Agouti historiography: the problem of widespread lexical forms and deep linguistic history in Central and South America

Throughout Central and South America, similar words for salient rodent species, the agoutis (Dasyprocta spp.), reconstruct to some of the most ancient known languages families in the species’ range, including Chibchan (*’kuri), Cariban (*akuri), Tupian (*akutˀi), and others. Unless these are dismissed as mere chance, this makes for as interesting as problematic evidence for deep linguistic history: what mechanisms account for the striking similarities that must go back far into prehistory? One possibility is that the words are lexical evidence for very old genealogical connections. While this is a possibility worth bearing in mind also in light of current and past suggestions that involve some of the relevant groups, we cannot conclude that this is indeed the case on the basis of the available evidence. Another, and perhaps less costly, hypothesis is very old borrowing involving relevant proto-languages or their ancestors However, is this a plausible scenario? On the basis of a large set of data featuring more than 400 languages from Central and South America, I show that neighboring languages have relatively frequently borrowed agouti-words from one another, involving reflexes of the above forms, but also others. This shows that it is natural to expect borrowing my. In either case, the similarities likely demonstrate very old interactions between the mentioned lineages, something with interesting implications on homeland hypotheses and scenarios of early spread.