The value of comparison

The value of comparison

More about crystallized knowledge.

 

This nice and readable paper presents a different perspective to "similarities" subtest and to learning activities involving comparison.

Boroditsky, Lera. Comparison and the development of knowledge.   Cognition 102 (2007) 118-128.

http://psych.stanford.edu/~lera/papers/comparison.pdf

Finding similarities between objects and concepts (like in the "similarities" subtest) helps us categorize (similar things belong to the same category), make inferences (if A behaves in a certain way, and A is similar to B, B probably also behaves in the same way), learn (this new thing is similar to what I already know), and retrieve information from memory (due to their similarity, retrieving A will facilitate retrieving B).

What about finding differences between objects and concepts?

Students were shown drawings of four relatively similar objects – a deer, a goat, a horse and a donkey.  They were then asked three questions.  The first question asked half of them to describe three similarities between two of these animals, and the other half to describe three differences between two of these animals.  The second question asked them to rate the similarity between the two animals they addressed in the first question.  The third question asked them to rate the similarity between two animals not addressed in the first question.

The act of comparison itself, whether it was through looking for similarities or through looking for differences, caused an animal pair to be rated as more similar compared to its rating without being compared first.  Apparently, while people are looking for differences, they also find similarities.  Finding meaningful differences between objects or concepts requires first to see the similarities between them.  If you say that a goat has a shorter tail than a donkey (a difference), this highlights the fact that both animals have a tail (a similarity).

Does this work the same with dissimilar objects and concepts?

This experiment was repeated using dissimilar objects (a pretzel, a phone, a hat, a football).  Thinking about differences between two dissimilar objects made them seem even more different (highlighted the differences between them).  Thinking about similarities between two dissimilar objects didn't change their similarity rating.

Comparing (finding similarities or differences) similar things increases the perceived similarity between them.  Finding similarities between dissimilar things does not increase their perceived similarity, but finding differences between dissimilar things makes them seem more different.

Differences between similar objects are usually alignable (the horse is larger than the goat).    An alignable difference indicates the existence of a common factor to both objects.  Discovery of this common factor increases the perceived similarity between the objects.  Differences between dissimilar objects are usually non alignable (a pretzel is a kind of food and a phone is a communication instrument, a pretzel has salt on it and a phone doesn't have salt on it).  Non aligned differences do not uncover any common factor to the two objects and thus don't increase the similarities between them.

To conclude, the process of comparison (finding similarities and differences) can change the way we perceive the world (the way we represent objects or concepts).  The comparison process makes similar objects or concepts seem even more similar.  Finding differences between dissimilar objects or concepts makes them seem more different.