Philosophy & Overview
The fact that our linguistic abilities are situated in a complex cognitive and cultural landscape is inescapable. Over the past 60 odd years, Generative Grammar and other programs of formal linguistic inquiry have provided a path to pull apart the domain-specific properties of language from those spanning multiple functional domains. As linguistics continues to develop as a science, the pursuit to understand how cognitive mechanisms interact has dissolved traditional and technical boundaries between fields. To this end, an openness to doing research by any means necessary is a central component of what drives my interests—be it with a pencil and paper at a desk, an acceptability judgment in the field, a brain scanner in the lab, or a computational model in the cloud. So far, I’ve worked on projects in the following areas:
- Eye tracking while reading
- Scaled acceptability judgments (Mechanical Turk)
- fMRI and MEG
- Similarity ratings and multi-dimensional scaling
- Child language acquisition
- Distributed Morphology
The major through line of my research has been uncovering on the nature of morphosyntactic knowledge. Two threads have emerged. The first thread has contributed to refining models of the relationship between the processing of morphosyntactic dependencies and the structure of working memory. From the broadest perspective, my question of interest is: what role do syntactic structures and morphological features play in memory? Under this umbrella, I’ve worked on:
- The role of clauses in defining the scope of antecedent accessibility in agreement processing
- Dissociating brain regions responsible for building syntactic structures, predicting upcoming structures, and forming semantic relationships
A second thread has examined the properties of the features we use to represent person, gender, and number. There are two fundamental questions here: (i) what are the primitives of morphosyntax? And (ii) what is the mapping between morphosyntactic structures and units of meaning? Major phenomena I’ve worked on are:
- Reconceptualizing the representation of first person “plural” pronouns
- Interpretation of grammatical gender at the semantic interface
- Cross-cultural/linguistic differences in conceptual categorization
- Small-world semantic networks in second language vocabulary acquisition
More information on each of these projects can be found below.
I’ve been involved in three projects under the umbrella of psycholinguistics.
Evidence for clause-bound retrieval from agreement attraction
Current models of the processing of dependencies do not predict that the accessibility of intervening nouns in agreement attraction configurations (e.g. the cabinets in the sentence the key to the cabinets are on the table) should differ depending on the syntactic structure in which they are embedded. With Brian Dillon and Lyn Frazier, I ran three experiments (one in progress as of Jan. 2016) that show attractor nouns are less accessible when embedded in clauses (e.g. the key that opened the cabinets are on the table) compared to phrasal modifiers. We show this occurs regardless of clause size, rendering a depth of embedding explanation implausible. We argue that clauses are therefore domains over which retrieval occurs, and extend Cue-Based models of retrieval to capture this fact.
Cross-cultural differences in conceptual categorization
With Ruxue Shao and Maria Sera, I examined differences in the organization of concepts across cultural and linguistic populations. We compared the behavior of speakers of Japanese, Mandarin, Hmong, and English in a similarity judgment task across a wide variety of categories, finding differences in how humans are categorized in relation to animate versus inanimate objects. This work is currently unpublished.
Semantic and phonological effects on L2 vocabulary learning
With Elizabeth Stephens and Maria Sera, I looked at second language acquisition, focusing on the initial trajectory of vocabulary learning. We explored both the phonological and semantic aspects of this process across different age groups. On the phonological side, we found that L2 words consisting of sounds common in L1 were easier to learn than those with less common sounds regardless of age. Surprisingly, we found no effect of the semantic groupings of the words learned in any case. This suggests that it is sound rather than meaning that has the biggest effect in the initial learning of words. This work was presented as a poster at SRCD in 2015, which can be found here.
What “other people” mean to “us”
First person plural has been construed as [speaker] + [group], or in languages with inclusive plurals, [speaker] + [addressee] + [group]. Using evidence from Ojibwe agreement and pronominal morphology, I show that a reconceptualization of first person plural is necessary, such that plurality is implied by the combination of [speaker] + [proximate] rather than represented as a feature. This differs from second and third person plurals, which utilize the group feature. I link this argument to morphological universals in person marking and the associative plural generalization.
Limiting gender: restrictions on gender specification and interpretation
Within a given language, it is a matter of empirical fact that not every nominal expression comes to be specified for every possible gender class. However, it is not clear where this restriction lies. I argue that this restriction in a failure to semantically interpret the root in context: All gender-root pairings are therefore produced by the generative engine, but only some are semantically coherent. This finding has consequences for our conception of (semantic) interpretability, in that it renders the interpretable-uninterpretable distinction irrelevant. I show all features must be semantically interpretable, even if they are not interpreted directly. The work of interpretability is therefore accomplished by the mechanism of interpretation, rather than the representation of the feature. A manuscript for this project, currently under review, can be downloaded here.
Neural correlates of semantic and syntactic combination
During my Baggett year at Maryland, I had the pleasure to work with William Matchin and Ellen Lau on a combination fMRI/MEG study. The goal of this study was to pin down the brain region and time course associated with the most elemental forms of syntactic and semantic combination. The results of the fMRI study have been published in Cortex (full paper available here). A manuscript presenting the results of the MEG study is in preparation, but have been presented in a poster at the Society for the Neurobiology of Language (available here).
Being indigenous is a central part of who I am: working on the revitalization of Anishinaabemowin and sharing the linguistic insights that the language provides is an important part of my long-term plan. Recently, I’ve been working on features in the pronominal system and how it links to the complex verbal morphology. Eventually, I hope to do experimental work with language learners in immersion schools throughout Minnesota, as well as with first speakers in my community, to help build stronger curricula, and to bring psycholinguistic methods to bear on a non-Indo-European language.