Neuroscience of Language and Language Recovery

 
 

What do we study?

My clinical neuroscience research moves the neurobiology of language field forward by examining the neural mechanisms and structures related to impaired language processes during discourse. Most of the research used to establish a neurobiology of language evaluated relationships between brain damage and impaired, isolated language processes (e.g., naming). Discourse is important to evaluate because the cognitive components required for lexical retrieval in this environment are more complex than in the isolated environment: future and past lexical alternatives must be suppressed during discourse, while other processes, like syntax, occur in parallel. Therefore, by evaluating discourse impairments and their relationship to brain damage, my research provides complementary and novel evidence for theory-building in the neurobiology of language. I typically utilize structural and functional MRI to investigate brain bases of language impairment and recovery.



Research from our lab

Early research, as a postdoctoral fellow at the Center for the Study of Aphasia Research at the University of South Carolina, evaluated the efficacy of transcranial direct current stimulation on naming improvements in chronic aphasia (Fridriksson et al., 2018a, 2018b), and the role of leukoaraiosis (white matter disease) on language recovery in chronic aphasia (Basilakos, Stark, et al., 2021).

From Stark et al’s Cortex paper

In one study, we compared similarities and differences in brain damage associated with paraphasias (words errors) made during discourse and picture naming. For example, a person with aphasia may be describing a picture and say “cat” when they meant to say “dog,” thus producing a semantically-related paraphasia. Using lesion symptom mapping to statistically relate lesion damage to paraphasia frequency, we identified that phonological paraphasias and real word paraphasias associated with unique areas of brain damage (Stark et al., 2019). These findings held true for paraphasias produced during both naming and discourse, suggesting a shared neural substrate for impaired lexical selection regardless of task. This work extends our knowledge about neural substrates necessary for lexical retrieval during discourse, extending theoretical understanding of the neural organization of language.

Fig 3 from Matchin et al’s 2022 paper

My research has also examined the relationship between brain damage and syntactic impairments made during spoken discourse. This research is in collaboration with Dr. William Matchin at the University of South Carolina. Agrammatism (a lack of, or impoverished, grammar) is a well-known, hallmark symptom of non-fluent aphasia, but little is known about the neurobiology of paragrammatism, or the misuse of grammar. Examples of paragrammatism include incorrect agreement (“the three sister”) and syntactically incorrect sentences (“She was handled to look at the books a bit” [from p. 14, Kleist 1914]). We found that damage to the left hemisphere posterior middle temporal gyrus was associated with speakers who demonstrated paragrammatic, but not agrammatic, speech, whereas damage to the left hemisphere inferior frontal gyrus pars triangularis (part of Broca’s area) was associated with speakers who demonstrated agrammatic, but not paragrammatic, speech (Matchin, Basilakos, Stark et al., 2020; Matchin, Basilakos, den Ouden, Stark, et al., 2022; Matchin et al., 2023). Notably, this distinction held when controlling for speech fluency, which is a major confounding factor between those with agrammatic versus paragrammatic speech. Given that there is little research evaluating the neural mechanisms of paragrammatism, our research reconceptualizes the neural organization of syntax in the brain. That is, the left posterior middle temporal gyrus may play a role in a syntactic process that tends to go awry in paragrammatic speech: hierarchical organization.

More recent research from my lab identified structural and functional disconnection underlying verbal fluency impairment in individuals with aphasia. This research is motivated by evidence demonstrating that focal lesions (after stroke) result in global brain changes. Some of these brain changes may support language recovery, such as a reorganization of functional connectivity in the right hemisphere after left hemisphere damage (Yourganov, Stark, et al., 2021). Dr. Yanyu Xiong, a postdoctoral fellow in my lab from 2021-2022, identified that left hemisphere white matter disconnection predicted impairments of semantic fluency after stroke, whereas damage to subcortical right hemisphere structures predicted impairments of phonological fluency (Xiong et al., 2023). This research underlines the importance of looking more globally at disconnection to explain language impairment post-stroke.

Currently, I am acquiring functional and structural MRI and detailed neuropsychological data focused on spoken discourse at IU from individuals with chronic stroke and a prospectively matched neurotypical group. Data collection is ongoing, and the analysis will focus on identifying the extent to which network-level disruptions (e.g., missing nodes or reduced edges) can explain spoken discourse level impairment.


Resources:

Multimedia presentations

Dr. Stark makes free, introductory lectures for ENIGMA-U, a free online school focused on neuroscience, neuropsychology, and basic science literacy. You can find her four-part lecture series there, as well as here, for free!

 

Funding