Pragmatic Reasoning Research Articles

Humans Learn Language from Situated Communicative Interactions. What about Machines?

Abstract Humans acquire their native languages by taking part in communicative interactions with their caregivers. These interactions are meaningful, intentional, and situated in their everyday environment. The situated and communicative nature of the interactions is essential to the language acquisition process, as language learners depend on clues provided by the communicative environment to make sense of the utterances they perceive. As such, the linguistic knowledge they build up is rooted in linguistic forms, their meaning, and their communicative function. When it comes to machines, the situated, communicative, and interactional aspects of language learning are often passed over. This applies in particular to today’s large language models (LLMs), where the input is predominantly text-based, and where the distribution of character groups or words serves as a basis for modeling the meaning of linguistic expressions. In this article, we argue that this design choice lies at the root of a number of important limitations, in particular regarding the data hungriness of the models, their limited ability to perform human-like logical and pragmatic reasoning, and their susceptibility to biases. At the same time, we make a case for an alternative approach that models how artificial agents can acquire linguistic structures by participating in situated communicative interactions. Through a selection of experiments, we show how the linguistic knowledge that is captured in the resulting models is of a fundamentally different nature than the knowledge captured by LLMs and argue that this change of perspective provides a promising path towards more human-like language processing in machines.

263 Public perspectives of calf handling: A survey of Canadians

Abstract In North American beef production handling and restraint of young calves is integral to animal welfare and management practices. This study used a mixed method approach to gather public perceptions of three handling and restraint methods common on beef operations in western Canada (RW - Roping and Wrestling, NF – roping and NordForks, and TT – Tilt Table). Canadian participants (n = 551) recruited by CloudResearch to represent the national demographics participated in an online survey that included videos of each handling method to ascertain preferences and acceptability. Prior to watching the videos, participants were given industry information about handling and restraint or generic information regarding hay as a control information statement within the topic of agriculture. The survey also collected information about knowledge of the beef industry, animal welfare, and empathy toward animals. The reasons for preferences were described as presence of a perceived positive attribute and absence of a perceived negative for most preferred methods, and inversely when explaining least preferred method. The findings focused on calf’s experience, perception of handler actions, and pragmatic balancing of handler needs and a good life for the calf. Methods were rated as more acceptable for participants that ate meat consistently, knew more about the beef industry, and, to a lesser extent, if the individual had a lower animal empathy score (Table 1). Acceptability was not affected by providing information about the practices; however, information about handling and restraint did elicit more pragmatic reasoning. Most participants preferred TT over NF and RW (P < 0.001) and found TT more acceptable as well (P < 0.001). The TT was the most preferred method due to calf experience and human handling, notably the absence of dragging a calf, which was predominant in why participants selected NF or RW as their least preferred method. Consistency of themes highlights that the Canadian public has a fundamental expectation of value for the quality of life of the calf, humane handling, and pragmatism which aligns with beef sustainability initiatives and represent points of connection for building trust between industry and the public.

Comparing Humans and Large Language Models on an Experimental Protocol Inventory for Theory of Mind Evaluation (EPITOME)

Abstract We address a growing debate about the extent to which large language models (LLMs) produce behavior consistent with Theory of Mind (ToM) in humans. We present EPITOME: a battery of six experiments that tap diverse ToM capacities, including belief attribution, emotional inference, and pragmatic reasoning. We elicit a performance baseline from human participants for each task. We use the dataset to ask whether distributional linguistic information learned by LLMs is sufficient to explain ToM in humans. We compare performance of five LLMs to a baseline of responses from human comprehenders. Results are mixed. LLMs display considerable sensitivity to mental states and match human performance in several tasks. Yet, they commit systematic errors in others, especially those requiring pragmatic reasoning on the basis of mental state information. Such uneven performance indicates that human-level ToM may require resources beyond distributional information.

(Im)precise personae: The effect of socio-indexical information on semantic interpretation

Abstract In this article, we address the following question: how do comprehenders reason about the persona embodied by the speaker to determine the referential meaning of numerical expressions such as ‘The price is $200’? Using a picture selection task, we show that descriptions uttered by speakers embodying a Nerdy persona, indexically associated with highly precise speech, are interpreted more precisely than those uttered by speakers embodying a Chill persona, indexically associated with imprecise speech. These findings contribute to building a more integrative perspective between the socio-indexical and the referential domain of signification, highlighting comprehenders’ social perception of the speaker as a crucial element informing pragmatic reasoning, and meaning interpretation more broadly. (Social meaning, personae, pragmatic reasoning, precision, numerals)*

Expertise Gained During the COVID-19 Pandemic: Exploring Ethical Implications Through Practitioners' Narratives.

The COVID-19 pandemic necessitated a transition to telehealth. Research supports the use of telehealth; however, there is a paucity of literature on the ethical considerations of this shift. This study explored ethics through narratives of practitioners' lived experiences of transitioning to telehealth. Semi-structured interviews were conducted virtually with 10 pediatric health care professionals. Transcripts were analyzed from a narrative phenomenological perspective. Discussed themes include responding to the rapid redirection, collaborating with colleagues and caregivers, applying therapeutic use of self, and evolving pragmatic clinical reasoning. Transitioning to telehealth resulted in ethical dilemmas pertaining to clients and practitioners. The need for additional support was salient to ensure clients received beneficial services that would not cause harm (non-maleficence). Communities of practice formed organically as forums to explore strategies for sharing comprehensive and equitable information (veracity and justice). Knowledge generated through providers' experiences may inform future guidelines on service delivery transitions.

Ad-Hoc Conversational Implicatures in Two-Year-Olds

ABSTRACT Young children’s (n = 108) and adults’ (n = 40) ability to compute ad-hoc quantity conversational implicatures was assessed using a new implicit task that relied on eye-tracking. The children were 2 and 5 years old. Looking times reveal that all participants interpreted simple references by relying on implicatures. However, 2-year-olds failed to derive standard ad-hoc implicatures and succeeded only on a novel type of implicature based on typicality, which was proposed with the aim of reducing the salience of non-target potential referents. These results suggest that even toddlers are able to derive conversational implicatures and that domain-general inhibitory demands, rather than domain-specific limitations in pragmatic reasoning, may play a crucial role in children’s difficulties with implicatures.

Bilingual children display comparative strength using prosodic cues for pragmatic word learning

Aims/objectives/research questions: Previous studies indicate differences in the way children who grow up with two languages use socio-pragmatic cues to help them identify referents and learn new words, yet the nature of these differences (executive control, better attention to social cues, or pragmatic reasoning) has not been investigated. Design/methodology/approach: This study examined 270 monolingual and bilingually exposed 4–6-year-old children’s performance in 2 tasks using different prosodic cues (contrastive stress and emotional affect) for fast mapping. It avoids a design where children have to inhibit an irrelevant cue, which would enhance the role of differences in executive control. Data/analysis: We performed statistical analyses using a logistic regression mixed model. Findings/conclusions: The bilingually exposed group performed lower than monolinguals in a control condition involving structural language (0.83 vs. 0.92). However, they performed on par with monolinguals in a pragmatic condition when considering only semantically correct answers in both groups (0.55 vs. 0.58), and even displayed significant comparative strength in the task once control performance and demographic variables were taken into account. This effect appeared when the task required reasoning about the speaker’s communicative intentions (contrastive stress) but not when children merely had to recognise a communicative cue (emotional affect). Originality: No study has so far investigated the socio-pragmatic abilities of bilingual children using a task that did not require inhibiting an irrelevant cue. Implications: These findings have implications for bilingual education and a better understanding of the impact of being educated in two languages. We also draw attention to implications regarding the existence of different types of pragmatic skills which may have differing developmental timelines and rely on different sets of abilities.

An individual differences perspective on pragmatic abilities in the preschool years.

Pragmatic abilities are fundamental to successful language use and learning. Individual differences studies contribute to understanding the psychological processes involved in pragmatic reasoning. Small sample sizes, insufficient measurement tools, and a lack of theoretical precision have hindered progress, however. Three studies addressed these challenges in three- to 5-year-old German-speaking children (N=228, 121 female). Studies 1 and 2 assessed the psychometric properties of six pragmatics tasks. Study 3 investigated relations among pragmatics tasks and between pragmatics and other cognitive abilities. The tasks were found to measure stable variation between individuals. Via a computational cognitive model, individual differences were traced back to a latent pragmatics construct. This presents the basis for understanding the relations between pragmatics and other cognitive abilities. RESEARCH HIGHLIGHTS: Individual differences in pragmatic abilities are important to understanding variation in language development. Research in this domain lacks a precise theoretical framework and psychometrically high-quality measures. We present six tasks capturing a wide range of pragmatic abilities with excellent re-test reliability. We use a computational cognitive model to provide a substantive theory of individual differences in pragmatic abilities.

When parsing and interpretation misalign: A case of WH-scope ambiguity in Mandarin

A large amount of sentence-processing work has focused on revealing how the parser incrementally integrates each incoming word into the current linguistic representation. It is often explicitly or implicitly assumed that the structure endorsed by the parser should determine the ultimate interpretation of the sentence. The current study investigates whether the interpretive bias in sentence comprehension necessarily tracks the parsing bias. Our case study concerns the locality bias in nonlocal dependencies, specifically Mandarin wh-in-situ scope dependencies. Our findings suggest a misalignment between parsing and interpretative decisions at the global level. In particular, for Mandarin wh-in-situ constructions that involve scope ambiguity, there is a locality bias in parsing, but an antilocality bias in interpretation. Building upon the rational speech act framework, we propose a Bayesian pragmatic analysis to account for these findings. Under our proposal, the seeming conflict between parsing and interpretation will ultimately disappear because parsing preferences will be naturally embedded under the pragmatic reasoning process to generate the ultimate interpretation. The current study therefore makes novel contributions, both empirically and theoretically, to addressing the broader question about the relationship between parsing and interpretation.

필자의 텍스트 반성에 대한 화용론적 접근

Objectives The purpose of this study is to suggest a way to overcome the limitations of process-centered writing education by identifying the cognitive action in which the writer reflects on his/her text during the writing process.
 Methods According to pragmatic semiotics, the interpreter of signs constitutes an immediate, dynamic and final interpretive body. In the writing situation, the writer is both a producer and an interpreter of signs, so he must be able to actively construct such interpretations. Each interpretive body can be specified as a text representation, a reader representation, and a discourse community representation. The cognitive strategies constituting each representation can be explained by the suitability theory, which is a model of pragmatic reasoning.
 Results The composition of text representation is related to bottom-up reading, while reader representation and linguistic community representation are related to connotation and contextual meaning, respectively, both of which involve top-down reading. When the author reads the text he is writing for the purpose of text reflection, a conscious effort is required to exclude the meaning and background knowledge he is trying to convey. Therefore, when reflecting on text, it is necessary to distinguish between bottom-up reading and top-down reading, and to divide top-down reading into reader representation and linguistic community representation, and select and perform one reading method at a time.
 Conclusions In the process of text reflection, the representations that can be constructed are related to the subject, purpose, reader, and language community, so it is necessary to teach them in sequence by rhetorical element. Just as the layers of reading are divided into factual, inferential, and critical understanding, it is possible to sequence the layers of my text reflection into three representations. Cognitive actions that form an inference model centered on each rhetorical element can be strategically made into educational content.

Pragmatics aid referent disambiguation and word learning in young children and adults.

How do children succeed in learning a word? Research has shown robustly that, in ambiguous labeling situations, young children assume novel labels to refer to unfamiliar rather than familiar objects. However, ongoing debates center on the underlying mechanism: Is this behavior based on lexical constraints, guided by pragmatic reasoning, or simply driven by children's attraction to novelty? Additionally, recent research has questioned whether children's disambiguation leads to long-term learning or rather indicates an attentional shift in the moment of the conversation. Thus, we conducted a pre-registered online study with 2- and 3-year-olds and adults. Participants were presented with unknown objects as potential referents for a novel word. Across conditions, we manipulated whether the only difference between both objects was their relative novelty to the participant or whether, in addition, participants were provided with pragmatic information that indicated which object the speaker referred to. We tested participants' immediate referent selection and their retention after 5min. Results revealed that when given common ground information both age groups inferred the correct referent with high success and enhanced behavioral certainty. Without this information, object novelty alone did not guide their selection. After 5min, adults remembered their previous selections above chance in both conditions, while children only showed reliable learning in the pragmatic condition. The pattern of results indicates how pragmatics may aid referent disambiguation and learning in both adults and young children. From early ontogeny on, children's social-cognitive understanding may guide their communicative interactions and support their language acquisition. RESEARCH HIGHLIGHTS: We tested how 2-3-year-olds and adults resolve referential ambiguity without any lexical cues. In the pragmatic context both age groups disambiguated novel word-object-mappings, while object novelty alone did not guide their referent selection. In the pragmatic context, children also showed increased certainty in disambiguation and retained new word-object-mappings over time. These findings contribute to the ongoing debate on whether children learn words on the basis of domain-specific constraints, lower-level associative mechanisms, or pragmatic inferences.

Social identity, precision and charity: when less precise speakers are held to stricter standard

Recent has begun to show systematic connections between social information and pragmatic reasoning. These findings raise the question of whether social information shapes comprehenders' assessments of the correctness of linguistic description in light of a single known and determined fact. We explore this question by testing the impact of speaker identity on T(ruth)-V(alue) J(udgment)s based on the interpretation of number words. We find that imprecise statements from speakers socially expected to be less precise – i.e. “Chill" ones – are rejected at a higher rate, and thus held to more stringent evaluation standards, than those from speakers socially expected to speak more precisely – i.e. “Nerdy" ones. We explain the new finding by appealing to the idea that, by virtue of generally being perceived to be more precise, Nerdy speakers are granted higher epistemic credibility than Chill ones. The emerging picture is one in which TVJ assessments are affected by social considerations in a different way from other experimental tasks, suggesting a nuanced interplay between social information and different interpretation tasks and processes

Effect of Risk Avoidance on Performance of Devolved Governments in Kenya

Purpose: Risk avoidance are critical techniques in the elimination of threats, undertaking and submissions that can negatively influence an institution or government gains and investment decision worldwide. By halting the activity that can lead to likely problems, the chance of incurring and loss is eliminated. However, devolved system of government studies on risk avoidance are scanty in Kenya. The main purpose of the study was to investigate the effect of risk avoidance and how it influences performance of devolved governments in Kenya.
 Methodology: The study was anchored on the agency and stakeholder theory. A correlational research design was used based on the nature of data collected. The study was analytical and hence it took a pragmatic philosophical reasoning. The study targeted 423 respondents in the 47 devolved governments departments in Kenya. A sample size of 381 respondents from the devolved governments departments was drawn from the targeted population using stratified sampling technique. The study used primary data from the use of questionnaires. Data was analyzed using Statistical Packages for Social Sciences version 24 utilizing inferential statistics which involved testing of the hypothesis at 95% confidence level and also descriptive statistics (frequencies, percentages, mean and standard deviation). Data was presented by use of tables.
 Findings: The outcomes indicated that risk avoidance indicator had a standardized Beta coefficient of 0.341, an absolute t-value of 5.609 and p˂0.05 (0.03) which indicated that risk avoidance had a statistically considerable effect on performance of devolved governments in Kenya.
 Unique Contribution to Theory, Practice and Policy: The study suggests that devolved governments and key stakeholders should upgrade, embrace and ensure that risk avoidance techniques are in place in the devolved governments so as to improve on performance of devolved government.

Predicting pragmatic cue integration in adults' and children's inferences about novel word meanings.

Language is learned in complex social settings where listeners must reconstruct speakers' intended meanings from context. To navigate this challenge, children can use pragmatic reasoning to learn the meaning of unfamiliar words. A critical challenge for pragmatic reasoning is that it requires integrating multiple information sources, which have typically been studied separately. Here we study this integration process. First, we experimentally isolate two sources of pragmatic information: expectations about informative communication and common ground. Next, we use a probabilistic model of conversational reasoning to formalize how these information sources should be combined and how this process might develop. We use this model to generate quantitative predictions, which we test against new experimental data from 3- to 5-year-old children (N = 243) and adults (N = 694). Results show close alignment between model predictions and data. Furthermore, the model provided a better explanation of the data compared with simpler alternative models assuming that participants selectively ignore one information source. This work integrates distinct sets of findings regarding information sources for early language learning and suggests that pragmatic reasoning models can provide a quantitative framework for understanding developmental changes in language learning. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

Probabilistic modeling of rational communication with conditionals

While a large body of work has scrutinized the meaning of conditional sentences, considerably less attention has been paid to formal models of their pragmatic use and interpretation. Here, we take a probabilistic approach to pragmatic reasoning about indicative conditionals which flexibly integrates gradient beliefs about richly structured world states. We model listeners' update of their prior beliefs about the causal structure of the world and the joint probabilities of the consequent and antecedent based on assumptions about the speaker's utterance production protocol. We show that, when supplied with natural contextual assumptions, our model uniformly explains a number of inferences attested in the literature, including epistemic inferences, conditional perfection and the dependency between antecedent and consequent of a conditional. We argue that this approach also helps explain three puzzles introduced by Douven (2012) about updating with conditionals: depending on the utterance context, the listener's belief in the antecedent may increase, decrease or remain unchanged.

Pragmatic reasoning and semantic convention: A case study on gradable adjectives

Gradable adjectives denote properties that are relativized to contextual thresholds of application: how long an object must be in order to count as long in a context of utterance depends on what the threshold is in that context. But thresholds are variable across contexts and adjectives, and are in general uncertain. This leads to two questions about the meanings of gradable adjectives in particular contexts of utterance: what truth conditions are they understood to introduce, and what information are they taken to communicate? In this paper, we consider two kinds of answers to these questions, one from semantic theory, and one from Bayesian pragmatics, and assess them relative to human judgments about truth and communicated information. Our findings indicate that Bayesian accounts can model human judgments about what is communicated better than they can model human judgments about truth conditions, and their weakness on the latter is mainly because they fall short in accurately predicting the relevant threshold distributions. We explore the possibility that the overall performance of the Bayesian accounts can be potentially improved when they are supplemented with the threshold conventions postulated by semantic theory. EARLY ACCESS

Great ape communication as contextual social inference: a computational modelling perspective.

Human communication has been described as a contextual social inference process. Research into great ape communication has been inspired by this view to look for the evolutionary roots of the social, cognitive and interactional processes involved in human communication. This approach has been highly productive, yet it is partly compromised by the widespread focus on how great apes use and understand individual signals. This paper introduces a computational model that formalizes great ape communication as a multi-faceted social inference process that integrates (a) information contained in the signals that make up an utterance, (b) the relationship between communicative partners and (c) the social context. This model makes accurate qualitative and quantitative predictions about real-world communicative interactions between semi-wild-living chimpanzees. When enriched with a pragmatic reasoning process, the model explains repeatedly reported differences between humans and great apes in the interpretation of ambiguous signals (e.g. pointing or iconic gestures). This approach has direct implications for observational and experimental studies of great ape communication and provides a new tool for theorizing about the evolution of uniquely human communication.This article is part of the theme issue ‘Revisiting the human ‘interaction engine’: comparative approaches to social action coordination’.

The regulation of RegTech and SupTech in finance: ensuring consistency in principle and in practice

PurposeThe paper’s aim is to consider how best to formulate sturdy regulatory frameworks for RegTech and SupTech. The paper appraises how key features of EU and UK regulatory and policy initiatives can contribute to a functional framework for RegTech and SupTech.Design/methodology/approachThe paper refers to the most comprehensive empirical findings within the EU and the UK on RegTech and SupTech, including reports released by the European Banking Authority and the Bank of England. As data is only gradually becoming available about the true rate of adoption of RegTech and SupTech, the paper identifies salient areas that warrant analysis from emerging findings. In light of the relatively restricted sources of empirical data, the article’s methodological approach is directed towards the most wide-ranging and detailed sources that are currently available at EU and UK levels.FindingsThe paper reveals distinct variations in how the EU and UK have pursued regulatory approaches towards RegTech and SupTech growth. However, there are many shared features in the respective approaches. The paper argues that a regulatory framework should ideally be imbued with overarching strategies and policy objectives, as well as with practical measures through innovation facilitators, such as sandboxes. Yet, legislative (top-down) intervention will be the significant ingredient in guaranteeing legal clarity for RegTech and SupTech.Originality/valueBy understanding the nuances in EU and UK approaches, the paper advocates for pragmatic reasoning when formulating a regulatory response. The importance of the article is in its focus on the elements of EU and UK regulatory approaches that are most capable of guaranteeing clarity on standards relating to RegTech and SupTech. The paper makes a vital contribution to existing commentary by determining how a balance can be struck between “top-down” and “bottom-up” types of regulation (i.e. should regulation be entirely concerned with industry-driven standards, such as codes of conduct?).

Context, precision, and social perception: A sociopragmatic study

AbstractIn two perception experiments we explore the social indexicality of numerical expressions, comparing the evaluation of three variants: precise (e.g. ‘forty-nine minutes’) vs. explicitly approximate (e.g. ‘about fifty minutes’) vs. underspecified (e.g. ‘fifty minutes’). We ask two questions: (i) What constellations of social meanings are associated with each of these variants? (ii) How are such indexical associations modulated by the conversational setting? We find that the choice of approximate vs. precise forms differentially impact speaker evaluation along the social dimensions of Status, Solidarity, and anti-Solidarity, with underspecified numbers showing a flexible behavior. Furthermore, these associations are to some extent affected by the conversational setting, in particular the demands on descriptive precision placed by the context and the interlocutors’ goals. These findings reveal an intimate connection between pragmatic reasoning and social perception, highlighting the importance of integrating pragmatic theory in the study of social indexicality. (Social meaning, pragmatic variation, social perception, numerals, (im)precision)*

What’s Your Laughter Doing There? A Taxonomy of the Pragmatic Functions of Laughter

Laughter is a crucial signal for communication and managing interactions. Until now no consensual approach has emerged for classifying laughter. We propose a new framework for laughter analysis and classification, based on the pivotal assumption that laughter has propositional content. We propose an annotation scheme to classify the pragmatic functions of laughter taking into account the form, the laughable, the social, situational, and linguistic context. We apply the framework and taxonomy proposed in a multilingual corpus study (French, Mandarin Chinese, and English), involving a variety of situational contexts. Our results give rise to novel generalizations about the range of meanings laughter exhibits, the placement of the laughable, and how placement and arousal relate to the functions of laughter. We have tested and refuted the validity of the commonly accepted assumption that laughter directly follows its laughable. In the concluding section, we discuss the implications our work has for spoken dialogue systems. We stress that laughter integration in spoken dialogue systems is not only crucial for emotional and affective computing aspects, but also for aspects related to natural language understanding and pragmatic reasoning. We formulate the emergent computational challenges for incorporating laughter in spoken dialogue systems.