Research

We enter into a preformed social environment, and through our social performance we restructure that environment. The social makes us, we make the social.

Language is, for humans at least, a foundational component of this dialogic self-formation. Langauage is how we form corporations, colleges, and co-ops. We build communities & relationships through cooperative action, coordinated in no small part by linguistic practice. Language is also necessarily shaped by the cognitive constraints & mechanisms that define what and who we are as thinking beings. The core of my research program is exploring the ways that language reflects this intersection of the cognitive and social.

Ongoing Projects:

Social constraints on interpersonal coordination.

As a conversation unfolds, partners become coupled as parts of a larger conversational system. This has been supported by evidence that individuals exhibit behavior-matching along various modalities and contribute unique but complementary behaviors to form synergies. Recently, there has been growing interest in how conversation dynamics are constrained. One high-level constraint that has been underrepresented in the study of interpersonal coordination is social status. Status has long been understood to play an important role in cooperative and linguistic behavior To better understand how status, the perception of status, and a host of other salient social variables influence conversational dynamics, we are applying CRQA to the CANDOR data set.

How (not) to use RQA.

Recurrence Quantification Analysis (RQA) and its many flavors have become increasingly popular in behavioral research over the last 20 years or so. RQA is a powerful and general method, and as such allows the researcher many degrees of freedom in application and interpretation. It is not always clear in either case whether the choices researchers have made are justified. The primary goal is to understand how behavioral researchers are applying RQA, and whether their applications are theoretically and methodologically sound. A secondary goal is to understand the relationship between RQA and its linear counterparts (e.g. autocorrelation) in the context of behavioral research.

How does interaction modality informs user trust in GenAI systems?

GenAI systems are now ubiquitous in tech - it's in your phone, your computer, and even your fridge. As we interact more and more with GenAI agents, it becomes increasingly important to understand how users interpret and trust the outputs of these systems. This project is aimed at understanding the role of presentation modality in this interpretation and trust.

Collaborators: Leigh Levinson, Skyler Wang

New norms for modal verbs.

There is much debate about the function and classification of modal verbs. The goal of this project is to develop a set of modality norms based on how they modulate comprehenders assesment of the strengh of a speakers belief in their claim.