[Lingtyp] Grammaticalization of past/resultative meaning from "stay"
wkj at uoregon.edu
Mon Mar 6 00:38:25 UTC 2023
There is a construction in Horokoi (a.k.a. Wasembo, [gsp], part of the Madang branch of TNG) in which a clause chain with the final verb "stay/exist" can have various past/resultative-like meanings. I am wondering where else such a construction has been found.
The form is: [V-SR stay-TAM], where SR means switch reference marking (same-subject or different-subject). With same-subject marking, it literally says "I [V] and I stay"; with different-subject, it says "I [V] and it (impersonal) stays".
So far I have found the following meanings for the construction. The different-subject marking tends to be associated with more distal meanings (past, far past, anterior).
* literal (he built a house and it stayed [didn't fall down])
* stative (the food is dry, lit. it dries and it stays)
* copula/stative (you are like me, lit. you become and you stay)
* Note that this meaning only occurs when the first verb is "become". It does not mean "you became like me" (eventive).
* resultative/stative ([you hit it and] it is broken, lit. it breaks and it stays)
* past (I went, lit. I go and I stay)
* far past (they [ancestors] got salt from trees, lit. they take and it stays)
* anterior (I had said it to you, [then something else happened], lit. I say and it stays)
I have been thinking that this is unusual because "stay" as an auxiliary usually grammaticalizes into continuative rather than past/resultative. Heine & Kuteva (2002) mention "sit" > copula, but not this path of "become and stay" > "become-past" > copula, nor any cases of "stay" (or similar) to these past-like meanings.
I've been attributing this pathway to the sequential semantics of the clause chaining construction (Horokoi does not mark simultaneous vs sequential in medial verbs, as far as I know). Thus, the sequence "I [V] and (then) I stay" implies that V is no longer happening and I am staying in whatever state endures at the end of V's action. But perhaps this is not right, and I received a comment that this implicature need not hold for the literal meaning.
I have received comments that similar constructions are found in Dani languages, Malay, and some others. I found mention of something very similar in Mian by Fedden (2020). I have found no cognate constructions or comparative evidence to shed light on this for Horokoi (presumably Mian constitutes a parallel innovation because of the vast time depth separating Madang from Ok).
Please let me know if you have seen something like this or if you know of references about this grammaticalization pathway, thank you!
Wesley Kuhron Jones
Ph.D. student, University of Oregon
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Lingtyp