Homing In on Human Language: More Than Semantics . . .

A closer look at the left inferior frontal gyrus

Acquiring language, by assigning meaning to and manipulating symbols, may be the most human of capabilities. When that process is impaired, it can be the most devastating of disabilities. Strokes that cause alexia, the loss or impairment of the ability to read, or aphasia, the loss or impairment of the ability to express or comprehend language by speech or sign, are examples of debilitating impairment. Early on, scientists made discoveries about dominant areas of the brain for particular activities, but recently, investigation has yielded identification of internal neural networks that permit communication between regions in different parts of the brain during language activities. These new discoveries of normal processes should illuminate potential approaches to language repair and recovery.

According to the May 2001 issue of Neuron, scientists at the National Institutes of Health have shown, using functional magnetic resonance imaging (fMRI), that different parts of the inferior frontal cortex of the brain participate in different functional networks during the processing of visually presented words and word-like stimuli. These networks seem to deal with different aspects of language processing.

For more than a century, scientists interested in human language have focused attention upon the left frontal cortex of the human brain, especially in and around the left inferior frontal gyrus (LIFG). Examining subjects who were aphasic as well as normal volunteers, they observed that parts of the LIFG are involved in mediating semantics (the meaning of words), syntax (the organizing system of language associated with putting things into phrases, clauses, and sentences), and phonology (the relation of speech sounds to linguistic units). Semantics, syntax, and phonology are called the "building blocks" of each specific language.

Based on current research, a team of scientists hypothesized that brain areas involved with processing these language functions must network, or link, in order to process each function. Barry Horwitz, supported by the National Institute on Deafness and Other Communication Disorders (NIDCD), and Arun Bokde, supported by the National Institute on Mental Health, with colleagues from the Functional Neuroimaging Laboratory at the Maryland Psychiatric Research Center, University of Maryland, and Georgetown University's Medical Center and NRH Research Center, collaborated to determine the nature of the functional connections linking the appropriate portions of LIFG and other brain regions in occipital and temporal cortex when an individual processes words and similar visually presented stimuli.

Barry Horwitz, Language Section, NIDCD, explained, "We tested the hypothesis that one part of the LIFG serves semantic processing and another part of the LIFG serves phonological processing by showing that the functional connections of each part differ depending on whether real words (which have meaning or semantics) were being processed, or whether word-like stimuli lacking semantics were being processed. We found different patterns of functional connections with other brain areas involved with reading in the occipital and temporal lobes, depending on which part of the LIFG we examined and which types of stimuli were used."

The team investigated the functional connectivity for patterns using word and word-like (pseudoword) stimuli, letter strings, and false fonts (visual symbols that could be letters, but aren't). Functional connectivity is based upon the notion that an experimental condition is mediated by a network of interacting brain regions and that different conditions use different functional networks. The investigators expected that the four different kinds of stimuli would elicit differing combinations of semantic and phonological processing.

To examine the response, the team looked at the correlation in brain activity, measured using fMRI, between different brain areas. This allowed for an identification of key nodes of activity in the network under study and, moreover, how these nodes are functionally connected.

As shown in figures in the paper, they found strong functional connectivity between activity in the lower part of LIFG and activity in occipital and temporal cortex only for words (the only stimuli type having semantics). They also found strong functional connectivity between activity in the upper part of LIFG and activity in occipital and temporal cortex for words, pseudowords, and letter strings (all of which have phonological features).

James F. Battey, Jr., M.D., Ph.D., Director of the NIDCD, noted that "Analysis of data supplied by powerful imaging strategies are giving us the power to not only identify activated functioning areas of the living brain, but permit scientists an elaborate view of how these activated areas interact with the rest of the brain."

The National Institute on Deafness and Other Communication Disorders is one of the Institutes of the National Institutes of Health. NIDCD is the focal point for conducting and supporting research and research training in the normal and disordered processes of human communication. NIDCD provides health information, based upon scientific discovery, to the public.

To interview Dr. Horwitz, or for additional information about the study or any aspect of human communication research, please contact the Office of Health Communication and Public Liaison, NIDCD, at 301-496-7243 (phone) or 301-402-0018 (fax).

To view data figures: www.nidcd.nih.gov/intram/scientists/materials/horwitz.asp.

Top