Persuasive Books Are Less Enjoyable

A few days ago, I bought the Cal Newport book Deep Work, about the benefits of intense focus on the work you’re doing: that means no responding to Slack messages as you’re in the middle of writing a report, no checking Twitter group chats to see if any of your colleagues have shared anything useful or interesting, no leaving your email client open and habitually checking, reading, and writing emails, and so on. I basically buy into the thesis of the book, but I think my reason for buying into it is that I already thought that minimising distractions was an important strategy for getting work done, and it’s fun to read positive stuff about beliefs you already hold (being aware of your own confirmation biases can be a recipe for figuring out what you’ll enjoy reading). I get much more done when I do ‘Deep Work’, and am very, very prone to being distracted and/or procrastinating (see my article on ADHD for more on that). I enjoyed reading the book a lot - it had anecdotes about Bill Gates isolating and doing nothing but think about how to solve Microsoft’s problems and Woody Allen refusing to own a computer so that he could write on his German Olympia SM3 manual typewriter. But I didn’t find it particularly persuasive.

Deep Work does cite a few studies, but not many. And certainly not enough to make an effective case for intense focus as a vastly superior way of getting stuff done to typical ‘Shallow Work’ (where you do a project but also reply to Slack messages and check emails and whatever else). In the first chapter, titled ‘Deep Work is valuable’, which is the chapter that makes the case for Deep Work’s superiority, most of the focus is on specific extremely high achievers who have some habit that resembles Deep Work. Nate Silver’s use of SQL and Stata is evidence that Deep Work is important - because those tools are ‘definitely not something you can learn intuitively after some modest tinkering’ (this was slightly funny to me because I basically did learn Stata and R through playing with it - I think messing around with R while working on other stuff can be a really effective way to learn it!). Another example that Newport gives is the Wharton Professor Adam Grant, who writes huge numbers of academic articles and incredibly popular books, apparently partially because he batches intellectual work into long, uninterrupted stretches.

I don’t think that these anecdotes make a convincing case for Deep Work. To be convinced (or at least, to adjust my view and believe in Deep Work more than I already do), the book would need to be filled with studies, basically a literature review. I would want to see some Randomised Control Trials, maybe some convincing observational data, some more methodologically rigorous qualitative data, and whatever else. Some accounts of highly successful people doing stuff that seems sort of like Deep Work (and I’m not even convinced that the Nate Silver example gives any evidence of Deep Work at all) are not convincing - they select on the dependent variable, and just aren’t persuasive. But if the book had been a literature review of the relevant evidence, I definitely wouldn’t have read it, because it would have been really bloody boring.

Deep Work was definitely enjoyable. I liked hearing about Bill Gates and Woody Allen and Nate Silver and Adam Grant. It was fun to read, and even though I didn’t think the evidence was particularly good, it scratched the itch I had to read about highly successful people using Deep Work to get stuff done. It was fun to read because it wasn’t persuasive. This seems like a bit of a problem. I like to write things, and I mostly want people to enjoy my writing. The blog post I am writing right now is supposed to be at least moderately enjoyable. And maybe it makes the same mistakes that Deep Work does - I’m proposing that there is a persuasive/enjoyable trade-off, that if a social science book is really enjoyable, it probably uses anecdotes and shallow case studies instead of relying on high-quality quantitative research (RCTs, natural experiments, etc.) and methodologically sound qualitative research, so it’s unpersuasive to me. And to demonstrate this trade-off, I talked about reading one book that was enjoyable but not persuasive. I didn’t trawl through the literature to get a good conclusion about whether there is a negative correlation between a book being enjoyable and being scientifically rigorous, because you probably don’t want to read about a ton of studies on that (I don’t know if those studies actually exist, maybe they don’t).

What’s a guy to do? I try to do very small social science literature reviews (well, not exactly literature reviews, but look at some of the evidence) like this one on the Gender Wage Gap, and this one on the Gender Equality Paradox. Spoiler Alert: these posts got way fewer hits than most of my other posts. My most viewed post ever is this post about advice, and it’s one of the minority of posts that doesn’t cite a single study. I want people to read my posts, and I want them to be enjoyable, but I also want to write about social science in a way that is persuasive. I think my next steps for figuring out how to do that are to look at the literature and see whether there really is a persuasiveness/enjoyable trade-off, and what the evidence says you can do to be both enjoyable and persuasive - but I bet delving into the literature won’t be 10% as enjoyable as reading Deep Work.