For example, the most prototypical exemplars of bird is
robin, the least ones are
penguin. But since it rejects the classical theory (aka the necessary and sufficient conditions), I think it should talk about conditions instead of exemplars to avoid comparing apple with orange. So for example,
eating worm are the most prototypical conditions, and
running fast or
living in Antarctica are less prototypical.
So why doesn't the prototype theory talk about features? I agree that the theory is meant to describe the reality that regular thinking isn't about features or boundaries, but being able to discuss with features is also convenient as well.
Advantages of using features can be seen in polysemy, meaning shift, or metaphor:
- In polysemy there is a basic sense that is frequently used than others. (A course in Cognitive Linguistics: Polysemy)
In 1980s, most people in the United States would consider a "family" consisted a legally married heterosexual couple and a child. Without a child they are just a couple, and without a legally marriage they just cohabit. Now the definition/understanding/perception/attitude is more flexible.
When we say an argument is like a war, we don't think much about typical arguments, but about features of
war. Only with features that we can map from the source domain to the target; a list of exemplars of argument is hard to map to a list of exemplars of war. Comparing a fiery debate between John and Mary with World War 2 sounds odd.