Can people with semantic variant primary progressive aphasia understand meaningful word parts in invented words?

Details

A morpheme is classically defined as the smallest meaning unit of language. Understanding derivational morphemes—word parts used to form derived words—helps people understand new or unfamiliar words. This study explored how semantic impairment affects comprehension of derivational morphemes in people with semantic variant primary progressive aphasia (svPPA). A few earlier studies had examined derivational morphology in svPPA (Kavé, Heinik, & Biran, 2012; Meteyard & Patterson, 2009), but had not specifically examined comprehension of derivational morphemes.

In this matched-group comparative task study, seven French-speaking people with svPPA and 14 controls matched for age, sex and education completed a word-to-definition matching task. Each target and its two foils were matched for token frequency and were morphologically complex, semantically transparent words formed from the base word included in the definition. There were six definition formats, with six items per format, for 36 items in total. For example, definitions in the format “De manière [mot de base]” (“In a [base word] way”) had targets formed with the suffix “-ment,” approximately equivalent to English “-ly.” Definitions were delivered orally through headphones and visually on a computer screen. Participants had to press a button to select the word matching the definition within eight seconds.

Because a derived word can be processed as a whole rather than as a base plus an affix, participants also completed a pseudo-word-to-definition matching task, in which choosing the correct answer required understanding the target morpheme. Each of the 36 pseudo-words was constructed in relation to a real-word item, and its foils used the same affixes as their real-word counterparts.

One svPPA participant’s pseudo-word results were excluded because it was uncertain whether she understood the instructions. People with svPPA were slower and less accurate than controls at selecting both words and pseudo-words that matched a definition. Within-group comparisons showed that controls selected correct answers faster for real words than for pseudo-words; the svPPA group showed no such advantage.

The real-word results alone cannot rule out that the svPPA group’s errors reflected difficulty understanding words rather than difficulty processing derivational morphemes. However, the lexical context in the real-word condition did not speed up performance in the svPPA group as it did in controls. Most importantly, the pseudo-word results showed that the svPPA group linked morphemes with their meanings less readily and reliably than controls. These findings support the involvement of semantic memory in morphological processing.