Putting the Pieces Together: Is There a Connection Between Weak Global Bias, Verbal Ability, and Object Categorization in Autism?

Background: Autism is characterized by disrupted language development, social deficits, and atypical patterns of interacting with objects. This may stem from a tendency for children with autism to not focus on relationships between details in their environments (e.g., Happé & Booth, 2008). Typically developing (TD) children, on the other hand, tend to focus on overall impressions, a bias that appears to follow a prescribed developmental course. For example, while early on TD children tend to focus on fine grain detail to categorize objects, as their productive count-noun vocabularies grow (herein: vocabulary size), they focus on an objects’ overall shape (Pereira & Smith, 2009; Smith, 2003). The development of this bias toward Gestalts is important for TD children as it may serve an adaptive function (e.g., Stephen, Dixon & Isenhower, 2009). Understanding how object categorization style relates to verbal abilities in autism may shed light on differentiated adaptive functioning.

Objectives: This study compares how vocabulary size relates to the ability to categorize objects on the basis of their overall shape (vs. fine-grain detail) for TD children and children with autism. This is a first step towards mapping out the relation between autism and the development of an adaptive tendency to detect higher-order Gestalts. 

Methods: Twenty-two children with autism were compared to 59 TD children using methods adapted from Pereira & Smith (2009). Parents completed standardized questionnaires (MacArthur-Bates Communicative Development Inventories; Fenson et al., 1994) to indicate their child’s vocabulary size. Children then participated in a force-choice task where they were asked to identify a target object from a field of two distractors. In each trial, the objects came from one of three categories. In one category, the objects possessed rich details of shape, color, and texture. In the second color and textural information were removed. In the third, objects were built from abstract, broad-stroke forms (geometric shapes), which only afforded categorization based on overall shape.

Results: Children from both diagnostic groups were subdivided based on their vocabulary sizes: groups of children with under 100 noun vocabularies and groups of children with between 100 and 200 count-nouns. TD children in the low vocabulary group demonstrated stratified performance across all object conditions. TD children in the high vocabulary group performed equally well on abstract objects and objects from the middle condition. For children with autism relative performance across each object condition did not differ between vocabulary groups. Further, children with autism performed at a high level compared to TD children across all conditions.

Conclusions: The current study makes several assumptions. The first assumption is that vocabulary size can be seen as a contextual factor. The second assumption is that categorization of shape abstraction objects translates to the ability to process Gestalt information. The third assumption is that Gestalt processing is an adaptive function that arises when contexts make tasks difficult. Under these assumptions, the fact that children with autism did not demonstrate stratified performance seen in TD children may suggest that they do not adapt to contextual changes in the same manner.