Detection of incongruent naming of occluded objects in 14-month-old infants

• In Event: Poster Session 06
  In Poster Session: PS 06 Topic - Language, Communication

Abstract

Once infants learn their first words, they can potentially also notice when those words are used incorrectly. The N400 is an ERP component sensitive to semantic violations such as mislabeling, or incongruent ending of sentences (Kutas & Hillyard, 1980). In Parise and Csibra (2012) 9-month-old infants showed an N400 effect when a labelling was followed by the fall of an occluder, revealing an incongruent (familiar) object. Reversely, in studies by Friedrich and Friederici (2005) pictures were presented first, followed by labelling. In this context it was at 14 months of age that infants showed an N400 effect in response to incongruent labels.
The present study aimed to probe infants’ detection of semantic mismatch when this requires relying on working memory. Specifically, we tested whether (in)correct naming referring to a sustained object representation can elicit semantic mismatch. We measured whether infants would show an N400 effect when a previously seen, but currently occluded familiar object is labelled incongruently. We presented 14-month-old Danish-speaking infants videos where a person first shows an object, then puts it in a box. This is followed by a hand pointing towards the box, and an auditory label (congruent or incongruent with the content of the box) presented concurrently with pointing. If infants detect a mismatch, a larger N400 is expected for the incongruent labels. Based on previous data suggesting that infants’ receptive vocabulary undergoes a radical increase around 14 months of age (Bergelson & Swingley, 2015), we expect 14-month-old infants to detect the mislabeling and therefore show an N400 effect.
In addition to the ERPs, gamma-band oscillations will be analyzed, which have previously been found to accompany sustained object representations (Kaufman, Csibra, & Johnson, 2003). Higher gamma-band activity in the incongruent condition would indicate enhanced processing potentially due to the detection of a mismatch between the auditory label and the sustained representation of the object.
The study is pre-registered with n=28 infants to be included. Visual inspection of preliminary data shows an effect of semantic mismatch at the N400 peak at the areas of interest (based on Parise & Csibra 2012; and Forgács et al, 2018). Data collection is almost done (n=27), statistical analyses will be carried out upon collection of the full sample (October 2020).
In case of positive results in the current study, a follow-up study (based on Forgács et al, 2018) is planned to investigate whether infants show a decreased n400 effect if the incorrect label is in fact congruent with what another person believes to be hidden at the object’s location (and thus from her perspective the labelling is congruent). This is motivated by the hypothesis that humans are influenced in their cognition by the presence and perspective of others (Kampis & Southgate, 2020), which may be particularly pronounced in young infants (Southgate, 2020).

Authors

• Dora Kampis, University of Copenhagen
  Presenting Author

• Emilie Poulsen, University of Copenhagen
  Non-Presenting Author

• Eugenio Parise, Lancaster University
  Non-Presenting Author

• Dimitris Askitis, University of Copenhagen
  Non-Presenting Author

• Victoria Southgate, University of Copenhagen
  Non-Presenting Author