Our Research

The mind represents abstract magnitude information, including time, space, and number, but in what format is this information stored? We show support for the bipartite format of perceptual magnitudes, in which the measured value on a dimension is scaled to the dynamic range of the input, leading to a privileged status for values at the lowest and highest end of the range. In six experiments with college undergraduates, we show that observers are faster and more accurate to find the endpoints (i.e., the minimum and maximum) than any of the inner values, even as the number of items increases beyond visual short-term memory limits. Our results show that length, size, and number are represented in a dynamic format that allows for comparison-free sorting, with endpoints represented with an immediately accessible status, consistent with the bipartite model of perceptual magnitudes. We discuss the implications for theories of visual search and ensemble perception.

A prominent strand of theorizing in linguistics models meaning in language by specifying an "interpretation function" which relates morphosyntactic objects (i.e., those representations whose properties are uncovered by research in morphology and syntax) to elements of non-linguistic experience. Such theorizing has, for the most part, proceeded in relative isolation from developments in the other cognitive sciences. A recent body of experimental work growing out of this tradition has, however, pressed the question of precisely how linguistic representations relate to other faculties of mind. We present the beginnings of a two-step formal proposal for how to do this, specifying: (i) the co-domain of the linguistic interpre- tation function as the Language of Thought (LoT); (ii) what this mental language is like, (iii) which expression of this language is the semantic value of a sentence like 'Most of the dots are yellow', and (iv) how that LoT expression is interpreted by other cognitive faculties, in ways that produce the choices of verification procedure that have been empirically observed.

English-acquiring children before 4 years of age show a fine-grained understanding of how the meaning of more interacts with the lexical semantics of nouns: if the noun expresses a concept of objects, the comparison is based on number; if it expresses a concept of substance, it is based on volume or area. Is the meaning that children have acquired sufficiently general to support parallel semantic sensitivities when more combines with verbs? We probe this question with 4-5 year olds. Our expectation, based in semantic theory, is that more combined with an 'event' verb like jump should be quantified by number, but with a 'process' verb like walk it should be more flexible. Our Experiment 1 tests this with adults and Experiment 2 with children. We find children’s understanding to be broadly consistent with that of adults, providing initial support for an early-acquired, highly general meaning for more.

On standard analyses, indicative conditionals like if A, then B behave in a Boolean fashion when interacting with and and or. We test this prediction by investigating probability judgments about sentences of the form If A, then B {and, or} if C, then D. Our findings are incompatible with a Boolean picture. This is challenging for standard analyses of indicative conditionals, as well as for several nonclassical analyses. Some trivalent theories, conversely, may account for the data.

Some linguistic features are more readily learned than others, and are thereby more likely to be maintained in diachronic language change, giving rise to typological universals. Less readily learned features may give rise to typological gaps. We consider an apparent typological gap—that a morphologically superlative determiner (e.g., gleebest in "gleebest of the cows") with a negative meaning is cross-linguistically unattested—and ask whether it reflects an underlying learning bias. We find 4-year-olds know that such determiners indicate quantity (replicating Wellwood, Gagliardi, & Lidz, 2016), but only when positive ('most'), but not negative ('least'). Importantly, the observed bias is not specific to the apparent typological gap: same-age children showed difficulty learning the negative meaning of a non-superlative determiner, though such meanings are attested. The data thus suggest that children are generally biased against negativity, consistent with much prior work on conceptual bias and language learning/processing.

Natural languages like English connect pronunciations with meanings. Linguistic pronunciations can be described in ways that relate them to our motor system (e.g., to the movement of our lips and tongue). But how do linguistic meanings relate to our nonlinguistic cognitive systems? As a case study, we defend an explicit proposal about the meaning of most by comparing it to the closely related more: whereas more expresses a comparison between two independent subsets, most expresses a subset–superset comparison. Six experiments with adults and children demonstrate that these subtle differences between their meanings influence how participants organize and interrogate their visual world. In otherwise identical situations, changing the word from most to more affects preferences for picture–sentence matching (experiments 1–2), scene creation (experiments 3–4), memory for visual features (experiment 5), and accuracy on speeded truth judgments (experiment 6). These effects support the idea that the meanings of more and most are mental representations that provide detailed instructions to conceptual systems.

This paper investigates the semantics of implicit comparatives (Alice is tall compared to Bob) and its connections to the semantics of explicit comparatives (Alice is taller than Bob) and sentences with adjectives in plain positive form (Alice is tall). We consider evidence from two experiments that tested judgments about these three kinds of sentence, and provide a semantics for implicit comparatives from the perspective of degree semantics.

In a degree semantics setting, comparative sentences express relations between degrees—formal objects representing extent along particular dimensions. Degrees and an ordering on them constitute 'scales', thought to be lexically specified by some adjectives (e.g., more intelligent) and adverbs (more quickly). Comparatives targeting nouns (more soup) and verbs (run more) work differently: here, scale selection depends in large part on the ontological implications of the NP or VP. Focusing on adverbial and verbal comparatives, we address three questions: (i) How well can speakers quantify and compare events (jump) versus activities (move) along different dimensions, as measured by their evaluation of adverbial comparatives?; (ii) To what extent does the event structure of a verb  determine the scale with bare more?; and, (iii) To what extent do features of the visual scene impact quantification when grammar leaves multiple dimensions open? Together, our four experimental studies show that speakers can easily quantify number, distance, and duration in dynamic scenes using more times, higher, and longer, and that dimensional selection for verbal more depends on the verb: given identical displays, our participants evaluated jump more (event verb) by number, but move more (activity verb) differently. So far, our results do not suggest that the dimension for move more depends on specific features of the visual scene. 

Semantic theories predict that the dimension for comparison given a sentence like A gleebed more than B depends on what the verb gleeb means: if gleeb expresses a property of events, the evaluation should proceed by number; if it expresses a property of processes, any of distance, duration, or number should be available. An adequate test of theories like this requires first determining, independently of language, the conditions under which people will understand a novel verb to be true of a series of events or a single ongoing process. We investigate this prior question by studying people’s representation of two cues in simple visual scenes: a) whether some happening is interrupted by temporal pauses, and b) whether and how the speed of an object’s motion changes. We measured representation by probing people’s choice of verb in free-form descriptions of the scenes, and how they segment the scenes for the purposes of counting. We find evidence that both types of cues shape people’s representation of simple motions as events or processes, but in different ways. 

Comparative constructions like More people have been to Russia than I have are reported to be acceptable and meaningful by native speakers of English; yet, upon closer reflection, they are judged to be incoherent. This mismatch between initial perception and more considered judgment challenges the idea that we perceive sentences veridically, and interpret them fully. It is thus potentially revealing about the relationship between grammar and language processing. This paper presents the results of the first detailed investigation of these so-called 'comparative illusions'. We test four hypotheses about their source: a shallow syntactic parser, some type of repair by ellipsis, an incorrectly-resolved lexical ambiguity, or a persistent event comparison interpretation. Two formal acceptability studies show that speakers are most prone to the illusion when the matrix clause supports an event comparison reading. A verbatim recall task tests and finds evidence for such construals in speakers' recollections of the sentences. We suggest that this reflects speakers' entertaining an interpretation that is initially consistent with the sentence, but failing to notice when this interpretation becomes unavailable at the than-clause. In particular, semantic knowledge blinds people to an illicit operator-variable configuration in the syntax. Rather than illustrating processing in the absence of grammatical analysis, comparative illusions thus underscore the importance of syntactic and semantic rules in sentence processing.

Additional materials related to the published experiments and our preliminary experiments can be found on Github.

Beginning at least with Bach (1986), semanticists have suggested that the objects/events and substances/processes that nouns and verbs apply to are strongly parallel. We investigate whether these parallels can be understood to reflect a shared representational format in cognition, which in turn underlies aspects of the intuitive metaphysics of these categories. We hypothesized that a way of counting (atomicity) is necessary for object and event representations, unlike substance or process representations. Atomicity is strongly implied by plural language (some gorps, for novel gorp) but not mass language (some gorp). We investigate the language-perception interface across these domains using minimally different images and animations designed to encourage atomicity ('natural' spatial and temporal breaks), versus those that should not ('unnatural' breaks). Testing preference for matching such stimuli with mass or count syntax, our results support Bach’s analogy in perception, and highlight the formal role of atomicity in object and event representation.

Additional materials related to the published experiments and our preliminary experiments can be found on Github.

Kotek et al. (2015) argue on the basis of novel experimental evidence that sentences like 'Most of the dots are blue' are ambiguous, i.e. have two distinct truth conditions. Kotek at al furthermore suggest that when their results are taken together with those of earlier work by Lidz et al. (2011), the overall picture that emerges casts doubt on the conclusions that Lidz et al. drew from their earlier results. We disagree with this characterization of the relationship between the two studies. Our main aim in this reply is to clarify the relationship as we see it. In our view, Kotek et al.'s central claims are simply logically independent of those of Lidz et al.: the former concern which truth condition(s) a certain kind of sentence has, while the latter concern the procedures that speakers choose for the purposes of determining whether a particular truth condition is satisfied in various scenes. The appearance of a conflict between the two studies stems from inattention to the distinction between questions about truth conditions and questions about verification procedures.

Recent proposals in the semantics literature hold that the negative comparative less and negative adjectives like short in English are morphosyntactically complex, unlike their positive counterparts more and tall. For instance, the negative adjective short might decompose into little tall (Rullmann 1995; Heim 2006; Büring 2007; Heim 2008). Positing a silent little as part of adjectives like short correctly predicts that they are semantically opposite to tall; we seek evidence for this decomposition in language understanding in English and Polish. Our visual verification tasks compare processing of positive and negative comparatives with taller and shorter against that of arguably less symbolically-rich mathematical statements, A > B, B < A. We find that both language and math statements generally lead to monotonic increases in processing load along with the number of negative symbols (as predicted for language by e.g. Clark and Chase 1972). Our study is the first to examine the processing of the gradable predicates tall and short cross-linguistically, as well as in contrast to extensionally-equivalent, and putatively non-linguistic stimuli (cf. Deschamps et al. 2015 with quantificational determiners).

Acquiring the correct meanings of words expressing quantities (seven, most) and qualities (red, spotty) present a challenge to learners. Understanding how children succeed at this requires understanding not only what kinds of data are available to them, but also the biases and expectations they bring to the learning task. The results of our word-learning task with 4 year-olds indicates that a "syntactic bootstrapping" hypothesis correctly predicts a bias towards quantity-based interpretations when a novel word appears in the syntactic position of a determiner, but leaves open the explanation of a bias towards quality-based interpretations when the same word is presented in the syntactic position of an adjective. We develop four computational models that differentially encode how lexical, conceptual, and perceptual factors could generate the latter bias. Simulation results suggest it results from a combination of lexical bias and perceptual encoding.

In acquiring a semantics, children relate their experience of their world to their experience of speakers. When we study this in the lab, we often presume to understand the first part of this relation: we take for granted how the child will experience the world of our experiment, and test for how she will experience an attendant event of speech. Such presumptions are fair. But they need to be justified, when the experience we impute to the child is much richer than what the world presents objectively. In this paper we discuss one such case, reporting on several experiments targeted at assessing event perception in pre-linguistic infants, following the lead of Gordon (2003). We begin by characterizing what a 'participant role' is, and how certain acquisition heuristics that depend on this notion are meant to facilitate verb learning.

Questions about the nature of the relationship between language and extralinguistic cognition are old, but only recently has a new view emerged that allows for the systematic investigation of claims about linguistic structure, based on how it is understood or utilized outside of the language system. Our paper represents a case study for this interaction in the domain of event semantics. We adopt a transparency thesis about the relationship between linguistic structure and extralinguistic cognition, investigating whether different lexico-syntactic structures can differentially recruit the visual causal percept. A prominent analysis of causative verbs like move suggests reference to two distinct events and a causal relationship between them, whereas non-causative verbs like push do not so refer. In our study, we present English speakers with simple scenes that either do or do not support the perception of a causal link, and manipulate (between subjects) a one-sentence instruction for the evaluation of the scene. Preliminary results suggest that competent speakers of English are more likely to judge causative constructions than non-causative constructions as true of a scene where causal features are present in the scene. Implications for a new approach to the investigation of linguistic meanings and future directions are discussed.

Acquiring the correct meanings of number words (e.g., seven, forty-two) is challenging, as such words fail to describe salient properties of individuals or objects in their environment, referring rather to properties of sets of such objects or individuals. Previous research has revealed a critical role for language itself in how children acquire number word meanings, however attempts to pinpoint precisely the strong linguistic cues has proved challenging. We propose a novel "syntactic bootstrapping" hypothesis in which categorizing a novel word as a determiner leads to quantity-based interpretations. The results of a word learning task with 4 year olds indicates that this hypothesis is on the right track.

No language has a determiner meaning something like 'less than half' (e.g., a fost to complement determiner most), nor does any language have a determiner meaning something like what most means, but which is non-conservative (e.g., a grfost). Like the hypothetical fost, but unlike the hypothetical grfost, every natural language determiner is "conservative"—i.e., it lives on the set denoted by its complement NP. Are these two typological gaps equally principled? We look at this question from the perspective of language acquisition, asking whether the two meanings (non-conservative or conservative 'less than half') are equally aquirable. Our experiments suggest that children are able to access the non-existent, yet conservative determiner meaning fost, but not a non-conservative counterpart like grfost.