The nonindependence of stimulus properties in human category learning

Abstract

Typically, models of category learning are verified through behavioral experiments with stimuli consisting of putatively independent dimensions such as shape, size, and color. The assumption of independence is critical in both the design of behavioral experiments and the development of models and theories of learning. Using the standard classification learning paradigm and a common stimulus set, the present work demonstrates that the assumption of independence is unwarranted. Systematic relations span stimulus dimensions and govern learning performance. For example, shape is not independent of size and color, because humans quantify size and color over shape when shape is relevant to the categorization. This quantification is reflected in natural language use (e.g., "blue triangle" as opposed to "triangle and blue"). In this example, color and size are predicates and shape is the argument. Across four experiments, the difficulty of mastering a classification rule can be predicted by the number of predicates that must be unbound in order to free rule-relevant stimulus dimensions.

Words can slow down category learning

“…Love and Markman (2003) suggested that shape is integral to categories and that other properties, such as color, are represented in relation to shape; in addition, developmental studies have noted the centrality of shape in determining category membership as a cue to function or kind (Gelman & Bloom, 2000;Prasada, Ferenz, & Haskell, 2002). The origin of the link between shape and novel words, however, may be developmental and rooted in the processes of attentional learning (e.g., Robinson & Sloutsky, 2007;Yoshida & Smith, 2003).…”

Induction with uncertain categories: When do people consider the category alternatives?

“…In contrast, Murphy and Ross have often used geometric exemplars varying in shape and color. There have been some suggestions that people may treat shapecolor combinations as integral stimuli (Love & Markman, 2003). It may require some effort to parse such stimuli into their components, making the assessment of the feature frequencies more difficult than in our text descriptions.…”

When learning to classify by relations is easier than by features

“…For example, Goldstone (1996) found that increasing the number of relational changes between two scenes could increase the rated similarity of the two scenes instead of decreasing it. In addition, Love and Markman (2003) showed that even classical features, such as shape, exhibit signs of relational processing by interacting with other features such as size, causing conjunctive rules involving shape and size to be harder to learn. In particular, they found that shape served as the argument to dimension predicates such as size and colour.…”