Skip to main content

The development of neural systems for language

Final Report Summary - DNLP (The development of neural systems for language)

The overall objective for the reintegration period was to build a neurolinguistic lab that integrates data from three main sources: diffusion tensor imaging, functional magnetic resonance imaging (fMRI) and cognitive measurements. This objective fits into an organisation wide goal, to build brain and language programmes at the graduate and undergraduate level, that bridge between knowledge fields within the brain sciences, cognitive sciences, and linguistics. These top level goals have all been achieved successfully. With the support of this grant, Dr Ben-Shachar now heads the neurolinguistics lab at the Gonda Brain Research Center (GBRC). The lab is staffed by one postdoctoral fellow, two mature and two beginning Doctor of Philosophy (PhD) students, two Master students near graduation and two Master students with approved proposals, and one research assistant. This is the only lab in Israel that combines DTI, fMRI and cognitive methods in the study of language, complementing a Transcranial Magnetic Stimulation (TMS) lab (Lavidor), fMRI lab (Faust) and Magnetoencephalography (MEG) lab (Goldstein) all studying language processing at the GBRC in Bar Ilan. Dr Ben-Shachar is the lead advisor for two new study programmes in Bar Ilan: A double major Bachelor in linguistics and brain sciences, and a linguistics Master with a focus on clinical research. The first programme has just opened, and we already have 8 registrants for next year. The second program has opened 3 years ago, growing from 3 initial students to 14 students this year. Both programs strive to give students a wide theoretical foundation in linguistics as well as a focused training in experimental science in order to promote neurolinguistic research in Bar Ilan specifically and in Israel and Europe broadly.

Based on the PIs expertise acquired through a large longitudinal developmental study which she led at Stanford University, the main research aim of this project was to study the development of linguistic skills in Hebrew speaking children. Children's language skills, particularly combinatorial aspects of language such as syntax and morphology, continue to develop during the school years. These ages are also a time of significant change in children's brains, both in white matter and gray matter. The current proposal aimed to apply a multimodal approach in acquiring and integrating data from three sources: fMRI, diffusion tensor imaging (DTI), and a cognitive testing battery, to study the neural circuits underlying language processes in Hebrew speaking school age children and young adults. We focus on the development of combinatorial aspects of language, specifically, syntax and morphology. This focus is motivated by scientific and clinical considerations: From a basic science perspective, combinatorial language skills constitute the core of human linguistic ability, combining structural units recursively into complex words, phrases and sentences. Very little is known about the development of neural systems for syntactic and morphological processing in European languages; almost nothing is known about the development of these systems in typically developing children who acquire Hebrew as their native language. In particular, no fMRI or DTI studies exist to date on the development of Hebrew syntactic and morphological knowledge, even though Hebrew's morphology in particular is quite different than European languages. Clinically, syntactic and morphological deficits have been detected in specific language impairments (SLI), but the neurological deficit underlying these impairments is not well understood.

We developed a functional localiser to identify brain regions involved in Hebrew language comprehension in individual children and adults. Similarly, we have developed and tested a functional localiser for verbal working memory. In developing an effective localiser for Hebrew language processing, we discovered that signal correlated noise (SCN) is a better baseline for speech processing than reversed speech. SCN separates out temporal language processing regions from adjacent primary auditory responses, while reversed speech evokes high activation in individual's language regions, almost as much as speech. These results were presented in the annual meeting of the organisation for human brain mapping (Barcelona, 2010), and have been accepted for publication in Brain and Behavior (in press).

To assess the relation between regional cortical activation and specific linguistic skills, we assembled a battery of cognitive tests probing semantic and phonological fluency, verbal working memory, phonological awareness, naming, production of inflectional and derivational morphology and syntactic processing. This comprehensive battery of tests in Hebrew is a unique tool in neuroimaging in terms of its specificity to linguistic components. We have applied this battery in several populations so far, including adults with developmental stuttering, typically developing, dyslexics and adults with little exposure to print, and are in the process of collecting similar measures in multiple sclerosis patients and stroke patients.

The next phase in the study, collection of imaging data in Hebrew speaking children, was severely hampered by the ethics authorities in Israel. Our protocol was considered in a Helsinki committee within the ministry of health. To our surprise and against all reasonable norms, after two rounds of corrections and one year after its initial submission, our protocol was rejected. We now understand that this decision was not specific to our study, but rather, a blanket refusal for scanning healthy children (probably due a change in the makeup of the committee), which put a stop to several projects submitted by other labs headed by established academics in Israel, even though similar protocols have been approved in Israel prior to 2008.

In light of this unfortunate turn of events, we took the following corrective measures. While we continued the appeal process to try and reach Helsinki approval (and acquired an approval to scan 5 children 18 month later), we diverted ongoing research towards 3 adult populations: stuttering, low literacy adults, and dyslexics. In parallel, we pursued further developmental studies on reading and language at Stanford, which have already produced two high profile publications (Ben-Shachar et al., 2011; Yeatman et al., 2011) and a recently awarded NIH grant (PI: Heidi Feldman, Stanford School of Medicine, subaward: Michal Ben-Shachar, Bar Ilan; grant No. 1 R01-HD069162-01A1).

The stuttering project is well underway (30 participants scanned), and is now financed by Israel Science Foundation (ISF) grant No. 513/11. The reading projects are ongoing, with twenty control participants scanned in preparation for the next phase to be implemented with dyslexic individuals. This study is now funded by the United States-Israel Binational Science Foundation grant No. 2011314. We have collected behavioural measurements in 22 low literacy adults, and are currently expanding one of these experiments to illiterate participants in Portugal, in collaboration with Paulo Ventura (University of Lisbon). The study in low literacy Ethiopian immigrants to Israel is now funded by the National Institute for Psychobiology Israel.

In sum, this grant has been the main source of funding in the first two years of the lab. With the support of this grant, we have now secured funding for the next four years of research in the lab (362, USD 500), and have established its role in bridging neuroscience and linguistic research in Bar Ilan and in Israel.