Papers by Adèle Hénot-Mortier
Proceedings of the 2024 Amsterdam Colloquium, 2024
Hurford Conditionals (HCs) involving scalemates appear felicitous, despite the fact that exh is n... more Hurford Conditionals (HCs) involving scalemates appear felicitous, despite the fact that exh is not predicted to rescue such structures from Super Redundancy, a principle introduced in Kalomoiros (2024) to capture HCs without scalemates. We thus propose an alternative to Super Redundancy based on two main ideas: (1) expressions evoke QuDs locally, and when they combine via logical operators so do their respective QuDs; (2) a local QuD Q that gets incrementally combined with an existing QuD Q ′ should have its maximal true answers "fit" the information structure already introduced by Q ′ . This predicts that the consequent of a conditional has to evoke a question that properly refines some question evoked by the antecedent. The pattern is then captured with the additional assumption that scalar items can evoke fine-grained enough questions (generated by their scalemates) out-of-the-blue, while non-scalar items conveying different degrees of granularity cannot.
Proceedings of the 54rd Annual Meeting of the North East Linguistic Society, 2025
Lingue e Linguaggio, 2025
French assigns grammatical gender (masculine or feminine) to nominals, and is endowed with a dimi... more French assigns grammatical gender (masculine or feminine) to nominals, and is endowed with a diminutive suffix -et/-ette. In most cases, the diminutive noun resulting from -et(te)-affixation will have the same gender as its base (Bally 1932), but there are a significant number of exceptions to this rule, that most of the previous literature Milner 1989 i.a.) took to be the result of lexicalization. In this study, we assess how frequent gender mismatches induced by et(te)-affixation are, in either direction (masculine to feminine and vice-versa), and what the exact semantic consequences turn out to be. In particular, we show that there exists a significant frequency asymmetry between -et-affixation and -ette-affixation, which affects both gender-matching and gender-mismatching base-derivative pairs, supporting the idea that gender-mismatching diminutives are to a certain extent morphologically transparent, but also that -ette-affixation may receive an analysis distinct from that of -et-affixation. We provide a analysis within the framework of Distributed Morphology (Halle & Marantz 1993) that is in line with the statistical data and with recent cross-linguistic findings on diminitive an augmentative affixes, according to which such elements may vary in place and manner of attachment, across, and also within, languages .
Proceedings of the 2023 CLASP Conference on Learning with Small Data (LSD), 2023
Large Language Models (LLMs) are often evaluated against massive benchmarks based on general-purp... more Large Language Models (LLMs) are often evaluated against massive benchmarks based on general-purpose tasks, which, despite being useful for concrete applications, tell us very little about the capacity of LLMs to learn specific and challenging aspects of the grammar. Here, we evaluate whether LLMs learn to identify a particular structure attested in Romance (and French in particular), called the pseudorelative. This structure, which is often surface-similar to a relative clause, is linked to robust syntactic and semantic restrictions. We present a series of experiments to test if LLMs pretrained on massive yet general corpora, manage to learn those various restrictions. Our results suggest that LLMs learn some but not all of these properties, but crucially fail at recognizing the most specific of them: cliticization. Pseudorelatives (PRs) ; ; Guasti (1988) a.o.) resemble relative clauses (RCs) but exhibit a specific cluster of properties: (1) their head noun can be cliticized; (2) they only feature subject gaps; (3) they only appear below perception verbs; (4) they require the matrix and embedded tenses to match; (5) they imply the existence/truth of the embedded event even under matrix negation . Those various properties are illustrated below. (1) Head noun cliticization Jean Jean la 3.SG.CL voit sees qui that sourit. smiles. 'Jean sees her smiling.' (2) Object gap (+cliticization) * Jean Jean la 3.SG.CL voit sees que that Marc Marc salue greets. . Intended: 'Jean sees Marc greeting her.' (3) Non-perception verb (+cliticization) * Jean Jean la 3.SG.CL pense thinks qui that sourit. smiles.
Proceedings of the 28th Sinn und Bedeutung Conference, 2024
Antonymic adjectives are subject to a variety of asymmetries regarding pragmatic inferences. The ... more Antonymic adjectives are subject to a variety of asymmetries regarding pragmatic inferences. The Inference Towards the Antonym in particular, consists in deriving the antonym of an adjective A when encountering its negation (not A). Within a given antonymic pair, this inference is supposed to apply to a greater extent to negated positive adjective, as opposed to negated negative adjectives. This is especially true when the latter is morphologically transparent. In this paper, we test if recent Large Language Models capture this contrast using different probing methods. We conclude that some but not all models exhibit a contrast between positive and negative adjectives regarding the target inference, although (i) the observed contrasts are not readily interpretable at the level of word processing (ii) part of it may be explained by frequency differences (iii) more general expectations about the models' behavior regarding antonymic adjectives (parsing, reversing effect of negation) are not met. This casts doubt on the ability of such models to abstractly encode the concept of antonymy.
Proceedings of the 25th Seoul International Conference on Generative Grammar, 2024
Tough-constructions (henceforth TC, Rosenbaum 1967 a.o.) involve a subjective predicate (like tou... more Tough-constructions (henceforth TC, Rosenbaum 1967 a.o.) involve a subjective predicate (like tough, easy, impossible, important, annoying) embedding an infinitival clause, and taking an optionally overt EXPERIENCER argument, as shown in (1) for English. In both English and French, TCs famously alternate between a "gapped" variant (1a)/(2a) and an "impersonal" variant (1b)/(2b). The former features a DP-like subject and a gapped infinitival clause. The latter features a subject proform (English it, French il/c'), and a gapless infinitive. Unlike English, which makes use of the clause-introducing preposition to across the board, French exhibits an alternation between two prepositions: à in gapped TCs (2a) vs. de in impersonal TCs (2b), the latter being optionally preceded by the complementizer que. 2 We dub gapped TCs like (2a) à-TCs, and impersonal TCs like (2b), de-TCs. What is the reason behind the prepositional alternation in French TCs, and what further differentiates them from English TCs?
Proceedings of the 47th Penn Linguistics Conference, 2024
The puzzle 1.1 The Plural Pronoun Construction and its inclusive reading When the associate is a ... more The puzzle 1.1 The Plural Pronoun Construction and its inclusive reading When the associate is a plural pronoun, as in (2), the comitative construction is dubbed Plural Pronoun Construction (PPC, Schwartz (1985, 1988); ; Aissen (1989) a.o.). In French, the PPC allows for two distinct readings: an exclusive ("E") reading and an inclusive ("I") one. Following Feldman (2001) we call PPCs under the I-(resp. E-) reading IPPCs (resp. EPPCs). IPPCs in particular, have been documented in Mokilese,
Proceedings of the 53rd Annual Meeting of the North East Linguistic Society, 2023
Proceedings of the 24th Seoul International Conference on Generative Grammar, 2023
Proceedings of the 27th Sinn und Bedeutung Conference, 2023
Hurford Disjunctions (HDs) are infelicitous disjunctions in which one disjunct entails the other ... more Hurford Disjunctions (HDs) are infelicitous disjunctions in which one disjunct entails the other . The infelicity of basic HDs has been successfully modeled by several competing approaches . As first noticed by Singh (2008) however, HDs involving entailing scalar items like all and some are subject to an asymmetry: when the weaker scalar item linearly precedes the stronger one, the sentence seems to be rescued from infelicity. This fact is not readily accounted for by standard approaches, which treat the disjuncts in a symmetric fashion. Fox and Spector (2018) and Tomioka (2021) proposed different solutions to that problem and extensions thereof, but at the cost of positing relatively heavy and complex machineries. Here we propose a novel analysis of Singh's asymmetry, based on the familiar process of alternative pruning Crnič et al. 2015 a.o.). In particular, we claim that exhaustification targeting the weak disjunct operates on a set of formal alternatives that is sensitive to previously uttered material. This leads us to propose a new dynamic constraint on alternative pruning, which ensures that the only remaining alternatives to a prejacent p are those which could be realistically entertained instead of p, given the eventualities previously and overtly raised by the speaker. Unlike other approaches, our account derives Singh's asymmetry via a direct computation, and not a global principle constraining either the insertion of the exhaustivity operator (Fox and Spector 2018), or the particular shape of the alternative set (Tomioka 2021).
A numerical algorithm of parameter estimation for dynamic model for HIV infection of CD4+ t cells
The problem of approximation the values of 6 coefficients of a HIV infection basic model by knowi... more The problem of approximation the values of 6 coefficients of a HIV infection basic model by knowing how viremia and target cells blood concentrations change as time goes on is considered. Above inverse problem for HIV dynamic model is reduced to the problem of minimization a least squares misfit function. A 6-dimensional Nelder-Mead algorithm is implemented to approximate this minimum. For a narrow interval to mean values of unknown coefficients (about 33% of the total bandwidth between experimental extreme coefficients values), proposed algorithm demonstrates a good convergence, with a probability of 17% for the Nelder-Mead algorithm. Results of numerical experiment are demonstrated and discussed.
Word-Embeddings Distinguish Denominal and Root-Derived Verbs in Semitic
Electronic Proceedings in Theoretical Computer Science
Why is our language vague? We argue that in contexts in which a cooperative speaker is not perfec... more Why is our language vague? We argue that in contexts in which a cooperative speaker is not perfectly informed about the world, the use of vague expressions can offer an optimal tradeoff between truthfulness (Gricean Quality) and informativeness (Gricean Quantity). Focusing on expressions of approximation such as "around", which are semantically vague, we show that they allow the speaker to convey indirect probabilistic information, in a way that gives the listener a more accurate representation of the information available to the speaker than any more precise expression would (intervals of the form "between"). We give a probabilistic treatment of the interpretation of "around", and offer a model for the interpretation and use of "around"-statements within the Rational Speech Act (RSA) framework. Our model differs in substantive ways from the Lexical Uncertainty model often used within the RSA framework for vague predicates.
Why is our language vague? We argue that in contexts in which a cooperative speaker is not perfec... more Why is our language vague? We argue that in contexts in which a cooperative speaker is not perfectly informed about the world, the use of vague expressions can offer an optimal tradeoff between truthfulness (Gricean Quality) and informativeness (Gricean Quantity). Focusing on expressions of approximation such as "around", which are semantically vague, we show that they allow the speaker to convey indirect probabilistic information, in a way that gives the listener a more accurate representation of the information available to the speaker than any more precise expression would (intervals of the form "between"). We give a probabilistic treatment of the interpretation of "around", and offer a model for the interpretation and use of "around"-statements within the Rational Speech Act (RSA) framework. Our model differs in substantive ways from the Lexical Uncertainty model often used within the RSA framework for vague predicates.
Conference Presentations by Adèle Hénot-Mortier
45th Generative Linguistics in the Old World (GLOW), 2022
* Many thanks to my current advisors Athulya Aravind, Martin Hackl and David Pesetsky for their p... more * Many thanks to my current advisors Athulya Aravind, Martin Hackl and David Pesetsky for their precious comments and advice on that project; thanks to my office-mate Christopher Baron for his useful English judgments on complex (tough!) sentences, and to my colleagues and office-mates Omri Doron, Filipe Kobayashi, Keely New, Margaret Wang for their semantic intuitions; thanks finally to all the people who attended the 04/20 LF-reading group at MIT for their insightful questions and comments. All mistakes are mine. 1 It has been shown that some nouns and verbs can behave like prettyor tough-predicates [Lasnik and Fiengo, 1974, Pesetsky, 1987, Gluckman, 2019], but for the sake of simplicity we will focus on adjectival predicates in that talk.
Uploads
Papers by Adèle Hénot-Mortier
Conference Presentations by Adèle Hénot-Mortier