Date of Award
Doctor of Philosophy (PhD)
L. R. Brooks
Determining how people identify objects and pictures in real world settings is one of the major challenges in psychology. Under normal conditions, objects and pictures are subject to an extraordinary amount of variability in perceptual characteristics. Within object categories, there are a wide variety of objects, and for any given object. there is a wide variety of possible viewing conditions. The ease with which people overcome the variability in the surface characteristics of objects and pictures is the phenomenon of interest here.
Current theories of object identification stress robust identification procedures. By the robustness view, identification proceeds by identifying reliable stimulus characteristics, and assigning category on the basis of these characteristics. Idiosynchratic, stimulus specific features are not considered a normal path to object identification. In abstractive theories of concept learning, this view of object identification is often implicit, but sometimes it is an explicit design feature of the theory. In contrast, episodic views of concept learning stress the use of prior items in categorization. Implicit in such a view is that adventitious stimulus characteristics that have been used previously to good advantage will be exploited in object identification. This is akin to Newell and Simon's view of expertise in which the expert adapts prior solutions to new problems rather than apply the same algorithmic solution again. This adaptation of prior solutions to new problems will be termed special expertise.
It is widely recognized that robust identification procedures do not exist for all object" and viewing conditions. Therefore. normal object and picture identification is a mixture or hybrid of robust procedures and special expertise. However, it is necessary to place some empirical constraints on the contribution of special expertise to determine whether special expertise is a special case reserved for unusual perceptual problems or whether it is a normal part of object perception.
The following series of experiments assesses the contribution of special expertise to picture identification by replicating and extending prior work. Additional work is reported that clarities factors potentially important to special expertise. Implications for theories of concept learning will he discussed.
Baker, John Gerard, "The Contribution of Special Expertise to the Identification of Pictures" (1990). Open Access Dissertations and Theses. Paper 3558.