Can negative existentials be referentially vindicated?

In The Theory of Objects, Alexius Meinong used true negative existentials to argue in favour of non-existent objects: in order to assert veridically that an object O does not exist, one has to refer to O itself. From Bertrand Russell’s “On Denoting” onwards, it has become a commonplace to say that this argument does not work. For every sentence apparently concerning non-existents one can provide a paraphrase which eliminates the singular term contained in it and therefore dispels the illusion of a reference to a non-existent item. As is well known, this strategy implies that such a singular term is a camouflaged definite description. However, Saul Kripke and other supporters of the so-called new theory of reference have claimed that singular terms directly referring to their designation (ordinary proper names, indexicals, mass and substance terms: from now on, DR terms) are not at all synonymous with descriptions. In their opinion, true negative existentials represent no exception to this claim. In what follows I would like to stress that, though Kripke’s critique of the Russellian treatment of negative existentials is convincing, his treatment of true negative existentials containing a DR term is flawed. Nor do other contemporary accounts that share Kripke’s critique as far as these sentences are concerned fare any better. In fact, I think that Meinong’s argument may be correct where true negative existentials containing DR terms are concerned.