Nice review, but I just want to note the difference in meaning between "training to beat the chimp as hard as they could" and "training as hard as they could to beat the chimp."
Obviously it's too late for the other entries this year, but I really enjoyed the 'Pairs Well With' section (and its title!) and would love to see others follow suit.
Formatting issue: The subheading starting 'The Importance of Methodology' repeats the '1.' counter, presumably an artefact of converting the document into Substack's system. There's also a space missing in 'Full disclosure, almosteverything I know'.
Is there any chance we could get a google form for voting on these as we read them, similar to what you've done for the non-finalists? I feel like by the end of this contest there is no way I'm going to remember any sort of ranking from previous reviews, but if you ask me to rate them immediately after reading them, you would probably end up with a much better indicator of true ranking.
This is one of my favorite non-fiction books, and I really can't encourage people enough to read it for themselves. It is a genuine joy to read for its own sake, especially the animal anecdotes - it's not one of those tedious-but-wise non-fiction books where you're better off saving your time and just banking a summary, even one as detailed as this.
Great review. I would also strongly endorse de Waal's book called "Chimpanzee Politics". It is the brilliant review of social organization and power dynamics of a group of chimps. I used to keep a copy of it in my office and tell everyone that it was the only business book that I would recommend everyone reads (though it was not necessarily a "how to" manual). A side benefit is that it also sheds light on the whole Trump phenomenon.
This may be controversial and I hope it doesn't break ACX's spirit of conviviality, but it did occur to me after reading this book that there may be an empathy divide, and therefore a potential progressive/conservative divide, in how likely we are to assign intelligence to the observed behavior of a non-human animal.
I suspect, without claiming to be certain, that those with low empathy are more likely to assume that a given animal is just a brute who responded to stimuli mechanistically unless it's decisively proven otherwise, whereas those with more empathy are more likely to start from the premise that the animal is "like me" until decisively proven otherwise.
Could I recommend adding the "The Bird Way" (https://amzn.to/3nceCfM) to your "pairs well with" at the end? :) It's what I thought of most while reading your review.
> like the primate research team that learned a chimpanzee in Japan performed significantly better than humans on a specific memory-related task and responded by training to beat the chimp as hard as they could.
Holy crap I first read this as meaning that the researchers had punished the chimp for performing better than them, by subjecting it to a beating (having first trained themselves to maximize the pain they could inflict).
I've been posting this everywhere this week, but I just listened to this podcast interview by Ezra Klein of Alison Gopnik. If you think of children and other humans as animals whose consciousness we only partially inhabit, and want to understand the intelligence of, it pairs very nicely with this: https://www.nytimes.com/2021/04/16/podcasts/ezra-klein-podcast-alison-gopnik-transcript.html
I don't know how it fairs for the reviewing process, but the first paragraph made me decide I didn't wouldn't read the review, because I'd much rather read the book. Well done?
> until fairly recently, and in some corners still today, we have had very smart people reasoning about thought as fundamentally being linguistic and nature, while some large chunk, maybe as many as ~15% of people just don't think in words.
Don't confuse the report for the reality. I suggest that exactly 100% of people do not think in words. That's not to say they have no internal monologue. But the monologue is an automatic process triggered by the thought, and the thought, having happened first, would be unaffected if the monologue were suppressed or absent.
I‘d like to point out that the additionally recommended book „The Hidden Life of Trees“ seems to be not without controversy. Years ago I felt intrigued by the thesis but finally refrained from reading it as a number of reviewers from the science community criticised it as being not particularly substantiated (and mixing fact, opinion and romantic metapher). I don’t remember too much detail. Just be on the lookout when reading it.
"One area where he sides with the skeptics is in asking for an end to inter-species intelligence contests and trying to grapple with big, messy topics like "self-awareness" or "consciousness". He points out that we don't even have good definitions for these concepts in humans, so why should we expect to apply them to other animals? "
The sentence "we don't have a good definition of consciousness" is outdated. This field has made enormous progress in the last ~15 years, and I think we have an answer on what consciousness is, though the answer is much more boring than philosophers anticipated. (Especially people who mix it up with self-awareness or meta-cognition.)
The key idea is that we have lots of different situations where people give firm answers to the question "did you see the picture on the screen?", and it depends on details whether the answer is "Yes." or "Which picture? The screen was all black!". Like whether the picture was shown 100 ms earlier or later. One can basically read of "Yes"-answers from an EEG because they correspond to totally different neural patterns: an avalanche of activity, a P3 wave, a brain-wide synchronization of activity, etc. Since we have quite a diverse set of situations/experiments (really! you would be surprised!) that show consistently the same patterns, we can read of from the outside whether something is perceived consciously.
And yes, animals like dogs, birds, dolphins etc are conscious according to that test, and babies are, too. As are some locked-in patients, but not others.
"for much of the twentieth century, many biologists went too far and would deny any human-like characteristics to animals"
That was the reigning view when I started reading National Geographic for Kids in 1966. For example, we were told that only humans used tools. Thus, it was a huge deal to 7-year-olds like me that Jane Goodall reported that her chimps used crude tools. Jane Goodall was an official Scientist, so she wasn't to be ignored.
As I grew older, I noticed that with just about anything we were told that Science had proven animals couldn't do, that somebody somewhere had an anecdote of an animal they'd known doing it. But anecdotes were not Science.
Today, in the Youtube Era, we of course have an absolutely immense body of evidence of individual animals doing all sorts of wild things that the Behaviorist Era denied they could do.
I second the recommendation of "War in Human Civilization". "The Horse the Wheel and Language", on the other hand, taught me more about the erosion of horse teeth than I ever wanted to know.
If chimpanzees were smarter than us, then we'd be the endangered species in their zoos.
Another book recommendation - for more about cephalopod cognition, Other Minds by Peter Godfrey Smith. There's a bunch of philosophy mixed in, but I found the descriptions of octopus behavior more interesting, and the thesis that since octopi are the smart animals that are least closely related to us, they're the closest we can get (for now?) to studying aliens.
There are also some claims that make me wonder if there is or will be a replication crisis in animal behavior studies similar to the one in the social sciences.
Big props for the structure and organization of the review (especially the "Pairs well with" section). While more stodgy and rigid than the woven masterworks from Scott and Reviewer #2, this is a structure that could actually be emulated and used as a guideline by others less skilled in the art. Given you're a finalist pulled from an already aggressively self-selected pool, that's...probably most people.
If I were ever to write a book review, I'd definitely be eyeing this format as my starting template. Unless the later entries source up something even better :)
Great choice of subject matter. I took a class on ethology in college (our main text was King Solomon's Ring by Lorenz, which I highly recommend) and had a similar reaction of "wow, here's this fascinating and distinctive field of science that I had no idea existed, what's up with that." I wasn't super-satisfied with the presentation, though. I wish there was more discussion of what we've actually learned from ethology and less of the meta-learnings, and the sections on predictive processing and other books felt a bit pointless. Still, the content was interesting and the writing, though somewhat dry by ACX standards, was quite clear. Thanks!
Another potential addition to the "Pairs Well With" section would be almost anything by Desmond Morris; but especially his *watching series, such as Catwatching. Contrasting his in-depth observational approach with the more rigorous discussion in this book is fun and interesting.
By the way, the “authors” tag is apparently wrong in Substack’s metadata, based on how this post shows up on a link-sharing site. Does Substack have a way to change it for guest posts? Maybe someone needs to file a feature request?
You said: Apes seem to learn sign language. This isn't quite accurate.
For a deep dive into all the attempts (pretty much all failed and falsified) to teach apes sign language, see Robert Sapolsky's lecture on trying to teach other animals language. https://youtu.be/SIOQgY1tqrU?t=4711
The tiny-narrator-guy-in-our-heads dichotomy isn't real. It'd be an impossibly slow way of processing the world if we had to put everything we sense into sentences. It's sometimes this, sometimes that.
Nice review, but I just want to note the difference in meaning between "training to beat the chimp as hard as they could" and "training as hard as they could to beat the chimp."
Obviously it's too late for the other entries this year, but I really enjoyed the 'Pairs Well With' section (and its title!) and would love to see others follow suit.
Formatting issue: The subheading starting 'The Importance of Methodology' repeats the '1.' counter, presumably an artefact of converting the document into Substack's system. There's also a space missing in 'Full disclosure, almosteverything I know'.
Is there any chance we could get a google form for voting on these as we read them, similar to what you've done for the non-finalists? I feel like by the end of this contest there is no way I'm going to remember any sort of ranking from previous reviews, but if you ask me to rate them immediately after reading them, you would probably end up with a much better indicator of true ranking.
This is one of my favorite non-fiction books, and I really can't encourage people enough to read it for themselves. It is a genuine joy to read for its own sake, especially the animal anecdotes - it's not one of those tedious-but-wise non-fiction books where you're better off saving your time and just banking a summary, even one as detailed as this.
Great review. I would also strongly endorse de Waal's book called "Chimpanzee Politics". It is the brilliant review of social organization and power dynamics of a group of chimps. I used to keep a copy of it in my office and tell everyone that it was the only business book that I would recommend everyone reads (though it was not necessarily a "how to" manual). A side benefit is that it also sheds light on the whole Trump phenomenon.
This may be controversial and I hope it doesn't break ACX's spirit of conviviality, but it did occur to me after reading this book that there may be an empathy divide, and therefore a potential progressive/conservative divide, in how likely we are to assign intelligence to the observed behavior of a non-human animal.
I suspect, without claiming to be certain, that those with low empathy are more likely to assume that a given animal is just a brute who responded to stimuli mechanistically unless it's decisively proven otherwise, whereas those with more empathy are more likely to start from the premise that the animal is "like me" until decisively proven otherwise.
Could I recommend adding the "The Bird Way" (https://amzn.to/3nceCfM) to your "pairs well with" at the end? :) It's what I thought of most while reading your review.
FWIW, Gwern wrote a bunch about Cat Sense here: https://www.gwern.net/Cat-Sense
> like the primate research team that learned a chimpanzee in Japan performed significantly better than humans on a specific memory-related task and responded by training to beat the chimp as hard as they could.
Holy crap I first read this as meaning that the researchers had punished the chimp for performing better than them, by subjecting it to a beating (having first trained themselves to maximize the pain they could inflict).
Best review yet!
This is certainly a joyful review! And it's very rare that one blog post immediately adds five books to my "to read" list.
I've been posting this everywhere this week, but I just listened to this podcast interview by Ezra Klein of Alison Gopnik. If you think of children and other humans as animals whose consciousness we only partially inhabit, and want to understand the intelligence of, it pairs very nicely with this: https://www.nytimes.com/2021/04/16/podcasts/ezra-klein-podcast-alison-gopnik-transcript.html
Excellent review, really gives me a sense of what the book is about.
I would like to apologize to all of the mice I have apparently been stressing out. I had no idea.
I don't know how it fairs for the reviewing process, but the first paragraph made me decide I didn't wouldn't read the review, because I'd much rather read the book. Well done?
Not related to the book review, but I'm posting it here so Scott sees it: The comments seem to have vanished from slatestarcodex.com since the formatting (blue header, gray background, &c.) was restored, & have been missing from unsongbook.com for several weeks at least. Under each post the comment section has "[number] responses to [article name]" but none of the comments are shown. This (https://fileleaks.com/e/wBaQVOCq58) is what the last pre-shutdown open thread (https://slatestarcodex.com/2020/06/21/open-thread-156-5/) looks like to me now, though the Wayback Machine (https://web.archive.org/web/20200622031835/https://slatestarcodex.com/2020/06/21/open-thread-156-5/) shows that there were a significant number of comments there.
Watching “what about bunny”'s videos on instagram/tiktok and https://www.theycantalk.org/ is pretty eye opening
> until fairly recently, and in some corners still today, we have had very smart people reasoning about thought as fundamentally being linguistic and nature, while some large chunk, maybe as many as ~15% of people just don't think in words.
Don't confuse the report for the reality. I suggest that exactly 100% of people do not think in words. That's not to say they have no internal monologue. But the monologue is an automatic process triggered by the thought, and the thought, having happened first, would be unaffected if the monologue were suppressed or absent.
I‘d like to point out that the additionally recommended book „The Hidden Life of Trees“ seems to be not without controversy. Years ago I felt intrigued by the thesis but finally refrained from reading it as a number of reviewers from the science community criticised it as being not particularly substantiated (and mixing fact, opinion and romantic metapher). I don’t remember too much detail. Just be on the lookout when reading it.
"One area where he sides with the skeptics is in asking for an end to inter-species intelligence contests and trying to grapple with big, messy topics like "self-awareness" or "consciousness". He points out that we don't even have good definitions for these concepts in humans, so why should we expect to apply them to other animals? "
The sentence "we don't have a good definition of consciousness" is outdated. This field has made enormous progress in the last ~15 years, and I think we have an answer on what consciousness is, though the answer is much more boring than philosophers anticipated. (Especially people who mix it up with self-awareness or meta-cognition.)
If you think otherwise, I recommend to read "Consciousness and the Brain" by Stanislas Dehaene. The wikipedia article already gives a summary: https://en.wikipedia.org/wiki/Consciousness_and_the_Brain .
The key idea is that we have lots of different situations where people give firm answers to the question "did you see the picture on the screen?", and it depends on details whether the answer is "Yes." or "Which picture? The screen was all black!". Like whether the picture was shown 100 ms earlier or later. One can basically read of "Yes"-answers from an EEG because they correspond to totally different neural patterns: an avalanche of activity, a P3 wave, a brain-wide synchronization of activity, etc. Since we have quite a diverse set of situations/experiments (really! you would be surprised!) that show consistently the same patterns, we can read of from the outside whether something is perceived consciously.
And yes, animals like dogs, birds, dolphins etc are conscious according to that test, and babies are, too. As are some locked-in patients, but not others.
"for much of the twentieth century, many biologists went too far and would deny any human-like characteristics to animals"
That was the reigning view when I started reading National Geographic for Kids in 1966. For example, we were told that only humans used tools. Thus, it was a huge deal to 7-year-olds like me that Jane Goodall reported that her chimps used crude tools. Jane Goodall was an official Scientist, so she wasn't to be ignored.
As I grew older, I noticed that with just about anything we were told that Science had proven animals couldn't do, that somebody somewhere had an anecdote of an animal they'd known doing it. But anecdotes were not Science.
Today, in the Youtube Era, we of course have an absolutely immense body of evidence of individual animals doing all sorts of wild things that the Behaviorist Era denied they could do.
I second the recommendation of "War in Human Civilization". "The Horse the Wheel and Language", on the other hand, taught me more about the erosion of horse teeth than I ever wanted to know.
If chimpanzees were smarter than us, then we'd be the endangered species in their zoos.
I thought "Cat Sense and Dog Sense" was a single book. Took me a couple searches to realize that they're two books.
Another book recommendation - for more about cephalopod cognition, Other Minds by Peter Godfrey Smith. There's a bunch of philosophy mixed in, but I found the descriptions of octopus behavior more interesting, and the thesis that since octopi are the smart animals that are least closely related to us, they're the closest we can get (for now?) to studying aliens.
There are also some claims that make me wonder if there is or will be a replication crisis in animal behavior studies similar to the one in the social sciences.
Big props for the structure and organization of the review (especially the "Pairs well with" section). While more stodgy and rigid than the woven masterworks from Scott and Reviewer #2, this is a structure that could actually be emulated and used as a guideline by others less skilled in the art. Given you're a finalist pulled from an already aggressively self-selected pool, that's...probably most people.
If I were ever to write a book review, I'd definitely be eyeing this format as my starting template. Unless the later entries source up something even better :)
Excellent read, though I could make neither heads nor tails of the section header formats and numbering.
Brief review-of-the-review:
Great choice of subject matter. I took a class on ethology in college (our main text was King Solomon's Ring by Lorenz, which I highly recommend) and had a similar reaction of "wow, here's this fascinating and distinctive field of science that I had no idea existed, what's up with that." I wasn't super-satisfied with the presentation, though. I wish there was more discussion of what we've actually learned from ethology and less of the meta-learnings, and the sections on predictive processing and other books felt a bit pointless. Still, the content was interesting and the writing, though somewhat dry by ACX standards, was quite clear. Thanks!
Another potential addition to the "Pairs Well With" section would be almost anything by Desmond Morris; but especially his *watching series, such as Catwatching. Contrasting his in-depth observational approach with the more rigorous discussion in this book is fun and interesting.
By the way, the “authors” tag is apparently wrong in Substack’s metadata, based on how this post shows up on a link-sharing site. Does Substack have a way to change it for guest posts? Maybe someone needs to file a feature request?
You said: Apes seem to learn sign language. This isn't quite accurate.
For a deep dive into all the attempts (pretty much all failed and falsified) to teach apes sign language, see Robert Sapolsky's lecture on trying to teach other animals language. https://youtu.be/SIOQgY1tqrU?t=4711
The tiny-narrator-guy-in-our-heads dichotomy isn't real. It'd be an impossibly slow way of processing the world if we had to put everything we sense into sentences. It's sometimes this, sometimes that.
Any ethical implications about this and how we treat other species?