Minimal Compositionality versus Bird Implicatures: two theories of ABC-D sequences in Japanese tits.

It was argued in a series of experimental studies that Japanese tits (Parus minor) have an ABC call that has an alert function, a D call that has a recruitment function, and an ABC-D call that is compositionally derived from ABC and D, and has a mobbing function. A key conclusion was that ABC-D differs from the combination of separate utterances of ABC and of D (e.g. as played by distinct but close loudspeakers). While the logic of the argument is arguably sound, no explicit rule has been proposed to derive the meaning of ABC-D from that of its parts. We compare two analyses. One posits a limited instance of semantic compositionality ('Minimal Compositionality'); the other does without compositionality, but uses instead a more sophisticated pragmatics ('Bird Implicatures'). Minimal Compositionality takes the composition of ABC and D to deviate only minimally from what would be found with two independent utterances: ABC means that 'there is something that licenses an alert', D means that 'there is something that licenses recruitment', and ABC-D means that 'there is something that licenses both an alert and recruitment'. By contrast, ABC and D as independent utterances yield something weaker, namely: 'there is something that licenses an alert, and there is something that licenses recruitment', without any 'binding' across the two utterances. The second theory, Bird Implicatures, only requires that ABC-D should be more informative than ABC, and/or than D. It builds on the idea, proposed for several monkey species, that a less-informative call competes with a more informative one (the 'Informativity Principle'): when produced alone, ABC and D trigger an inference that ABC-D is false. We explain how both Minimal Compositionality and Bird Implicatures could have evolved, and we compare the predictions of the two theories. Finally, we extend the discussion to some chimpanzee and meerkat sequences that might raise related theoretical problems.

Inconsistent effects of components as evidence for non-compositionality in chimpanzee face-gesture combinations? A response to Oña et al (2019).

Using field observations from a sanctuary, Oña and colleagues (DOI: 10.7717/peerj.7623) investigated the semantics of face-gesture combinations in chimpanzees (Pan troglodytes). The response of the animals to these signals was encoded as a binary measure: positive interactions such as approaching or grooming were considered affiliative; ignoring or attacking was considered non-affiliative. The relevant signals are illustrated in Fig. 1 (https://doi.org/10.7717/peerj.7623/fig-1), together with the outcome in terms of average affiliativeness. The authors observe that there seems to be no systematicity in the way the faces modify the responses to the gestures, sometimes reducing affiliativeness, sometimes increasing it. A strong interpretation of this result would be that the meaning of a gesture-face combination cannot be derived from the meaning of the gesture and the meaning of the face, that is, the interpretation of chimpanzees' face-gesture combinations are non compositional in nature. We will revisit this conclusion: we will exhibit simple compositional systems which, after all, may be plausible. At the methodological level, we argue that it is critical to lay out the theoretical options explicitly for a complete comparison of their pros and cons.

The ABC-D of animal linguistics: are syntax and compositionality for real?

In several animal species, an alarm call (e.g. ABC notes in the Japanese tit Parus minor) can be immediately followed by a recruitment call (e.g. D notes) to yield a complex call that triggers a third behaviour, namely mobbing. This has been taken to be an argument for animal syntax and compositionality (i.e. the property by which the meaning of a complex expression depends on the meaning of its parts and the way they are put together). Several additional discoveries were made across species. First, in some cases, animals respond with mobbing to the order alarm-recruitment but not to the order recruitment-alarm. Second, animals sometimes respond similarly to functionally analogous heterospecific calls they have never heard before, and/or to artificial hybrid sequences made of conspecific and heterospecific calls in the same order, thus adding an argument for the productivity of the relevant rules. We consider the details of these arguments for animal syntax and compositionality and argue that, with one important exception (Japanese tit ABC-D sequences), they currently remain ambiguous: there are reasonable alternatives on which each call is a separate utterance and is interpreted as such ('trivial compositionality'). More generally, we propose that future studies should argue for animal syntax and compositionality by explicitly pitting the target theory against two deflationary analyses: the 'only one expression' hypothesis posits that there is no combination in the first place, for example just a simplex ABCD call; while the 'separate utterances' hypothesis posits that there are separate expressions (e.g. ABC and D), but that they form separate utterances and are neither syntactically nor semantically combined.

Formal Models at the Core.

The grammatical paradigm used to be a model for entire areas of cognitive science. Its primary tenet was that theories are axiomatic-like systems. A secondary tenet was that their predictions should be tested quickly and in great detail with introspective judgments. While the grammatical paradigm now often seems passé, we argue that in fact it continues to be as efficient as ever. Formal models are essential because they are explicit, highly predictive, and typically modular. They make numerous critical predictions, which must be tested efficiently; introspective judgments do just this. We further argue that the grammatical paradigm continues to be fruitful. Within linguistics, implicature theory is a recent example, with a combination of formal explicitness, modularity, and interaction with experimental work. Beyond traditional linguistics, the grammatical paradigm has proven fruitful in the study of gestures and emojis; literature ("Free Indirect Discourse"); picture semantics and comics; music and dance cognition; and even reasoning and concepts. We argue, however, that the grammatical paradigm must be adapted to contemporary cognitive science. Computational methods are essential to derive quantitative predictions from formal models (Bayesian pragmatics is an example). And data collection techniques offer an ever richer continuum of options, from introspective judgments to large-scale experiments, which makes it possible to optimize the cost/benefit ratio of the empirical methods that are chosen to test theories.

Evidence for compositionality in baboons (Papio papio) through the test case of negation.

Can non-human animals combine abstract representations much like humans do with language? In particular, can they entertain a compositional representation such as 'not blue'? Across two experiments, we demonstrate that baboons (Papio papio) show a capacity for compositionality. Experiment 1 showed that baboons can entertain negative, compositional, representations: they can learn to associate a cue with iconically related referents (e.g., a blue patch referring to all blue objects), but also to the complement set associated with it (e.g., a blue patch referring to all non-blue objects). Strikingly, Experiment 2 showed that baboons not only learn to associate a cue with iconically related referents, but can learn to associate complex cues (composed of the same cue and an additional visual element) with the complement object set. Thus, they can learn an operation, instantiated by this additional visual element, that can be compositionally combined with previously learned cues. These results significantly reduce any claim that would make the manipulation and combination of abstract representations a solely human privilege.

The influence of polarity items on inferential judgments.

Polarity items are linguistic expressions such as any, at all, some, which are acceptable in some linguistic environments but not others. Crucially, whether a polarity item is acceptable in a given environment is argued to depend on the inferences (in the reasoning sense) that this environment allows. We show that the inferential judgments reported for a given environment are modified in the presence of polarity items. Hence, there is a two-way influence between linguistic and reasoning abilities: the linguistic acceptability of polarity items is dependent on reasoning facts and, conversely, reasoning judgments can be altered by the mere addition of seemingly innocuous polarity items.

Children's Acquisition of Homogeneity in Plural Definite Descriptions.

Plural definite descriptions give rise to homogeneity effects: the positive The trucks are blue and the negative The trucks aren't blue are both neither true nor false when some of the trucks are blue and some are not, that is, when the group of trucks is not homogeneous with respect to the property of being blue (Löbner, 1987, 2000; Schwarzschild, 1994; Kriz, 2015b). The only existing acquisition studies related to the phenomenon have examined children's comprehension only of the affirmative versions of such sentences, and moreover have yielded conflicting data; while one study reports that preschoolers interpret definite plurals maximally (Munn et al., 2006, see also Royle et al., 2018), two other studies report that preschoolers allow non-maximal interpretations of definite plurals where adults do not (Karmiloff-Smith, 1979; Caponigro et al., 2012). Moreover, there is no agreed upon developmental trajectory to adult homogeneity. In this paper, we turn to acquisition data to investigate the predictions of a recent analysis of homogeneity that treats homogeneous meanings as the result of a scalar implicature (Magri, 2014). We conducted two experiments targeting 4- and 5-year-old French-speaking children's interpretations of plural definite descriptions in positive and negative sentences, and tested the same children on standard cases of scalar implicature. The experiments revealed three distinct subgroups of children: those who interpreted the plural definite descriptions existentially and failed to compute implicatures; those who both accessed homogeneous interpretations and computed implicatures; and finally, a smaller subgroup of children who appeared to access homogeneous interpretations without computing implicatures. We discuss the implications of our findings, which appear to speak against the implicature theory as the adult-like means of generating homogeneous meanings.

Constraints on the lexicons of human languages have cognitive roots present in baboons (Papio papio).

Using a pattern extraction task, we show that baboons, like humans, have a learning bias that helps them discover connected patterns more easily than disconnected ones-i.e., they favor rules like "contains between 40% and 80% red" over rules like "contains around 30% red or 100% red." The task was made as similar as possible to a task previously run on humans, which was argued to reveal a bias that is responsible for shaping the lexicons of human languages, both content words (nouns and adjectives) and logical words (quantifiers). The current baboon result thus suggests that the cognitive roots responsible for regularities across the content and logical lexicons of human languages are present in a similar form in other species.

Orangutans' Comprehension of Zoo Keepers' Communicative Signals.

Zoological institutions often encourage cooperative interactions between keepers and animals so as to promote animals' welfare. One useful technique has been conditioning training, whereby animals learn to respond to keepers' requests, which facilitates a number of, otherwise sensitive, daily routines. As various media have been used to convey keepers' instructions, the question remains of which modality is best to promote mutual understanding. Here, we explored this question with two captive female orangutans. In the first experiment, we compared orangutans' understanding of previously acquired instructions when those were performed with verbal signals only, gazes only, gestures only, and when all those modalities were combined. Our results showed that gestures were sufficient for successful comprehension by these two apes. In the second experiment, we asked whether this preference could be driven by the non-arbitrary relationship that gestures bear to what they refer to, through iconicity or pointing. Our results revealed that neither iconicity nor pointing helped the subjects comprehend the keepers' instructions. Our results indicate a preference for instructions given through gestural signals in two captive female orangutans, although its cause remains elusive. Future practice may encourage the use of gestures in communication between keepers and orangutans in general or potentially other animals.

Linguistic inferences without words.

Contemporary semantics has uncovered a sophisticated typology of linguistic inferences, characterized by their conversational status and their behavior in complex sentences. This typology is usually thought to be specific to language and in part lexically encoded in the meanings of words. We argue that it is neither. Using a method involving "composite" utterances that include normal words alongside novel nonlinguistic iconic representations (gestures and animations), we observe successful "one-shot learning" of linguistic meanings, with four of the main inference types (implicatures, presuppositions, supplements, homogeneity) replicated with gestures and animations. The results suggest a deeper cognitive source for the inferential typology than usually thought: Domain-general cognitive algorithms productively divide both linguistic and nonlinguistic information along familiar parts of the linguistic typology.

Mouse tracking as a window into decision making.

Mouse tracking promises to be an efficient method to investigate the dynamics of cognitive processes: It is easier to deploy than eyetracking, yet in principle it is much more fine-grained than looking at response times. We investigated these claimed benefits directly, asking how the features of decision processes-notably, decision changes-might be captured in mouse movements. We ran two experiments, one in which we explicitly manipulated whether our stimuli triggered a flip in decision, and one in which we replicated more ecological, classical mouse-tracking results on linguistic negation (Dale & Duran, Cognitive Science, 35, 983-996, 2011). We concluded, first, that spatial information (mouse path) is more important than temporal information (speed and acceleration) for detecting decision changes, and we offer a comparison of the sensitivities of various typical measures used in analyses of mouse tracking (area under the trajectory curve, direction flips, etc.). We do so using an "optimal" analysis of our data (a linear discriminant analysis explicitly trained to classify trajectories) and see what type of data (position, speed, or acceleration) it capitalizes on. We also quantify how its results compare with those based on more standard measures.

Revealing abstract semantic mechanisms through priming: The distributive/collective contrast.

Sentences such as The bags are light allow both collective (they are light together) and distributive interpretations (each bag is light). We report the results of two experiments showing that this collective/distributive contrast gives rise to priming effects. These findings suggest that collective and distributive readings involve different interpretative mechanisms, which are at play during real comprehension and can be targeted by priming, independently of the specific verification strategy associated with each interpretation.

Competition and Symmetry in an Artificial Word Learning Task.

Natural language involves competition. The sentences we choose to utter activate alternative sentences (those we chose not to utter), which hearers typically infer to be false. Hence, as a first approximation, the more alternatives a sentence activates, the more inferences it will trigger. But a closer look at the theory of competition shows that this is not quite true and that under specific circumstances, so-called symmetric alternatives cancel each other out. We present an artificial word learning experiment in which participants learn words that may enter into competition with one another. The results show that a mechanism of competition takes place, and that the subtle prediction that alternatives trigger inferences, and may stop triggering them after a point due to symmetry, is borne out. This study provides a minimal testing paradigm to reveal competition and some of its subtle characteristics in human languages and beyond.

Cross-linguistic regularities and learner biases reflect "core" mechanics.

Recent research in infant cognition and adult vision suggests that the mechanical object relationships may be more salient and naturally attention grabbing than similar but non-mechanical relationships. Here we examine two novel sources of evidence from language related to this hypothesis. In Experiments 1 and 2, we show that adults preferentially infer that the meaning of a novel preposition refers to a mechanical as opposed to a non-mechanical relationship. Experiments 3 and 4 examine cross-linguistic adpositions obtained on a large scale from machines or from experts, respectively. While these methods differ in the ease of data collection relative to the reliability of the data, their results converge: we find that across a range of diverse and historically unrelated languages, adpositions (such as prepositions) referring to the mechanical relationships of containment (e.g "in") and support (e.g. "on") are systematically shorter than closely matched but not mechanical words such as "behind," "beside," "above," "over," "out," and "off." These results first suggest that languages regularly contain traces of core knowledge representations and that cross-linguistic regularities can therefore be a useful and easily accessible form of information that bears on the foundations of non-linguistic thought.

Priming methods in semantics and pragmatics.

Structural priming is a powerful method to inform linguistic theories. We argue that this method extends nicely beyond syntax to theories of meaning. Priming, however, should still be seen as only one of the tools available for linguistic data collection. Specifically, because priming can occur at different, potentially conflicting levels, it cannot detect every aspect of linguistic representations.

What Do Monkey Calls Mean?

A field of primate linguistics is gradually emerging. It combines general questions and tools from theoretical linguistics with rich data gathered in experimental primatology. Analyses of several monkey systems have uncovered very simple morphological and syntactic rules and have led to the development of a primate semantics that asks new questions about the division of semantic labor between the literal meaning of monkey calls, additional mechanisms of pragmatic enrichment, and the environmental context. We show that comparative studies across species may validate this program and may in some cases help in reconstructing the evolution of monkey communication over millions of years.

What Homophones Say about Words.

The number of potential meanings for a new word is astronomic. To make the word-learning problem tractable, one must restrict the hypothesis space. To do so, current word learning accounts often incorporate constraints about cognition or about the mature lexicon directly in the learning device. We are concerned with the convexity constraint, which holds that concepts (privileged sets of entities that we think of as "coherent") do not have gaps (if A and B belong to a concept, so does any entity "between" A and B). To leverage from it a linguistic constraint, learning algorithms have percolated this constraint from concepts, to word forms: some algorithms rely on the possibility that word forms are associated with convex sets of objects. Yet this does have to be the case: homophones are word forms associated with two separate words and meanings. Two sets of experiments show that when evidence suggests that a novel label is associated with a disjoint (non-convex) set of objects, either a) because there is a gap in conceptual space between the learning exemplars for a given word or b) because of the intervention of other lexical items in that gap, adults prefer to postulate homophony, where a single word form is associated with two separate words and meanings, rather than inferring that the word could have a disjunctive, discontinuous meaning. These results about homophony must be integrated to current word learning algorithms. We conclude by arguing for a weaker specialization of word learning algorithms, which too often could miss important constraints by focusing on a restricted empirical basis (e.g., non-homophonous content words).

Classification and automatic transcription of primate calls.

This paper reports on an automated and openly available tool for automatic acoustic analysis and transcription of primate calls, which takes raw field recordings and outputs call labels time-aligned with the audio. The system's output predicts a majority of the start times of calls accurately within 200 milliseconds. The tools do not require any manual acoustic analysis or selection of spectral features by the researcher.

An Attentional Effect of Musical Metrical Structure.

Theories of metrical structure postulate the existence of several degrees of beat strength. While previous work has clearly established that humans are sensitive to the distinction between strong beats and weak ones, there is little evidence for a more fine grained distinction between intermediate levels. Here, we present experimental data showing that attention can be allocated to an intermediate level of beat strength. Comparing the effects of short exposures to 6/8 and 3/4 metrical structures on a tone detection task, we observe that subjects respond differently to beats of intermediate strength than to weak beats.