Wolfgang Schwarz

Amie Thomasson's Fiction and Metaphysics

Brian has made so many puzzling remarks about fictional characters being real but abstract that I've decided to read Amie Thomasson's Fiction and Metaphysics. Here is my little review.

Thomasson's theory, in a nutshell, is that the Sherlock Holmes stories are not really about the adventures of a detective who lives at 221B Baker Street, but rather about the adventures of a ghostly, invisible character who lives at no place in particular and never does anything at all. We don't find this written in the Sherlock Holmes stories because, according to Thomasson's theory, Arthur Conan Doyle simply doesn't tell the truth about Holmes. In fact the only thing he gets right is his name: That ghostly character he is telling wildly false stories about is really called "Sherlock Holmes".

In support of this theory, Thomasson distinguishes two kinds of statements we would, in appropriate circumstances, regard as true about Holmes:

1. Fictional Statements: Holmes is a detective; he lives at 221B Baker Street; he is human; etc.

2. Serious Statements: Holmes was invented by Arthur Conan Doyle; he doesn't really live at 221B Baker Street; he is what the Sherlock Holmes stories are about; etc.

Since some of these statements contradict each other, they can't all be literally true. Thomasson suggests that we take the Serious Statements at face value and reinterpret the Fictional Statements by prefixing "according to the story": According to the story, Sherlock Holmes lives at Baker Street; but in reality, he doesn't. According to the story, he is human; but in reality, he isn't. This is how we arrive at the ghostly characterisation of the real Sherlock Holmes mentioned above.

What's controversial about this is that we should take the Serious Statements at face value. Thus to a large extent the book consists of arguments why we should.

The main positive argument (to be found in chapter 7) is simply that there is no adequate and elegant paraphrase of all Serious Statements. Clearly, prefixing "according to the story" won't help in this case: It's not that according to the story, Conan Doyle invented Sherlock Holmes. Thomasson makes a fairly good case that any alternative paraphrases will be unpleasantly piecemeal and ad hoc.

There is also a positive argument in chapter 6, suggesting that reference to fictional entities is needed for a good theory of intentionality. This argument is however a little odd. Thomasson there discusses under which circumstances two thoughts are about the same fictional character. She notes that this is not simply a matter of internal features of the thoughts' content, nor is it simply a matter of suitable external relations to the litarary text. Realism about fictional entities is supposed to solve these questions: That two thoughts are about the same fictional character is due to the fact they are really about the same fictional character. -- Well, yes, but does this answer the question? Certainly being about Sherlock Holmes isn't a primitive property of thoughts, independent of their content and causal history. So by postulating fictional characters we can't avoid spelling out which features of content and causal history determine whether or not a thought is about Sherlock Holmes.

The remaining arguments are defensive. In chapters 4 and 5 Thomasson tries to defuse worries that fictional characters lack clear identity conditions, and that reference to fictional entities might be incompatible with the direct reference theory of proper names. I'm willing to grant her these points, not because her arguments are very convincing, but mainly because I neither believe we need clear identity conditions for all entities nor that incompatibility with the direct reference theory would be a problem, given that this theory is a hopeless project anyway.

More interestingly, in most of the remaining chapters (1, 3, 9 and 10) Thomasson argues that postulating fictional characters does not involve commitment to a new kind of ontological category because, she says, many ordinary things like litarary works, governments or even chairs more or less belong to the very same category, namely to the category of things whose existence depends in a certain way on mental and physical states. I don't think this argument works at all, but it deserves saying more about it, so I'll postpone this discussion until I've thought about it a little more.

The main problem I see for Thomasson's view is that the Serious Statements can't be sparated from Fictional Statements in the way she requires. Some examples:

• Arthur Conan Doyle invented a detective who is always lucky in his hazardous inductive inferences.
• The hero of the Sherlock Holmes stories consumes drugs that are illegal nowadays.
• Daniil Charms used to dress and behave like Sherlock Holmes, whom he admired much.
• The Sherlock Holmes stories are not about a ghostly, invisible character who lives at no place in particular and never does anything.

What are we to do about these statements? Thomasson's ghostly Holmes doesn't make inductive inferences, I guess, nor does he consume drugs or dress and behave. But we can't just prefix the "according to the story" operator, since this would only make the statements false. Maybe we have to put the operator somewhere inside the statements: "Arthur Conan Doyle invented a character who, according to the stories, is lucky in his hazardous inductive inferences". Except that the stories never mention that these inferences are hazardous this might work. Things get worse with the second example, because here the quantification into the fiction operator has to be taken even further: "The hero of the Sherlock Holmes stories is such that there are drugs which are illegal nowadays and according to the stories, the hero consumes these drugs". But what if careful reconstruction of what the Holmes stories say about these drugs show that they don't have the properties of any known real drug? Then we presumably have to replace the "drugs" in our paraphrase by "fictional drugs", and we're in trouble again because fictional drugs aren't illegal nowadays. Even if such paraphrases can be made to work, it seems clear that they'll turn out to be quite inelegant, piecemeal and ad hoc. So we haven't really gained anything by admitting Thomasson's fictional characters.

Another problem: Thomasson agrees that if you write the biography of a real person, you don't create your fictional character. After all, a biography of, say, Oliver Cromwell is definitely about Oliver Cromwell and not about some abstract, ghostly entity. But the difference between biography and fiction is a matter of degree. Take any biography of Oliver Cromwell and continuously alter it into a Sherlock Holmes story. If you think it all depends on the author's intentions, then continuously alter the author's intentions too. But you're wrong anyway: Even if Arthur Conan Doyle really and sincerely intended to tell the biography of Oliver Cromwell in his Sherlock Holmes stories, he didn't succeed. Now if Thomasson is right, there is a certain point in this series at which the story is no longer about Cromwell but about a ghostly thing. Note that Thomasson can hardly say that it is indeterminate exactly where this point lies: It's not that there independently are both Cromwell and the ghostly thing, and we have to decide about whom the story is. On Thomasson's view, the ghostly thing gets created by the fictional story. So if it is indeterminate whether or not the story is still about Cromwell, then it is indeterminate whether or not the ghostly character really exists. But can existence be indeterminate?