A Recalcitrant Problem for Abstract Creationism

Abstract

Abstract creationism about fictional characters is a popular view. But its popularity is undeserved. In Brock (), I explained why, outlining my case against all forms of creationism about fictional characters. The argument was never intended to give a decisive refutation of creationism—like many others, I do not think it is possible to give knock‐down arguments in philosophy—but it was meant to outline what I see as a serious and underappreciated challenge to the view. In a recent issue of this journal, David Friedell () makes an admirable attempt to answer that challenge. But his answer is inadequate, and my aim in this note is to explain why. In Section I, I briefly outline, in very general terms, my objection to creationism. In Sections II–IV, I articulate what I take to be three different but related responses to the objection advocated by Friedell and explain why I think each response ultimately fails. Creationism about fictional objects is the view that there are (merely) fictional objects, brought into existence by the authors of the works in which their names (or designating descriptions) first appear. Abstract creationism is a variety of creationism according to which (merely) fictional objects are abstract entities. In order for abstract creationism to count as an adequate theory, it should, at a minimum, give us the resources to answer the following very general questions: Ontology. Are there any fictional objects? Identity. Under what conditions is a fictional object x identical to a fictional object y? Plenitude. How abundant is the domain of fictional objects? How many fictional objects are there? In this respect, abstract creationism is held to the same standard as any other theory of fictional objects. On the basis of a theory's answers to these general questions, the theory will help us answer questions about particular cases. We should, for example, be able to apply the theory to answer questions such as: Are there more than 753 fictional objects? Is Sherlock Holmes one of them? How does Sherlock Holmes differ from actual detectives? Is Sherlock Holmes identical to James Bond? Meinongians and fictionalists give clear answers to each of the general questions posed above. So, for example, naïve Meinongians accept a comprehension principle according to which, for any combination of properties, there is a unique fictional object made up of (or associated with) that combination of properties. Consequently, on this view, there are fictional objects, there is an infinite number of them, and fictional object x is identical to fictional object y just in case they share all of the same properties. Fictionalists, on the other hand, tell us there are few fictional objects. In fact, there are none (we only make‐believe that there are). Consequently, it is trivially true that every fictional object is identical to every other fictional object. On the basis of these answers, fictionalists and Meinongians can answer questions about particular cases. Sometimes those answers are counterintuitive. Meinongians, for example, have suggested that Holmes is an incomplete and nonexistent object who existed before Conan Doyle was born; fictionalists have suggested that Holmes is trivially identical to James Bond (though we do not make‐believe that “he” is). Abstract creationists do not tend to say things that conflict with commonsense opinion in the same way that Meinongians and fictionalists do. Consequently, abstract creationism can seem preferable to alternatives such as Meinongeanism and fictionalism. But in Brock (), I argued that this appearance is illusory. The reason is that abstract creationists do not tend to say enough about fictional objects to answer the important general questions about fictional objects. Consequently, creationists are also not in a position to answer salient questions about particular cases at all, let alone in a way that coheres or complements our commonsense opinions about such cases. Creationists avoid the problem of counterintuitive commitments by dodging making solid commitments at all. They should answer the very general questions, but they choose instead not to. My claim in Brock () was that leaving it a mystery how such basic questions might be answered is not a virtue of the view; to the contrary, it disqualifies the view from consideration in the list of theories to be compared and evaluated. My claim was not that abstract creationists fail to answer any of the general questions above. To be sure, they give a clear answer to the ontological question. According to abstract creationists, there are indeed fictional objects. The trouble comes when they attempt to answer any of the other questions. Sometimes they grapple with questions about the identity of fictional objects, but as Anthony Everett () has demonstrated, they have not given us the resources to answer such questions adequately. In Brock (), though, my focus was squarely on questions about the plenitude or abundance of fictional objects.1 My claim was that abstract creationists have universally refrained from attempting to provide a substantial answer to this question. This is surprising. Given that creationists claim the number of fictional objects is a contingent matter, questions about how abundant the realm of fictional objects is at a time and in a world should be more pressing and more salient for them than it is for adherents of alternative views. In order to answer this sort of question, creationists must give us a way of determining when a fictional object comes into (and goes out of) existence. But creationists either do not attempt to do this or alternatively merely gesture at a woefully underspecified method for doing so. Here is how I put the point in my original paper: In order to learn when a fictional individual is created, we must first learn something about how it is created. One explains how a character is created in the relevant sense either by describing the act of creation in terms already well understood or by specifying the act's proximate cause. With respect to how they answer this how question, creationists come in two varieties. First, the majority simply refrain from addressing the question at all. These creationists keep silent on the issue that seems to me most crucial in understanding and assessing their position. Because they refrain from answering the question, we are left with nothing but the barest sketch of a theory that really can't be evaluated until the details are filled in. Second, there are those who echo John Searle (1974–75; 1979, 73), who famously answers the question as follows: “By pretending to refer to people and to recount events about them, the author creates fictional characters and events.” (, 356) Searle's answer to these questions is insufficient. Searle could mean many different things here, depending on what counts as an act of pretend reference and which acts of pretend reference involve the creation of a fictional individual. In Brock () I considered the following three interpretations of Searle's thesis: Any Use Account (AUA). A fictional object is created whenever an author uses a fictional name when writing (or telling) her story. First Use Account (FUA). A fictional object is created whenever an author uses a fictional name for the first time. Intended Creation by Pretense View (ICP). A fictional object is created whenever (i) an author intends to create a new fictional character and (ii) as a causal consequence of that intention, she pretends to refer to it. Each of these interpretations comes with counterintuitive consequences. But what those consequences are varies between the views. In Brock (), I explained in detail the specific problems associated with each interpretation. My objection to creationists thus draws inspiration from David Lewis's famous objection to ersatz (or abstract) modal realism. Lewis complained that ersatz realism—that is, abstract realism about possible worlds and individuals—seems more plausible than it should because the view conflates three importantly different theories, with different problems and different virtues.2 Once the views are distinguished carefully and the problems associated with each are made salient, the cluster of views becomes less plausible. Like Lewis's complaint against ersatzers, then, I maintain that creationists “keep silent, or explicitly decline to commit themselves, on the questions that seem … crucial in dividing the versions. … Since the different versions meet with different objections, that makes it hard to pin any objection at all on an author who keeps neutral. His position may seem to have the advantages of all the versions and the drawbacks of none!” (Lewis , 141, my emphasis). There are only three ways one could respond to this challenge. Direct Response. One might give a defense of one of the answers canvassed in Brock (): AUA, FUA, or ICP. The defense might concede that there is something counterintuitive about the view defended but claim that on balance the view is preferable to others, or alternatively it might explain why there is nothing counterintuitive in the answer at all. Indirect Response. One might refuse—despite my challenge to do so—to answer questions about the abundance of fictional objects and specific questions about when and how they are created. Such a defense should be accompanied by either an explanation as to why these questions are unimportant or, alternatively, an explanation as to why the prospects for answering them are good. Ambitious Response. Finally, one might respond by presenting an alternative answer to the general questions, an answer that differs in substance from AUA, FUA, and ICP. This answer might involve giving an alternative interpretation of Searle's thesis, or it might involve giving an entirely different account altogether. Friedell does not consider an ambitious response. Friedell initially considers the prospects of a direct response (see Sections III–V). But in the final main section of the paper (Section VI), he opts instead for an indirect response. Both of Friedell's responses fail, and it is the aim of this note to explain why. Let us consider first Friedell's (attempted) direct response. It is the one that he spends most of the paper developing. Yet it is also the response that most clearly fails, even by his own admission. Friedell has no sympathy for AUA and FUA. He concedes that the counterexamples in Brock () succeed to refute these views. Friedell is, however, more sympathetic to ICP (Friedell , 130). In Brock (), I presented what I take to be three counterexamples to ICP. The majority of Friedell's article is devoted to a consideration of the first two counterexamples, in an attempt to show that they are not genuine counterexamples at all. Although I am not convinced by Friedell's response to the cases, let me focus on what he says about the third counterexample, because it is the one he has most trouble with. Imagine that J. K. Rowling is not only a famous author but also an amateur philosopher. In particular, we might suppose that she has nominalistic tendencies and believes that abstract entities of any kind do not exist. As a consequence, she does not believe that by telling her stories and engaging in any acts of referential pretense, she will thereby create a fictional character. We might even suppose further that if she did have such a belief, she would not have written the stories because her disdain for abstract objects in general is so great she would never do anything to bring about their proliferation. Realists about fictional characters should agree, first, that Rowling's attitude is possible (others have certainly shared it) and, second, that even if she were to hold such a view, the existence of the fictional characters she has written about—Harry Potter, Hermione Granger, Ron Weasley, and others—is under no threat at all. ICP, however, is inconsistent with this intuition (Brock , 362). What does Friedell say in response? He concedes the force of the thought experiment. He suggests that under such circumstances we would say that Rowling accidentally created fictional characters. Consequently, he says the counterexample “is genuine—it refutes ICP” (, 130, see also 134, 135). So, in the final analysis, Friedell concedes that AUA, FUA, and ICP all fail to provide an answer to the challenge in Brock (). Friedell's main response to the argument in Brock () is an indirect response. Rather than give precise answers to questions about when fictional objects come into existence, how they come into existence, or how abundant the realm of fictional objects is, he draws an analogy with carpentry that he hopes will demonstrate that the prospects of answering these questions is good. Outlining a thought experiment of his own, he says: Suppose Peter van Inwagen has stayed faithful to his views about composition (van Inwagen 1990). He thinks there are no tables. There are merely simples arranged tablewise. His musings have sparked an interest in carpentry. Every weekend he carves wood in his garage. To any nonphilosophical observer it would appear he makes tables. But he is not intending to make tables. He is trying to arrange simples tablewise. Intuitively, van Inwagen still creates tables. Just as a nominalist storyteller may create fictional characters, a nihilist carpenter may create tables. (Friedell , 134) Friedell concludes, on the basis of this thought experiment, that one can make concrete artifacts such as tables and chairs without intending to make anything. On this point, I am entirely in agreement with Friedell. But Friedell also thinks that this fact somehow undermines the argument against abstract creationism in Brock (). He explains why in the following passage: Recall that Brock's argument relies on the claim that abstract creationism is more mysterious than the phenomena it is trying to explain, namely, our intuitions about certain sentences (for example, ‘Rowling made Harry Potter’ and ‘Harry Potter is a fictional character’). His argument should not persuade us to reject abstract creationism if the mystery surrounding fictional characters applies also to tables. Brock is trying to show that fictional characters are uniquely mysterious, or at least that they are more mysterious than mundane concrete artifacts. He has not succeeded. (Friedell , 134) But this objection is misguided. I was never trying to show that fictional characters are uniquely mysterious or that their existence is more mysterious than mundane concrete artifacts. What I was trying to show was that Creationism about Fictional Objects is a mysterious view in the sense that it gives rise to more questions than it answers. Friedell's thought experiment, therefore, cannot be taken as a tu quoque argument. Friedell might have had a point if Meinongians, fictionalists, and others also refused to answer principled questions about the identity and number of fictional objects. But they do not! As we saw in Section I, Meinongians and fictionalists give clear and straightforward answers to these questions (even if they are counterintuitive). Moreover, the analogy I raised with carpenters and their acts of creation was not intended to imply that tables and chairs are essentially created artifacts. To be clear, I do not think they are. Instead, I am tempted by the view that the property of being‐a‐table is a functional kind. Something is a table, on this view, just in case it has a flat surface, which serves the purpose of resting items used for eating, writing, working, and so forth. It is consistent with this view that tables come into existence in a variety of different ways, some of which involve no agency or intentions. On this view, whether or not a table exists is something that metaphysically depends on the ways it is used (or would be used). This is not an answer a creationist about fictional characters can give. So, if fictional characters can come into existence without being the direct product of an agent's intention to create them, then the creationist owes us some account of what grounds their existence. On what does their existence depend? The analogy with carpenters and tables is of no help here. In the first paragraph of the final section of his paper, Friedell says: I claimed the number of characters in a fiction depends in some way on authorial intentions. I [also] conceded that … authors can accidentally create characters. There is some tension between these claims. I will resolve the tension by saying more about how the number of characters depends on intentions. (2016, 134, my emphasis) The passage not only promises to resolve this tension, it also promises to say something helpful to address questions about the plenitude of fictional objects. Generalizing a suggestion made by Amie Thomasson, Friedell defends the following principle: Nothing New. Generally, when a fiction's author intends for a property to be ascribed to a particular pre‐existent character—whether that character is originally from that work or another—the author represents that character without creating a new one. (Friedell , 135) The remainder of Friedell's paper is a defense of Nothing New. Whatever its merits, though, Nothing New is not a principle that can be used as the basis for a response (ambitious or otherwise) to the argument in Brock (), for two reasons. First, Nothing New is not a Principle of Plenitude. It does not help us establish how abundant the realm of fictionalia is. At best, it can be used to ground answers to questions about the identity rather than the plenitude of fictional objects. Second, and more importantly, Nothing New is false if it is understood as a metaphysical principle grounding identity claims. If it were such a principle, it would have to be exceptionless—universally true in all possible worlds. Zsófia Zvolenszky () argues persuasively that fictional objects (and other abstract artifacts) are often created inadvertently. She illustrates the idea by asking the reader to consider a number of thought experiments, the first of which is the following: While writing War and Peace, Tolstoy was under the mistaken impression that the protagonist, Prince Bolkonsky, like Napoleon (also featured in the novel), was a real person. Introducing the name “Andrei Bolkonsky,” Tolstoy intended to refer to a historical figure he thought existed quite independently of his novel. … Quite independently of what these theorists say about mythical objects, it is overwhelmingly plausible to think that in Scenario T, Tolstoy created Bolkonsky as an abstract artifact, and did so inadvertently. And the reason why he did so is because of the non‐cooperation of the world to provide the relevant entity. Further, as a result of Tolstoy's writing the novel, the range of actual fictional characters plausibly came to include Bolkonsky also. (319) Of course, the same point would apply, mutatis mutandis, if Tolstoy, while writing Anna Karenina, was under the mistaken impression that Anna was a character in Middlemarch, created by George Elliot just a few Let us to fact, that when Tolstoy of the name he intended to refer to a fictional character. I will to say that Tolstoy also created but did so inadvertently. If that is Nothing New has Friedell that there might be to Nothing New. with cases like the in he New is just a of Sometimes an author makes a character when intending for a property to be ascribed to a (, 135). But this makes clear that Friedell's principle is of no help to the (though it may be of to the consider what we say about questions about identity or when a I see to a I saw the I saw is identical to the I see when events of kind are with and tend to events of kind events of kind events of kind These Nothing give us for but they cannot ground a theory of identity or a theory of simply does not in of or should, clearly and the existence and identity conditions of the

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 35 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.