Our lab is devoted to the study of how language is processed in the normal brain; how language breaks down in patients with stroke, head injury and dementia; and how the brain recovers language functions -- with or without therapy -- in the months following a stroke or head injury. Our research program may be divided into five related areas of interest:

1. Understanding acquired disorders of language (aphasia) and reading (alexia) subsequent to stroke and head injury.

One goal of this project is to use data obtained from brain damaged patients to inform our current models of language and reading processing. A second goal is to use these data to gain a better understanding of the types of reading and language deficits that can occur, so that treatment can be tailored accordingly. This project uses the detailed case study or multiple case study approaches to learn about the breakdown of cognitive processing skills in the damaged brain.

2. Rehabilitation of reading following stroke or head injury.

The goal of this project, the largest in our research program, is the development and testing of cognitive treatments for specific acquired reading disorders, based upon a neuropsychological model of reading. Studies are designed to determine which approach to rehabilitation is most successful: stimulation (reactivation); re-training; or re-organization of function, for the different types of alexias. A proposed new project would expand the current rehabilitation efforts by incorporating current notions of types of memory and learning into the work, and by expanding the focus of the rehabilitation to include anomia (word finding problems) as well as alexia.

In addition to behavioral responses from patients, we are also able to record their eye-movements while reading. This allows us to assess whether they are using an 'optimal' reading stratgey. For example, it is known that some patients with pure alexia initally focus too far to the left of a given word. We are currently testing whether it is possible to train these patients to focus in a more optimal position, i.e. further towards the center of the word.

3. The neural basis of reading in healthy readers and alexic patients,

- assessed by functional magnetic resonance imaging (fMRI) studies.

This project makes use of functional magnetic resonance imaging (fMRI) studies to: 1) determine the potential usefulness of the technique of fMRI in predicting the outcomes of the various treatments for alexia; 2) elucidate the mechanisms of recovery from alexic symptoms, by examining the shifting loci of brain activation as different compensatory cognitive strategies are adopted both before and after successful cognitive therapy; and 3) provide data that will help refine current neuropsychological models of reading

- assessed by event related potential (ERP) studies.

We also record neural activation using event related potentials (ERPs). Whereas fMRI provides us excellent information about the location of neural activation, ERPs provide superior information about the timing of those neural events. Thus, the two techniques provide differing and complementary information about the nature of neural activation in healthy volunteers and patients (both before and after treatment).

Previous studies using ERPs have identified a negative component -the N400- thought to reflect semantic processing. Research has generally found a reduced N400 for closed class compared to open class words. In a current study we consider the effects of context and timing on reading of open and closed class words, in both healthy volunteers and patients with phonological alexia.

4. Rehabilitation of anomia following stroke or head injury.

This study incorporates current notions from the literature on learning and memory and amnesia in exploring the effects of two paradigms of learning - errorless and errorful learning - upon rehabilitation of anomia.

5. Semantic memory deficits in dementia.

We are using a priming paradigm in an attempt to learn more about the semantic processing deficits of Alzheimer's disease. In this paradigm, a target word that follows a prime word is read faster when the preceding prime word is related in some way. Work is focusing on the distinction between priming that results from the semantic relationship between the concepts underlying the words and priming that results from the associative relationship between the words themselves.