P1 Words represent concepts.

P2 Concepts can be made of more concepts. (dragon=lizard with wings)

C1 Therefore if a word purports to represent a combination of many concepts, but those concepts fail to cohere into a new concept then the word fails to represent a concept.


For a example of my conclusion "Buglump" is a Colorless red existing dog that is a cat and doesn't exist.

Buglump's concepts fail to cohere and thus do not form a new idea. Unlike wings and lizard, which form the new thought dragon

You need a D1 or something like that. Some people use "concept" to refer to "idea", and others more narrowly use it to refer to "idea that is so important that it is conventionalized as a word (or other symbol)". I would use "concept" in the latter manner (granting that "concept" is becoming a meaningless term in current usage). There are cases of words that don't represent a single concept, such as "up". As a particle in V-particle constructions, (look up, chat up, screw up) it doesn't represent a concept, instead it's a somewhat arbitrary component of a lexical item made up of two words. You could fix this by fine-tuning the concept "word", but we don't want to beg the question.

"Made of" suggests something along the lines of language acquisition, where "animal" is "made of" the concepts "dog, cat, rat, horse...", and dog et al. are cognitively more basic (acquired earlier), and "animal" is an abstraction that subsumes a number of similar concepts. In a way, you glue concepts together. However, "husky, poodle, chihuahua" further differentiates the first-order concept "dog" so those concepts are not "made of" other concepts. But this is not a huge objection, so I think P2 is probably correct.

I will leave C1 alone for the moment, and turn to the matter of what a concept is good for. A concept, like "dog", serves a cognitive function, that it identifies an open-ended class of facts that you can perceive. A concept abstracts away from a specific instance, and allows you to identify all of those things that are of the same cognitive category, while being distinct from other things. So to have a concept, you have two or more things that are similar in a particular way yet different in irrelevant ways, and the group is distinct from some third thing.

The problem with "Buglump" is that there is no thing, much less a class of things, that are "Colorless red existing dog that is a cat and doesn't exist". The stimulus for a concept is some actual thing: it is useful to invent the concept "transistor", "syllable", "recursion" because there are real things out there that these concepts are about. There is no thing for the concept to be about; this broken concept fails to fulfill the cognitive function of a concept.

This may be what you mean when you say that the concepts of Buglump "fail to cohere and thus do not form a new idea".

There are words (especially in the political sphere) which don't actually refer to anything, rather they have an emotive function. You would not say that such words represent concepts: so it's not that all words represent concepts, rather, all concepts are represented by words.

I would suggest starting with a good grasp of what a concept is (and how it relates to ideas; what the function of a concept is, etc), for which I recommend How we know by Harry Binswanger.

I mentioned the tomato example in the linked slides, as well as others. Consider gold, which retained a narrower meaning vs pyrite, compared to jade, which always referred to two distinct minerals that are still now treated as one category. We as laymen don't always know what things mean and sometimes defer to expert judgments.

As for the last point and "we", that is discussed by Putnam (also source for the gold and jade examples) as discussed in my slides. There is some socially held aspect of lexica meaning in that individuals do not always know how to precisely determine membership into a category but act as if they do and defer to experts as relevant. Anyway, I meant "we" in a disjunctive sense: we may individually have some incoherent concepts, possibly. Especially if you consider variation. Like the student who thought penguins are mammals, maybe they did know about the eggs but not the lack of hair, hence a somewhat mixed concept? (Platypuses aside!) But even if not, I can imagine such a self-contradictory meaning existing for someone parallel to "oh I didn't know that spelling was the same as that other word I know how to pronounce", etc.

As for your last examples, good point. But are those really incoherent (self-contradictory) or just vague and/or polysemous?