Outside they had a display with a bunch of nonfiction books, including one by Sapolsky, one by Gladwell, and . . . “Noise”!
I was trying to figure out what bugged me so much about all this, and I think the problem is not just that these books make misleading or false claims, but that by representing popular science, they displace more interesting/serious/accurate popular science that could be out there.
I think my reaction is similar to that of academic historians who get annoyed by popular history books of the Band of Brothers / Greatest Generation variety. It’s not that they mind popular history—I’m guessing that pop history is what got many of them interested in the topic as kids—it’s just frustrating when bad pop history is taking the place of the good stuff.
On the other hand, maybe bad pop history and bad pop science are the gateway drug that leads people to the good stuff. Start with Gladwell, then move to pop science that’s actually interesting, not just interesting-if-true.
Also, to be fair, even the bad stuff has lots of good stuff in it. Simplified WW2 books still have lots of information on WW2, and even Gladwell, who gets so much deserved criticism, has lots of interesting ideas and tells lots of good stories.
P.S. The bookstore isn’t really trolling me. I know this because they’re not displaying a copy of Why We Sleep.
What’s good popular statistics in your view?
Does the Bill James Baseball Abstract fit here? That is a book for a general audience written by a stats person. There are certainly lots of numbers in the book, though I don’t think that’s the focus. Should it be numbers first to be a popular stats book?
Bernoulli’s Fallacy by Aubrey Clayton is a great recent one. Really well written, unapologetically Bayesian, with really well-laid out examples & explanations as well as what seems to me like a pretty cohesive treatment of the history of statistics. I also remember liking Drunkard’s Walk & Fooled by Randomness. Also Infinite Powers is a really good recent book, and, although it’s about calculus not statistics, there is actually a fair amount of statistics-relevant content from what I remember.
I’ve never read that book, but I want to draw attention to the following article by Clayton: https://nautil.us/the-flawed-reasoning-behind-the-replication-crisis-8365/
It’s not all bad, but in the first few paragraphs he seems to make the mistake of conflating Bayes’ Theorem with Bayesian philosophy. He seems to not realise that Bayes’ Theorem is a thing for non-Bayesians as well, and that the Sally Clarke ruling was based on false reasoning no matter what school of thought you fall under. What makes it especially galling is that the derivation of his prior for Clarke’s innocence replicates how a non-Bayesian would derive the marginal probability of Clarke’s innocence. IMO It’s overall a very dishonest slant in favour of Bayesianism. And I say this as a hardline Bayesian myself.
If he genuinely believes that non-Bayesians don’t know or don’t agree with Bayes’ Theorem and that they seem to have no way of resolving the Prosecutor’s Fallacy, then that might be one of the most absurd statistical errors I’ve ever read from an actual qualified statistician in a (presumably) curated piece of writing. If that’s not what he meant then it’s very poorly written, because that’s how it reads to me.
I feel a bit bad pointing this out because I realise it’s not fair to judge an author based on one article. And like I said, the article isn’t all bad. Maybe the book you suggested is really good. But the errors he made in the above article strike me as a major red flag regarding his expertise.
You link to Sapolsky being overly credulous of the marshmallow study in a WSJ article, but don’t give a reason to believe his book might be bad. I however can provide one. He wrote a book titled “Why Zebras Don’t Get Ulcers: A Guide to Stress, Stress-Related Diseases, and Coping” in 1994 (and republished multiple times since then). By that point Barry Marshall had already shown that ulcers are caused by Helicobacter pylori by dosing himself with them and then curing himself with antibiotics (nurses had cured ulcers with antibiotics back in the 40s, but people just forgot this happened). Marshall won the Nobel Prize in Medicine for that, and says “Everything that’s supposedly caused by stress, I tell people there’s a Nobel Prize there if you find out the real cause.”
Not all ulcers are caused by H pylori.
I don’t get your disdain for Sapolsky, Andrew. Yes, perhaps he’s not sufficiently credulous about the marshmallow study. Nonetheless, he has written some very good books.
More generally: Yes, there are bad pop-science books out there, and there are legitimate grounds for criticizing many books and many authors. The bar for criticizing should be higher than “there’s one wrong thing in it,” though. I recently wrote a pop-science book and I put a lot of effort into making sure I accurately represented all of the many studies, discoveries, and insights I wrote about, so I definitely share the annoyance with pop-science that’s sloppily done. I also spent a lot of time — many, many hours — thinking about “what do people actually want to read about?” (Or: How can I write a book that someone actually reads, that might find itself on a bookstore display case?) These are *much* tougher questions.
It looks like you failed to actually link to your pop-science book. And form an ironic juxtaposition to my comment directly above yours :)
That’s odd — it shows up blue as a link, but you’re right that it’s not actually a link! That’s what I get for trying to use tags, instead of just writing https://eighteenthelephant.com/2021/06/20/so-simple-a-beginning-a-popular-science-book-about-biophysics/
or
https://smile.amazon.com/So-Simple-Beginning-Physical-Principles/dp/0691200408/
My guess is that part of this is this is what the publishing companies are most comfortable promoting because it maximizes revenue per book. Kind of like how the movie theater chains always have posters of the next marvel flick, even if that foreign language romcom is objectively better and much more profitable on a per screen basis. The workers at book stores are often heavily influenced by the thoughts of the publishing industry about what the customer “should” like. To me this makes an even better case for why Amazon Kindle despite all it’s flaws is a superior product.
There’s a popular book for healthier eating habits titled: “Eat This, Not That.” I think this sort of book is more helpful to people than if the book had been just about (and had been titled) “Don’t Eat That.”
If the above pop-sci books are bad, what are the good ones that people should be reading? (It’s summer and would love some good reading material!)
Good pop-sci book: How about “Shape” by Jordan Ellenberg?
Science books I’ve enjoyed in last couple of years:
Lee Smolin – The Trouble with Physics (this sounds like it should be quite dry – but I found it compelling and it’s an excellent example of how to criticise an academic science field constructively)
The Gene – Siddhartha Mukherjee
George Dyson – Turings Cathedral (Historical account of the origins of the digital universe)
Genome – Matt Ridley (OK he misrepresents the science on global warming – but it’s a very good read – maybe a tad dated now)
David Wootton – The Invention of Science (this will swallow up a good two months of summer reading – a tour de force)
Earth – an Intimate History by Richard Fortey (quite dated but it’s about Geology and so dates slowly!)
If anyone has some ideas for really good new popular science books would be great to hear them
For environmental science, anything by Vaclav Smil. I have not yet read his latest, but his earlier books are outstanding.
“The Gene” is excellent!
It’s too dense to really be popular, but “Energy and Civilization: A History” by Vaclav Smil (2017) is great, and he apparently has a new, more accessible book out. I note “Energy and Civilization” and Oliver Morton’s “The Planet Remade” in my “Books about Energy…” post here.
How far back does “Energy and Civilization: A History” go? I recently bought a book on the history of energy and it disappointingly started with coal in England. Interesting topic, well written, but I was looking for the bigger picture going back to the earliest civilizations – how much and what kinds of energy did they use? How much total energy and what types were used – including human energy – to build the great Pyramids or other ancient structures?
@chipmunk “Energy and Civilization” goes back to prehistory — estimates of power and energy use for hunting, gathering, and early agriculture, including more than you wanted to know about draft animals. There’s a lot about early civilizations — Egyptians, etc., with the Chinese parts especially fascinating and new to me. It’s thorough!
FWIW, the book reviews in Science are a great source of ideas for pop sci books to read*. I was quite amused by a review of a book on the science in Moby Dick: the punch line was that other than inventing a few non-existent whale species, Melville largely got the biology and behavior of whales right.
*: Truth in advertising: Or to not read. I find that the reviews tell me what I need to know about the subject, and I’ve never actually purchased one. Anyway, the book reviews almost make Science worth the price. If you are at an institution that gets it, check them out.
I will say that Gladwell and Freakonomics were gateway drugs for me to good science. They were very good in getting me to think critically and not accepting everything at face value. Of course, the further I got into that, the more I realized that they were full of shit. I don’t know how many of their readers take that next step,
I was introduced to Gladwell via mockery of him, but Freakonomics (the book, not the blog, or the sequel to the book) served a similar role for me as it did you and I don’t know if I would have started reading Gelman if not for it.
> On the other hand, maybe bad pop history and bad pop science are the gateway drug that leads people to the good stuff. Start with Gladwell, then move to pop science that’s actually interesting, not just interesting-if-true.
Hey that was me! Stumbled onto Gladwell and other pop psych stuff in high school which introduced me to the ideas / difficulties in measurements and designing experiments. Which led me to major in statistics in undergrad and eventually to this blog and doing a stat PhD. It’s funny how that works!
Sorry, I’m too old for this discussion. My gateway drug was Velikovsky.
My grandmother subscribed to “Science Digest” for me in the nineteen fifties, along with various little golden books, and I think that has made all the difference. I remember being fascinated with Velikovsky’s works, too, and wondering “could there be something there?, but that may have been more early 60’s.
“Also, to be fair, ” their books are not the worst books, even some academic books / articles are much worse.