Similarity and Symbols in Human Thinking
edited by Steven A. Sloman and Lance J. Rips

"What cognitive processes account for the success of science in making predictions and increasing our understanding of the natural world? This question underlies much of the modern history and philosophy of science. It should also be one of the foundational questions of cognitive science." -- F. C. Keil et al. (p. 20)
Similarity and Symbols in Human Thinking is a reprint from COGNITION: International Journal of Cognitive Science, Volume 65, Numbers 2-3, 1998. Those who aren't really interested in cognitive processes will find this book to be much ado about nothing. Those treading deeply in the areas of cognition, AI, and/or psychology will find it to be a very worthwhile resource in the ongoing rules (taking one or more mental representations as input, carrying out a finite number of internal steps, and producing one or more representations as output) vs. similarity (forming conceptual categories as clusters held together by the similarity of their instances) debate.

The book includes seven different essays which all basically deal with the same issue. The conclusions are also similar. Essentially, the various authors conclude, extreme views in the debate are unwarranted as hybrid processes are at work. People sometimes 'compute' using rule application while in other instances using exemplar similarity. E. E. Smith et al. used positron emission tomography providing empirical data to add weight to this hypothesis.

Although I'm not recommending this to anyone interested only in popular science titles, some of the testing performed--particularly the details surrounding the methodologies employed--can be useful to those wishing to more fully understand the means by which scientific understanding is derived.

From the publisher:
Much of current cognitive science is a debate between two views of thinking--thinking as governed by mental rules and thinking as governed by similarity among ideas. Contributors to this volume explore these contrasting views in research on reasoning and concepts, and consider their merits from the perspectives of cognition, development, neuroscience, and artificial intelligence. The book evaluates the potential of each view to describe human cognition and examines whether systems compatible with these different perspectives might work together in explaining thought. While maintaining a high level of scientific sophistication, the book remains accessible to undergraduates and researchers in other fields.