Zompist Bboard Again

I seem to remember reading somewhere that there is archaeogenetic evidence of two prehistoric migrations from the Near East back to North Africa, one near the end of the last Ice Age, and one in the Neolithic, but I can't find any reference. Can anybody around here help me find it? Of course I tried Google but it was no help (as everybody knows, Google is quite good for finding where you can buy something, but sucks at anything else).

Thank you in advance.

Have you tried asking ChatGPT?

I asked it and it did come up with one study:

"Yes, you're referring to the research on ancient DNA that suggests multiple migrations from the Near East to North Africa. Studies have shown genetic connections between populations in these regions during different periods, such as the migrations you mentioned. One notable study you might be interested in is "Ancient genomes from North Africa evidence prehistoric migrations to the Maghreb from both the Levant and Europe" by van de Loosdrecht et al. (2018). This paper discusses the genetic evidence supporting migrations during both the late Pleistocene and the Neolithic period."

ChatGPT is a bad research tool because it doesn't know anything and will fabricate citations if you ask it for citations. van de Loosdrecht et al. (2018) happens to exist, but it's not titled "Ancient genomes from North Africa evidence prehistoric migrations to the Maghreb from both the Levant and Europe" - that's a different paper by Fregel et al..

Thank you, that looks useful!

The migration of Y-DNA haplogroup I may be the earliest back-migration that leaves large-scale evidence in modern populations of North Africa.
https://genome.cshlp.org/content/18/5/830.full.pdf
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1181996/
I haven't found anything specifically dating the migration into Africa, but I'm sure you can find it if you dig.

Ketsuban wrote: ↑Sun May 19, 2024 9:42 am ChatGPT is a bad research tool because it doesn't know anything and will fabricate citations if you ask it for citations.

If you want to search for papers, you have to use the Consensus GPT: https://chatgpt.com/g/g-bo0FiWLY7-consensus I haven't looked at the results, but this is what you get when you enter the query: https://chatgpt.com/share/eda883c9-f13d ... ca84b6e411

If I have to go to academia.edu and type "African pleistocene migration" to get the paper, how is this saving me any steps? That's what I was doing before.

Glass Half Baked wrote: ↑Mon May 20, 2024 5:56 pm If I have to go to academia.edu and type "African pleistocene migration" to get the paper, how is this saving me any steps? That's what I was doing before.

Consensus GPT will give you names of the papers by a semantic analysis of the abstracts in case keywords aren't cutting it. It has been useful for highly granular searches that aren't captured by keywords. You can also customize the output with natural language queries like how far from your search parameters is still ok, how many papers you want to see, and so on.

OK, but again, whatever mining it's doing of abstracts, I have to replicate it to get the actual paper. It's basically performing a search, then giving me hints about what I should be typing into the search bar in my other browser tab. No link, no title, not even a DOI! Why did they design it like this?

Glass Half Baked wrote: ↑Mon May 20, 2024 8:28 pm OK, but again, whatever mining it's doing of abstracts, I have to replicate it to get the actual paper. It's basically performing a search, then giving me hints about what I should be typing into the search bar in my other browser tab. No link, no title, not even a DOI! Why did they design it like this?

Only the shareable link is like this. For me, clicking the first link sends me to this page: https://consensus.app/papers/pleistocen ... ce=chatgpt