Happy new year, everyone! Remember in that "Altaica" languagehat thread back in April when I mentioned that one of James Unger's students, Alex Ratte, was allegedly preparing some sock-offknocking new arguments for a genetic relationship between Japanese and Korean? It turns out that one Alexander T. Francis-Ratte posted several papers along these lines to academia.edu in 2015. Here are a few words about a couple.

"Importation or inheritance? Thoughts on the Japanese lexicon" is a short paper about the "bifurcation of Korean-Japanese lexical matches." The argument is that some "lexical matches" can be organized into regular groups that make them look like cognates, while others exhibit no such regularity and so look more like loanwords. This, AFR argues, is inconsistent with an Alexander Vovinesque hypothesis of "extensive borrowing," and in any case, how was this mass borrowing supposed to have taken place when, "as Unger (2009: 16) points out, there is 'no compelling historical evidence that Korean and Japanese stood on equal sociolinguistic footing for a sustained period of time'"?

I suppose the obvious weakness here is the lack of even a semi-objective heuristic for sorting words into "cognate" and "loanword" piles. It isn't obvious on the face of it why the non-corresponding segments in OJ poye "howls" 〜 MK pullu "calls out" are qualitatively less problematic than the ones in OJ kusiro "bracelet" 〜 MK kwusul "jewel," for example. I think that AFR's argument against mass borrowing on practical grounds is more convincing.

Much longer and meatier is "Morpho-lexical Evidence for Proto-Korean-Japanese". Here AFR makes two specific proposals:

a functional element *(w)o- that reveals striking correspondences in Japanese and Korean noun-modifying structures, and a verb *pә- that shows that identical verb-compounding structures exist in both languages.

This one really deserves a more expert treatment than I can offer, but I was not overwhelmed. It's strange, for example, that AFR argues for his "verb *pә-" without addressing Whitman and Frellesvig's more recent (2014) paper about /e/ 得 as OJ lower bigrade formant ("The Historical Source of the Bigrade Transitivity Alternations in Japanese"). This contains a more sophisticated analysis of /e/ than the simple "transitivity flipper," and offers some support for AFR's theory in analyzing bigrade intransitives as the result of "suppressing the nonagentive experiencer/goal in secondary predicate + -e-". Sounds very compatible with a hypothetical verb meaning "watch things (inexorably) happen."

On the other hand, F+W 2014 also observes that:

[c]omparative Ryūkyūan evidence indicates that this pattern may not be reconstructible to proto-Japanese. For example, while transitive yak- 'burn (tr.)' and tak- 'burn/cook (tr.)' have corresponding verbs in Yonaguni (Ikema 2003) and prewar Yaeyama (Miyara 1930), their bigrade intransitive counterparts appear to be unattested

... which would obviously be a problem for an argument based on reconstructing this pattern for proto-Korean-Japanese (let alone proto-Japanese!).

Anyway, these papers are definitely worth reading for those craving new linguistic arguments in the are-Japan-and-Korean-related wars. (Also of possible interest: Vovin's "Out of Southern China?", which proposes some "lexical parallels between Japonic and Tai-K(r)adai" and even a few between Old Japanese and the language of Chu 楚.)