The Curious History of the Gap/Overlap Procedure

Background: Impaired attentional disengagement is widely considered a key cognitive mechanism in the pathogenesis of autism (Elison & Reznick, 2012; Keehn, 2013).  Disengagement is often examined with the gap/overlap procedure (also referred to as a shift/disengagement procedure or competition/non-competition procedure). The translational potential of this construct holds much promise, and as such, requires careful scrutiny.

Objectives: To trace the use of this procedure since its inception (Saslow, 1967) to identify interpretational pivot points within and between separable bodies of literature (i.e., an infant cognition line and an adult cognition line).  More specifically, I will attempt to identify papers that offered interpretations that subsequently altered researchers conceptualization of performance in this procedure within a given line of research.  I will also identify convergence/divergence between disparate bodies of literature that have employed this task.

Methods: As of November 1, 2014, Web of Science identified 382 papers that cited the original report of the gap/overlap paradigm (Saslow, 1967).  Each abstract was examined and dichotomously classified according the age of the sample studied. Those papers that examined infants and/or children were tagged for further examination in order to characterize citation patterns in the developmental line of research, which I suspected to be the line of research adopted by the autism field.

Results: Two, essentially non-overlapping bodies of work (with very few exceptions, e.g., Farroni et al., 1999; Masuzawa & Shimojo, 1997; Ross & Ross, 1980; Senju & Hasegawa, 2005) with different emphases stemmed from the original Saslow (1967) report, one in the adult cognition literature (encapsulating cognitive/systems neuroscience) and one in the infant cognition literature.  In the adult literature, effort was primarily focused on understanding/characterizing the mechanism(s) responsible for reduced latencies in the gap condition (e.g., disengagement of attention facilitating express saccades, the premotor theory, etc.).  On the other hand, with very few exceptions (c.f. Aslin & Salapatek, 1975), the infancy literature focused on the overlap condition and its service to understanding 1) the extent of the visual field during infancy, and 2) the phenomenon of obligatory attention observed in very young infants.  The emergence of cognitive neuroscience and its merger with developmental science in the late ‘80’s and early ‘90’s yielded new interpretations of performance in the gap/overlap paradigm that were adopted by the autism field (with few exceptions, see Goldberg et al., 2002) at the expense of including emerging data from the adult line of research.

Conclusions: Building a cumulative developmental science requires conceptual and methodological precision coupled with innovation.  This inquiry into the history of the gap/overlap procedure yields evidence that important findings from adults (and non-human primates) have been neglected in current interpretations of gap/overlap performance.  Whether performance in the overlap condition (or difference between overlap condition and baseline condition) reflects “attentional disengagement” as originally conceived remains open for debate and requires further examination.