Maria Mavridaki, Natalie Bleijlevens, Tanya Behne and Ágnes Melinda Kovács
Affiliation: Central European University, Georg-August University of Göttingen, Georg-August University of Göttingen, Central European University
Curiosity is a hallmark of childhood, with children actively seeking and creating learning opportunities for themselves. Traditionally, curiosity has been defined as non-instrumental, driven primarily by uncertainty and novelty (Berlyne, 1960; Oudeyer et al., 2007). Indeed, children seek information to reduce uncertainty and they explore when their knowledge is incomplete or inconsistent with new evidence (Kidd et al., 2012; Siegel et al., 2021; Cook et al., 2011; Schulz et al., 2008). Similarly, in the word learning domain, when children are confronted with uncertainty about the referent of a word (referential ambiguity), they selectively seek information from the speaker via social referencing (Hembacher et al., 2020). However, it remains unclear how children’s information-seeking is influenced by other factors, namely by considerations that could be characterized as socially instrumental, such as whether the to-be-acquired information will be useful during future interactions (social utility of information). Here we examine whether four-year-olds’ social information-seeking and word learning are modulated by the degree of referential ambiguity and the information’s prospective social utility (i.e. whether the novel word will be useful and relevant to everyday communicative contexts). Children played a tablet-based labeling game in the presence of an actively involved Experimenter. Within the game, they encountered labeling events in which they had to identify the referents of novel labels in situations with varying levels of a) referential ambiguity and b) prospective relevance of the information. In high ambiguity trials, children had to find the referent of a novel label in the presence of two novel objects, and in low ambiguity trials, they had to do so in a situation resolvable via mutual exclusivity (by excluding a referent whose name is well-known). In high relevance trials, objects were introduced by the Experimenter as commonly found in the child’s environment, while in low relevance trials objects were introduced as commonly found in a faraway country (see Henderson et al., 2013). We manipulated both factors in isolation, as well as in interaction with each other (resulting in all possible combinations of ambiguity and relevance). Across trial types, we measured implicit (social referencing towards the Experimenter) and explicit (verbal question asking) information-seeking during children’s referent identification. Additionally, we measured their retention of novel labels after a delay. We predicted greater information-seeking in trials with high ambiguity compared to low ambiguity, and sensitivity to the information’s relevance, with children showing higher information-seeking in high compared to low relevance trials. This result would suggest that children’s information-seeking is influenced not only by their attempt to reduce uncertainty, but also by instrumental considerations. Additionally, we will assess the impact of both manipulated factors on children’s learning success. Data collection is ongoing and will be completed by September. This study will provide new insights into the complexity of social information-seeking and its role in learning. The implications for theories of curiosity and social learning will be discussed.