How Children Identify Events from Visual Experience

ABSTRACT

Three experiments explored how well children recognize events from different types of visual experience: either by directly seeing an event or by indirectly experiencing it from post-event visual evidence. In Experiment 1, 4- and 5- to 6-year-old Turkish-speaking children (n = 32) successfully recognized events through either direct or indirect visual access. In Experiment 2, a new group of 4- and 5- to 6-year-olds (n = 37) reliably attributed event recognition to others who had direct or indirect visual access to events (even though performance was lower than Experiment 1). In both experiments, although children’s accuracy improved with age, there was no difference between the two types of access. Experiment 3 replicated the findings from the youngest participants of Experiments 1 and 2 with a matched sample of English-speaking 4-year-olds (n = 37). Thus children can use different kinds of visual experience to support event representations in themselves and others.

Acknowledgments

This research has been supported by NSF grant BCS0749870 to A. P. We thank Yasemin Sandıkçı, Didem Aykut, Taylor Maher, and Sam Katz for their assistance in data collection, the preschools in Istanbul, Turkey and Newark, DE, and all of the children that participated in these studies.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Our choice for using static pictures as stimuli is motivated by two reasons. First, our aim was to test children’s inferences about events on the basis of a single snapshot of the event. Second, prior developmental work has revealed that static depictions of events can elicit rich interpretations about events both in language production tasks and non-linguistic tasks from children of similar ages (Bunger et al., 2016; Göksun et al., 2010; Nappa, Wessell, McEldoon, Gleitman, & Trueswell, 2009; Ünal et al., 2017).

² For the selection of stimuli for the indirect visual access events, a list of events that would be depicted by an indirect visual cue was created. The familiar visual access items were constructed by using the most typical object on which this action would be performed (e.g., for “cracking,” an “egg” was picked). For unfamiliar visual access items, these objects were replaced by another object that would not be easily identified and/or associated with the action by children (e.g., a walnut, little pieces of wood, etc.)

³ Due to the binary nature of the data, Tables 1 and 3 only report the proportion of selecting one of the options, the visible photograph, split across Verb Types, Types of Access, and Age groups. The proportion of selecting the other option, the facing down card, can be deduced by subtracting these proportions from 1. Since the data that contributes to Matching and Mismatching Verb conditions come from different items, the proportions of selecting the visible photographs for Matching and Mismatching Verb trials do not add up to 1.

⁴ A different analytical strategy in which we started with the most inclusive model and gradually removed the non-significant effects and interactions to reach to the most parsimonious model also revealed the same findings in Experiment 1 and all subsequent experiments.

Additional information

Funding

This work was supported by the National Science Foundation [BCS0749870].