Poster #31 - Infants Fail to Categorize Forms by Kinds

• In Event: 3-053 - Poster Session 10
  In Poster Session: PS 10 Section - Cognitive Processes

Integrative Statement

Only by around their first birthday do infants begin to categorize objects by what kinds of things they are, using properties like shape and function to identify such object kinds (Futó et al., 2010; Xu, 2007). Nevertheless, even very young infants identify and track objects as cohesive bodies that move rigidly, are in only one place at a time, and act on one another only when they come into contact (e.g., Spelke, 1990). Young infants also discriminate visual forms, i.e., connected figures with distinct shape properties like relative lengths, angles, and orientations (e.g., Cohen & Younger, 1984; Schwartz & Day, 1979; Slater et al., 1991). However, prior to their first birthday, infants do not automatically bind form information to objects (see Ullman et al., 2017). Xu & Carey (1996) showed that 10-month-old infants expected only one object to be present behind an occluder despite seeing two different objects emerge in alternation from either side. While infants detected that these objects had two different forms (as indicated by their longer looking time to the two different objects versus two of the same object during initial viewing), they did not bind that distinguishing form information to the object motion to individuate two versus one object. Infants at this age only succeed in identifying two different objects under these conditions when each object is given a verbal label (Xu, 2002; Dewar & Xu, 2007). Indeed, infants even expect two different object kinds versus two examples of the same object when they hear two different verbal labels in the absence of any visual stimuli (Xu, Cote, & Baker, 2005; Dewar & Xu, 2007). Can forms alone bind to verbal labels to grant infants access to kinds of forms, like angles?
In this study, 26 10-12-month-old infants (median = 11.49 months, range = 10.54 - 12.61, 11 females) were presented with a series of animations displaying a box alternately occluding and revealing a pair of rotating 2D visual forms of same or different angle sizes (Fig. 1). During test trials, infants saw the exact same visual stimuli, but critically, before the box revealed the angles, infants heard either two identical or different verbal labels. If simple 2D visual forms of different shapes act like objects, then verbal labels should allow infants to develop kind categories binding distinguishing form information to distinct labels. If forms cannot serve this role, then infants might distinguish them, but have no expectations about their relation to two versus one kind of label.
Data were consistent with the second possibility. Infants looked longer at two different versus two identical objects during test trials (linear mixed-effects model, P = 0.033), but did not differ in looking to unexpected versus expected trials (P = 0.675). A follow-up study will provide motion and dimensionality cues that indicate that these forms are also objects and evaluate if the expectancy effect emerges. Such future adaptions may elucidate infants’ use of geometric forms and language in categorizing objects by kind.

Authors

• Nicole Loncar, New York University
  Presenting Author

• Holly Huey, New York University
  Non-Presenting Author

• Moira Rose Dillon, New York University
  Non-Presenting Author