2018 Friday Poster 6627

Friday, November 2, 2018 | Poster Session I, Metcalf Small | 3pm

Temporal contingency augments attention to a referent in a word learning task
S. Tsuji, R. Mazuka, D. Swingley

The contingent nature of social interactions has been considered central for early language acquisition. Socially contingent human teachers who are relevant and appropriate in content enhance word learning [1], but establishing the mechanism by which such contingency might boost learning remains elusive.

We developed a novel, gaze-contingent eye-tracking method to systematically manipulate the contingency behavior of an abstract 3D entity in a fast-mapping task. As such, we were able to measure the effect of pure temporal contingency on word learning in the absence of a rich social environment. Temporal contingency has been shown to elicit gaze following, and suggested to serve as an ostensive signal indicating communicative intent [2]. However, whether and how it would affect the word learning process is an open question.

We examined the effect of a temporally contingent communicator on two aspects of the word learning process: Toddlers’ attention to the object being taught, and toddlers’ learning of the taught word-object associations.

Forty 14-17 month-old Japanese-learning toddlers were taught a novel word-object association in a contingent condition and a matched (non-contingent) condition, within participants (Fig.1).

The entity’s contingent reactivity (or lack thereof) was demonstrated in two scenarios. First, the entity was observed to interact with (or ignore) a second agent on-screen (Observation in Fig.1). Second, the entity reacted contingently to toddlers’ own gaze (or displayed yoked, pre-programmed motion; Interaction).

Subsequently, the entity repeatedly oriented towards one of two objects (6x target, 2x distracter), naming the target (Naming). The only way for toddlers to map the label and the target object was to register the entity’s orientation.

Finally, we assessed whether toddlers had learned the new word-object association by presenting both objects side by side, and playing either the label they just learned (4x) or a novel label (3x; Test).

To assess toddlers’ looking behavior during the test phase, we analyzed proportion of looks to the target object with condition and label type (learned, novel) as predictors, allowing random effects of trial and subject, in a growth curve analysis [3]. Toddlers looked more at the learned object than the novel object, and much more so in the contingent condition [Fig. 2; interaction condition x label type χ2 (1) = 17.20, p < . 001]. This looking preference did not depend on which word they heard [χ2 (1) = 0.00, p = .990]. Thus, a communicator’s contingency led to increased attention to the object it was referencing. It did not, however, lead to superior word learning.

Our study, for the first time, examined the effect of an otherwise non-social, purely temporally contingent “teacher” on word learning. Our findings propose a specific role for temporal contingency in the learning process: The temporal contingency of interaction partners might guide attention to particular elements in complex environments. As such, it might play a mediating role in the word learning process. Its contribution might be a prerequisite, but not a sufficient condition, to augment language learning.

References

  1. Roseberry, S., Hirsh-Pasek, K., & Golinkoff, R. M. (2014). Skype me! Socially contingent interactions help toddlers learn language. Child Development, 85(3), 956-970.
  2. Deligianni, F., Senju, A., Gergely, G., & Csibra, G. (2011). Automated gaze-contingent objects elicit orientation following in 8-month-old infants. Developmental Psychology, 47(6),
  3. Mirman, D. (2014). Growth curve analysis and visualization using R (pp. 109-112). Boca Raton, FL: CRC