My takeaway was that Concepts have Contexts. The problem is that the Context may vary from person to person. Kinda like an n-gram, where n can vary between peop

My takeaway was that Concepts have Contexts. The problem is that the Context may vary from person to person. Kinda like an n-gram, where n can vary between people. Later, I read Lotfi Zadeh who treated Concepts as a kind of probability distribution. Such as mapping someone's height, described as "short", "tall", "very tall", to a range measured in centimeters. Pygmies and Belgians may vary. Now days, I guess we'd use word embeddings.