Can someone with a background in this area speak to why the DNA data is taken to be a strong signal about the origin of language families? Like, clearly when people move around they generally take their language with them, but trade, war/political power, cultural exchange etc also move languages around.
This was much less common in the pre-modern era. One of the largest learnings of the latter 20th century in this field is that when language and culture are moving, it typically means a previous population was displaced, not merely that the same people adopted a new culture.
My comment was perhaps overgeneralizing specifically from the early neolithic transition in Europe more broadly than I should have been. One example of the work I am discussing is the work of Cavalli-Sforza (ie https://www.pnas.org/doi/10.1073/pnas.94.15.7719). My understanding of the history of academic work on this subject was the prior to the 1970s or 80s or so, it was largely thought the spread of agriculture did not imply genetic population replacement and subsequent genetic research has since proven that is false.
I don't think the article you've linked actually disagrees with me. Excerpting briefly
> Essentially, in the Pueblo Southwest, language became partially decoupled from genes and culture, probably because language is a stronger convention (Young 1966) than are the other two, and can therefore better withstand the blending forces discussed below. A second force for creating zones of cultural similarity, more applicable to the case of phylogenetically distinct groups considered here, is through mixture of groups via movement of people (“demic” diffusion) that also entrained sharing of culture. To the extent that this describes the Pueblo case, once again, such mixing could not to have been so extreme as to undermine the linguistic differences that still survive. Recent research (e.g., Kandler 2009) provides some guidance as to the factors determining which language will prevail when linguistically different groups come into contact. The relative sizes of the populations, their relative status, and the duration of the flow are all relevant. To the extent that emigrating populations were relatively low in status, small in number, and arriving sporadically, it is more likely that they would adopt, rather than displace, the language of any group they were entering. Of course cultural similarities among groups can also emerge without population movement by copying neighboring groups, a process that anthropologists once simply called diffusion (e.g., Sahlins 1960). In the terms used by Collard et al. (2006), both movement of people among groups and this horizontal cultural transmission among groups are blending processes, as opposed to the branching processes discussed in the previous paragraph.
To me, this article seems to be attempting to explain the cultural similarity between (at least originally) disparate linguistic Pueblo groups (that largely, afaict from the article, share common genetics). In the paragraph above, you can see that the author is describing population migration and displacement as one of the key ways for cultural/linguistic transfer (especially when the populations moving are large) that 'can also' emerge without displacement, but I understand the article to be describing this as a minoritarian current - an exception to the broader pattern of pre-modern linguistic & cultural spread through displacement.
As an aside, Cavalli-Sforza was a legendary figure, but a lot of this stuff has benefited from powerful new tools like eDNA and effective aDNA that he never had much of a chance to speak on.
For Kohler's paper, I'd point to the summary as a better example of what I was trying to communicate:
Evidently, though, powerful blending forces were also at work. These included local exogamy, which, on the borders of linguistic groups, led to blending, movement of traders among increasingly sedentary groups, movements forced on populations increasingly reliant on agriculture by changing climates, and movements forced on increasingly large and sedentary groups by anthropogenic depression of local resources. One of my goals in this article has been to add convergent evolution to this list. This would have been driven by adaptive considerations, possibly bootstrapped by additional blending processes such as indirectly biased cultural transmission across groups (Mesoudi and O’Brien 2008). Together, these blending forces created the Pueblo culture area and Sprachbund.