Cross-linguistically, the choice of the embedding predicate systematically affects the behavior of the complement:
(a) how transparent it is for operations across its boundary (e.g., movement), and
(b) how semantically dependent it is (e.g., temporal dependence on the matrix predicate).
One way to capture this is to say that those difference correlate with different complement sizes (Wurmbrand 2001, et seq.):
- complements of ‘say’-type verbs are CPs;
- complements of ‘want’-type verbs can be smaller, i.e., TPs/ModPs;
- complements of ‘try’-type verbs are vPs/VPs.
According to such approaches, smaller complements are (a) more syntactically transparent with respect to e.g., clitic climbing, binding, NPI-licensing and (b) semantically dependant regarding, e.g., temporal interpretation. Thus far, the clause-size diagnostics have been successfully applied to Indo-European languages. It has been shown that factors such as agreement and temporal morphology in the complements do not matter (Todorović and Wurmbrand 2020).
In this course, we will explore Gitksan, an endangered Tsimsianic language of British Columbia, Canada. Gitksan is typologically interesting because it has agreement morphology but lacks temporal morphology. What patterns do we see?
First, we will see that above syntactic diagnostics do not exactly work for Gitksan. Instead, we will need a tailor-made set of diagnostics. Second, in the absence of temporal morphology, it is not immediately clear what kind of temporal relations hold between the matrix and embedded domain. We will thus need to fine-tune the existing temporal diagnostics to fit Gitksan.
As a result, we will see that Gitksan complements also come in 3 different sizes, albeit with a twist. We will also see that the syntactic differences between these complements capture their peculiar agreement patterns.
Gitksan is by no means typologically unique, so this course opens the door to comparing it with other languages without temporal morphology, as well as with languages with temporal morphology.
Finally, we will learn how to create a novel set of syntactic and semantic diagnostics that fit a language under study. We will also see how to precisely use the fieldwork-related techniques to elicit the complex data.