The J.T. Leroy case raises some interesting philosophical issues. For those who haven't been paying attention: three novels have been published under the name of J.T. Leroy, and many people take themselves to have had phone conversations with Leroy and to have met and talked with him in person. Leroy is put forward as a reclusive 25-year-old transgendered former child prostitute and drug addict. It now seems plausible that novels attributed to J.T. Leroy were written by Laura Albert, a 39-year-old musician who the official Leroy narrative holds to have rescued Leroy from the streets since 1993, and to have lived with him and her husband Geoffrey Knoop since then. It also seems plausible that most phone conversations apparently with Leroy were with Laura Albert. However, the person identified as J.T. Leroy in public appearances and in face-to-face conversation appears to be Savannah Knoop, Geoffrey Knoop's half-sister.
(For background, see the New York article "Who is the Real JT LeRoy? A search for the true identity of a great literary hustler", and the more recent New York Times article "The Unmasking of J.T. Leroy", as well as the Wikipedia link above. Or for another perspective, see the official J.T. Leroy website and blog.)
Let us assume for the sake of discussion that the facts are roughly as suggested by these articles. Then we can still ask the question: Who is J.T. Leroy? Four answers seem to be possible:
(1) J.T. Leroy is Laura Albert
(2) J.T. Leroy is Savannah Knoop
(3) J.T. Leroy does not exist
(4) It is indeterminate who J.T. Leroy is
Set aside (4) for now, and try to choose between the first three if you can. (Stop here to form your own judgment before reading on.) (2) seems to me to be clearly the least plausible, so that the leading contenders are (1) and (3). I can see someone making a case for (1). But my own intuition is that the best thing to say is (3): if the facts are as described, then J.T. Leroy does not exist. I'm interested to hear others' judgments, though.
The naturalness of this judgment gets some support from other famous literary hoaxes, such as the Ern Malley case in Australia in the 1940's. Here, the poems attributed to Ern Malley (who like J.T. Leroy was put forward as having a specific life story) were in fact written by James McCauley and Harold Stewart, in an attempt to debunk modernist poetry. Since the facts were revealed, the standard description of the case usually involves the claim that Ern Malley did not exist. Even the official Ern Malley website characterizes him as "the poet who never lived". And as distinguished a source as David Lewis tells us that "Ern Malley was a nonexistent object".
If (3) is correct, and J.T. Leroy does not exist, then 'J.T. Leroy' does not refer. Likewise for 'Ern Malley'. This has consequences for the causal theory of reference. The dominant causal source for 'J.T. Leroy' is clearly Laura Albert, I think. Savannah Knoop complicates things, but it appears that she only entered the picture late in the day, around 2001. So one might think that if a simple causal theory of reference were correct, the name would refer to Laura Albert, and we could correctly say that Leroy is Albert. The same goes, mutatis mutandis, for Ern Malley. (Again, this case is complicted by the two authors, but it seems likely that people would still say that Ern Malley did not exist even if there were just one author.) So our intuitive judgments here tend to cast doubt on a simple causal theory of reference.
Of course there are more subtle versions of the causal theory of reference, which allow speaker's intentions to play a central role, although such theories have not been developed in much detail. But at the least, cases like this can help show the form that such a theory would have to take. The obvous suggestion here is that there are also descriptive as well as causal constraints on reference: for example, it might be required that for someone to qualify as the referent of 'J.T. Leroy', they have to fit Leroy's purported life-story well enough, and no-one does. Of course this is tricky: no-one says that James Frey did not exist, because his purported life-story was greatly exaggerated. And even with the largely fabricated life-story of Helen Demidenko, people are inclined to say that Helen Demidenko is Helen Darville. If the Leroy case were more like these cases, then presumably we would say that J.T. Leroy exists and wrote the novels, but that Leroy did not do most of the things that he/she claims to have done.
Here it is interesting to examine what sort of modifications to the Leroy case lead to a difference here, and to see what are the relevant differences between the Leroy and Malley cases on the one hand, and the Demidenko and Frey cases on the other. The name of the actual author seems to matter: If one modifies the Leroy case so that Albert's name had in fact been 'J.T. Leroy' all along, while leaving other facts the same, then I think my judgment would incline toward the view that Leroy exists. But the Demidenko case suggests that this is not all that is going on. Some similarity in actual and purported characteristics seem to matter: both Demidenko and Frey at least share an age and a gender with the characters they describe. In the Leroy case, if one modifies the case so that the Albert figure is a 25-year old male, I think I would incline much more in the direction of (1). So there seems to be a strong sensitivity of our judgments to certain descriptive information, but it is far from obvious what the relevant information is, or why it should be relevant. I'd say that this is a good case for experimental philosophers!
Of course one might just throw up one's hands and say (4), or perhaps decline to answer altogether. There is certainly some appeal to the suggestion that the whole issue as to who counts as "J.T. Leroy" is terminological. On the other hand, it is possible that the issue matters at least for some legal issues. For example, if Leroy's name appears in contracts, then the issue could matter a great deal -- though I suspect that Albert has designed any contracts very carefully! And, like other terminological issues, the question at least has consequences for the philosophy of terminology -- here, in its embodiment as the theory of reference. So it would be nice to get clearer on just what's going on here.