t know what I think. In addition, with every query — “We would be delighted to have you review this …” — I include a caveat along these lines: “… as long as you don’t foresee any conflicts of interest.” In other words: (a) You’re not a friend or colleague of, or have had any other form of close relationship with, the writer; (b) you’re not a mortal enemy of the writer; and (c) you haven’t had any kind of transactional relationship with the writer that would amount to a real or perceived conflict of interest. (For instance: You have not, in the past, received a blurb from, or publicly expressed effusive admiration or distaste for, or corresponded heatedly online with, the person whose work you’re about to assess.)
Once a reviewer passes the conflict-of-interest test, off the book goes. And off I go to wait, in suspense. Why suspense? Because I have no idea what I’m going to get back.
But all this is prelude. On to the “good” review.
A good review should assess the actual book, and here I mean a couple of things. For one, it should consider the version offered by publishers for review, not an early, unfinished draft, not a copy acquired through shadowy means. Second, it should evaluate the book the author wrote, not the book the reviewer wishes the author had written. A history on political assassinations in the Philippines should not be faulted for failing to chronicle the assassinations of leaders around the world. A novel meant to be a family farce should not be criticized for being unserious. (If it has failed as farce, that’s another matter: Have at it.)
A good review does not simply reveal whether a reviewer “likes” or “dislikes” a book. It measures a book’s arguments, ideas, and artfulness, and is ideally well argued itself. Perhaps it is forceful and suffused with certainty. Perhaps it is questioning, allowing you to eavesdrop as the critic works out her thinking on the page. But regardless of whether it praises, affirms, disputes, or censures, its assertions should be supported with evidence: quotations from the text, accurate summation, concrete illustrations from other sources. If a review declares the writing beautiful (or dreadful), it should include an example from the book to help readers judge for themselves. If it says the book contains errors of fact or logic, it had better prove it via clear exposition and well-made counterargument.
A good review comments perceptively on a book’s literary accomplishment (though this may be weighted differently according to genre). In some nonfiction, stylish writing is perhaps less important than the author’s mastery of subject and display of reason — although if bad writing distracts from sound reasoning, the review should say as much. If the author has wielded fancy writing in an effort to obscure intellectual bankruptcy, this, too, is fair game.
A good review avoids lengthy plot summary. (“This happened, then this happened” — this is a book report.) It also avoids indulging in spoilers. To give away endings or crucial twists is a reviewing sin.
A good review provides context, which requires the reviewer to come to the work well informed: Is this the novelist’s first book or sixth? Can or should the work be interpreted through a certain historical, social, political, or biographical lens? Is the book the first of its kind on a given topic? If it addresses a question others have confronted, how does it compare?
A good review can say a book is bad, but shouldn’t do so with gratuitous venom. The American poet Donald Hall has said that when a reviewer “obviously hates you, it doesn’t do any good.” Yet Hall does seek out those “annoyed” but sage critics who can suggest “new occasions for scrutiny, for crossing out” — for seeing the words fresh, for tinkering and improving.
A good review is also more than just content: We must consider its style. If a review is so dull or dense that no one will want to read it, what’s it worth? A good review should avoid cliché, especially book-reviewese, the empty adjectives and exclamations you find plastered on cover after cover. A good review should be written with flair — lucidity, elegance, panache — and indeed may be so well composed that it becomes its own lesson in writing. Yet flair should not overwhelm substance. Showiness for the sake of showing off is obnoxious.
Now, all this is a lot to accomplish in a single review. But the better critics keep these tenets in mind whether they’re writing 600 words or 6,000. In doing so, they perform a potentially powerful service: enlightening the reader and perhaps even transforming the artist — and, by extension, literature itself.
Jennifer B. McDonald is an editor at The New York Times Book Review and a 2012-13 Nieman Fellow at Harvard University.