| Summary: | Numerous studies have documented difficulties in verbal expression and auditory processing of morphosyntax, including verb morphology, in persons with agrammatic aphasia (PWA) (e.g., Bastiaanse & Thompson, 2012; Faroqi-Shah & Dickey, 2009; Tyler et al., 2002). It is not known if difficulties with verb morphology are a downstream effect of early (sub)lexical impairments, or are restricted to higher level syntactic-semantic processes. Moreover there is little research on orthographic decomposition in agrammatic PWA. Evidence regarding sub(lexical) sensitivity to morphological complexity in PWA is equivocal and is restricted to auditory processing (Longworth et al., 2005; Tyler et al., 2000). The main question of this study was whether agrammatic PWA differ from neurotypical adults in neural processing of morphological structure and lexical access. This question was addressed by investigating spatiotemporal response for visually presented words using magnetoencephalography (MEG) – a technique sensitive to lexical subroutines (Pylkkanen & Marantz, 2003).
In neurologically healthy young adults, visually presented words undergo early and automatic pre-lexical decomposition, even pseudoaffixes such as “brother” (Marslen-Wilson & Tyler, 2007). Neural sensitivity to morphological complexity emerges at 200-600 milliseconds (ms) in left frontal regions, and to lexical retrieval in left posterior regions around 300-400 ms (e.g., Pylkkanen & Marantz, 2003; Lavric et al., 2012). There is some equivocal evidence for pre-200ms effects in posterior regions for morphological complexity (Fruchter et al., 2013).
Methods
The participants were 17 young (mean age=21.5 years, 10 females) neurotypical adults and 7 English-speaking agrammatic PWA following left perisylvian stroke (mean age=51.3 years, 2 females; mean Western Aphasia Battery AQ= 72.7, Mean proportion grammatical utterances=0.15). The experimental procedure involved visual lexical decision of four word types: real words and pseudowords that were morphologically simple or inflected (e.g., ride, zide, riding, ridest). Neuromagnetic (MEG) responses were measured using a 160-channel whole-head axial gradiometer and analyzed for the effects of morphological status (inflected[riding, ridest] > simple[ride, zide] and lexical status (real[ride, riding] > pseudowords[zide, ridest]) at Bonferroni-corrected p<.05.
Results and Discussion
The spatial distribution of the neuromagnetic response was similar for neurotypical adults and PWA, with significantly greater bilateral posterior activity pre-200ms followed by left posterior activity during 200-400ms. However, the two groups differed in sensitivity to morphological complexity and lexical status. For morphological complexity, the neurotypical effect first appeared in left anterior sensors at 150-200ms; PWA showed this left anterior sensitivity at 200-400ms, indicative of a delayed response. Further, a right lateralized response to morphological complexity found in neurotypical adults at 200-400ms was absent in PWA. For lexical status, the effect appeared at 150-200ms in neurotypical adults in bilateral anterior and right posterior sensors while PWA showed no effect during this time window. At 200-400ms, lexical status effects were left-lateralized for neurotypical adults and were absent for PWA. To summarize, agrammatic PWA show delayed left anterior response for morphological complexity, and insensitivity to all other contrasts. This finding of early morpho-lexical parsing abnormalities needs to be integrated with findings of sentence level morpho-syntactic abnormalities (e.g., functional categories) to develop a more comprehensive theoretical account of agrammatism.
|
|---|
