As an ornery aporetician, I want ultimately to say that an equally strong case can be made both for and against the thesis that ficta are impossibilia. But here I only make (part of) the case for thinking that ficta are impossibilia.
Every human being is either right-handed or not right-handed. (But if one is not right-handed, it doesn't follow that one is left-handed. One could be ambidexterous or ambisinistrous.) What about the fictional character Hamlet? Is he right-handed or not right-handed? I say he is neither: he is indeterminate with respect to the property of righthandedness. That makes him an incomplete object, one that violates the law of Excluded Middle (LEM), or rather one to which LEM does not apply.
Hamlet (the character, not the play) is incomplete because he has all and only the properties ascribed to him by the author of the play, and the author left Hamlet's handedness unspecified. It is worth noting that Hamlet the play is complete and this holds for each written token of the play, the type of which they are tokens, and each enactment of the play. This is because the play and its enactments are actualia.
But don't we say that Hamlet the play is fictional? We do, but what we mean is not that the play is an object of fiction, but that the people and events depicted therein are fictional. The play is not fictional but entirely real. Of course, there could be a play that is a mere object of fiction: a play within a play. The same holds for novels. My copies of Moby Dick are each of them complete and actual, hence full-fledged citizens of the real, with all the rights and privileges pertaining thereunto; but Ishmael, Queequeg, and Ahab are not. They are objects of fiction; those books are not. And presumably the type of which they are tokens, though an abstract object, is also actual and complete. A person's reading or 'enactment' of the novel is typically a long, interrupted process; but it too is complete and actual and resident in the real order.
Back to the character Hamlet: he is an incomplete object, having all and only the properties ascribed to him in the play (together with, perhaps, entailments of these properties). London Ed balks at this:
I don't follow this at all. I don't agree with the second sentence "He has all and only ….". Of course Shakespeare said that there was a person called ‘Hamlet’ who had certain properties (e.g. he said that Hamlet was a prince of Denmark. It doesn’t follow that there is someone who has or had such a property. For example, legend says that there was a horse called ‘Pegasus’ that flew. It doesn’t follow that there are or were flying horses.
This objection shows misunderstanding. I did not say or imply that there exists in actuality, outside the mind, a man named 'Hamlet.' The point is rather that when I read the play there appears before my mind a merely intentional object, one that I know is fictional, and therefore, one that I know is merely intentional. If Ed denies this, then he denies what is phenomenologically evident. And, as a matter of method, we must begin with the phenomenology of the situation.
Suppose I write a two-sentence novel:
It was a dark and rainy night. Shakey Jake, life-long insomniac, deciding he needed a nightcap, grabbed his flashlight and his raincoat and headed for the Glass Crutch bar and grill, a local watering hole a half a mile from his house.
Now I couldn't have written that, and you can't understand it, without thinking about various intentional objects that do not exist. Am I saying that there exist objects that do not exist? No, that would be a contradiction. Nor am I committed to saying that there are objects that have mind-independent being but not existence. Furthermore, I am not committed to Meinong's doctrine of Aussersein.
All I am doing is holding fast to a phenomenological datum: when I create a fictional character as I just did when I created Shakey Jake the insomniac, I bring before my mind an intentional object. (The act-object schema strikes me as having pretty good phenomenological credentials, unlike the adverbial schema.) What can we say about this merely intentional object? First, it is no part of the acts through which I think it. My acts of thinking exist in reality, but Shakey Jake does not exist in reality. (This point goes back to Twardowski.) When I think about Hamlet or Don Quixote or Shakey Jake, I am not thinking about my own mind or any state of my mind. I am not thinking about anything real. But it doesn' t follow that I am not thinking of anything.
If Ed denies that there are merely intentional objects, then he is denying what is phenomenologically evident. I take my stand on the terra firma of phenomenological givenness. So for now, and to get on with it, I simply dismiss Ed's objection. To pursue it further would involve us a in a metaphilosophical discussion of the role of phenomenological appeals in philosophical inquiry.
Ficta are Impossibilia
Let us confine ourselves to purely fictional objects and leave out of consideration real individuals who are partially fictionalized in fables, legends, apocryphal stories, so-called historical novels that blend fact and fiction, and the like. One of my theses is that purely fictional objects cannot exist and thus are broadly logically impossible. They are necessarily nonexistent, where the modality in question is broadly logical. It does not follow, however, that pure ficta have no ontological status whatsoever. They have a mode of being that could be called existential heteronomy. On this point I agree with Roman Ingarden, a philosopher who deserves more attention in the Anglosphere than he receives here.
Earlier I gave an argument from incompleteness: the incomplete cannot exist and so are impossible. But now I take a different tack.
Purely fictional objects are most plausibly viewed as made up, or constructed, by novelists, playwrights, et al. It may be that they are constructed from elements that are not themselves constructed, elements such as properties or Castaneda's ontological guises. Or perhaps fictional objects are constructed ex nihilo. Either way, they have no being at all prior to their creation or construction. There was no Captain Ahab before Melville 'cooked him up.' But if Ahab were a merely possible individual, then one could not temporally index his coming to be; he would not come to be, but be before, during, and after Melvlle's writing down his description.
The issue could be framed as follows. Are novels, plays, etc. which feature logically consistent pure ficta, something like telescopes that allow us to peer from the realm of the actual into the realm of the merely possible, both realms being realms of the real? Or are novels, etc. more like mixing bowls or ovens in which ficta are 'cooked up'? I say the latter. If you want, you can say that Melville is describing something when he writes about Ahab, but what he is describing is something he has made up: a merely intentional object that cannot exist apart from the acts of mind trained upon it. He is not describing something that has ontological status apart from his mind and the minds of his readers. He is also not descrbing some real feature or part of himself as subject. So we could say that in describing Ahab he is describing an item that is objectively but not subjectvely mind-dependent.
Here is an Argument from Origin:
1. Pure ficta are made up or constructed via the mental acts and actions of novelists, playwrights, et al.
2. Ahab is a pure fictum.
3. Ahab came into being via the mental activity of a novelist or playwright. (from 1,2)
4. No human being comes into being via the mental activity of novelists, et al., but via the uniting of human sperm and human egg.
5. Ahab is not a human being. (from 3, 4)
6. A merely possible human being is a human being, indeed a flesh-and-blood human being, though not an actual flesh-and-blood human being.
7. Ahab is not a merely possible human being, but a fictional human being where 'fictional' unlike 'merely possible' functions as an alienans adjective.
This argument does not settle the matter, however, since it is not compelling. A Meinongian or quasi-Meinongian could, with no breach of logical propriety, run the argument in reverse, denying (7) and denying (1). One man's modus ponens, etc.