Many congratulations are in order for Lydia Newkirk for successfully defending her dissertation! Her defense took place on August 24.
Title: Be flexible, but not too flexible: limited variable-force modals in Kinande and the typology of modal force
Committee: Simon Charlow, Troy Messick, Ken Safir, Lisa Matthewson
This dissertation has two goals: First, to describe and analyze a new kind of variable-force modality in Kinande, a Bantu language spoken in the Democratic Republic of Congo and in Uganda, contextualizing it within the existing typology of variable-force modals and expanding that typology descriptively; second, to take the new information gained from examining Kinande and use it to refine our typology of modal force, re-examining previously attested cases of variable-force modals and creating a more refined typology of modal force, and opening new avenues of further research into modal force variation cross-linguistically.
Prior work in modal semantics typically divides modals according to a binary of possibility and necessity, formalized in terms of existential or universal quantification over possible worlds, respectively. While such systems cover the empirical ground of English and several other languages reasonably well, variable-force languages such as St’at’imcets (Rullmann et al., 2008), Nez Perce (Deal, 2011), and Ecuadorian Siona (Jeretič, 2021), among others, bring this binary into question with modals that, depending on their context, are sometimes interpreted akin to English can and sometimes akin to English must or have to. Since these first attestations there have been numerous efforts to account for this variation in modal force.
In this work I present novel data from Kinande (Bantu J, DRC) illustrating a previously unattested kind of variable-force modal; Kinande anga varies between possibility and weak necessity interpretations (akin to English should) but never to strong necessity interpretations. I call this Limited Variable-Force. The existence of such limited variability demonstrates that just as modal force itself is not merely a binary between possibility and necessity, there is not a simple binary between fixed and variable force either. The typology is expanded to languages that have variable-force modals that are more constrained than the modals in Nez Perce (for example), but still less constrained than the modals in say, English.
I present an account of Kinande variable-force anga in terms of exhaustification, using a combination of domain restriction and “scaleless implicatures” as proposed by Jeretič (2021). Scaleless implicatures allow an item to be strengthened in its meaning when it lacks a stronger scalar alternative. I show that the lack of a weak necessity modal in Kinande (and the presence of a strong necessity one) allows anga to be variable to weak necessity, but not to strong necessity, because the strong necessity interpretation is blocked by scalar implicature through the same mechanism that derives the scaleless implicature.
This account does not only provide a satisfactory explanation for the behavior of Kinande modals, but it also generates a useful typology of modal force as a whole: Languages with variable-force modals are predicted to be languages with sparse modal paradigms, while languages with dense modal paradigms are expected not to show variable-force behavior. I give a typology situating previously-described modal sys- tems (both variable-force and not) and discuss the implications of this typology for our understanding of modal force as a whole.