Portail HAL EHESS (École des hautes études en sciences sociales)
Not a member yet
103682 research outputs found
Sort by
Brain size reduction in dogs was already established at least by the Late Neolithic of western Europe, 5,000 years ago
No full textThe timing and causes of brain size reduction in domestic dogs remain uncertain. Using endocast's volume as a proxy for brain size, this study provides a first insight into long-term brain size evolution in the wolf-dog lineage. We compared endocranial volumes of 185 modern and 22 prehistoric wolves and dogs ranging from Western Europe to Australia, and spanning the Pleniglacial (35 Ky BP) to the Late Neolithic (5 Ky BP). Our results reveal that Pleistocene so called "protodogs" show no brain size reduction compared to coeval Pleistocene wolves. Instead, we observed a slightly larger relative endocranial volume in the 35,000-year-old 'protodog' from Goyet, which could suggest increased behavioural flexibility in the presence of humans. This hypothesis needs to be tested further. In contrast, Late Neolithic dogs show a drastic 46% brain size reduction with an endocranial volumes comparable to modern small terrier and toy breeds. The anxious and wary temperaments of these Late Neolithic dogs, induced by the brain tissue reorganization associated with such a size reduction, could have served an alerting purpose, among the many other potential roles dogs could have played within this Late Neolithic socio-ecosystems.</div
Évaluer la traduction automatique à l’ère de l’IA : de la métrique à l’évaluation humaine (étude de cas français–chinois dans le domaine médical)
No full textInternational audienc
Philosophie du langage et de l’esprit: [résumé des cours et travaux : 2021-2022]
No full textInternational audienc
Préface - Imaginaires, technologies et enjeux éthiques à l’ère hypermoderne
No full textInternational audienceÀ l’heure où la transformation numérique reconfigure en profondeur les structures sociales, économiques et culturelles de nos existences, il semble impératif de réinterroger les régimes d’appropriation symbolique et critique des technologies. Ce que nous entendons ici explorer, sous un angle socio-philosophique et éthique, est la façon dont nos sociétés hypermodernes, prises dans l’étau d’une innovation effrénée, peuvent encore préserver les conditions d’une autonomie du jugement et d’une capacité de réinvention collective. Pour cela, il est essentiel de rappeler que les technologies ne sont jamais indépendantes de pratiques sociales qui s’inventent sans cesse et qui échappent aux conceptions souvent déterministes inhérentes à certains « grands récits ». Un enjeu consiste donc à reconnaître que si la stupéfaction l’emporte souvent face à l’expansion des technologies dans nos vies, les termes qui désignent ces technologies doivent quant à eux être questionnés. Ainsi, et de manière souvent abusive, le terme d’intelligence prospère dans de nombreuses sphères de l’innovation, non sans créer une confusion dans l’appréciation des impacts des technologies sur nos modes d’existence. Un effort de décryptage des techno-discours s’impose donc, mais il implique de bénéficier de clés de lecture des objets technologiques eux-mêmes qui sont par excellence des objets sociaux, imbriqués dans des jeux d’interactions complexes et plurielles. [...
针对法语母语学习者的汉语近义词教学
No full textInternational audienceCette étude vise à examiner les quasi-synonymes en chinois à partir d’une double perspective : celle des locuteurs natifs du chinois et celle des apprenants du chinois langue étrangère. En s’appuyant sur des analyses portant sur les morphèmes, le sens lexical et les différentes acceptions, elle analyse les types d’erreurs liées aux quasi-synonymes produites par des apprenants francophones dans le processus d’acquisition du lexique chinois, ainsi que leurs causes. Sur cette base, un cadre d’analyse combinant les dimensions sémantique, syntaxique et pragmatique est proposé et illustré par des exemples concrets. Enfin, en s’appuyant sur des travaux antérieurs en didactique ainsi que sur des recherches en acquisition actuellement en cours, l’étude propose plusieurs pistes pédagogiques pour l’enseignement des quasi-synonymes chinois auprès d’apprenants francophones.本研究旨在从汉语母语者视角与汉语作为外语学习者的视角出发，对汉语近义词进行探讨。研究以语素、词义与义项等层面为切入点，分析法语母语学习者在汉语词汇习得过程中所产生的近义词偏误类型及其成因，并在此基础上构建语义、句法与语用相结合的辨析框架，辅以具体实例加以说明。最后，结合前人的相关教学实验成果以及正在进行的二语习得研究，提出若干针对法语母语学习者的汉语近义词教学思路
Do Your Own Research (?) A Weak Link Between Conspiracism and Preference for First-Hand Evidence in a Perceptual Task
No full textInternational audienceConspiracy theorists often prompt others to “Do your own research.” Could a general desire for such epistemic autonomy—to make up one’s own mind rather than deferring to others’ testimony—explain why some people are attracted to conspiracy theories? In four pre-registered studies (United States and United Kingdom, N = 1196), we test whether participants more likely to believe in conspiracy theories have a stronger preference for forming their own beliefs independently. Participants chose between doing a difficult perceptual task themselves or relying on an expert’s answer. Internal fixed-effect meta-analyses revealed a weak but statistically significant relationship between belief in conspiracy theories and preference for first-hand evidence. By contrast, the relationship between epistemic individualism and this preference was stronger and more robust. This suggests that, although individuals endorsing conspiracy theories express a stronger preference for “doing their own research,” their behavior mostly does not match this preference in non-conspiratorial contexts
Numerical Analysis of Test Optimality
No full textIn nonstandard testing environments, researchers often derive ad hoc tests with correct (asymptotic) size, but their optimality properties are typically unknown a priori and difficult to assess. This paper develops a numerical framework for determining whether an ad hoc test is effectively optimal-approximately maximizing a weighted average power criterion for some weights over the alternative and attaining a power envelope generated by a single weighted average power-maximizing test. Our approach uses nested optimization algorithms to approximate the weight function that makes an ad hoc test's weighted average power as close as possible to that of a true weighted average power-maximizing test, and we show the surprising result that the rejection probabilities corresponding to the latter form an approximate power envelope for the former. We provide convergence guarantees, discuss practical implementation and apply the method to the weak-instrument-robust conditional likelihood ratio test and a recently-proposed test for when a nuisance parameter may be on or near its boundary
Acquiring numerical equality
No full textInternational audienceWe propose to analyze the notion of numerical equality as a complex, Numerical Equality, constituted of three facets ("same number" statements, numeral labels, and one-to-one pairings of collections) all mapping onto a same relation defined on collections, herein referred to as the relation of equal numerosity. With this framework in hand, we review experimental studies assessing whether children correctly map the three facets of Numerical Equality onto the relation of equal numerosity, and whether they understand that these three facets are interrelated. By and large, these studies indicate that children still struggle with various aspects of the Numerical Equality complex as they enter school -and provide no evidence that children represent the relation of equal numerosity itself before the age of 5 years. Our review also identifies significant gaps in the literature, leaving many aspects of children's acquisition of numerical equality open to research
"ICS for complex data with application to outlier detection for density data"
No full textNational audienceInvariant coordinate selection (ICS) is a dimension reduction method, used as a preliminary step for clustering and outlier detection. It has been primarily applied to multivariate data. This work introduces a coordinate-free definition of ICS in an abstract Euclidean space and extends the method to complex data. Functional and distributional data are preprocessed into a finite-dimensional subspace. For example, in the framework of Bayes Hilbert spaces, distributional data are smoothed into compositional spline functions through the Maximum Penalised Likelihood method. We describe an outlier detection procedure for complex data and study the impact of some preprocessing parameters on the results. We compare our approach with other outlier detection methods through simulations, producing promising results in scenarios with a low proportion of outliers. ICS allows detecting abnormal climate events in a sample of daily maximum temperature distributions recorded across the provinces of Northern Vietnam between 1987 and 2016