Where can I find recent research on the role of context effect in perception of similarity?

Question

BACKGROUND. I’m writing a paper on the role of similarity in the theories of concept empiricism developed by the British Empiricists Locke, Berkeley, and Hume, and how the idea of context effect from modern cognitive psychology could be used to better understand these theories. The basic idea I’d like to develop in my paper is that similarity is contextual, relative to a frame of reference, and this can explain how a general abstract idea (on Locke’s theory of abstraction) or a set of similar ideas held together in the mind that are called up by awareness of an exemplar (on Berkeley and Hume’s theory of general ideas) can be formed from particular instances none of which are identical: while not identical, they are similar to each other in a certain context, relative to a certain frame of reference, and thus can be classified together by the mind (e.g. scarlet and crimson are not identical but are similar to each other, in context, relative to aquamarine; or mammals are all different from each other but similar to each other relative to birds; or birds and mammals are different but relative to plants are similar. Etc.)

QUESTION. I am unfamiliar with cognitive psychology and the literature on context effect. Where can I find recent papers on context effect? What is the best way to go about finding references, especially with experimental data, on context effect? I am aware of a paper which discusses context effect very well (Tversky, A. (1977). Features of similarity. Psychological review, 84, 327-352 doi: 10.1037/0033-295X.84.4.327). However, I’d appreciate any help I could get on finding more recent work on this phenomenon, preferably (but not necessarily) which further develops or supports Tversky.

Answer 1

A classic reference in this space is Goodman (1972) "Problems and projects", which develops exactly this idea that similarity is poorly defined unless you specify in what respect. From your description, you might also enjoy Barsalou's (1983) "Ad-hoc categories", which emphasizes how much coherent groupings depend on context. These are not exactly recent, but you should turn up some relevant stuff by looking for recent papers that cite these.

There are a few different lines of work building on Tversky's results. The most prominent problem with that account is the combinatorial explosion you get out of conjunctive features. The relations between features are also important to people, but considering relations as just another kind of feature explodes the number of features you have to consider in an utterly horrifying way.

Probably the most direct solution is to consider structured representations that assign roles to features and judge similarity by the goodness-of-fit of a mapping between these structured representations. If you had to pick one entry-point to this literature, it'd probably be Gentner & Markman (1997). "Structure mapping in analogy and similarity", but again, if you look for work citing this you'll get stuff up to this year, it's an ongoing project. Goldstone has also done a bunch of work in this area, check out Goldstone & Son (2005). "Similarity" in the Oxford handbook of thinking and reasoning for a nice review.

That's probably the most popular/well known approach, but these folks are not the only game in town. There's a more or less independent angle of attack from people interested in Bayesian property induction. Try Tenenbaum, J. B., & Griffiths, T. L. (2001). "Generalization, similarity, and Bayesian inference." This is a BBS paper, so you'll see commentary from all kinds of positions there. Again, search-citing-articles is your friend for the most recent work in this line.

Getting more out into the fringes (where the really fun stuff happens) you could also consider things like transformational similarity (Hahn, U., Chater, N., & Richardson, L. B. (2003). "Similarity as transformation") or the quantum cognition account (Pothos, E. M., Busemeyer, J. R., & Trueblood, J. S. (2013). "A quantum geometric model of similarity"). Don't be turned off by the word 'quantum', the math is quantum but these folks have their feet firmly on the ground. I think you'll enjoy the intro section to this paper, it's the most recent and engages most directly with the idea of context and frames of reference.

Hope this is the kind of thing you were after!

Full refs:

Goodman, N. (1972). Problems and projects.

Barsalou, L. W. (1983). Ad hoc categories. Memory & cognition, 11(3), 211-227.

Gentner, D., & Markman, A. B. (1997). Structure mapping in analogy and similarity. American psychologist, 52(1), 45.

Goldstone, R. L., & Son, J. Y. (2005). Similarity. In The Oxford handbook of thinking and reasoning.

Tenenbaum, J. B., & Griffiths, T. L. (2001). Generalization, similarity, and Bayesian inference. Behavioral and brain sciences, 24(4), 629-640.

Hahn, U., Chater, N., & Richardson, L. B. (2003). Similarity as transformation. Cognition, 87(1), 1-32.

Pothos, E. M., Busemeyer, J. R., & Trueblood, J. S. (2013). A quantum geometric model of similarity. Psychological Review, 120(3), 679.