The size of domains: verbs in Sinhala and Japanese

Zoom link: https://us06web.zoom.us/j/89465925613?pwd=esJUxnrLCOZr77doBj3aXbDjYr1JKo.1
Meeting ID: 894 6592 5613
Passcode: 774763

Many analyses assume that domain effects in grammar exist where different types of linguistic phenomena cannot apply across or outside a particular domain. For example, most syntactic operations such as movement or agree are constrained by locality domains; in the morphology there are suppletion asymmetries; and in the phonology stress assignment or hiatus resolution are sensitive to different locality domains as well. Since domains seem to play a role in each module, the question is whether these domains are (i) inherited from one module to the next (I call this the `inheritance question’), and (ii) the same size in one language and across languages (I call this the `size question’).

In order to answer these questions, it is important to combine in-depth investigation into a single language with cross-linguistic work. To this end, I will look at the verbal domains in Sinhala and Japanese, as these languages provide a rich testing ground for both the inheritance and the size question. Both languages have rich morphological verbal paradigms, which can be used to test predictions in the phonology, morphology, and syntax.

I show that remarkable parallelisms exist in both languages between the application of phonological rules in different word-domains, and the domains that are picked out for certain syntactic processes, such as the choice of negation or verb fronting. I argue that these correlations are to be expected if word-internal domains are inherited from the syntax. I account for this by assuming that word-building can be timed at different stages of the derivation, where earlier processes can be sensitive to syntactic domains (phases) and that later word-formation processes can mask the output of the syntax.

Secondly, I discuss cases where inheriting the domain from a module to the next one can lead to minimal differences in the size of the domain. This means that the size of the verbal domain can differ, but based on the results from Sinhala and Japanese it is not the case that anything goes, and there are systematic differences between the two languages. I argue that various structure manipulating operations can account for this.