'What is X?' is always an ambiguous question. Ambiguity in language is generally resolved by context and and clarification. For instance, 'What is X?' might be understood as:
- What is the nature of x? (ontological)
- What is the purpose of x? (explanatory)
- What is the behavior of x? (description)
- What is the name of x? (reference)
- I have the name of x; to what does it refer? (ostension)
Think about it. When you are having a discussion about a word in the context of grammar, 'What is X' becomes a question you answer with a dictionary. It is a noun. It is a verb. It is a particle. If you ask the question while repairing a car, then the response might include the name and explanation. 'What is that?' 'It's a butterfly valve, and it regulates the flow of oxygen in coordination with the accelerator.' In philosophy class, of course, one becomes very ontological with the question by seeking what makes something real or allows us to know we have this and not that. The real definition is a typical answer to the question. 'What is a person?' 'Man is a featherless biped.' 'No, it is the recognition of a human mind in society. According to Kant, we know...' So on and so forth. If the same conversation is had in geometry class, we might not want to ponder the ultimate nature of a line, but simply seek out the definition we use in class to find a piece of proof that has been assigned to us. 'What is a line?' 'A line is that which is determined by two points.' This would be in line with a lexical definition. There are other types of definitions, like operational definitions, precising definitions, and ostensive definitions. Each answers 'what is x' in a different context to a different end.
Metaontological positions are described by such positions as realist, essentialist, anti-realist, nominalist, and constructivist, etc.
You mention "the inherent nature of things". This is might be met with one of essentialist doctrines (SEP). From SEP:
There are at least four fairly standard ways of characterizing essentialism, and by considering two extreme views, we can easily see the differences among these four characterizations.
And yet, as I many others do, we accept that ontological properties are ambiguous, and there may no way to know a thing's ultimate nature since we have the problem of dealing with representations in our phenomenological experience rather the thing itself. Thus, we simply are using language to construct descriptions, and that those descriptions may even prejudice our observations as in conceptualism. From WP:
In metaphysics, conceptualism is a theory that explains universality of particulars as conceptualized frameworks situated within the thinking mind.2 Intermediate between nominalism and realism, the conceptualist view approaches the metaphysical concept of universals from a perspective that denies their presence in particulars outside the mind's perception of them.
Thus, how to answer 'What is X' in this sense is given by your metaphilosophical dispositons. Is there a best way? To me, such arguments ultimately degenerate into boo-yay contests that can never be settled in the same way the meta-ethical theory of non-cognitivism dominates ethical discussions.