Research

Broadly, I am interested in early online language comprehension and word learning, and in how children’s developing event representations contribute to these processes. I employ a wide range of experimental methodologies including eye-tracking, language production, and corpus analysis techniques.  I also regularly employ advanced statistical techniques and have experience with longitudinal designs and analysis. My cross-disciplinary training in both Psychology and Communication Sciences and Disorders enables me to ask theoretically rich questions with concrete clinical implications. 

Reference Comprehension

In a few short years, children learn thousands of words and a productive language system that allows them to produce and comprehend sentences that they have never heard before. One challenge that children face in acquiring this complex web of linguistic knowledge is to link words to their real-world referents. For instance, how do children learn who “he”  is or what a new word means? In these situations, adult listeners use knowledge of linguistic structure as well as real world knowledge about the events under discussion to determine reference. My research explores how children utilize both information sources to make these links in real time as sentences unfold. Broad questions in this area include:

1. To what extent might situation models be involved in early language comprehension and learning?

2. How can we disentangle the tight correlation between referential patterns that can be learned through experience with language versus what is inferred on the basis of a cognitive representation of the message being conveyed? 

3. What sources of individual variation are there in children's ability to make event-based pragmatic inferences during online language comprehension? 

Publications and Manuscripts

Yuile, A.R. & Hartshorne, J.K. (in review). The Winograd schema without world knowledge: Minimal post-pronoun semantics modulate the implicit causality pronoun bias.

Yuile, A.R., Fisher, C., & Borovsky, A. (in review). Situation models support referential prediction at 36 months of age.

Yuile, A.R. & Hartshorne, J.K. (in review). Age-related changes in implicit causality and consequentiality pronoun biases. 

Yu, Y., Yuile, A.R., Ishak, D., & Fisher, C. (2025). 4- and 5-year olds integrate verb knowledge with situation models in online reference resolution. Proceedings of the 49th annual Boston University Conference on Language Development. ed. Aditya Yedetore, Rebecca Dufie Bonney, and Yuanyuan Zhang, 746-759, Somerville, MA: Cascadilla Press.

Yuile, A.R. & Fisher, C. (2021). 4- and 5-year-olds use mental models of events in reference resolution. Proceedings of the 45th annual Boston University Conference on Language Development. ed. Danielle Dionne and Lee-Ann Vidal Covas, 829-844, Somerville, MA: Cascadilla Press.

In Preparation:

Yuile, A.R. & Fisher, C. (in prep.). 4- and 5-year-olds use mental models of events in reference resolution.

Yuile, A.R., Fisher, C. & Wagner, L. (in prep.). When perfect isn't enough: Children use verb semantics -- but not aspect -- to interpret pronouns. 

Yuile, A.R. & Hartshorne, J.K. (in prep.). The role of linguistic experience in implicit causality and consequentiality pronoun biases.

Word Learning

In another line of research, I explore how the nature of children's existing semantic/conceptual representations impact their word learning.  The size and semantic structure of children’s early productive vocabularies reliably predict later language processing skill (e.g., Marchman & Fernald, 2008), yet relatively little is known about how children’s existing lexico-semantic structure supports vocabulary growth and language abilities. Some work suggests that children leverage existing semantic representations to learn new words (Borovsky et al., 2016; Borovsky, 2020). In my work in this area, I aim to better understand the relation between early vocabulary and future language skill. Broad questions in this area include:

1. What semantic information specifically are children leveraging to support word learning during the toddler years?

2. How does the approach researchers take to defining and operationalizing existing semantic knowledge affect study outcomes? (This represents a methodological aim intended to refine and support future work in this area.)

This line of work has clear clinical implications: Prior work suggests that children with language delays (i.e., late-talking toddlers and children with Developmental Language Disorder) may have sparser and more superficial lexico-semantic networks (e.g., Beckage, Smith, & Hills, 2011). Thus, better understanding the nature of the relation between semantic structure and vocabulary growth can help shed light onto why children develop early language delays, potentially pointing to a route for effective early intervention.

Publications and Manuscripts

Yuile, A.R., Kueser, J.B., Outzen, C., Christ, S., Stiegler, R., Adams, M., Brown, B., & Borovsky, A. (in press). Lexical leveraging in novel word learning: Different semantic properties support learners at different stages of development. Developmental Science.

Yuile, A.R., Kueser, J.B., Outzen, C., Christ, S., Stiegler, R., Adams, M., Brown, B., & Borovsky, A. (2025). Lexical leveraging across the vocabulary spectrum: Different semantic properties support delayed and advanced learners. Proceedings of the Cognitive Science Society.

In Preparation:

Yuile, A.R., Kueser, J.B., Outzen, C., Christ, S., Stiegler, R., Adams, M., Brown, B., & Borovsky, A. (in prep.). In toddler novel word learning, existing knowledge matters -- but so does how you measure it. 

Infant Cognition

Theory of Mind

Under the supervision of Dr. Renee Baillargeon, I developed two new video-taped tasks designed to reliably measure false-belief understanding (FBU) in toddlers. Adults routinely generate inferences about others’ mental states (goals, beliefs, desires, etc.) in order to make sense of their actions and/or predict future behavior – an ability termed theory of mind. Investigators have long sought to determine at what age children first show FBU, which is widely perceived to be an important facet of theory of mind. Initial studies using explicit tasks suggested that children are not capable of FBU until about 4 years of age. Subsequent investigations using implicit tasks suggested that some capacity for FBU is already present in infancy. However, non- or partial-replications of implicit FBU findings have led a number of researchers to question the reliability and validity of implicit measures of FBU. Thus, the aim of this work is to (1) contribute evidence to the ongoing debate on implicit tests of FBU using open and transparent best practices in the field, and (2) provide researchers with easy and reliable resources for assessing early FBU in their own labs. 

Publications and Manuscripts

It is my sincere hope to get these two projects written up soon, so researchers can have access to this resource!

Physical Reasoning

During a Summer internship at the Allen Institute for AI, I helped develop a benchmark for assessing whether popular computer vision models could perform causal reasoning over simple physical events they way infants do.

Publications and Manuscripts

Weihs, L., Yuile, A.R., Baillargeon, R., Fisher, C., Marcus, G., Motteghi, R., & Kembhavi, A. (2022). Benchmarking progress to infant-level physical reasoning in AI. Transactions on Machine Learning Research.