Presented at the 11th International Science of Aphasia Conference; Potsdam, Germany, 27 August–1 September 2010.

Introduction: Deficits in verb inflection by aphasic speakers constitute a valuable source of information towards understanding the language system by addressing issues and assumptions in theoretical linguistics. Previous studies (Friedmann & Grodzinsky, 1997; Wenzlaff & Clahsen, 2004; Burchert, Swoboda-Moll & de Bleser, 2005; Varlokosta, Valeonti, Kakavoulia, Lazaridou, Economou & Protopapas, 2006; Nanousi, Masterson, Druks & Atkinson, 2006) have suggested that grammatical morphemes are not equally disrupted in aphasia; rather, deficits appear to be selective, evidenced by poor performance in certain functional categories while other categories appear relatively spared. The present study was conducted as a follow-up of Varlokosta et al. (2006). In the previous study, Varlokosta et al. (2006) investigated production and perception of subject-verb agreement, tense, and aspect in Greek aphasic speakers, using a sentence completion task and a grammaticality judgement task. They provided evidence for selective deficits, in that participants performed relatively poorly in tense and aspect while performance on subject-verb agreement was comparatively higher. Varlokosta et al. proposed a structural-representational explanation based on the hypothesis of Wenzlaff & Clahsen (2004). However, a weakness in their experimental design precludes establishment of a highly confident conclusion. Specifically, the tests were not balanced in length across the three functional categories, thus confounding category with processing load, and allowing an alternative explanation for the significant differences among conditions in terms of sentence length rather than linguistic structure. The present study was designed to address this weakness, by specifically equating testing materials in the three functional categories on as many properties as possible, while also expanding the number of verbs used. We aimed to confirm the functional category differences having ruled out the possibility of length effects.
Participants: Ten individuals clinically diagnosed with aphasia participated in the study. They all had suffered a (unilateral) left hemisphere lesion at least four months prior to testing. A control group of 10 non-impaired participants were recruited, matched with the aphasic speakers in sex, age and education.
Materials and procedure: The experimental procedure consisted of two tasks: sentence completion and grammaticality judgement. There were three conditions in each task, addressing the participants’ performance on the three basic functional categories of Greek verb morphology: subject-verb agreement, tense, and aspect. Ten verbs were used in the sentences, controlled for phonological properties, regularity, and frequency (estimated via subjective familiarity). Further details about the design and interaction of the tasks may be found in Varlokosta et al. (2006). The new materials were balanced across functional categories for length of phrase (number of characters, M=48, SD=6.3, and number of words, M=8.6, SD=1.1) and number of words preceding the verb (M=4.9, SD=0.6).
Results: As a group, aphasic speakers made more errors than controls in each of the three conditions in both the sentence completion (by Mann-Whitney U test, 1-tailed exact significance; agreement: U=0, p=.000; tense: U=12, p=.003; aspect: U=0, p=.000) and the grammaticality judgement task (by Mann-Whitney U test, 1-tailed exact significance; agreement: U=12, p=.003; tense: U=0, p=.000; aspect: U=3, p=.000). As far as the selective deficit is concerned, the aphasic speakers’ group differences between categories did not reach statistical significance in the sentence completion task (by Wilcoxon signed ranks test, 2-tailed; agreement-tense: z=-1.072, n.s; tense-aspect z=-.308, n.s.; aspect-agreement: z=-1.485, n.s.). Nevertheless, in grammaticality judgement, aphasic speakers made more errors in tense and aspect than in agreement (by Wilcoxon signed ranks test, 2-tailed; agreement-tense: z=-2.395, p=.017; tense-aspect z=-.561, n.s.; aspect-agreement: z=-2.497, p=.013). Proportion of errors (per cent, relative to the total number of test items in each condition) made by each group in each condition of the sentence completion task and the grammaticality judgement task are shown in table 1.

Table 1. Proportion of errors (percent)
	Sentence completion			Grammaticality judgment
Group	Agreement	Tense	Aspect	Agreement	Tense	Aspect
Aphasics
M	32.3	40.3	40.6	19.4	42.5	40.0
SD	20.7	26.1	10.8	17.0	9.8	12.0
Controls
M	2.0	8.8	7.0	2.0	8.1	9.8
SD	2.2	15.7	5.6	2.3	6.0	6.7

Discussion: Our findings are not compatible with structural approaches of agrammatism, as they all predict a selective deficit. Furthermore, if the deficit found was due to a structural-representational account, it should be expected to appear not only in the grammaticality judgement task, but also in the sentence completion task for the same categories (Dickey, Milman & Thompson, 2008). These findings are compatible with a computational approach of agrammatism. One possible interpretation of the results, compared with the previous study (Varlokosta et al., 2006), is that if the sentences were balanced in terms of length, they were balanced in terms of computational load and thus the deficit was balanced across the categories.

References:
Burchert, F., Swoboda-Moll, M., & De Bleser, R. (2005). Tense and agreement dissociations in German agrammatic speakers: Underspecification vs. hierarchy. Brain and Language, 94, 188–199.
Dickey, M. W., Milman, L. H., & Thompson, C. K. (2008). Judgment of functional morphology in agrammatic Broca’s aphasia. Journal of Neurolinguistics, 21, 35–65.
Friedmann, N., & Grodzinsky, Y. (1997). Tense and agreement in agrammatic production. Pruning the syntactic tree. Brain & Language, 56, 71–90.
Nanousi, V., Masterson, J., Druks, J., & Atkinson, M. (2006). Interpretable vs. uninterpretable features: Evidence from six Greek-speaking agrammatic patients. Journal of Neurolinguistics, 19 (3), 209–238.
Varlokosta, S., Valeonti, N., Kakavoulia, M., Lazaridou, M., Economou, A., & Protopapas, A. (2006). The breakdown of functional categories in Greek aphasia: Evidence from agreement, tense, and aspect. Aphasiology, 20 (8), 723–743.
Wenzlaff, M., & Clahsen, H. (2004). Tense and agreement in German agrammatism. Brain & Language, 89, 57–68.