Special Coverage

Capturing the Aging Lexicon Using Network Science Techniques: Implications for Dementia Aphasia (SYM34)

Nichol Castro (University of Buffalo)

Summary by Taylor Curley, Digital Content Associate Editor

What Can the Science of Word Networks Tell Us About Dementia?

Network science, which quantifies the connectivity of related items using space-based models, has become an increasingly important tool for cognitive psychologists, particularly for language research. Much previous work in this domain was limited to theoretical investigations of connectivity in lexical networks.

For Nichol Castro, network science provides an opportunity to detect and measure dementia and age-related language disorders.

The key is to understand how lexical networks change with normal aging. For example, as we age, word networks are more tightly clustered together, causing the exclusion of some related words. As shown in the figure below, people around 30 years old show large word clusters while people around 70 years have much smaller word clusters, with many related words on the periphery. Thus, older adults show normative differences in the structure of lexical networks compared to young adults.

Researchers can use these normal changes in lexical networks to locate significant deviations, such as rapid declines in letter and category fluency, that could signal dementia for aphasia. Network science estimates can then identify the structural changes in lexical networks that give rise to abnormalities, such as degradations in the network structure.

Castro’s work represents a new way of understanding aging (both normal and abnormal). It also provides implications for future clinical uses.