On 26 February 2016 at 09:50, Patrik Austin <[log in to unmask]> wrote: > At this point I would conclude that it's safe to say that all meaningful > languages have nouns. Only if "noun" is defined in such a way that this statement is part of the definition of "noun". It's no sort of logical or empirical conclusion. > And: on nouns vs. predicates: when predicate is used as a term for a > flexible class, the perspective is western/English. Your argument is not > logically motivated. For simplicity, let's look at a language that has a > flexible class for (or makes no formal difference between) nouns and verbs. > On the surface level there are in fact three logical possibilities: > > 1) the language expresses them all as nouns > 2) the language expresses them all as verbs > 3) the language expresses them all as a class distinct from noun or verb; > one that can e.g. be called "predicate". > Right. Now all this is irrelevant to whether predicates are nouns, because while you could indeed call the class that neutralizes Noun and Verb "predicate", you could equally well call it "sausage", and in neither case is this going to tell us what the normal meaning of "predicate" or "sausage" is. "Predicate" is the term for a fundamental element in the structure of the proposition encoded by syntax. Where there is homoeomorphism between the syntactic encoding and the logical structure encoded, it would be legitimate to extend "predicate" to the syntactic counterpart. Noun and Verb, OTOH, are, when used cross-linguistically, vague family-resemblance categories. In the description of a particular language, the definitions can be much tighter, and dependent on a complex of language-specific criteria. --And.