Research Training: Bilingualism and Executive Control
I began my research career as an undergraduate research assistant in the Department of Psychiatry at the University of Vermont, where I conducted my undergraduate honors thesis. My thesis investigated how the interactions of executive control and language differ among the two different scripts of Japanese, kanji and kana. To do so I used functional magnetic resonance imaging (fMRI), a cutting-edge neuroimaging technique at the time. This research represented the start of my interest in language processing in the brain: a theme which has formed the backbone of my entire research career. Following my graduation from UVM, I worked as a full-time research assistant in this lab for two additional years before pursuing my interests in language and executive control through a Master of Science in Cognitive Neuroscience and Neuroimaging and a Ph.D. in Cognitive Psychology at the University of Nottingham in the U.K., under the mentorship of Dr. Walter van Heuven.
My graduate work focused on two main themes. The first explored the interaction of language and executive control in the Stroop task, a classic executive control paradigm which asks participants to ignore a written color word and name the mismatching ink color in which it is printed (e.g., the word “red” printed in blue ink). In my Master’s thesis, I used electroencephalography (EEG) to demonstrate that presenting the irrelevant word before the target color shortened the latency of the event-related potential (ERP) component associated with Stroop conflict (Coderre et al., 2011). In a follow-up fMRI study performed during my dissertation, I also showed that pre-exposing the irrelevant word recruited brain regions involved in response inhibition, suggesting a proactive recruitment of executive control to suppress the word meaning (Coderre et al., 2013). These studies helped to further understand the temporal and spatial correlates of Stroop conflict.
The second main theme of my graduate work explored the “bilingual advantage”: the phenomenon that bilinguals often out-perform their monolingual counterparts on executive control tasks, which is thought to stem from an increased need for executive control during language processing. Using the Stroop task again, I showed that executive control abilities and language processing speed interact to determine the magnitude of Stroop conflict in bilinguals and that a number of factors, such as proficiency, language immersion experience, and script similarity, influence this interaction (Coderre et al., 2013; Coderre & van Heuven, 2014). In two neuroimaging studies using EEG and fMRI, respectively, I also demonstrated that the bilingual advantage may be specifically related to an enhanced ability to suppress distracting information (Coderre & van Heuven, 2014), since bilingualism alters the functional overlap of the language and executive control networks in the brain (Coderre et al., 2016). This last fMRI study was performed during a travel scholarship to the NIH, where I trained for 6 months as a guest researcher with Dr. Barry Horwitz in the Brain Imaging and Modeling Section at the National Institutes of Deafness and Other Communication Disorders (NIDCD).
Postdoctoral Work: Autism and Language
Following my graduate training, I returned to the U.S. to gain new expertise in clinical applications of cognitive neuroscience through a postdoctoral fellowship in the Division of Cognitive Neurology/Neuropsychology at Johns Hopkins University, under the guidance of Dr. Barry Gordon. In this lab I continued my research on language processing in the brain, but to gain more clinical experience I switched from studying populations with superior language abilities – bilinguals – to populations that often experience language difficulties – autistic individuals*. My postdoctoral training provided me with extensive experience in working with autistic individuals, including adults with high levels of autistic traits that significantly impact their daily functioning. Individuals who are severely affected by autism are often drastically underrepresented in cognitive neuroscience studies; therefore some of the most important skills I obtained during my postdoc revolved around how to adapt EEG testing protocols to accommodate the sensory and cognitive needs of autistic individuals to ensure that individuals across the entire autism spectrum are represented in research. I also gained experience with new techniques including eye-tracking and pupillometry, which I used in conjunction with EEG to demonstrate that these three “implicit” measures reliably distinguished between known and unknown vocabulary in a population of non-autistic adults (Ledoux et al., 2016) as well as in autistic individuals with high levels of autistic traits (Coderre et al., 2019). During the second half of my postdoc, I was also successful at securing independent funding through a Distinguished Science of Learning Fellowship with the Science of Learning Institute at Johns Hopkins. Through this fellowship I gained additional experience in translational and interdisciplinary research, as well as new techniques of neuromodulation (specifically transcranial direct current stimulation, tDCS). My fellowship project, performed in collaboration with Dr. Nancy Madden in the School of Education at Johns Hopkins, paired an empirically based literacy intervention with tDCS to improve reading comprehension in autistic individuals.
In addition to the new skills and clinical experience I gained during my postdoc, I also began to develop two independent lines of research which have provided a springboard for my independent research career. Specifically, I sought to understand why semantic processing (understanding meaning) in autistic individuals is sometimes found to be localized to linguistic stimuli like words or sentences but spared for non-linguistic stimuli like pictures or sounds. In two preliminary studies performed during my postdoc, I investigated this phenomenon in single words (Coderre et al., 2017) as well as more complex narratives (Coderre et al., 2018). These two studies provided preliminary data for two areas of research that I have continued to explore in my independent career.
Independent Career: Autism and Semantic Processing
In 2017 I returned to UVM to join faculty as a tenure-track Assistant Professor in the Department of Communication Sciences and Disorders. As an independent investigator I have built on my graduate training as a cognitive neuroscientist and my more applied clinical postdoctoral training in autism to establish a research agenda focused on understanding the semantic processing of language in autistic individuals. To do so, I take an approach of contrasting language to non-linguistic modalities like pictures in order to understand whether differences in semantic processing among autistic individuals are specific to language or are representative of a more domain-general processing style. My independent research has focused on two main themes investigating 1) semantic processing and 2) narrative comprehension in autistic individuals using neuroimaging techniques.
1) Semantic processing in linguistic and visual modalities
Prior research has suggested that autistic individuals often have difficulties with understanding the meaning of linguistic stimuli like words and sentences. However, some research suggests that such semantic processing difficulties may be absent for non-linguistic stimuli like pictures or sounds. In an EEG study I performed during my postdoc, I observed similar neural responses to semantic processing for linguistic stimuli (written words) and non-linguistic stimuli (pictures) among autistic and non-autistic adults (Coderre et al., 2017), suggesting that adults with ASD do not experience difficulties with semantic processing of language at the level of single words. This result contradicts previous literature, and we speculated that it may be related to task demands (we used an explicit task which drew attention to semantic processing, whereas a more implicit task may accentuate semantic processing differences) and/or age (we tested an adult population, whereas language processing differences may be more pronounced in children). In a follow-up study, my Master’s student and I varied task demands using an implicit semantic processing task and showed that individuals with higher levels of autistic traits showed reduced sensitivities to semantic relatedness in both pictures and words, suggesting domain-general differences in semantic processing (O’Rourke & Coderre, 2021). I am also extending this work to examine implicit and explicit semantic processing in children and adolescents to explore the possibility that age also modulates effects.
I have also expanded this line of research to explore neural connectivity during semantic processing and how it might differ in autistic individuals. In a secondary analysis of an existing dataset, my undergraduate student and I demonstrated that EEG coherence, a proxy for functional connectivity in the brain, is reduced for word stimuli in autistic individuals, suggesting differences in neural connectivity between language and semantic regions of the brain (Curl & Coderre, 2022). Even more intriguingly, we found that these differences in neural connectivity during language processing were localized to extremely early time windows (within 100-300 ms), which implies fundamental differences in the neural connections of semantic processing in autism. I am currently preparing to continue this work by employing fMRI to investigate the specific neural correlates of connectivity during semantic processing in linguistic and visual domains.
2) Narrative comprehension in autism
My most active line of independent research involves contrasting narrative comprehension, or understanding a story, between linguistic and non-linguistic modalities in autism. Difficulties with narrative comprehension among autistic individuals have been reported for linguistic narratives (i.e., stories told through words). However, there is a widespread assumption in the autism field that pictures are “easier” for autistic individuals to understand, which has led to the prevalent use of pictures and visual supports to support text comprehension in this population. In a recent review of visual narrative processing in clinical populations, I have termed this the “Visual Ease Assumption” (Coderre, 2020). However, this assumption is based largely on anecdotal evidence and has little empirical backing. With a collaborator of mine who studies the neurocognition of comics, Neil Cohn, I have established a line of research examining whether difficulties with narrative comprehension in autistic individuals are indeed spared in visual modalities (which would support the Visual Ease Assumption) or are present across modalities (which would contradict this assumption and suggest a domain-general difficulty with narrative comprehension).
This line of research began with an EEG study conducted during my postdoc, in which I demonstrated differences in narrative comprehension abilities for both linguistic and visual narratives in autistic adults (Coderre et al., 2018). As an independent investigator I have explored this finding through several follow-up studies using visual narratives to explore various aspects of narrative and language processing in individuals with ASD. In one EEG study, which was recently completed for a student’s undergraduate thesis in my lab and is currently being prepared for publication, we found that autistic individuals are less sensitive to narrative structure, or grammar, during visual narrative comprehension (Pellegrino-Wood et al., in preparation). Alongside previous literature suggesting differences in grammatical processing in language, these results suggest that structural processing may be a domain-general cognitive skill that is challenging for autistic individuals and contributes to difficulties with narrative comprehension. In another EEG study, which is serving as an undergraduate student’s honors thesis in my lab, we have found that predictive abilities, which are proposed to be different in autism, affect narrative comprehension abilities in similar ways across both linguistic and visual narratives (Kubinski et al., under review). Finally, in a behavioral study that is currently being analyzed and prepared for publication by a Ph.D. student in my lab, we have found that inference generation abilities are modulated by level of autistic traits during visual narrative comprehension (Medeiros et al., in preparation).
These various studies examining different cognitive domains of narrative comprehension – prediction, grammatical processing, inference generation – have formed the basis of a larger study for which I recently secured external grant funding through a Department of Defense (DoD) Autism Research Program (ARP) Career Development Award. In this new project, we will be examining these various aspects of narrative comprehension, along with several others, across both linguistic and visual modalities in individuals with and without autism. By looking at all of these cognitive domains at once, we aim to better characterize the strengths and challenges that autism brings to narrative comprehension by determining which narrative domains pose difficulties for autistic individuals, whether they work differently between visual and verbal narratives, and how they contribute to narrative comprehension overall. Our general hypothesis for this project, based largely on the results of the preliminary studies mentioned above, is that all of the cognitive functions that contribute to narrative comprehension are affected in autism, and affected in similar ways between linguistic and visual modalities. If this hypothesis is confirmed, it would directly contradict the Visual Ease Assumption and call into question the pervasive use of visual supports to support literacy for this population. Through this project and my broader examinations of visual narrative processing in autistic individuals, I aim to challenge the autism field’s assumption that visual processing is always “easier” for autistic individuals and instead prompt a shift toward more evidence-based practices.
* A note on language:
I recognize the preferences of the autistic community in using identity-first language (e.g., "autistic individual" instead of "individual with autism") and avoiding ableist language (e.g., "difference" rather than "disorder"). While I strive to respect these preferences in my current writings, I sincerely apologize that my older papers use outdated, and potentially harmful, language.
