The general theory seems coherent, but the only reason to shoehorn Eve into it is Genesis. The Bible isn't particularly old, and Eve seems even less likely when considering Indian or Sumerian myths.
Yeah, onus on me to show that Genesis is doing much better than chance, and additionally that the elements that are hits are shared in other traditions.
I think, particularly, we don't know how far the oral tradition in the Bible stretches back. We do have pretty good evidence that _some_ stories go back about 13,000 years, which I'll cover next.
Just published the piece about snakes. I think it's another thing that Genesis might have gotten right
Also, coming back to the Sumerian, Shamhat in the Epic of Gilgamesh is a fine Eve character. It is she that civilizes Enkidu. After their liaison, he is no longer one with the animals.
What are the reasons to assume there was *one* transition - language, consciousness, recursion, selfhood, theory of mind, "unicameralism", whatever else, all a package deal? Can we easily discard the possibility that (for instance) language and the gods arose 12000 BC, and then theory of mind and "unicameralism", and the gods died out, 1200 BC? And maybe something happened 50000 BC as well, to produce the diagonal lines in the rock. (Furthermore, ought we suppose the transition or transitions were necessarily *discrete*?)
The best bet for a discrete transition is recursion, which many argue is a package deal including self-awareness, ability to count, language with grammar, mental time travel, ability to imagine, and sometimes even subjectivity. It's definitely an assumption that becoming human was a) recursion b) a package deal. But I'm at least in good company. If you haven't I recommend reading my post Deja-you, as that makes the case for recursion.
One additional reason to want a single transition contained all is that it's really hard to argue for global diffusion on these time frames. Best to just have to do that once.
At present I'm also trying to build the simplest model possible. It's obviously not fleshed out yet, and I'm still curious how much the recursion-maximalist position can explain.
Thanks, I'd only read the Snakes post previously. Having now read it (a couple of days ago) the "package deal hypothesis" is still basically where it was at "plausible" for me though. (Just as a random point that happens to still be in-cache, I'd be surprised if many/most social animal species don't behave in accordance with something like the golden rule, even if they can't articulate it.)
More likely, consciousness is related to Openness/Intellect/Imagination, as that's the factor most clearly and obviously involved with thought, reflection, and engagement with philosophy, emotion, and the human experience. Arthropods behave prosocially and even self-sacrificially; no consciousness is necessary for Agreeableness/Alpha.
Without doing second-order statistics on survey data, PC1 explains ~80% of the variance. Not only that, but Openness takes as much variance from PC1 as PC5. It is one of those quirks of varimax rotation that it becomes an important factor (and only very inconsistently so, many studies produce different fifth factors).
The % variance is important from the evolutionary perspective. How does the fifth factor, which explains a very small portion of the variance, have to do with what made us human?
No, not in my own private datasets nor any published dataset I have ever encountered. In 1997, Digman found 35% of variance for factor I, and Rushton and Irwing in 2008 did find as much as 45% of the variance in factor I. But recent survey of mine with n > 1000 has % of Var. 25% for the first unrotated factor. So percent variance may indeed be "very important," but even if so, the importance you attribute to the unrotated first factor of personality is out of proportion to the variance it can be verified to explain.
Well, one caveat! If you do PCA on the correlation matrix, it will be under 50%. But if you do it on the covariance matrix, it really is 80%. This is because the latter does not remove the mean or norm each dimension to be unit length. A priori, no real reason to choose on or the other. Forgive me for only citing the one that makes a better case for me :)
Similarly, if you have word vectors instead of survey results, then PCA on even the covariance matrix will yield about 80%. One has to ipsatize the data (which to me is not justified), to get that down to 30%.
So with very different data sources (surveys and vectors) the "naive" processing yields a massive leading eigenvalue. It's also important to note that the processing decisions described do not change the direction of PC1 at all. The two results correlate at >0.95 whether the covariance or correlation matrix is used.
"Indeed, applied to history it has essentially zero predictive power."
Oh, bullshit. The theory is based on the transition between orc and socdem (see my most recent post). It correctly predicts the origin of tableware. And if we were all JavaScript Canvas all along (and we definitely seem to be after 1950), it would explain the orc gods as being real figures existing outside the brain. The most plausible way for orcs to have transitioned to SocDems was rising food intake per capita due to rising infectious disease burden -a shift from a malthusian world to a clarkian one.
"If you say it affected language, then show me the language family it founded."
Aramaic.
"We would see a phase change in creativity, planning, and searching for meaning."
We did -ancient Greece, India, Persia, and Judah (all with bad writing directions) and China (with a good writing direction, but oracle bones are fundamentally socdemmy rather than orcish). Orcs are not bad at trig!
The general theory seems coherent, but the only reason to shoehorn Eve into it is Genesis. The Bible isn't particularly old, and Eve seems even less likely when considering Indian or Sumerian myths.
Yeah, onus on me to show that Genesis is doing much better than chance, and additionally that the elements that are hits are shared in other traditions.
I think, particularly, we don't know how far the oral tradition in the Bible stretches back. We do have pretty good evidence that _some_ stories go back about 13,000 years, which I'll cover next.
Just published the piece about snakes. I think it's another thing that Genesis might have gotten right
Also, coming back to the Sumerian, Shamhat in the Epic of Gilgamesh is a fine Eve character. It is she that civilizes Enkidu. After their liaison, he is no longer one with the animals.
What are the reasons to assume there was *one* transition - language, consciousness, recursion, selfhood, theory of mind, "unicameralism", whatever else, all a package deal? Can we easily discard the possibility that (for instance) language and the gods arose 12000 BC, and then theory of mind and "unicameralism", and the gods died out, 1200 BC? And maybe something happened 50000 BC as well, to produce the diagonal lines in the rock. (Furthermore, ought we suppose the transition or transitions were necessarily *discrete*?)
The best bet for a discrete transition is recursion, which many argue is a package deal including self-awareness, ability to count, language with grammar, mental time travel, ability to imagine, and sometimes even subjectivity. It's definitely an assumption that becoming human was a) recursion b) a package deal. But I'm at least in good company. If you haven't I recommend reading my post Deja-you, as that makes the case for recursion.
One additional reason to want a single transition contained all is that it's really hard to argue for global diffusion on these time frames. Best to just have to do that once.
At present I'm also trying to build the simplest model possible. It's obviously not fleshed out yet, and I'm still curious how much the recursion-maximalist position can explain.
Thanks, I'd only read the Snakes post previously. Having now read it (a couple of days ago) the "package deal hypothesis" is still basically where it was at "plausible" for me though. (Just as a random point that happens to still be in-cache, I'd be surprised if many/most social animal species don't behave in accordance with something like the golden rule, even if they can't articulate it.)
Diffusion is a good point.
More likely, consciousness is related to Openness/Intellect/Imagination, as that's the factor most clearly and obviously involved with thought, reflection, and engagement with philosophy, emotion, and the human experience. Arthropods behave prosocially and even self-sacrificially; no consciousness is necessary for Agreeableness/Alpha.
Without doing second-order statistics on survey data, PC1 explains ~80% of the variance. Not only that, but Openness takes as much variance from PC1 as PC5. It is one of those quirks of varimax rotation that it becomes an important factor (and only very inconsistently so, many studies produce different fifth factors).
The % variance is important from the evolutionary perspective. How does the fifth factor, which explains a very small portion of the variance, have to do with what made us human?
> PC1 explains ~80% of the variance
No, not in my own private datasets nor any published dataset I have ever encountered. In 1997, Digman found 35% of variance for factor I, and Rushton and Irwing in 2008 did find as much as 45% of the variance in factor I. But recent survey of mine with n > 1000 has % of Var. 25% for the first unrotated factor. So percent variance may indeed be "very important," but even if so, the importance you attribute to the unrotated first factor of personality is out of proportion to the variance it can be verified to explain.
Well, one caveat! If you do PCA on the correlation matrix, it will be under 50%. But if you do it on the covariance matrix, it really is 80%. This is because the latter does not remove the mean or norm each dimension to be unit length. A priori, no real reason to choose on or the other. Forgive me for only citing the one that makes a better case for me :)
Similarly, if you have word vectors instead of survey results, then PCA on even the covariance matrix will yield about 80%. One has to ipsatize the data (which to me is not justified), to get that down to 30%.
So with very different data sources (surveys and vectors) the "naive" processing yields a massive leading eigenvalue. It's also important to note that the processing decisions described do not change the direction of PC1 at all. The two results correlate at >0.95 whether the covariance or correlation matrix is used.
"Indeed, applied to history it has essentially zero predictive power."
Oh, bullshit. The theory is based on the transition between orc and socdem (see my most recent post). It correctly predicts the origin of tableware. And if we were all JavaScript Canvas all along (and we definitely seem to be after 1950), it would explain the orc gods as being real figures existing outside the brain. The most plausible way for orcs to have transitioned to SocDems was rising food intake per capita due to rising infectious disease burden -a shift from a malthusian world to a clarkian one.
"If you say it affected language, then show me the language family it founded."
Aramaic.
"We would see a phase change in creativity, planning, and searching for meaning."
We did -ancient Greece, India, Persia, and Judah (all with bad writing directions) and China (with a good writing direction, but oracle bones are fundamentally socdemmy rather than orcish). Orcs are not bad at trig!
I'll stick with Dennett's vastly more sensible explanation: https://www.youtube.com/watch?v=IZefk4gzQt4