Dawson Petersen

About Me

I am a Ph.D. candidate in the Linguistics Program at the University of South Carolina under Dr. Amit Almor. My research focuses the relationship between linguistic framing and reasoning. Currently, I am investigating in the ways that the semantic-syntactic features of animacy, agency, and subjecthood change reader perceptions of AI responsibility in my dissertation entitled “The Effects of Anthropomorphic Linguistic Framing on the Online Sentence Processing of Texts about AI”.

Dissertation Abstract

Linguistic framing effects show that subtle differences in how information is presented can dramatically change how readers evaluate and respond to the situation depicted in a text, even when the propositional content is unchanged (Thibodeau & Boroditsky, 2011; McGlynn & McGlone, 2019). This is especially true with regards to the assignment of agency (Fausey & Boroditsky, 2010; Dragojevic et al., 2010). Grammatical metaphors (e.g., placing a non-agent in subject position of an active clause, Devrim, 2015; McGlynn & McGlone, 2019) allow agency to be assigned to non-agents resulting in linguistic anthropomorphism (i.e., the ascription of human characteristics to non-humans). However, insufficient research has investigated 1) the effects of linguistic framing on online text processing and 2) how this linguistic phenomenon interacts with the psychological phenomenon of anthropomorphism (Heider & Simmel, 1944; Epley et al., 2007). Therefore, my dissertation project seeks to fill these gaps in the literature by using an eye-tracking reading paradigm to investigate how anthropomorphic linguistic framing affects the online processing of texts about LLMs compared to other AIs, and computer programs.

MA Thesis Abstract

Talmy argues that causal sentences are understood by reference to basic image schemas, such as Starting and Stopping. The current study employs a priming paradigm in which participants experience a force dynamic prime followed by self-paced reading of a sentence with a causative verb which referenced a compatible or incompatible force dynamic schema. In order to strengthen the priming manipulation, the primes used were short interactive 2D computer games created using Unity and the Unity Experiment Framework as primes (Brookes et al., 2019). In each game, participants had to either prevent or cause an object to move. Following the prime, participants read sentences that described events with the same or the opposite force dynamic schema. Sentences were presented region by region, and the response time for each region was recorded. I predicted that participants would respond more quickly when the target sentence matches the interactive force dynamic prime. However, the results show no such priming effects, which suggests either that force dynamic schemas are not used in online sentence comprehension or that their influence is weak enough that an even stronger manipulation is required to uncover it.

News

09/12/24 - I received funding from the UofSC Graduate School to present my paper “Anthropomorphism Mediates the Algorithm Outrage Deficit” at the 65th meeting of the Psychonomic Society in New Yark City.

07/02/24 - I received the NSF Linguistics Program’s Doctoral Dissertation Improvement Research Grant.

Grants

Doctoral Dissertation Improvement Research Grant
NSF Linguistics Program, 2024
$17,817

Russell J. and Dorothy S. Bilinski Dissertation Fellowship
University of South Carolina, 2024
$40,000

SPARC Graduate Research Grant
University of South Carolina, 2024
$4,800

Publications

Petersen, D., Almor, A. & Shalin, V. L. (2024). Linguistic framing affects moral responsibility assignments towards AIs and their creators. Proceedings of Agency and Intentions in Language 4.
https://ail-workshop.github.io/AIL4-Workshop/program.html
AIL4 Proceedings Paper.pdf

Petersen, D. & Almor, A. (2023). Anthropomorphism, not depiction, explains interaction with social robots. [Peer commentary on the paper “Social robots as depictions of social agents” by H. H. Clark & K. Fischer]. Behavioral and Brain Sciences, 46.
https://doi.org/10.1017/S0140525X22001698

Petersen, D. (2022). The role of force dynamic schemas in the comprehension of causal language [MA thesis].
https://scholarcommons.sc.edu/etd/6800/

Petersen, D. (2018). Embodied simulation in response to semantic priming [Unpublished undergraduate thesis].

Courses Taught

Linguistics 300: Introduction to Language Sciences, University of South Carolina
Upper level survey of Linguistics, Cross-listed in Psychology and Anthropology

Linguistics 101: Introduction to Language, University of South Carolina
Introductory level survey of Linguistics

English 101: Critical Reading and Composition, University of South Carolina

English 102: Composition and Rhetoric, University of South Carolina

Conference and Workshop Organizing

Lead Workshop Organizer
Experimental Pragmatics: Bridging Theory and Praxis
with invited speakers: Mira Ariel and Anna Papafragou
American Pragmatics Association, November 11, 2023