No abstract

No abstract

Languages show differences in how they encode motion in discourse: Verbframed languages lexicalize Path in the verb, leaving Manner peripheral or implicit; Satellite-framed languages lexicalize Manner together with Path adjuncts. The present study investigates: 1) the extent to which such typological constraints affect the verbalizations of second language learners (English learners of French) and of aphasic speakers (English and French speakers with agrammatism) -who typically show dissociations between lexical and syntactic knowledge -in comparison to controls (English and French native speakers); as well as 2) the role of language-independent factors (level of acquisition, syndrome type). Despite some similarities between learners and speakers with aphasia due to language-independent factors, the findings suggest typologically constrained verbalizations in all groups, as well as diverging strategies that may reflect distinct underlying conceptualization processes.

In this paper, we report automatic pronunciation assessment experiments at phone-level on a read speech corpus in French, collected from 23 Japanese speakers learning French as a foreign language. We compare the standard approach based on Goodness Of Pronunciation (GOP) scores and phone-specific score thresholds to the use of logistic regressions (LR) models. French native speech corpus, in which artificial pronunciation errors were introduced, was used as training set. Two typical errors of Japanese speakers were considered: /ö/ and /v/ often mispronounced as [l] and [b], respectively. The LR classifier achieved a 64.4% accuracy similar to the 63.8% accuracy of the baseline threshold method, when using GOP scores and the expected phone identity as input features only. A significant performance gain of 20.8% relative was obtained by adding phonetic and phonological features as input to the LR model, leading to a 77.1% accuracy. This LR model also outperformed another baseline approach based on linear discriminant models trained on raw f-BANK coefficient features.

No abstract