15 Comments

There are two books which you *have to* read (if already haven't done so):

1. *Descartes' Error: Emotion, Reason, and the Human Brain* by Antonio R. Damasio. Besides being an incredibly informative book about human neuroscience, it also features Damasio's *somatic marker hypothesis*, which is in a nutshell: when reacting to a situation, our brains 'project' past images of the arising/developing situation 'into' our bodies (soma) & our reactions are guided by the sense data coming 'upward' from our bodies into our brains -- essentially we poll our bodies: "how do I *feel* about this (developing) turn of events?"

2. *Vehicles: Experiments in Synthetic Psychology* by Valentino Braitenberg. A most insightful thought experiment about brains based on the neurological/cybernetic research conducted by (the late) Braitenberg. I can assure you that you'll be royally entertained: it is pure mind candy. For a teaser, here are the chapter headings:

1. Getting around

2. Fear and Aggression

3. Love

4. Values and Special Tastes

5. Logic

6. Selection, the Impersonal Engineer

7. Concepts

8. Space, Things, and Movements

9. Shapes

10. Getting Ideas

11. Rules and Regularities

12. Trains of Thought

13. Foresight

14. Egotism and Optimism

(Both books can be found on the net in PDF forms; I'm not going to link to them for copyright reasons, but with reasonable effort, you can start reading them after 10 minutes of searching.)

Expand full comment

I have to say I disagree with a lot of what is presented.

In one of your more recent articles, you talk about consciousness beginning around the end of the last ice age or about 15,000 years ago, and claimed that the spread of this can be seen via the spread of the "na" first person pronoun, which is perfectly linguistically situated to be preserved.

Here, however, you are claiming that recursion is the basis of self-awareness, which seems completely analogous to what was mentioned before. You say that recursion was needed for a language with grammar (all languages have grammar), but this means proto-sapiens, the language that evolved before modern humans, would have required recursion, moving consciousness back to at minimum 50K years ago.

Please clarify the differences you're making between the terms, or clear it up for me some.

Additionally, why would you need to be able to introspect in order to do the Pythagorean theorem or learn how to count? None of these seem to require introspection of any sort, only access to outside information.

And finally, I think that your view that the beginnings of recursion requiring a few "broken eggs" seems incorrect. Schizophrenia seems to be mainly related to overactive agent detection (seeing agents, such as gods or motivations or patterns, where none exist) which doesn't seem to have a strong bearing on introspection. The "mirror test", the ability to recognize oneself in the mirror, seems a more reasonable step towards introspection, that is, identifying oneself as separate from other members of the group.

And yes, we teach children to introspect all the time, to stop and think about their actions.

Expand full comment
author

Fren,

Yeah I think it spread about 15kya. I'm treating self-awareness and recursion as a package, and that must have started earlier, at least sporadically. Venus statues 40kya, counting stick 45kya (in Africa); both indicate recursion. But importantly the theory is that recursion was not the base state as humans left Africa. I link a couple of linguists who also say language was not recursive when we left Africa. One of them I honestly find very silly. He says, with basically no evidence, that there was another OoA event 20-10 kya and recursive language diffused from Africa. But he's a real linguist who's published a bunch on Zulu and such. He has some reasons for thinking recursion is a recent addition. The other I find more sophisticated, and he basically says that there is very little evidence of complex social structure/thinking and our skulls were changing shape; recursive language seems like a good explanation for those changes circa 10 kya. He really downplays the phenomenological aspect of recursive language.

Anyway, I'm not alone on the recursive language date. And indeed part of the project is just to ask "how early can we go, deferring to experts" when so many have asked "how far back can we go"

>additionally, why would you need to be able to introspect in order to do the Pythagorean theorem or learn how to count?

It's not about introspection, it's about recursion (which is also required for introspection/self-awareness). The connection between recursion and abstract thought + counting I'm taking from Corballis in his book The Recursive Mind. It's also strange that the Piraha, who don't speak a recursive language, apparently could not learn to count in 8 months of lessons (and 300 years being swindled by Portuguese traders).

>And finally, I think that your view that the beginnings of recursion requiring a few "broken eggs" seems incorrect.

This is actually a point I'm confident about! How would schizophrenia be a phenotype before the construction of the self (which I take to be recursive). More generally, recursion is a load-bearing infinite loop. At least computer science says that is trouble.

>The "mirror test", the ability to recognize oneself in the mirror, seems a more reasonable step towards introspection, that is, identifying oneself as separate from other members of the group.

Lots of animals can pass the mirror test but show no signs of other recursive thinking. I think some fish pass? I guess I'm a bit in the minority but I find it hard to believe even dogs have an inner life.

Expand full comment
Apr 9·edited Apr 9Liked by Andrew Cutler

15kya that's too tight given the inter-glacial isolation of Tasmania | lutruwita. I suspect the selection environment for this is more social, and the delay is in integrating recursions across mental faculties (tools and roles) within a nuturing society that can maintain them. And thus cultures as transmitted across massive geographies. I guess I am a proponent of the long ages of the Paleolithic as it origin and the take-off as a could-have-been failure too (like the others before it we have no clue about). As such your origin story is the latest possible, not the first opportunity. Some re-phrasing might be in order.

Expand full comment
Aug 23, 2023Liked by Andrew Cutler

I'm reading The Language Game by Morten H. Christiansen and Nick Chater, and after reading about EToC over the past couple of days I got to the part in the book where the authors mentioned that Chomsky put forward recursion as the precondition for language, the book refutes this by bringing up the Piraha and also expanding on the point that recursion is hard.

The authors mention center embedding recursion, where sentences any more layered than "The cat [that the dog scared] ran away" make our brains melt. A footnote does mention linguistic tail recursion, exemplified in the nursery rhyme "The house that Jack built", which we can handle really well.

I didn't know about linguistic tail recursion, and I found it interesting how this mirrors the computer science method of tail call optimization so that the stack doesn't explode during recursion.

At a later point the book brings up the FOXP2 gene, which a while back was touted as the language gene, and this gene is essential for nerve cell growth. The book mentions an interesting experiment where researchers bred mice with the human version of the FOXP2 gene and the result was mice with faster learning through faster chunking of sequential information.

This makes me wonder if the FOXP2 gene codes for recursion, what most animals have is primitive recursion, and humans have the unbounded recursion, which expands available recursive constructions given that primitive (or iterative) recursion cannot describe all the functions that full recursion can.

It's like most animals are stuck with thinking that allows only for loops where the number of iterations is known in advance but humans have access to the `while` loop and can ponder infinite sequences.

Expand full comment
author

Refute, is a pretty strong word for what the Piraha can do with recursion. My understanding is something like 90% of linguists still think recursion/embedding are fundamental.

So there's my appeal to (90% of) authority. But the Piraha are actually super interesting for EToC. If self-awareness diffused, then maybe you could find some hold-out tribe. The place to look would deep in the Amazon. It's not just recursive language that Piraha don't have; it's anything that requires recursion. He tried teaching them to count for months. They enjoyed coming to his counting classes. But nobody could learn to count at all. The Grammar of Happiness documentary emphasizes that they are always in the moment, suggesting less mental time travel. They borrowed their pronoun system from a neighboring tribe, and tend not to use pronouns. They don't have creation myths, or any cultural memory more than a couple generations. This is all from Everett's book "Don't Sleep, There are Snakes." The one ritual he concedes they do have is to dance with venemous snakes.

I don't know what to make of all this, but it doesn't seem evidence _against_ EToC or the Snake Cult.

>This makes me wonder if the FOXP2 gene codes for recursion, what most animals have is primitive recursion, and humans have the unbounded recursion, which expands available recursive constructions given that primitive (or iterative) recursion cannot describe all the functions that full recursion can.

I'm open to their being degrees, and truthfully don't know what to make of animal consciousness. What is the best way to separate consciousness from higher-order self-awareness?

But as for a single gene, complex traits usually implicate hundreds or thousands. I forget the exact statistic but if you chop the genome into 1,000 (100?) chunks each will still have a statistically significant effect on schizophrenia. Schizophrenia is often implicated in the evolution of language and consciousness, so it stands to reason all three have that degree of polygenicity.

Expand full comment
Aug 23, 2023Liked by Andrew Cutler

I think the book just argues against Chomsky's idea that recursion was a one time mutation that made humans human and gave them language.

But I don't think this has any direct bearing on EToC. As I understand it, humans have had language since 100k (and hominids even earlier) but lacked that certain something which made humans modern behaviorally, which EToC aims to explain.

Now, if there actually is a hard line between primitive and unbounded recursion (in the sense of what an animal equipped with either can do), and the latter is much more difficult to implement properly so that the human brain can run it efficiently, then to me it sort of makes sense that it took a long time for the efficient implementation to be found and diffused. I can definitely see women having developed this optimized recursion first, and given the arguments and the mechanisms you described, I can also see how they could have sped up the spread of this optimization throughout men, or rather, selected for it.

And I kind of see the point you make about schizophrenia. This thing, recursion, type of recursion, or whatever it turns out to be, offered advantages that even such a negative drawback couldn't offset.

Also, when I read the Wikipedia page for the Piraha I did see a couple of points that I could see as directly linked to EToC, or that can be viewed in its context:

- they had no concepts of drawings, reminds me of art before the whole snake tripping rituals

- the highest Piraha value is no coercion, one does not tell other people what to do, it reminds me of the Golden Rule you mentioned in PFP

- the strongest point, which you mentioned, is that their entire pronoun set might have been borrowed recently, you should put that in bold in any next post about pronouns, it was a really cool thing to learn

I don't think the Piraha lack anything that other humans have, they're just as genetically "modern" as we are. I think it's just memetic persistence of that old mode of being. The fact that the adults could not learn to count but children did could be related to how we cannot develop language after a certain age if we're not exposed to it.

I didn't say it in the last post, but thank you for this theory, it's been an exciting read, and I can almost picture what a hypothetical Victorian would have felt the day Darwin's theory came to public light. What is this madness? Could it be true?

Expand full comment
author

>they had no concepts of drawings, reminds me of art before the whole snake tripping rituals

Yeah, it's one of the most bizarre things because it is thought to be a human universal, as are creation myths and recursion. I'm really not satisfied with Everett's claim that they have a cultural aversion to recursion (he calls it the Immediacy of Experience Principle), and then refuse to practice anything that requires recursion (grammar, counting, thinking about the future). They must have very good intuition for this class of abilities that it took scientists thousands of years to identify. It seems more likely they have more limited recursive abilities, and thus find counting and Portuguese grammar more difficult (or even impossible) to learn. Very hard to explain it *just* with memetic differences. And in fact he refused to do genetic sequencing as to avoid the appearance of evil.

>As I understand it, humans have had language since 100k (and hominids even earlier) but lacked that certain something which made humans modern behaviorally, which EToC aims to explain.

Very hard to estimate, and they usually do so with behavioral complexity. So it could be 100k, but other people say 2 million, and still others say 10k (for full recursive/embed language).

EToC line is that the first recursion was self-awareness, and once that was in place recursive grammar could be invented.

>The fact that the adults could not learn to count but children did could be related to how we cannot develop language after a certain age if we're not exposed to it.

According to Everett nobody could learn, even kids! This is over like 8 months of classes. Those kids must be really clever to figure out that the taboo about recursion so young.

I'd definitely like to write a post on the Piraha, but it's not politically palatable so on the back burner for now.

>I didn't say it in the last post, but thank you for this theory, it's been an exciting read, and I can almost picture what a hypothetical Victorian would have felt the day Darwin's theory came to public light. What is this madness? Could it be true?

Thanks! Spread the word if you know communities that would be interested and can give feedback!

Expand full comment
Aug 24, 2023Liked by Andrew Cutler

Hm, I just went with the Wikipedia description, which says that some kids did manage to make progress but were taken out of the classes by their parents for it.

Since the description doesn't go into detail regarding specific ages I can't be sure, but I think a reasonable base assumption would be that there is a specific interval in which humans can pick up on the cultural package that allows them to use unbounded recursion, given, of course, that they have the genetic mechanism to allow for it.

If I'd have to bet I'd say there's only an extremely slim chance, or a mathematical improbability, that humans without the genetic mechanism for unbounded recursion have survived to this day. But not slim enough that I wouldn't be curious to want to know about it if it was testable.

I understand why you'd say that writing about the Piraha is not politically palatable, since perceived differences between humans have been used, at least superficially, as the basis for horrors, but I think that we do have a discourse environment where we understand that whatever observations science may make, that does not imply that empathy and valuing life go out the window. Sometimes I have difficulty separating, at least at a root motivation level, the people who engaged in those horrors based on pseudoscience and the people who object to science looking into places they find uncomfortable.

I am curious why the women of the Piraha didn't jump-start recursion use in the tribe. Too small a group with too little interaction with others, added to a no coercion environment with assured sustenance?

An interesting tidbit; I asked ChatGPT to consider the case of two early humans, one capable of thinking using only primitive recursion, the other with unbounded recursion. I then asked it to come up with scenarios where the two would differ. I could only come up with an example where they are asked to answer how many glasses worth of water would they have when given glasses that are full based on the 1 + 1/2 + 1/4 ... series that converges to two. The one with primitive recursion cannot even begin to ponder the question. Basically I asked it to come up with a similar scenario but set in the environment of early humans and comment on the differences.

One of the examples it provided was that primitive recursion humans would have no plan for preserving the food or considering future availability, which instantly reminded me of this bit from Wikipedia: "Piraha have ignored lessons in preserving meats by salting or smoking".

This is not very rigorous, true, and it hinges on there being an actual translation of the computer science concepts and their implications to human terms. But it is so appealing, primitive recursion is bound by P in terms of what it can express and always terminates, while full recursion I think goes even beyond EXP, not to mention it can bump into the problem of decidability.

Expand full comment

Schizophrenia can be a phenotype because if you have overly sensitive agent detection, you get fewer false-negatives about agents. That is, you notice EVERY tiger trying to eat you and run away (true positive), but sometimes you run away when there is no tiger (False positive).

Second, in the bicameral mind arrangement you talked about before, all humans were acting under a kind of schizophrenia. And I didn’t mention it previously, but it is true that even mow human minds are not fully integrated. There are different mechanisms and attitudes that dominate in different circumstances.

Expand full comment
Apr 13, 2023Liked by Andrew Cutler

Great stuff. Will read again in a few days' time to build on my understanding of it. Will then read again in a few more days to build on my understanding of it. Will then...

Expand full comment
Apr 13, 2023Liked by Andrew Cutler

Interesting reading.

Curious where the rest of this series will go.

In the meantime, I thought you might like this (not mine, but it feels tangentially related): https://praxtime.com/2023/04/08/intelligence-is-prediction/

Expand full comment
author

Intelligence as prediction means that amoebas or plants are intelligent, just less so. To me it seems they, and even dogs, are not on the same scale as humans. That humans have an intelligence of an entirely different kind.

Abstract thinking does not seem to be continuous, it just leaps out of the archeological record. Corballis is a linguist who needles Chomsky about being almost creationist with his theory that recursion evolved with a single gene 50k years ago. In his Ted Talk there is this little aside where he mentions archeologists also tend to believe abstract thinking came almost out of nowhere. It looks like a phase change.

Expand full comment

thanks for the refresher

Expand full comment