On Tue, 29 Jan 2019 at 14:44, Aidan Aannestad <[log in to unmask]> wrote: > > Japanese arguably doesn't have an 'infinitive' though, or at least > anything analogous to an IE infinitive. Do you mean 'base uninflected > form'? That form is 100% valid as the verb of a main clause. That would seem to make doing a comparable test rather difficult--but English has similar issues, and while they make it harder to collect significant amounts of usable data, it's not impossible. The base uninflected form of English verbs is also a valid finite form, but you can still do the test on English L1 learners by identifying places where a different form *should've* been used in adult speech, but the child substituted the base form instead. -l. > On 2019/01/29 15:42, Logan Kearsley wrote: > > On Tue, 29 Jan 2019 at 13:28, Aidan Aannestad <[log in to unmask]> wrote: > >> It's a bit surprising to me to hear that about verb inflections - > >> Japanese is /very/ pro-drop, and yet has no agreement marking on the > >> verb at all. > > That's a good point. I wonder what the infinitive usage statistics are > > for children acquiring Japanese as their L1.... > > > > -l. > > > >> On 2019/01/28 18:22, Logan Kearsley wrote: > >>> This is a fairly old paper, but I only recently discovered it, and > >>> found it terribly interesting: > >>> http://web.uconn.edu/snyder/papers/EB_WS_FASL.pdf > >>> > >>> For one thing, it introduced me to an interesting developmental test > >>> for the "pro-drop" classification: If kids use a lot of infinitive > >>> forms before learning proper finite verb inflections, your language is > >>> *not* actually pro-drop. No particular reason for why that should be > >>> is presented, but it kinda makes sense to me: if the adults around you > >>> don't actually drop subject pronouns very often, then inflectional > >>> information in the verb is less important; but, if they do, verbal > >>> inflection becomes much more important for encoding thematic subjects, > >>> so it makes sense that it would be learned faster. > >>> > >>> Possibly of more direct relevance to conlangers: a distinction is made > >>> between dropping expletive subjects vs. thematic subjects (where only > >>> dropping thematic subjects "counts" as pro-dropping). I think this > >>> intuitively makes sense to a lot of people (i.e., all of the > >>> conlangers who include something in the grammar to the effect of "if > >>> you want to say 'it's raining', my language leaves out the 'it', > >>> because it's dumb."), but I haven't before seen it formally > >>> distinguished as a separate syntactic parameter. > >>> > >>> -l.