Slower-than-normal syntactic processing in agrammatic Broca's aphasia: Evidence from Dutch

Abstract

Studies of agrammatic Broca's aphasia reveal a diverging pattern of performance in the comprehension of reflexive elements: offline, performance seems unimpaired, whereas online—and in contrast to both matching controls and Wernicke's patients—no antecedent reactivation is observed at the reflexive. Here we propose that this difference characterizes the agrammatic comprehension deficit as a result of slower-than-normal syntactic structure formation. To test this characterization, the comprehension of three Dutch agrammatic patients and matching control participants was investigated utilizing the cross-modal lexical decision (CMLD) interference task. Two types of reflexive-antecedent dependencies were tested, which have already been shown to exert distinct processing demands on the comprehension system as a function of the level at which the dependency was formed. Our hypothesis predicts that if the agrammatic system has a processing limitation such that syntactic structure is built in a protracted manner, this limitation will be reflected in delayed interpretation. Confirming previous findings, the Dutch patients show an effect of distinct processing demands for the two types of reflexive-antecedent dependencies but with a temporal delay. We argue that this delayed syntactic structure formation is the result of limited processing capacity that specifically affects the syntactic system

Processing Clitic Pronouns in Bulgarian — Evidence from Normal and Agrammatic Comprehension

“…Ruigendijk et al 2006). On the other hand, we also find cross-linguistic evidence that aphasic individuals do not experience more difficulties in comprehending personal than reflexive pronouns, both being quantitatively similarly impaired (Edwards and Varlokosta 2007) The results of the present study support the hypothesis that the obtained pattern of clitic processing is based on a slowed-down implementation of narrow syntax knowledge (Burkhardt et al 2008) in Bulgarian Broca's aphasics. Syntactic binding of reflexive clitics provides for immediate reference establishment.…”

“…The overt realisation of a higher or lower copy depends on phonological constraints like enclitisation requirements. In line with the slow-syntax-hypothesis (Burkhardt et al 2008) and with the assumption that prosody-related processes may also compete for the same limited processing resources of Broca's aphasics , we test sensitivity to the phonosyntactic constraints negation imposes on the word order of personal and reflexive clitics. Results suggest that the pattern of agrammatic processing of clitic clusters resembles normal comprehension but proceeds in a protracted manner.…”

“…The comparison of normal processing data with language breakdown data from agrammatic aphasics may yield valuable insights in two respects. First, under the assumption that agrammatism is due not to a general syntactic deficit, but to a slowed-down implementation of syntactic operations (Burkhardt et al 2008;Burkhardt 2005;Haarman and Kolk 1994;and Piñango 2002, inter alia), we expect the agrammatic data to work as a magnifying lens to the healthy system. Through such a comparison, we can learn more about the relative influence of syntactic and prosodic features on clitic placement.…”

Reference assignment in Dutch: Evidence for the syntax–discourse divide

“…13 For the analyses of Experiment 1, we collapsed data from two studies that were carried out separately. The first study was conducted with students from Utrecht University and the second study served as a control for research with agrammatic aphasic speakers (see Burkhardt et al, 2008). Both studies used the same experimental design and stimuli, the only difference being that the latter study included 35 experimental items per condition.…”

“…We refer to reflexives of this sort as d-linked reflexives. Note that in these latter cases, a discourse dependency is established only after syntactic information has been processed and the absence of coargumenthood has been determined (see Burkhardt et al, 2008 for evidence from impaired language comprehension for the key role of syntactic processing).…”