School for Advanced Studies in the Social Sciences

About: School for Advanced Studies in the Social Sciences is a facility organization based out in Villejuif, France. It is known for research contribution in the topics: Politics & Context (language use). The organization has 1230 authors who have published 2084 publications receiving 57740 citations. The organization is also known as: Ecole des Hautes Etudes en Sciences Sociales & EHESS.

Phonological deficits in specific language impairment and developmental dyslexia: towards a multidimensional model

Abstract: An on-going debate surrounds the relationship between specific language impairment and developmental dyslexia, in particular with respect to their phonological abilities. Are these distinct disorders? To what extent do they overlap? Which cognitive and linguistic profiles correspond to specific language impairment, dyslexia and comorbid cases? At least three different models have been proposed: the severity model, the additional deficit model and the component model. We address this issue by comparing children with specific language impairment only, those with dyslexia-only, those with specific language impairment and dyslexia and those with no impairment, using a broad test battery of language skills. We find that specific language impairment and dyslexia do not always co-occur, and that some children with specific language impairment do not have a phonological deficit. Using factor analysis, we find that language abilities across the four groups of children have at least three independent sources of variance: one for non-phonological language skills and two for distinct sets of phonological abilities (which we term phonological skills versus phonological representations). Furthermore, children with specific language impairment and dyslexia show partly distinct profiles of phonological deficit along these two dimensions. We conclude that a multiple-component model of language abilities best explains the relationship between specific language impairment and dyslexia and the different profiles of impairment that are observed.

Partially Verifiable Information and Mechanism Design

Abstract: In a principal-agent model with adverse selection, we study the implementation of social choice functions when the agent's message space is a correspondence which depends on this true characteristic. We characterize such correspondence for which the Revelation Principle is valid. returns must not contradict easily observable elements of the taxpayer's lifestyle. In this paper we present a model of two-person organizations in which these considerations are important. We ask whether the restrictions on fallacious statements inherent in the system are sufficient to achieve the goals of the organization. We show that the optimal way of eliciting the agent's response may not be to establish incentives for truthtelling in all instances. In our model there is one player, the "agent", who observes the state of the economic system. Another player, the "principal", takes an action based on the information presented to him by the agent. This class of models is often referred to as the principal-agent problem. In the extensive and growing literature on this problem, it has always been assumed that the agent could lie to the principal, and therefore that he had to be properly motivated to act at least partially in the principal's interest. Our formal way of modelling the partial verifiability of information is to introduce the restriction that the agent's responses must lie in a set M(O) that varies with the true state 0. It is assumed that this set always admits the possibility of responding with the true state. The way in which the allowable response set varies with the true state is known to the principal. In other respects we retain the usual principal-agent formalization. In particular, the principal is the Stackelberg leader in this game. He can commit himself to choose a collective action as a function of the message transmitted by the agent. Whether the variation of the message space with the true observation is purely technological, or whether it is induced by the severity of potential actions of the principal,

Linguistic Constraints on Statistical Computations The Role of Consonants and Vowels in Continuous Speech Processing

Abstract: Speech is produced mainly in continuous streams containing several words. Listeners can use the transitional probability (TP) between adjacent and non- adjacent syllables to segment ''words'' from a continuous stream of artificial speech, much as they use TPs to or- ganize a variety of perceptual continua. It is thus possible that a general-purpose statistical device exploits any speech unit to achieve segmentation of speech streams. Alterna- tively, language may limit what representations are open to statistical investigation according to their specific lin- guistic role. In this article, we focus on vowels and con- sonants in continuous speech. We hypothesized that vowels and consonants in words carry different kinds of infor- mation, the latter being more tied to word identification and the former to grammar. We thus predicted that in a word identification task involving continuous speech, learners would track TPs among consonants, but not among vowels. Our results show a preferential role for consonants in word identification.

A robust method to study stress "deafness".

Abstract: Previous research by Dupoux et al. [J. Memory Lang. 36, 406–421 (1997)] has shown that French participants, as opposed to Spanish participants, have difficulties in distinguishing nonwords that differ only in the location of stress. Contrary to Spanish, French does not have contrastive stress, and French participants are “deaf” to stress contrasts. The experimental paradigm used by Dupoux et al. (speeded ABX) yielded significant group differences, but did not allow for a sorting of individuals according to their stress “deafness.” Individual assessment is crucial to study special populations, such as bilinguals or trained monolinguals. In this paper, a more robust paradigm based on a short-term memory sequence repetition task is proposed. In five French–Spanish cross-linguistic experiments, stress “deafness” is shown to crucially depend upon a combination of memory load and phonetic variability in F0. In experiments 3 and 4, nonoverlapping distribution of individual results for French and Spanish participants is observed. The paradigm is thus appropriate for assessing stress deafness in individual participants.