Abstract
The current study examines the impact of conceptual information on competition from shared structural features during an object recall task. Previous studies with category specific visual agnosia (CSVA) patient ELM show that structural dimensions such as tapering and pinching are stored in a distributed fashion, and that integration of these dimensions can fail during recall due to competition from objects that share values on these dimensions (Dixon, Bub & Arguin, 1997). Visual similarity is therefore determined not only by proximity among distributed diagnostic structural dimensions, but also by the number of values shared by objects within a category. Importantly, ELM was able to use distinctive conceptual information to resolve structural competition during recall. This interaction between structural similarity and conceptual relatedness was replicated in normal recall of newly learned attributes to known objects (Bukach et al., 2004). However, it is difficult to define the relevant diagnostic features of real objects. In the current study, we use novel stimuli to manipulate both the number of shared features and conceptual relatedness of objects while controlling for similarity due to proximity. Participants associated either conceptually related or unrelated labels to object sets that shared few or many values on diagnostic structural dimensions. Participants in the conceptually related and shared structural values condition made more errors than any other group. These results are consistent with the pattern of errors from CSVA patient ELM and provide strong evidence for the influence of conceptual information on resolution of competition from shared values on structural dimensions during normal recall.
the University of Richmond.