Plato was a realist, in the same way the term is used today by philosophers. Reality isn't mind-dependent, hence the Forms. Again, Derrida and all those guys are far more the idealists than Plato or Platonic thinkers like Frege.
I still don't find Charles' logic as expressed persuasive, but I recently finished an article by one Edward C. Moore which makes the case for the term "realism" clearer. Said article introduces a collection of the work of Charles Sanders Pierce, and argues that "realism" is predicated on the question as to whether or not human concepts of reality correspond to anything in reality:
"All of our knowledge consists of concepts... But objects in the external world appear to be particular determinate individuals. The question, then, is whether anything in the external world corresponds to our concepts of it... If one holds that the concepts in the mind correspond to something in the external world, then he thinks that the concepts are real, not fictional, and hence he is a realist."
Moore then goes on to present four gradations of reaction to this question of realism, with Platonic realism/idealism at one extreme and nominalism at the other. In between are "conceptualism," which argues that the concepts do have reality within the mind alone, and "moderate realism," the position of Pierce, which argues that "each external object has an essential nature, or an essence. This essence is neither universal nor particular; it just is. It is neutral. It cannot exist in a separate realm by itself, but it can exist either in an object or in a mind."
After all this, I would still say that this use of the term "realism" makes sense ONLY if one has in mind the sort of propositional construction Moore presents here. In addition, I still find the pairing "rationalism/empiricism" preferable to "realism/idealism" in any context.
With that in mind, going on Moore's essay alone I suspect I'll find Pierce's interpretation of semiology restrictive in terms of its empirical background. Moore points out that since Moore was a chemist, he tended to ground the reality he found in abstract concepts in terms of what is termed "the consequent." This refers to the notion that the potentiality represented by a concept is real if it can be demonstrated that it will invariably lead to a repeatable actuality.
This is a significant insight, but I think it is overdetermined by the paradigm of scientific investigation. In THE BURNING FOUNTAIN Philip Wheelwright finds the essence of the abstract conceptual symbol less in *praxis* than in *theoria*:
"...the attitude which a symbol represents and to which it appeals is contemplative rather than directive or pragmatic. A symbol refers to what supposedly is, not (or at least not directly) to what one is to do. It is the logos theoretikos, not the logos praktikos..."
I assume that Kant would substantially agree with this emphasis on contemplation over practicality given his assertions re: judgments of taste and "the beautiful," and obviously I agree that the symbol's appeal is first to mankind's expressive aspects, rather than the practical ones. At least I would hope having written umpteen essays on Cassirer would make that clear, but one never knows for sure.