07 February 2009


In field methods class we’re working with Paulina Yourupi on Pollapese, the language of Pollap Island in Chuuk State, Federated States of Micronesia. According to Ethnologue Pollapese isn’t a distinct language but is instead a dialect of Puluwatese. I did some hunting around and found Sam Elbert’s grammar and dictionary of Puluwatese, and put together a Blust’s Swadesh list for it to compare with what we’re doing with Paulina on her language. Paulina thinks that Pollapese is distinct from Puluwatese, and she says the intelligibility is not all that great. I get the impression it’s something like Swedish and Nynorsk Norwegian, or maybe Galician and Spanish. The differences between the two Swadesh lists are fairly numerous, so I’m inclined to agree with her on that basis.

Since it’s only the beginning of the semester we’ve only done phonetics, and our writeups are due next week. Pollapese has a pretty ordinary consonant system, although it does have labio-velarized bilabial stops (something most Micronesian languages have), /pʷ/ and /mʷ/. These two have some interesting effects on the vowels, and word finally they’re more like [pᶭ] and [mᶭ] where only velarization is preserved, e.g. /uːmʷ/[ʔuːmᶭ] “oven”. The really interesting thing is that it has initial geminates, which are almost impossible for me to hear. I figured that speakers would differentiate them via visual cues, since that’s what I’ve been doing for them. When watching a speaker you can see the articulatory muscles shift below the jaw, and the length of time between this movement and the release can reliably lead to differentiating them. But Paulina doesn’t rely on visual cues, because I had my back turned to her and tried two forms, one with initial geminate and one without, and she knew the difference purely by sound. Ken Rehg suspects there’s a difference in the release that is audible, but we need some spectrograms to really tell.

The vowels in Pollapese are pretty difficult. It’s got the usual cardinal vowels, but also a couple of uncommon ones. There’s a high central rounded vowel /ʉ/, but no unrounded one (Puluwatese is supposed to be the reverse). There’s a close-mid central rounded vowel /ɵ/, which is apparently comparable to Puluwatese’s /ə/ (or more likely /ɘ/). Figuring out /ɵ/ was challenging, since the central position makes it seem to wiggle all over the chart, and the rounding makes me want to hear [œ] or [ø] instead. Especially frustrating for me is the three vowels /ɛ/, /æ/, and /a/, which seem to have overlapping ranges so that it’s really hard for a nonnative speaker to even tell them apart. Long vowels diphthongize readily, making it confusing for me given that there are also two vowel sequences that aren’t diphthongs.

Being a Micronesian language I’m sure the morphology is going to be pretty cool. We’ve already figured out direct and indirect possession, which requires noun classifiers for the indirect forms. Pronouns are pretty simple, with just 1{SG, PL.INC, PL.EXC}, 2{SG, PL}, and 3{SG, PL}. I actually expected duals, but there aren’t any. Reduplication is pretty apparent in a lot of the words we’ve gathered, and it’s not exactly the usual form but is instead the rather complicated Micronesian reduplication.

Labels: , ,

01 February 2009

Pronominal argument hypothesis

One of my current projects is applying generative syntax to Tlingit, something which Seth Cable has been at for a while now. He's been treating Tlingit much like any other language, where NPs are arguments of Vs. But Tlingit is not really that simple (surprise!), and it seems to be amenable to an analysis according the Pronominal Argument Hypothesis. This idea was first advanced by Eloise Jelinek back in 1984 in an article in Natural Language and Linguistic Theory titled "Empty categories, case, and configurationality", and it has been argued continuously since then.

What she and Ken Hale termed "non-configurational" languages are those which have a number of features that set them apart from the more typical "configurational" languages. In particular, non-configurational languages have free phrase order (inaccurately called "free word order"), where the various phrases in a sentence can come in any order rather than simple SOV, SVO, VSO, or the like. Supporting this free phrase order is the extensive use of what appear to be agreement affixes on the verb, indexing at least the S and O arguments if not others like instrumentals, locatives, etc. There are a couple of other things but I'm going to ignore them for the moment.

The Pronominal Argument Hypothesis (PAH) says that instead of having independent NPs which are the arguments of the verb, the agreement affixes in the verb are actually the real arguments. These are thus not agreement affixes, but actual pronouns. Furthermore, the sentence can consist of just a verb and no other words, and there are no invisible constituents like pro which take the place of arguments.

The PAH has profound implications for languages like Tlingit, and for theoretical linguistics as well. What it means for Tlingit is that the verb is no longer a thing formed by morphological rules and processes, but is instead constructed by syntax. That is, in Tlingit the verb morphology is the syntax, aside from the other stuff that floats around the core of a sentence.

I'm only in the early phases of exploring how the PAH can be applied to Tlingit, so I don't have anything interesting to report yet. But I'll put up some examples in the near future, after I've finished with my literature survey.


Cable, Seth. "Covert A-scrambling in Tlingit". In Lyon, John (ed.) Papers for the 43rd International Conference on Salish and Neighboring Languages. UBC Working Papers in Linguistics. Vancouver: University of British Columbia.

Carnie, Andrew; Harley, Heidi; & Willie, MaryAnn. 2003. Formal approaches to function in grammar: In honor of Eloise Jelinek. Vol. 62 in Linguistik Aktuell/Linguistics Today. Amsterdam: John Benjamins. ISBN 1-58811-348-5.

Jelinek, Eloise. 1984. "Empty categories, case, and configurationality". Natural Language and Linguistic Theory 2.1:39-76.

Jelinek, Eloise; & Demers, Richard. 1994. "Predicates and pronominal arguments in Straits Salish". Language 70.4:697-736.

LeSourd, Philip. "Problems for the pronominal argument hypothesis in Maliseet-Passamaquoddy". Language 82.3:486-514.

Labels: , ,