(1) Taking account of currently live topics in epistemology, why would one suppose that any specific epistemological problem can or should be seen as ‘finished’ (e.g. whether perceptions count non-inferentially as a priori, whether ‘justification’ is necessary)? This seems analogous with the idea of the possibility of a ‘completed’ ideal science.
The point is that your assertions containing epistemology is your responsibility apart from the logical argument/structure.
(2) A specific object given to a child, through normative language use, may have a name (‘pink Sphinx’) conceptually equivalent to that of an imagined object (the ‘mythical Sphinx’). The ontological status of the toy object and that of the naming can be seen as different, but the argument from causation is surely wrong, suggesting that the concept is communicable because the toy is perceptible. How so?
The ontology of the toy is definite
as opposed to the metaphysical Sphinx which according to the myth may be able to do a lot of stuff, like flying, very much unlike the Sphinxes of Egypt. Ok?
(3) I don’t follow the argument about logic. Logic concerns relations, and can be surely be relevant only to propositions, not to ‘individuals’ (e.g. ‘mythical Sphinx’, ‘toy pink Sphinx’)?
It's a small task to go from "individuals" to propositions, like "mythical Sphinx" nicely turns to "there is a mythical Sphinx"! Are you playing with me? Don't you think I understand what I write?
Besides, the OP relates to three articles from AP Martinich's book of Phil. of Lang. and these are B. Russell's "On Denoting", PF Strawson's reaction to it and K. Donnellan's Reference and Definite Descriptions (1966) in that vein. I'm uncertain about this third article, but I'll make it clear to you later. [Edit2, 13.08.2010:] You can also add Quine's (Ch. 1:) On What There Is (from From a Logical Point of View). [End of edit.] [Another edit.] Except Quine, these articles file under Reference and Descriptions by Martinich. [End edit.]
[Edit, 12.08.2010:] You may also think that Logics is all about symbols, but what I like to point to is the Universe of Discourse of Logics that may be easily overlooked by some, ie. beginners, popular readers. So the Epistemology point enters here, UoD. I'll clarify in greater detail as this may look a little confused and blurred. [End of edit.]