The goal of today’s lecture is to introduce the necessary ingredients to discover and compare units of language that serve to manage interaction.

Here are the slides.

i) I start by introducing what’s at stake, namely to develop a formal typology of discourse markers, something that hasn’t been done in the literature.  To develop such a typology, we need a framework that allows us to compare categories, however different, to each other. I argue that the Universal Spine Hypothesis provides us with exactly the right kind of approach.

ii) The universal spine hypothesis (Wiltschko 2014) is a framework I have used in my investigation of discourse markers thus far. It is designed as a heuristic for discovery and comparison of (understudied) categories across languages. Since discourse markers are understudied within the generative tradition, this is a framework ideally suited to discover and compare them across languages. I will introduce the motivation for this framework, as well as its core assumptions and then move on to presenting two case-studies: one on subjunctives and the other one on aspectual categories. For subjunctives we shall see that what we might refer to as a subjunctive can be derived in different ways. For the aspectual categories, we see that they can be based on temporal distinctions (as in English) but also on participants only, deriving direct/inverse marking in Algonquian languages, and finally they can be based on situations alone without focussing on any one component of a situation. This derives control marking in Squamish Salish.

iii) The syntacticization of interaction. Based on the logic of  the Universal Spine Hypothesis I will extend the spine to  exclude several more layers that serve to encode the management of interaction. Specifically, drawing on the literature on the syntactizication of speech acts, I introduce a proposal that overcomes the challenges identified in Lecture #1. I argue that there is an articulated grounding layer, where speakers relate their attitudes towards the proposition and their assessment of the addressee’s attitude towards the proposition. In addition, I propose an articulated response layer where speakers identify what they are responding to and let the addressee know what they want them to do with what they are saying (this has been referred to as the Call on Addressee in the work of Beyssade and Marandin. I show what the typology might look like.

iv) Methodologies. Before we turn to our case studies I introduce the methodologies we use in the eh-lab to collect data.  Specifically, we adopt the story-board elicitation techniques developed at the department of linguistics at UBC where contexts are presented to consultants in the form of story-boards (essentially little cartoons). This allows us to set up the context so the consultant can see who knows what at the context of the conversation.