Events / MindCORE Seminar: Carlos Santana

MindCORE Seminar: Carlos Santana

February 7, 2025
12:00 PM - 1:15 PM

111 Levin Building

We will also stream this seminar via Zoom.
For the link, please email: pennmindcore@sas.upenn.edu

 

Carlos Santana
Department of Philosophy
University of Pennsylvania

 

Brown bear, brown bear, what do you tree? Discordance in linguistic phylogenetics

 
What do language trees mean?
Easy answer: they’re models of linguistic evolutionary history

 

In this talk I complicate that easy answer by exploring two issues.
          First, the ways language change can fail to be treelike.
          Second, discordance, a phenomenon where different data generates non-equivalent trees

 

To help us understand what these facts mean for what language trees mean, I explore how phylogeneticists have handled these issues in biology. I argue that language change can more closely resemble cloudlike microbial evolution than the tree-like phylogenies of multi-cellular sexually-reproducing organisms. Regarding discordance, linguists can follow recent phylogeneticists in taking discordant trees to represent distinct but compatible evolutionary histories—this is where the brown bears come in.

 

I draw from three lessons from this analysis.
          1 – Linguistic phylogenetics, is exciting and informative, but is a compliment to rather than a replacement for traditional methods in historical linguistics like the comparative method.
          2 – We should be pluralists about language trees. When two trees disagree, it’s not necessarily the case that only one is right. They could both adequately represent part of the evolutionary history of a language.
          3 – The inferences we can use language trees to make are highly contingent on the kind of data used and how the phylogenetic algorithm handles non-treelike changes. This means that, for instance, a tree might be a good representation of a real evolutionary history, but not good evidence for the geographic origins of a language family. Researchers who want to use a language tree as evidence for a particular historical claim need to argue that their choice of data type and phylogenetic algorithm are the right ones for that type of historical claim.

 

A pizza lunch will be served. Please bring your own beverage.