Abstract
Facial configural processing may be preferentially vulnerable to face inversion, and there are reports that patients with an apperceptive variant of acquired prosopagnosia have impaired configural processing. However, configuration has several meanings, including the perception of spatial structure (particularly second-order relations) or holistic styles of processing. To distinguish between these two, we recruited ten patients with acquired prosopagnosia to test their processing of abstract objects, in which whole-object processing demands could be manipulated more systematically than faces.
First, we used an oddity paradigm in which subjects had to compare three dot patterns to detect the odd one out. This consisted of a series of blocks in which we manipulated stimulus complexity (number of dots = 2, 4 or 8), variations in size and orientation between the three patterns, or the regularity of global structure (octagons versus random structure in the 8-dot stimuli) generated effects that would suggest a failure to process at a whole-object level. Second, we used hierarchical letters, to determine if reaction times reflected perception of global structure.
With dot-patterns, patients were impaired even for simple 2-dot stimuli, but did better with more complex patterns, when size or orientation varied, or with a regular global structure. In experiment 2, they demonstrated normal latency effects of global-level processing.
We conclude that apperceptive prosopagnosia is associated with a configural deficit that impairs perception of spatial structure, not just for faces but also for non-facial patterns. While we cannot conclude that holistic processing is entirely normal in these subjects, their performance shows significant modulation by whole-object structure, indicating that some whole-object processing is occurring in these patients.
NIMH 069898, CIHR MOP-77615, Canada Research Chair and MSFHR Senior Scholarship .