[Note: This page will be updated soon.]

Neurological Investigation of the Use of Auditory Information in Sentence Comprehension

This project investigates how a particular speech property, prosody, aids listeners in processing massive amounts of linguistic information so quickly and efficiently. Prosody includes features such as intonation, pause and pitch height. There is ample evidence in the psycholinguistic literature showing that sentence comprehension proceeds incrementally, i.e., word by word, as soon as each word is received. Recent behavioural studies conducted mostly in English (e.g., off-line questionnaires and a few eye-tracking experiments), have revealed the significant role that prosody plays in the incremental processing of sentences. In particular, prosodic boundaries (i.e., grouping of words into several speech units) help listeners process potentially ambiguous sentences such as “Put the doll on the mat in the box,” which can be uttered with a pause before or after “the mat.” Yet the precise nature of the prosodic boundary used during on-line sentence comprehension and its neurological correlates (or brain responses) remain to be investigated. In addition, the research will focus on Japanese, a strict verb-final language which has prosodic boundaries similar to those of English, but also has a unique prosodic property, lexical accent, which appears to have an effect on Japanese speakers’ perception of prosodic boundaries.

The main goals of the proposed research project are to: (i) investigate which prosodic parameters influence listeners’ perception of prosodic boundaries (chunking of speech into several units) when processing sentences on-line, using Event Related Potentials (ERPs) to measure the brain’s electrophysiological response to particular prosodic cues in sentence comprehension; (ii) examine what brain responses (ERPs) are elicited, corresponding to distinct prosodic parameters that appear in the sentence; and (iii) investigate the universality of (i) and (ii) by comparing sentences in different languages such as English and Japanese.


Cognitive Modeling of Eye Movements during the Reading of Non-alphabetical Texts

This project investigates the human cognitive and information processing system by examining how readers read non-alphabetical texts, using an eye-movement technique. Previous eye-movement research, mostly based on English and other alphabetical languages, indicates that readers’ eye movements are highly controlled by the amount as well as the type of information being processed. Studying a language whose properties are different from those of alphabetical languages should resolve some controversies in current theories of eye-movement control during reading. This research project will focus on Japanese, which is unique both in its orthographic system and sentence structure, compared to well studied languages such as English.

The objectives of this project are to: (i) identify which features (i.e., visual complexity, semantic and phonological codes) of Japanese characters, along with the location of their appearance within a sentence, influence readers’ eye movements, (ii) examine how the findings in (i) differ from English and other well-studied alphabetical languages, so that we can specify which reading processes are common to all languages and which are language specific, and (iii) incorporate our findings based on Japanese to an existing information processing model on reading, and build a unified account of the cognitive mechanisms involved in reading.


Neural Investigation of Language Processing

Language plays a critical role in our daily life and having full command of our language faculties provides us with many valuable social advantages. In fact, so important are these faculties that, if we lose or experience difficulty with any one of them, the consequences to our mental health can be catastrophic. Therefore, it is essential that reliable interventions be developed to counter these difficulties when they occur.

We believe that our research into language processing will contribute significantly to the development of these interventions and has the potential to benefit a wide variety of people experiencing language difficulties.

Specifically, our research will investigate the neurological basis for computing the meaning of a sentence.

Using Event Related Potentials (ERP) or brain waves, a temporal measure of brain activity, and neuroimaging (fMRI) techniques, a standard method of examining brain regions recruited during sentence comprehension, we will carry out two sets of experiments. This work will examine: 1) the comprehension of modal sentences and 2) the interpretation of negation. We have chosen to target these two measures because they bring together a distinct and particularly useful interaction between structure and context in the interpretation of sentences.

Lastly, in addition to the actual experimentation itself, this research effort seeks to create an interdisciplinary network of researchers, both domestic and international (Germany, Japan, and US), who are interested in the same topic (linguists, psycholinguists, psychologists, neuroscientists, and speech-language pathologists). Such a diverse research team is indispensable when tackling the complex and multi-faceted questions found in the field of language processing.