As adults, we use polite speech on a daily basis. What do chil-dren understand about polite speech? Looking at children’s po-lite speech comprehension can help examine children’s prag-matic understanding more generally, and can be informativefor caregivers who want to teach children what it means to bepolite. Even though children start to produce polite speechfrom early on, there is little known about whether they under-stand intentions behind polite language. Here we show that by3 years, English-speaking preschool children understand that itis more polite and nicer (and less rude and mean) to use polite-ness markers such as “please” when making requests, and by4 years, they understand that the use of these politeness mark-ers indicates that the speaker is more socially likeable and ismore likely to gain compliance from their conversational part-ners. This work can help lay the foundation for future work onchildren’s understanding of polite speech and pragmatic devel-opment more generally.

Despite their considerable communicative abilities, youngchildren often have difficulty interpreting complex linguisticstructures in context. Two examples of this phenomenon arenegation and pragmatic implicature, both of which pose some-times surprising difficulties for preschoolers. Both of thesestructures require children to resist a more salient alternativeinterpretation; since executive function abilities develop ex-tensively during childhood, perhaps failures are due to prob-lems in inhibition. To test this hypothesis, we designed tasksto measure inhibitory control, negation, and implicature com-prehension in children and adults. Using standard analyses aswell as drift diffusion models, we found different patterns ofprocessing on all three tasks, and no support for the hypothesisthat inhibitory control per se is playing a role in either adults’or children’s negation or implicature processing. Instead, ouranalyses reveal qualitatively different developmental trajecto-ries, suggesting task-specific factors driving these changes.

Language comprehenders routinely make pragmatic inferences that go beyond the literal meanings of utterances. If A said “I ate some of the cookies,” B should infer that A ate some but not all. Children perform poorly on experimental tests of scalar implicatures like this, despite their early-emerging sensitivity to pragmatic cues. Our current work explores potential factors responsible for children’s successes and failures in computing pragmatic inferences. In two experiments, we used an eyetracking paradigm to test children’s ability to compute implicatures when they have access to contextual alternatives to the target word (Experiment 1), and when they hear prosodic cues that emphasize the contrast between the target and alternative (Experiment 2). We found that by the time children are four years old, they successfully identify the inferential target referent in this paradigm; with supportive prosodic cues, we saw evidence of success in three-year-olds as well. In sum, with sufficient contextual support, preschool children are capable of making online pragmatic inferences.

Why are we polite when we talk to one another? One hypoth-esis is that people expect others to choose what to say basedon their goals both to transfer information efficiently (an epis-temic goal) and to make the listener feel good (a social goal).In our previous work, we found that when these two goals con-flict, they sometimes produce white lies. In the current work,we expand on this theory to consider another prominent case ofpolite speech: indirect remarks using negation (e.g., “It wasn’tamazing”). With minimal extensions from our previous frame-work, our formal model suggests that a pragmatic speaker willproduce more indirect remarks when the speaker wants to beinformative and seem considerate at the same time. Thesepredictions were borne out in a language production experi-ment. These findings suggest that the conflict between socialand epistemic goals can account for a broad range of politenessphenomena.

Our actions shape what we learn. Because of this dependency,learners are proficient at choosing their actions to maximizetheir information gain. However, learning often unfolds insocial contexts where learners have both informational goals(e.g., to learn how something works) but also social goals (e.g.,to appear competent and impress others). How do these goalsshape learners’ decisions? Here, we present a computationalmodel that integrates the value of social and informationalgoals to predict the decisions that people will make in a simpleactive causal learning task. We show that, in a context wherethe informational and social goals are in conflict, an empha-sis on performance or self-presentation goals leads to reducedchances of learning (Exp. 1) and that social context can pushlearners to pursue performance-oriented actions even when thelearning goal is highlighted (Exp. 2). Our formal model ofsocial-active learning successfully captures the empirical re-sults. These findings are first steps towards understanding therole of social reasoning in active learning contexts.

Conveying information in a false or indirect manner in consid-eration of listeners’ wants (i.e. being polite) seemingly contra-dicts an important goal of a cooperative speaker: informationtransfer. We propose that a cooperative speaker considers bothepistemic utility, or utility of providing the listener new and ac-curate information, and social utility, or utility of maintainingor boosting the listener’s self-image (being polite). We for-malize this tradeoff within a probabilistic model of languageunderstanding and test it with empirical data on people’s infer-ences about the relation between a speaker’s goals, utterancesand the true states of the world.