Jennifer Sclafani

Talk Title: Can linguistics contribute to solving complex social problems?
The case of public health and housing in Boston

Abstract: In this talk, I describe an ongoing collaborative interdisciplinary project between sociolinguists and public policy researchers that has set out to tackle a complex nexus of public policy issues in the city of Boston. The problems converge, both literally and figuratively, at an intersection known as “Mass and Cass”, which has become the epicenter of the city’s intersecting opioid, mental health, and homelessness crises. The issue has received heightened media attention in recent years, with local residents, business owners, and community activists calling on city officials to resolve the issues plaguing the neighborhood. Policy debates surrounding Mass and Cass also featured as one of the central issues in the city’s last mayoral election.

I share some findings from our critical intertextual and multimodal analysis of a Boston Globe news corpus (~378,000 words) and 150 videos of local broadcast coverage (~10 hours) on the events and initiatives surrounding Mass and Cass over the past decade, focusing on whose voices are represented in news coverage, which interests these voices represent, and what solutions are promoted and rejected through the strategic selection of quoted sources, and lexicogrammatical and visual aspects of journalistic speech representation.

I conclude by discussing some implications of the study for public policy practitioners, underscoring how media discourse informs public opinion on urban health policy and political action, and I share our plans for engaging and collaborating with local stakeholders to ultimately improve policy action and outcomes in Boston. I also reflect on some personal discoveries, insights, and challenges of engaging in this type of interdisciplinary work as an applied linguist.

Peter Lasersohn

Talk Title: Coordination of Unmatched Clause Types and Dynamic Look-Ahead 

Abstract: In coordinate structures, the conjuncts normally must all belong to the same logical type. A major class of apparent exceptions exists, however: declarative, interrogative, imperative, and exclamative clauses can all be coordinated with one another, even though they are frequently analyzed as belonging to different logical types:

(1)                John left town, and after all, why wouldn’t he? (declarative-and-interrogative)

(2)                Leave now or I will shoot! (imperative-or-declarative)

(3)                John’s dissertation defense is tomorrow, and is he ever nervous! (declarative-and-exclamative)

(4)                Try to make sense, or are you incapable of that? (imperative-or-interrogative)

(5)                What a fool I’ve been, or am I just human? (exclamative-or-interrogative)

(6)                Look in this box, and what a nice surprise you will find! (imperative-and-exclamative)

A related observation is that conditional clauses can antecedent clauses of any type (although the conditional clause itself must be declarative):

(7)                If you are ready, we can proceed. (if-declarative-declarative)

(8)                If John is so smart, then why can’t he answer the question? (if-declarative-interrogative)

(9)                If he tries to escape, kill him! (if-declarative-imperative)

(10)            If John were here, what a good time we could have! (if-declarative-exclamative)

To account for these facts, I propose a dynamic semantics in which all clause types, when occurring as main clauses, denote functions from contexts to contexts. Each clause type effects a different kind of alteration to the context — mostly in ways already proposed in the literature: declaratives add a proposition to the Common Ground, interrogatives add a question to the Questions-Under-Discussion set, imperatives add an individual property to the addressee’s To-Do List, and exclamatives add a degree property to the speaker’s Affective Trigger Registry. Propositions, questions, individual properties and degree properties may still be treated as belonging to different logical types, but this does not block coordination since these items do not serve directly as the semantic values of main clauses.

In such a system, conjunction can be treated as simple function composition. Disjunction and conditionalization require a more complex treatment, and are used to motivate a representation of contexts which “looks ahead” to possible future states of the Common Ground, QUD set, To-Do Lists, and Affective Trigger Registries. This allows for temporary indeterminacy as to the identifies of these sets; for example Bring your own bicycle, or we will rent you one for $10/hour updates the context so that all future developments either have the property of bringing one’s own bicycle on the addressee’s To-Do List, or the have the proposition that the speaker will rent one out for $10/hour in the common ground. Some of these potential developments may later be pruned from the context as more information is received, returning the context to a determinate state. The availability of indeterminacy differentiates the system proposed here from similar “look-ahead” frameworks such as that proposed by Cohen and Krifka (2014).

I argue that the dynamic system proposed for main clause connectives should not be extended to subordinate clauses, which are treated in a more traditional static style. A series of “staticization” operators is proposed, to relate subordinate clause denotations to their main clause counterparts. These operators may be seen as contributing to the semantics of complementizers, as well as relating the dynamic connectives to the traditional truth-functional connectives of standard propositional logic.

Silvia Perez-Cortes

Talk Title: The Future is Within: Centering Heritage Language Research Around Intra-speaker Variability

Abstract: Despite the recent theoretical and methodological advancements in the field of heritage language acquisition, the task of how to best represent HL grammars in an all-encompassing model has remained rather elusive (Putnam & Sánchez, 2013; Polinsky & Scontras, 2019; Lohndal, Rothman, Kupisch & Westergaard, 2019). In this presentation, which features an overview of my work on heritage language acquisition, I will focus on a series of data-driven puzzles to argue the importance of zooming in and examine the nature of optionality and ‘differential’ outcomes when modeling HL grammars. A particular emphasis will be given to the study of intra-speaker variability and how language internal as well as experiential factors modulate their representation and use of the HL.