Why is the meaning of a sentence better remembered than its form? An fMRI study on the role of novelty-encoding processes.

Episodic memory is based primarily on meaning. This is behaviorally well documented in studies on memory for prose, in which the meaning of novel sentences is typically well remembered but information pertaining to exact wording and syntax is not. The neural basis of this 'verbatim effect' is poorly understood. In the current fMRI study, we manipulated the novelty of sentences at different levels to test whether medial temporal lobe (MTL) regions that are known to play a critical role in verbal episodic encoding would respond preferentially to the novelty of sentence meaning. Fifteen participants were pre-familiarized with auditory sentences describing unique episodes. During scanning, they encountered sentences that were old, that contained a change in (i.e., were novel in terms of) syntactic relationships, that contained a change in semantic relationships, or that described an entirely novel episode. Subsequently, participants performed a recognition memory test for the different types of novel information encountered. Behavioral data confirmed the typical verbatim effect. Analyses of fMRI data revealed differential MTL activation in the left hippocampus and entorhinal cortex with a response profile across conditions that paralleled the behavioral results; the identified region responded selectively to those conditions that contained semantic novelty. Other regions, by contrast, showed a novelty response that did not share this selectivity. Our findings suggest that the verbatim effect can be linked to hippocampally-based novelty-assessment processes that operate based on semantic relationships.

Language deficits in dyslexic children: speech perception, phonology, and morphology.

We investigated the relationship between dyslexia and three aspects of language: speech perception, phonology, and morphology. Reading and language tasks were administered to dyslexics aged 8-9 years and to two normal reader groups (age-matched and reading-level matched). Three dyslexic groups were identified: phonological dyslexics (PD), developmentally language impaired (LI), and globally delayed (delay-type dyslexics). The LI and PD groups exhibited similar patterns of reading impairment, attributed to low phonological skills. However, only the LI group showed clear speech perception deficits, suggesting that such deficits affect only a subset of dyslexics. Results also indicated phonological impairments in children whose speech perception was normal. Both the LI and the PD groups showed inflectional morphology difficulties, with the impairment being more severe in the LI group. The delay group's reading and language skills closely matched those of younger normal readers, suggesting these children had a general delay in reading and language skills, rather than a specific phonological impairment. The results are discussed in terms of models of word recognition and dyslexia.

Impairments in verb morphology after brain injury: a connectionist model.

The formation of the past tense of verbs in English has been the focus of the debate concerning connectionist vs. symbolic accounts of language. Brain-injured patients differ with respect to whether they are more impaired in generating irregular past tenses (TAKE-TOOK) or past tenses for nonce verbs (WUG-WUGGED). Such dissociations have been taken as evidence for distinct "rule" and "associative" memory systems in morphology and against the connectionist approach in which a single system is used for all forms. We describe a simulation model in which these impairments arise from damage to phonological or semantic information, which have different effects on generalization and irregular forms, respectively. The results provide an account of the bases of impairments in verb morphology and show that these impairments can be explained within connectionist models that do not use rules or a separate mechanism for exceptions.

Specific language impairment: a deficit in grammar or processing?

Specific Language Impairment (SLI) is observed in children who fail to acquire age-appropriate language skills but otherwise appear to be developing normally. There are two main hypotheses about the nature of these impairments. One assumes that they reflect impairments in the child's innate knowledge of grammar. The other is that they derive from information-processing deficits that interfere with several aspects of language learning. There is considerable evidence that SLI is associated with impaired speech processing; however, the link between this deficit and the kinds of grammatical impairments observed in these children has been unclear. We suggest that the link is provided by phonology, a speech-based code that plays important roles in learning linguistic generalizations and in working memory.