Reviewing reviews

Most nonfiction book reviews suffer from the same problems. If they’re critical they generally go one of two ways.

  1. The author misses out topic X, I know about X and here are some other books on X that should have been addressed.((X can be another author, or a whole other tradition, etc.))
  2. The author is seemingly a part of Intellectual Trend Y and I don’t like Y, here are its antecedents, whom I also dislike.

It won’t have escaped your notice that neither review assesses whether or not the author’s argument was valid. ((Quick calculation tells me that in my blessed life, benefiting from the internet, most notably, I have already read some 1000 nonfiction book reviews, at least 2000 fiction reviews. Maybe that’s more than any other genre of writing… I guess news articles probably beat them out, even though I’ve been avoiding them for 5 years.))

And neither of them say whether the book is well written.

For number 1 the issue is that every book ever written fails to address almost everything, so it’s a somewhat moot criticism to begin with. If a book is purporting to be about 21st century US presidents and doesn’t mention Obama, fair enough, but this type of review normally expects complete coverage of all possible angles of a topic. This predominates in academic reviewing, certainly, where the goal is of course to signal one’s own superior knowledge of Topic X.

For number 2, the problem is just as obvious. Occasionally the reviewer might go on to argue why it is that Intellectual Trend Y is wrongheaded, but often it stops short even of doing that, as though the mere labelling or classifying of a book as part of a preexisting trend (which is also impossible to avoid) is itself a takedown.

What does a good review look like? Well, consider what we should want from a book review. ((What we actually want is probably some worldview reinforcement, or justification that we don’t need to read a whole book to debunk it or dismiss it. We might as well at least try and do better.))

  • I think we want to know if it’s worth reading, i.e. spending our precious, limited time on. So we want a bit of a summary and indeed most reviews do give us this.
  • We also want to know if it’s well written: that’s massively important. On basically every topic, there’s no shortage of books; spend your time on the ones that are a pleasure to be with.((N.B. this isn’t exactly the same as the author’s tone, which often seems to piss off reviewers so much they ignore the content. I guess no one likes someone being smarmy, arrogant, or shrill when they disagree with them. When they agree with them these qualities are transmogrified into confidence, panache and passion. In reality reviewers very rarely separate out the matter from the manner. It’s hard to do.))
  • We shouldn’t be interested in what the book excludes, unless it’s something truly neglectful. Instead, what does it include: what is the book’s argument?
  • Finally, is that argument internally consistent. ((Naturally that includes whether facts have been accurately reported or studies accurately interpreted, etc. But this isn’t the same as, “They overlooked Smith and Wong’s obscure article only I know about…”))

I think that’s the best we can do. No argument about any topic can be totally complete, consistent and convincing. So instead, we should worry if it works on its own terms. Because even a book we disagree with is worth reading if it’s a good example of the POV we don’t — at this time — agree with. We have to assume that our views will change over time based, hopefully, on compelling arguments we encounter. How can we encounter the arguments unless we expose ourselves to them? We should therefore aim to read things we currently disagree with, as long as they’re well done. This means that criticism number 2 above is totally misguided. John Gray is one of the worst offenders. He hates every intellectual trend, so whenever he reviews a book he ignores whether it’s well written or a good version of its argument and instead just starts pointing out that such-and-such circle of thinkers from the 19th century already said this, and so-and-so mystics from the medieval period already said that. Irrelevant.

People who do number 1 type reviews might say the book doesn’t achieve everything it set out to do, so it failed on its own terms. Be careful. Most broadly, a nonfiction book offers an argument about how the world is, or how it should be, or how you can improve (or a combo). If a book is saying how the world is, it can only ever cover a part of the world, obviously. So even if it purports to be a worldview kind of book, if you expect it to literally cover everything, you’re in the wrong universe.

Two recent examples. Here’s a review of Jordan Peterson’s book. I read the book and didn’t really like it. But that review made me want to buy it because it was so obviously not engaging with the book’s argument and instead clearly just irked the reviewer. There would have been no way for anyone to say the things the book said in an acceptable way for that reviewer. Terrible review, much worse than the book. Anyone who doesn’t share the reviewer’s worldview isn’t helped.

Here’s one about Dan Dennett’s latest book, which I did like. Again, the reviewer repeatedly says Dennett is flat wrong but I agree with all the things the reviewer says are wrong and disagree with the things he says instead (i.e. right from his POV). This gives us no sense of whether Dennett made a wrong (from the reviewer’s POV) argument in a good way. Indeed he hints that it’s more narrative than argument, rambling, and repetitive. But if he agreed with the argument, perhaps it would be a great story, littered with fascinating details, with enough reiteration that the message sticks. Anyone who doesn’t share the reviewer’s worldview isn’t helped. Bad review.

So I don’t expect reviewers to magically change their ways, because they’re all working on different premises to me of what book reading and book reviewing is all about. But with this kind of framework in place, I now read reviews with a mind to what I want out of the books they’re reviewing.

Bonus heuristic. If a book has all 5- and 1-star reviews on Amazon, it’s probably good because it’s clearly just pissed people off ideologically. If it has a whole heap of 2s and 3s, then it’s probably no good. If you think a book is truly bad, you give it a 2. If it riles you and you think the whole idea it represents needs to be squished then you give it a 1.

Addendum [29-05-2018]. Any nonfiction book subjected to intense scrutiny will yield up multiple problems. This is like how any true crime podcast finds fault with a police investigation and how any major event sufficiently complex will have enough gaps in the official account that a conspiracy theory will find fertile ground. So here’s another possible heuristic: a measure of a nonfiction book’s quality is how much effort you have to go to to discredit it.((I’ve been toying with the inverse for fiction, where I feel like how much effort you have to go to to praise a novel is inversely proportional to literary quality. Hence there will be a very ingenious way, utilising all your sophisticated cultural knowledge and tools of analysis, to do a laudatory reading of a pulp novel. But to read a literary classic will actually require less effort, less contrivance, to provide praise for its sophistication. In a sense, literary works are easier to read analytically, albeit often harder to read syntactically.))