Respects for similarity

| How do we determine whether two things are similar? |

There are many ways in which to describe things as being the same or different, but the similarity of two items cannot be understood without knowing how / in what sense / in what respect the items are being compared. That is, similarity goes beyond simply comparing two items. It also involves the context in which the comparison is made. As suggested in this research, that context is predetermined by the nature of the relationships we perceive between the items being compared.

Goodman claimed that the similarity of A to B is in an ill-defined, meaningless notion unless one can say “in what respects” [context] A is similar to B. He argued that similarity, like motion, requires a frame of reference [context]. Just as one has to say what something is moving in relation to, one also must specify in what respects [context] two things are similar. For example, if Mary says that John is similar to Bill, one may have no idea what she means until she adds the observation [context] that they both are avid chess players (p. 254) …

[Goodman] called similarity a chameleon, but we believe that similarity is more like two yoked chameleons: The entities entering into a comparison jointly constrain one another and jointly determine the outcome of a similarity comparison. (That is, paradoxically, two chameleons may behave in a more orderly manner than one.) Thus, [the concept of] similarity is changeable and context dependent [across various instances] but systematically fixed in context [for a specific instance] (p. 272) …

[O]ur observations suggest that the effective representations of [two terms being compared] are determined in the context of comparison, not prior to it. It is as if the two terms were dancers: Each dancer may have a repertoire of stylistic preferences, but the actual performance depends on an interaction between the two (p.275).

Medin, D. L., Goldstone, R. L., & Gentner, D. (1993). Respects for similarity. Psychological Review, 100, 254-278.