Stephen Carlson wrote:It is fruitless to use semantic criteria to distinguish between Greek nouns and adjectives. Adjectives can act as noun when they take an article, and noun can act as adjectives when they are used in attributive apposition. [...]
Since you say that nouns can function as adjectives when in attributive position, I will take it, but will you then say that only context can distinguish between appositive and attributive use of a noun? That is one reason I thought nouns and adjectives should not be interchangeable. In my view adjectives only modify explicit or implicit nouns, and therefore substantive use of adjectives are actually simply normal use of an adjective with an implicit noun. And if a word can function as an adjective in attributive position, then I would rather say that it is indeed an adjective there. Otherwise, how will you distinguish between these two possible meanings of "T X T Y" where X and Y are both nouns and T is the article:
(1) "the X, which is also the Y," (apposition of "the X" and "the Y"; "the X" and "the Y" uniquely refer to the same entity)
(2) "the X that is a Y" ("Y" qualifies "the X"; "T X T Y" merely refers to a specific "X" that is a "Y" but there may be other "X"s)
Your other example αἰώνιος is inflected for gender, so it's an adjective, though it can be used substantively like a noun. In addition, contrary to your intuition, its comparative form, αἰωνιώτερον, is attested in Greek literature, so there's no issue there. But, even if they weren't, it doesn't need to have every distinctive syntactic feature of the prototypical adjective to be an adjective under linguistic prototype theory, just enough of them to distinguish them from nouns. In the case of Σαμαρῖτις, you have not given any syntactic reason that it is not a noun.
Okay thanks for finding it for me. I could not find it on LSJ so I had guessed, wrongly, that it might not exist. Anyway, my reason for considering "σαμαρειτις" in John 4:9 is precisely as above, that "T N T N" and "T N T A" are different syntactic patterns with different meanings. "T A N" is the correct order and not "T N A", which means that nouns and adjectives are syntactically different classes. Thus I did use semantic meaning to identify which class "σαμαρειτις" in John 4:9 is under. Anyway, to clarify what I mean, I am trying to describe what actual grammatical functions a word has based on its usage, syntactically and semantically, and I do not consider inflection for gender a criterion, though there must be corresponding words for the applicable genders. I never once said that "σαμαρειτις" is not a noun but an adjective. Rather I said that it is functioning exactly as an adjective in John 4:9 and exactly as a noun in other places.
Now my last question is, can every single noun function as an adjective? My answer is, no, because I believe each word's possible grammatical functions with their associated possible semantic meanings are inherent. So if a noun is never used adjectivally, I will conclude that it cannot function as an adjective, unless it later acquires such a meaning. The examples I mentioned and the references you quoted from Smyth suggest that such nouns typically have certain semantic meanings, such as denoting nationality and occupation, while other nouns cannot be used in attributive position. Would you agree?