Shared mechanisms in pragmatic enrichment with contextual and lexical alternatives [ with Nadine Bade, Léo Picat, Rachel Dudley, and Salvador Mascarenhas; accepted with minor revisions at Journal of Memory and Language ]

We explore the role of contextual versus lexical alternatives in pragmatic strengthening using a novel training-with-feedback paradigm. In two experiments, we investigate whether training with inferences over contextual alternatives affects pragmatic strengthening with lexical alternatives, and the other way around. In both Experiments, we find that training that encouraged (or discouraged) pragmatic strengthening of simple disjunctions carried over to complex disjunctions of an unfamiliar kind to our experimental participants. This shows that our novel methodology is effective in training general mechanisms for activating alternatives. In Experiment 2, we showed that this methodology can be made to work across different kinds of alternatives, if certain salience conditions are met.

The Conjunction Fallacy: Confirmation or Relevance? [ with Kevin Dorst, Matthew Mandelkern, and Salvador Mascarenhas; Thinking and Reasoning ]

The conjunction fallacy is the well-documented empirical finding that subjects sometimes rate a conjunction A&B as more probable than one of its conjuncts, A. Most explanations appeal in some way to the fact that B has a high probability. But Tentori et al. (2013) have recently challenged such approaches, reporting experiments which find that (1) when B is confirmed by relevant evidence despite having low probability, the fallacy is common, and (2) when B has a high probability but has not been confirmed by relevant evidence, the fallacy is less common. They conclude that degree of confirmation, rather than probability, is the central determinant of the conjunction fallacy. In this paper, we address a confound in these experiments: Tentori et al. (2013) failed to control for the fact that their (1)-situations make B conversationally relevant, while their (2)-situations do not. Hence their results are consistent with the hypothesis that conversationally relevant high probability is an important driver of the conjunction fallacy. Inspired by recent theoretical work that appeals to conversational relevance to explain the conjunction fallacy, we report on two experiments that control for this issue by making B relevant without changing its degree of probability or confirmation. We find that doing so increases the rate of the fallacy in (2)-situations, and leads to comparable fallacy-rates as (1)-situations. This suggests that (non-probabilistic) conversational relevance indeed plays a role in the conjunction fallacy, and paves the way toward further work on the interplay between relevance and confirmation.

Modality, expected utility, and hypothesis testing [ with Salvador Mascarenhas; Synthese 202(11) ]

We introduce an expected value theory of linguistic modality that makes reference to expected utility and a likelihood-based confirmation measure for deontics and epistemics, respectively. The account is a probabilistic semantics for deontics and epistemics, yet it proposes that deontics and epistemics share a common core modal semantics, as in traditional possible-worlds analysis of modality. We argue that this account is not only theoretically advantageous, but also has far-reaching empirical consequences. In particular, we predict modal versions of reasoning fallacies from the heuristics and biases literature. Additionally, we derive the modal semantics in an entirely transparent manner, as it is based on the compositional semantics of Korean modal expressions that are morphosyntactically decomposed into a conditional and an evaluative predicate.

Local context is calculated domain by domain, for each maximal projection [ Glossa 8(1) ]

Theories of local contexts that either rely on strict linear order (Schlenker 2009) or hierarchy (Ingason 2016, Romoli and Mandelkern 2017) face a number of challenges, most notably in nominal modification (Schlenker 2020) and partitives (Anvari and Blumberg 2021) but also, depending on the theory, in accounting for belief reports and coordination in head-final languages. To resolve such issues, I propose a hybrid theory which takes into account both linear order and syntactic structure: local context is calculated domain by domain, and for each maximal projection.

Question-answer dynamics in deductive fallacies without language [ with Nadine Bade, Sam Blanc-Cuenca, and Salvador Mascarenhas; In Proceedings of the 44th Annual Cognitive Science Society Meeting. Presented as a talk at CogSci 2022 ]

We introduce purely visual paradigms that convey the logical structure of illusory inferences from disjunction: (a ∧ b) ∨ c, a ⊢? b. Although the logical information was conveyed entirely via non-linguistic means, we found that the visual paradigmsinduce reasoning fallacies, though less attractive than their linguistic counterparts. The visual paradigms highlight the role of alternative-based reasoning, or question-answer dynamics, as they control for narrowly interpretive processes that confound the study of their linguistic counterparts. To our knowledge, this is the first work to develop visual paradigms that representreasoning fallacies committed by adults and involve multiple logical operators non-trivially embedded. Previous studies focused on pre-verbal children or non-human animals, and for this reason limited the scope of research to visually representing logically simple, valid inferences.

Alternatives and attention in language and reasoning: A reply to Mascarenhas & Picat (2019) [ with Nadine Bade, Léo Picat, and Salvador Mascarenhas; Semantics and Pragmatics 15(2) ]

In this paper, we employ an experimental paradigm using insights from the psychology of reasoning to investigate the question whether certain modals generate and draw attention to alternatives. The article extends and builds on the methodology and findings of Mascarenhas & Picat (2019). Based on experimental results, they argue that English epistemic modal might raises alternatives. We apply the same methodology to English modal allowed to to test different hypotheses regarding the involvement of alternatives in deontic modality. We find commonalities and differences between the two modals we tested. We discuss theoretical consequences for existing semantic analyses of these modals, and argue that using reasoning tasks can serve as a diagnostic tool for exploring the issue of which natural language expressions involve alternatives.

Intervention in Deontic Reasoning [ Semantics and Pragmatics 13(16) ]

The 'if p, ought p' problem, famously known as Zvolenszky’s puzzle (Zvolenszky 2002), questions how possible worlds semantics can systematically distinguish non-defective cases of ‘if p, ought p’ from defective ones. This paper suggests that it is not a problem of possible worlds semantics of modality, but rather, the ‘if p, ought p’ problem reveals the counterfactual nature of deontic modals which otherwise would have gone unnoticed. I propose that a counterfactual-based formulation of deontic necessity that implements intervention, jointly with the assumption that indicative conditionals facilitate backtracking, offers a deterministic account of the ‘if p, ought p’ problem. I also present empirical evidence in favor of an interventionist approach to counterfactuals as opposed to similarity-based theories, at least in the domain of deontic reasoning.

Decomposing Deontic Modality: Evidence from Korean [ Journal of Semantics 36(4) ]

Korean deontic modal expressions inform about the composition of modality which is not evident from languages that express modal concepts via an auxiliary or a verb. They are expressed in terms of a conditional and an evaluative predicate: permission and obligation respectively translate to ‘even if φ, good’ and ‘only if φ, good’, where φ is the prejacent. Their transparent morphosyntax lets us probe deeper into the interior of deontic modal concepts. I propose that Korean deontic modal expressions do not set up the domain of quantification in which the prejacent is evaluated. Instead, they assess the goodness of the prejacent and its alternatives, all else being equal. I allude to the possibility that English obligation also involves similar reasoning by offering a counterfactual-based account of the Professor Procrastinate puzzle. On the other hand, Korean permission gives rise to a specfic interpretation of strong permission, conveying that the truth or falsity of the prejacent does not affect the deontic status. I additionally point out that Korean has a separate copula-based construction for expressing modal possibility, which hints that there are in fact two types of permission in natural language.

The Lifted Matrix-Space Model for Semantic Composition [ CoNLL 2018 Proceedings ]

Tree-structured neural network architectures for sentence encoding draw inspiration from the approach to semantic composition generally seen in formal linguistics, and have shown empirical improvements over comparable sequence models by doing so. Moreover, adding multiplicative interaction terms to the composition functions in these models can yield significant further improvements. However, existing compositional approaches that adopt such a powerful composition function scale poorly, with parameter counts exploding as model dimension or vocabulary size grows. We introduce the Lifted Matrix-Space model, which uses a global transformation to map vector word embeddings to matrices, which can then be composed via an operation based on matrix-matrix multiplication. Its composition function effectively transmits a larger number of activations across layers with relatively few model parameters. We evaluate our model on the Stanford NLI corpus, the Multi- Genre NLI corpus, and the Stanford Sentiment Treebank and find that it consistently outperforms TreeLSTM (Tai et al., 2015), the previous best known composition function for tree-structured models.