The Matter of Serial Fiction

The Matter of Serial Fiction[1]

1. Introduction

Recent work on the problem of truth in serial fiction has focused on the semantics of certain sentences used to talk about serial fictions, as in Ross Cameron’s (2012) ‘How to Be a Nominalist and a Fictional Realist’ and Andrew McGonigal’s (2013) ‘Truth, Relativism, and Serial Fiction’, or semantic properties of works themselves, as in Ben Caplan’s (2014) ‘Serial Fiction, Continued’. In this paper I argue that these proposed solutions are mistaken. But, more importantly, this general approach to the problem is mistaken: the problem of truth in serial fiction is an instance of the problem of change. In section 2, I introduce the problem of truth in serial fiction. In section 3, I consider a reply motivated by temporal eternalism, the view that future events can affect past or present content. In sections 4-6, I consider Cameron’s contextual view, the relativist view McGonigal presents, and Caplan’s work contextualism, respectively. Finally, in section 7, I present my preferred solution to the problem of truth in serial fiction. On the solution I prefer, fictions can undergo change, much like you and me in certain respects. As a result, what is true in or according to them changes as well.

2. The Problem of Truth in Serial Fiction

1977 saw the release of Star Wars: A New Hope.[2] Luke Skywalker was the protagonist and Darth Vader was the antagonist. And according to Hope, the two had a relationship, of sorts: Luke’s mentor, Obi-Wan Kenobi, tells Luke that Vader betrayed and murdered his father.

1980 saw the release of Star Wars: The Empire Strikes Back.[3] Luke was again the protagonist and Vader was again the antagonist. But according to Empire, the two had a surprising relationship: Luke’s nemesis, Darth Vader, reveals that he is Luke’s father.

In 1977 and even in early versions of the script for Empire, it seems that Star Wars creator George Lucas had no intention of having it turn out according to the story that Vader was Luke’s father. The plot twist was an addition to later versions of the script. And it became true, according to the Star Wars canon, that Vader is Luke’s father.

Let’s suppose further, as is plausible, that in 1977 Lucas didn’t merely lack a definite intention as to whether Vader is Luke’s father. Rather, let’s suppose that in 1977 Lucas definitely intended for Vader not to be Luke’s father. According to final drafts of the script, Luke was an only child whose father was dead. And in the movie, Obi-Wan tells Luke that Vader betrayed and murdered his father. So when we confine our attention to Hope, it seems that it is true according to Hope that Vader is not Luke’s father. So it’s false, according to Hope, that Vader is Luke’s father. Similarly, if Lucas intended that according to the story Luke was not to be raised on Yavin IV, then it’s false according to Hope that Luke was raised on Yavin IV.

Hope and Empire are both parts of the Star Wars canon. And the Star Wars canon seems to involve a consistent story. But it seems one of its parts, Hope, is inconsistent with the canon. That’s strange.

We can present the problem a bit more carefully as follows:[4]

1. In 1977, according to Hope, it is false that Vader is Luke’s father.

2. According to Empire, it is true that Vader is Luke’s father.

3. Hope and Empire are parts of the Star Wars canon.

4. If (1-3), then, Hope is a part of the Star Wars canon that is inconsistent with other parts of the Star Wars canon.

5. So, Hope is a part of the Star Wars canon that is inconsistent with other parts of the Star Wars canon.

6. The Star Wars canon is consistent.

The argument is valid. Premises (1) and (2) are supported by a plausible claim about the relationship between an author’s intentions and truths in a fiction created by that author. Though it is notoriously difficult to spell out this relationship precisely, it’s plausible that on any promising account of the relationship between authorial intent and meaning (hence truth) in fiction, Lucas had what it took to make it true according to Hope in 1977 that Vader was not Luke’s father.[5] Premise (3) is plausible because Star Wars appears to be an extended fiction comprised of a number of parts. Hope and Empire are exemplary candidates for being among those parts.[6] Premise (4) is plausible since it is true according to part of the Star Wars canon that Vader is Luke’s father. But by (1) and (3), it is true according to Hope that Vader is not Luke’s father and Hope is part of the canon. Premise (6) is plausible since the Star Wars story in its entirety does not seem to entail any proposition along with its negation. Suppose you were to relate the Star Wars story to the unenlightened. Doing so successfully would not involve contradicting yourself. This is in contrast to, say, relaying the Sherlock Holmes canon, according to which John Watson’s only war wound is in his leg and Watson’s only war wound is in shoulder, or Graham Priest’s (1997) “Sylvan’s Box”, according to which there is a box that is completely empty yet with something inside of it.

From these premises we can conclude that a consistent whole has a part that is inconsistent with it. The conclusion is not quite contradictory, but it is strange. A satisfactory account of truth in serial fiction must resist this strange conclusion by rejecting a premise or embrace it and explain away its apparent strangeness. Since the argument is valid, it is unsound only if it has a false premise. So let’s turn to the premises.

3. Against (1): The Temporal Eternalist Reply

According to temporal eternalism, future events can affect past or present content.[7] If temporal eternalism is true, then in 1977 it was true that there were future events involving Empire in 1980. And if the existence of Empire is sufficient to make it true in the canon that Vader is Luke’s father, perhaps in 1977 the existence of Lucas’s later intention in 1980 makes it true according to Hope that Vader is Luke’s father. Maybe it would have been inappropriate in 1977 to say that according to the relevant fiction, Vader is Luke’s father. But that would be due to a lack of evidence rather than a lack of a fact of the matter.

3.1 Against the Temporal Eternalist Reply

On this view, the content of a work at any particular time can depend on what hasn’t happened yet. So it’s more or less impossible to know, prior to a work’s completion, what its content is at any particular time. So no one—including Lucas—knew what the Star Wars story was so far as of 1977. And even now no one knows what the Star Wars story is so far since it is ongoing. Of course, we can’t know everything about a work until it is complete. But it at least seems that some people—like Lucas—are in a position to know what the Star Wars story is so far.[8] But even that is precluded if future facts retroactively determine past content. That doesn’t seem impossible, but it does seem implausible. It would be nice if we could avoid this consequence.

Another, related, consequence we might like to avoid is an objectionable sort of backward causation: what Lucas intends in 1980 affects what is true in Hope in 1977. But we shouldn’t suppose that backward causation ever occurs, perhaps apart from very special circumstances like those involving time travel. And, presumably, Lucas’s 1980 intentions did not time travel to 1977.[9]^{, [10]}

4. Against (3) or (4): The Contextualist Reply

Ross Cameron (2012) has endorsed a contextualist solution to the problem of truth in serial fiction. On his view, roughly, ‘Luke’ refers to a character that does not have Vader as a father in some contexts, but in other contexts ‘Luke’ refers to a different character that does have Vader as a father.

Less roughly, in some contexts, ‘Luke’ refers to Luke- and ‘Vader’ refers to Vader-, and Luke- and Vader- are both characters in the fiction Hope-. And ‘Vader is Luke’s father’ is not true according to Hope-. According to that fiction, it is not the case that Vader- stands in the is a father of relation to Luke-.[11]

However, in other contexts, ‘Luke’ refers to Luke+ and ‘Vader’ refers to Vader+. And Luke+ and Vader+ are both characters in the fiction Hope+. And ‘Vader is Luke’s father’ is true according to Hope+. According to that fiction, it is the case that Vader+ stands in the is a father of relation to Luke+.

Crucially, while Hope- and Hope+ are visual and sonic duplicates of each other, Hope+ is part of the Star Wars canon while Hope- is not. So utterances of ‘according to the relevant fiction, Vader is Luke’s father’ are not true in 1977, where the singular terms refer to Hope-, Vader-, and Luke-, respectively. In these contexts, an utterance of ‘Hope and Empire are parts of the Star Wars canon’ is false. So in these contexts, Cameron would reject (3).

Accordingly, utterances of ‘according to the relevant fiction, Vader is Luke’s father’ are true in 1980, since they are about the “+” counterparts. In these contexts, an utterance of (3) is true while an utterance of (4) is false.[12]

4.1 Against Contextualism

As McGonigal (2013) and Caplan (2014) point out, it seems as if someone who watches Hope in 1977 and someone who watches it now watch the same movie. But if Cameron’s view is right, the first watches Hope- while the second watches a distinct movie, Hope+. It’s not the case that both watch part of the same series, parts of the Star Wars canon. But that seems incorrect.

Cameron’s view is also susceptible to familiar problems that plague other contextualist views. Being surprised by Empire, I might say ‘I used to think Vader was not Luke’s father, but it turns out that’s false.’ And it seems what I say is true; it’s hard to see how else we could make sense of the surprise that the plot twist elicited. But on Cameron’s view, ‘Vader’ and ‘Luke’ do not refer to particular characters; Vader- is the referent in some contexts, while Vader+ is the referent in others. And there is no single proposition encoded by ‘Vader was not Luke’s father’. In a context in which my utterance of ‘Vader was not Luke’s father’ expresses the “-” proposition, then on Cameron’s view my utterance is false since my utterance of ‘that’ does not refer to something that turned out false. In a context in which my utterance of ‘Vader was not Luke’s father’ expresses the “+” proposition, then on Cameron’s view my utterance is false since it is false that that is a proposition I used to think is true. So if Cameron’s view is true, my claim that seems to repudiate what I used to think is false. As a corollary, we are left with no account of why Empire’s revelation is surprising.[13]

Additionally, we could correctly describe those who only saw Hope in 1977 along with those who have seen Empire as knowing who Vader is, being familiar with that character, or agreeing that Vader is Sith. But this description is false on Cameron’s view, since there is no one Vader that they all know or are familiar with or agree is Sith.[14]

It would be good to avoid these consequences if we could. What seems right to me about this reply to the problem of truth in serial fiction is that it identifies (4) as the culprit. (Well, at least in certain contexts.) So the remainder of the paper considers alternative ways of resisting premise (4).

5. Against (4): Relativism

Andrew McGonigal (2013) has advanced a relativist solution to the problem of truth in serial fiction.[15] On the view he describes, ‘According to the relevant fiction, Vader is Luke’s father’ expresses a single proposition in every context. But that proposition can be evaluated differently with respect to different circumstances. According to a standard, non-relativist semantic account, circumstances include possible worlds.[16] On the relativist account McGonigal sets out, circumstances also include sets of salient fictional episodes. In circumstances in which only Hope is salient, the proposition is false. In circumstances in which Hope and Empire are salient, the proposition is true. The proposition expressed by ‘According to the relevant fiction, Vader is not Luke’s father’ is true when ‘Vader is not Luke’s father’ is evaluated according to Hope, but it is not inconsistent with the Star Wars canon. It’s just false in some circumstances and true in others, much as the proposition that Lucas is a director is true in actual circumstances but false in some counterfactual circumstances.

Relativists of this sort reject (4) since, for semantic purposes, Hope, Empire, and the Star Wars canon are treated as “points” of evaluation which, when paired with other features of contexts, yield differing truth values for one and the same proposition. The proposition expressed by ‘According to the relevant fiction, Vader is Luke’s father’ is just false in some circumstances and true in others. It has no “settled” truth value apart from worlds and salient fictional episodes.

5.1 Against Relativism

A more metaphysically serious view of fictions takes them not just to be “points” that are relevant to semantic evaluation. Fictions include propositions, perhaps by having them as parts or as parts or constituents of their contents. And, plausibly, those propositions that are included in a fiction are those that are true according to that fiction. But it seems Hope includes the proposition that Vader is not Luke’s father and Empire includes its negation. So on a more metaphysically serious view of fictions, Hope does seem to contradict the Star Wars canon while being a part of it. So it appears relativism at best offers a solution to the problem of truth in serial fiction that crucially depends on abstracting away from a plausible account of the metaphysical nature of fictions.

The attempt to bring analytic semantic relativism to bear on the problem of truth in serial fiction is surprising in another way. Analytic semantic relativism is motivated in part by an attempt to make sense of genuine disagreement. So it would be natural for a relativist account of truth in serial fiction to maintain that Hope and the Star Wars canon do genuinely disagree. On a relativist account of predicates of personal taste, for example, the proposition expressed by ‘Star Wars is not awesome’ is the negation of the proposition expressed by ‘Star Wars is awesome’ (as opposed to a contextualist account on which ‘is awesome’ expresses different properties in different contexts). If we add standards of personal taste to circumstances of evaluation, then the proposition that Star Wars is awesome is true when evaluated in my circumstance but false when evaluated in the philistine’s circumstance. On such an account, ‘What he said is true iff what I said is false’ comes out true for each of us. And thus even on the relativist’s view it seems there is genuine disagreement between us.

Similarly, on a relativist account of the problem of truth in serial fiction, the proposition expressed by ‘Vader is not Luke’s father’ is the negation of the proposition expressed by ‘Vader is Luke’s father’. If we add salient fictional episodes to circumstances of evaluation, then the proposition expressed by ‘according to the relevant fiction, Vader is not Luke’s father’ is true when evaluated in a circumstance in which Hope is salient but false when evaluated in a circumstance in which Empire is salient. On such an account, if fictions in some sense say things, then ‘What Hope says is true iff what Empire says is false’ comes out true. And thus it seems there is genuine disagreement between Hope and the Star Wars canon. So it would seem, even by the relativist’s own lights, that Hope contradicts the Star Wars canon in the sense that there is genuine disagreement between the two. And this leads to a close variant of the original puzzle to which the relativist lacks the resources to reply.

Relativist semantics have further problems that extend to a relativist account of truth in serial fiction.[17] On such a semantics, typically, the proposition that x knows P is true when the proposition that x knows P is true for x. But then if Hope is salient for Hamill while Empire is salient for Lucas, then P1 is true in Hamill’s circumstance:

P1 Lucas knows that according to the relevant fiction, Vader is Luke’s father.

But then since knowledge is factive, it seems to follow that P2 is also true in Hamill’s circumstance:

P2 According to the relevant fiction, Vader is Luke’s father

But since Hamill’s circumstance is one in which Hope is salient, that proposition is supposed to be false, according to the relativist. So it’s difficult to see how relativism helps here.[18]

It would be good to avoid these consequences if we could. So far we have considered replies to (4) that are motivated mainly by linguistic considerations. In the last two sections, we will turn to replies to (4) that are motivated by more metaphysical considerations.

6. Against (4): Work Contextualism

Ben Caplan (2014) prefers a different solution to the problem of truth in serial fiction—one that is more “metaphysical”. Caplan endorses what he calls “work contextualism”. On his view, serial fictions have contents, in much the same way sentences do. That is, they express contents relative to contexts. To grossly oversimplify, just as a context-sensitive sentence expresses different contents relative to different contexts, serial fictions express different contents relative to different contexts. In 1977, the Star Wars canon expressed some propositions that included P3:

P3 Vader is not Luke’s father

But in 1980, the Star Wars canon expressed some propositions that included P4:

P4 Vader is Luke’s father

As a result, Hope does not contradict the Star Wars canon. It used to contradict what would be the Star Wars canon, but there is no time at which the Star Wars canon has as a part of its content any propositions that contradict propositions that are the content of Hope. So, according to the work contextualist view, (4) is false.

6.1 Against Work Contextualism

Work contextualism is an improvement over contextualism and relativism. Since it does not propose an explicitly linguistic solution to the problem, it does not inherit the attendant problems of the broader semantic contextualist and relativist research programs. But as Caplan (2014) notes, on his view works themselves behave in important ways like context-sensitive expressions. But there seems to be an important disanalogy between context-sensitive expressions and serial fictions.

Context-sensitive expressions are associated with Kaplanian characters. These can be thought of as functions from contexts to contents, and are associated with a sort of rule that “tells” us which content gets associated in which context.[19] For example, in a context in which I am the speaker, the content of ‘I’ is me, whereas in a context in which you are the speaker, the content of ‘I’ is you. The associated rule for ‘I’ tells us to assign the speaker of a given context as the content of ‘I’ in that context. But there does not seem to be a correspondingly simple character that assigns propositions to the Star Wars canon relative to contexts. This problem becomes especially difficult if we try to factor in forms of retconning, such as in Return of the Jedi where Obi-Wan says that his remarks in Hope were true “from a certain point of view”. It seems that, given sufficient artistic license, almost any serial fiction could have practically any propositions as its content at any given time.

Caplan (2014) discusses this problem.[20] He suggests that the following rule might be analogous to “‘I’ refers to the speaker”: ‘Hope refers to the set of propositions that are true in it’. But Caplan (2014) acknowledges that this rule might ultimately not be entirely satisfactory.

One explanation of this apparent fact may be that serial fictions aren’t really analogous to context-sensitive expressions—they are not themselves context-sensitive entities. But if that’s so, work contextualism does not provide a satisfactory solution to the problem of truth in serial fiction.

7. Changing Your Story

On my view, when considering the nature of serial fiction, we should move even farther away from linguistic solutions. Let’s start by considering a case that has nothing to do with serial fiction: the Height Problem. In 1947, Harrison Ford was less than 6 feet tall. In 2015, Ford is over 6 feet tall. Call ‘Ford’s Life’ the collection of true propositions about Ford. Call ‘1947-Ford’ the collection of true propositions about Ford that are restricted to what’s going on in 1947. Now we can present a problem that closely parallels the problem of truth in serial fiction:

7. 1947-Ford includes the proposition that Ford is less than 6 feet tall.

8. 2015-Ford includes the proposition that Ford is not less than 6 feet tall.

9. 1947-Ford and 2015-Ford are parts of Ford’s Life.

10. If (7-9), then 1947-Ford is a part of Ford’s Life that is inconsistent with Ford’s Life.

11. So, 1947-Ford is a part of Ford’s Life that is inconsistent with Ford’s Life.[21]

12. Ford’s Life is consistent.

The argument is valid. Premises (7) and (8) express matters of fact. Premise (9) is true because Ford’s Life is a collection of true propositions that begins with true propositions concerning Ford in 1942 (1942-Ford) and includes true propositions concerning Ford in 2015 (2015-Ford). Premise (10) seems true since, given (7-9), the claim that Ford is less than 6 feet tall is true according to 1947-Ford. But its negation is true according to Ford’s Life. Finally, premise (12) is true since it does not appear that any proposition and its negation is true according to Ford’s Life. The claim that Ford is less than 6 feet tall is false in Ford’s Life, given that 2015-Ford is part of Ford’s Life that in some sense supersedes 1947-Ford, and it is false according to 2015-Ford that Ford is less than 6 feet tall. So it is false according to Ford’s Life that Ford is less than 6 feet tall. So it is true according to Ford’s Life that Ford is not less than 6 feet tall. So, it seems that Ford’s Life does not contradict itself.

The conclusion is not contradictory, but it does seem to be strange. It seems a satisfactory account of the Height Problem must resist this strange conclusion by rejecting a premise or embrace it and explain away its apparent strangeness. Since the argument is valid, it is unsound only if it has a false premise.

The Height Problem parallels the problem of truth in serial fiction. But it also parallels a familiar problem of change over time. (In the Height Problem, propositions are really just acting as surrogates for changing properties of Ford over time. Any problem of change can be similarly re-cast as a problem about certain truths.) My proposal is to treat the parallel problems—truth in serial fiction and change over time—in a parallel fashion: serial fictions are things that change over time.[22]

According to the solution to the Height Problem I prefer, Ford is distinct from his “matter”.[23] Plausibly, Ford is the sort of thing that has cells as matter. And Ford changes over time in part by having different matter at different times.[24] Ford’s having considerably less matter in 1947 explains why it was true then that he was less than 6 feet tall. His having considerably more matter in 2015 partly explains why he is now over 6 feet tall. So (10) is false—that 1947-Ford was less than 6 feet tall does not contradict Ford’s Life since the former fact concerns the matter he had in 1947, which, as it happens, is not the matter he has today.

Note that this does not mean in any sense that Ford is a sort of context-sensitive being “expressing” or somehow referring to different matter at different time. What’s going on is instead a matter of an ordinary object enduring ordinary change.[25]

In the case of serial fictions, we can offer a parallel solution. Serial fictions have “matter” as well, though they’re not the sorts of things that have cells as their matter. Rather, they have propositions as their matter. And what matter they have can change over time. In 1947, Ford’s matter determined that he was less than 6 feet tall, and in 1977 the Star Wars canon’s matter included the proposition that Vader was not Luke’s father. In 2015, Ford’s matter determined that he was not less than 6 feet tall. In 1980, the Star Wars canon’s matter included the proposition that Vader was Luke’s father. Recall the parallel premise to (10), premise (4), in the original argument:

4. If [(in 1977, according to Hope, it is false that Vader is Luke’s father), (according to Empire, it is true that Vader is Luke’s father), and (Hope and Empire are parts of the Star Wars canon)], then Hope is a part of the Star Wars canon that is inconsistent with other parts of the Star Wars canon.

(4) is false, but not because of the linguistic properties of utterances concerning the Star Wars canon. Nor because the canon itself is a context-sensitive entity. Rather, (4) is false because the problem, at bottom, is the familiar problem of change over time, though it appears here in an unfamiliar place. The Star Wars canon changed over time in such a way as to make (4) false in virtue of a part of that canon, Hope, changing over time. Much as Ford has the property being over 6 feet tall at one time but not another, Star Wars has the property of being a fiction in which Vader is not Luke’s father at one time but not another.[26] And much as this feature of Ford is due to his changing over time, those features of Star Wars are due to its changing over time.

Many philosophers are skeptical of the claim that so-called abstract objects can undergo change. But, as Gideon Rosen (2012) points out, how we should characterize abstract objects is controversial, and it is controversial whether there even is any deep distinction between the abstract and the concrete. Some, of course, characterize abstracta as those entities incapable of certain sorts of change. Insofar as we accept the abstract/concrete ideology, proponents of the proposed solution clearly cannot accept that account of abstracta. And while many philosophers are skeptical of the claim that abstract objects can undergo change, many outside philosophy, including computer scientists and practitioners of library and information science, do not seem to have similar qualms.[27] On the face of it, “repeatable” works like documents, certain musical works, and other social constructions, like checking accounts and marriage, all enjoy hallmarks of the abstract but seem to undergo change.[28] It would be a methodological mistake to rule out by fiat the possibility that apparently abstract objects can undergo change.

Though my main claim is that the problem of serial fiction is a special case of the problem of change, I believe my preferred solution provides a framework for understanding how an apparently abstract object could undergo change. On the hylomorphic picture I prefer, Harrison Ford is a compound of form and matter: his matter may consist of his cells while his form may be a structural arrangement of his cells. And Ford may change by changing his matter—coming to have different cells among his matter. Or perhaps Ford may change by changing his form—coming to have his cells arranged in a different way. Similarly, on the proposal I prefer, Star Wars is a compound of form and matter: its matter may consist of propositions while its form may be a structural arrangement of those propositions. And Star Wars may change by changing its matter—coming to have different propositions among its “matter”. (It’s important not to get too hung up on the term “matter” here. What’s important is that the term is a place-holder for an important element of a hylomorphic whole.) Or perhaps Star Wars may change by changing its form—coming to have certain propositions arranged in a different way. The differences between cells on the one hand and propositions on the other are not especially important. What is important to the account is how the cells and the propositions, respectively, figure into the hylomorphic wholes that are Ford and Star Wars, respectively. Clearly, there are a number of ways in which the proposal just sketched could be fleshed out. But as it stands I believe the sketched proposal at least allows conceptual room for a metaphysics of apparently abstract objects that change.

Finally, I believe my proposal has a distinct advantage over the others we have considered. The problem of change over time demands a solution, and whatever solution is offered can, it seems, be extended to the problem of truth in serial fiction.[29] We are not similarly beholden to a particular view in the philosophy of time, a particular semantic view, or even a particular view on which entities exhibit semantic behaviors. We are perhaps beholden to solving the problems these views are meant to address, but we are not beholden to those particular solutions. However, we are beholden to solving the problem of change over time, somehow or other. And the problem of truth in serial fiction is, at bottom, just an instance of this more general problem. Serial fictions (and their parts) are just things that change, more or less as you and I (and our parts) do. As a result, what’s true in or according to them changes as well. Apart from the largely irrelevant differences between the matter of you and me and the matter of serial fiction, there is no deep difference between the problems or their solution. Regardless of which solution to the problem of change you prefer, the problem of truth in serial fiction is an instance of the problem of change. And if so, whatever solves the latter also solves the former.[30]

Chris Tillman

University of Manitoba

chris.tillman@gmail.com

References

Almog, Joseph, John Perry, and Howard Wettstein. Eds. 1989. Themes from Kaplan. Oxford: Oxford University Press.

Braun, David. 1995. “What is Character?” Journal of Philosophical Logic 24: 227-40.

Braun, David. 2005. “Empty Names, Fictional Names, Mythical Names.” Noûs 39: 596-631.

Braun, David. 2015. “Wondering about Witches.” In Brock and Everett (eds) 2015, pp. 71-113.

Brock, Stuart and Anthony Everett. Eds. 2015. Fictional Objects. Oxford: Oxford University Press.

Cameron, Ross. 2012. “How to be a Nominalist and a Fictional Realist.” In Mag Uidhir (ed) 2012, pp. 179-96.

Caplan, Ben. 2014. “Serial Fiction, Continued.” British Journal of Aesthetics 54: 65-76.

Cappelen, Herman and John Hawthorne. 2009. Relativism and Monadic Truth. Oxford: Oxford University Press.

Cappelen, Herman and Ernest Lepore. 2005. Insensitive Semantics: A Defense of Semantic Minimalism and Speech Act Pluralism. Oxford: Wiley-Blackwell.

Carrol, Noël. 1992. “Art, Intention, and Conversation.” In Iseminger (ed) 1992, pp. 97-131.

Colaco, David, Wesley Buckwalter, Stephen Stich, and Edouard Machery. 2014. “Epistemic Intuitions in Fake-Barn Thought Experiments.” Episteme 11(2): 199-212.

Davies, Stephen. 2006. “Authors’ Intentions, Literary Interpretation, and Literary Value.” British Journal of Aesthetics 46: 223-47.

Gendler, Tamar Szabó and John Hawthorne. 2005. “The Real Guide to Fake Barns: A Catalogue of Gifts for Your Epistemic Enemies.” Philosophical Studies 124: 331-352.

Heller, Mark. 1990. The Ontology of Physical Objects: Four Dimensional Hunks of Matter. Cambridge: Cambridge University Press.

Huitfeldt, Claus, Fabio Vitali, and Silvio Peroni. 2012. “Documents as Timed Abstract Objects.” Proceedings of Basilage: The Markup Conference 2012.

Iseminger, Gary. Ed. 1992. Intention and Interpretation. Philadelphia: Temple University Press.

Jackman, Henry. 1999. “We Live Forwards but Understand Backwards: Linguistic Practices and Future Behavior.” Pacific Philosophical Quarterly 80: 157-77.

Kaplan, David. 1989. “Demonstratives: An Essay on the Semantics, Logic, Metaphysics, and Epistemology of Demonstratives and Other Indexicals.” In Almog, Perry, and Wettstein (eds) 1989, pp. 481-563.

Kaplan, David. 2011. “Words on Words.” Journal of Philosophy 108: 504-29.

Koslicki, Kathrin. 2008. The Structure of Objects. Oxford: Oxford University Press.

Levinson, Jerrold. 1980. “What a Musical Work Is.” Journal of Philosophy 77: 5-28.

Levinson, Jerrold. 1996. The Pleasures of Aesthetics: Philosophical Essays. Ithaca, NY: Cornell University Press.

Livingston, Paisley. 2005. Art and Intention: A Philosophical Study. Oxford: Clarendon Press.

Mag Uidhir, Christy. Ed. 2012. Art and Abstract Objects. Oxford: Oxford University Press.

McFee, Graham. 1980. “The Historicity of Art.” Journal of Art and Art Criticism 38: 307-24.

McFee, Graham. 1992. “The Historical Character of Art: A Reappraisal.” British Journal of Philosophy 32: 307-19.

McGonigal, Andrew. 2013. “Truth, Relativism, and Serial Fiction.” British Journal of Aesthetics 53: 165-79.

Mellor, D. H. 1998. Real Time II. London: Routledge.

Nolan, Daniel and Carrie Ichikawa Jenkins. 2008. “Backwards Explanation.” Philosophical Studies 108(1): 103-15.

Priest, Graham. 1997. “Sylvan’s Box: A Short Story and Ten Morals.” Notre Dame Journal of Formal Logic 38: 573-82. Reprinted in Priest 2005, pp. 125-33.

Priest, Graham. 2005. Towards Non-Being: The Logic and Metaphysics of Intentionality. Oxford: Oxford University Press.

Rohrbaugh, Guy. 2003. “Artworks as Historical Individuals.” European Journal of Philosophy 11: 117-205.

Rosen, Gideon. 2012. “Abstract Objects.” The Stanford Encyclopedia of Philosophy (Fall 2014 edition.) http://plato.stanford.edu/entries/abstract-objects/.

Sider, Theodore. 2001. Four Dimensionalism: An Ontology of Persistence and Time. Oxford: Oxford University Press.

Sosa, Ernest. 2007. A Virtue Epistemology: Apt Belief and Reflective Knowledge Volume I. New York: Oxford University Press.

Stoneham, Tom. 2003. “Temporal Eternalism.” Philosophical Papers 32(1): 97-107.

[1] Many thanks to two anonymous referees for this journal and to Ben Caplan, David Doerksen, Curtis Kehler, Kathrin Koslicki, Christy Mag Uidhir, Carl Matheson, Andrew McGonigal, Lewis Powell, David Sanson, Joshua Spencer, Amie Thomasson, Rebecca Wardell, commentators on Aesthetics for Birds, participants at the 2014 seminar on Current Research in Metaphysics at the University of Alberta, participants at the 2015 Auburn Ontology of Art Workshop, and the editorial staff at this journal.

[2] When it was released it was titled Star Wars, which was later changed to Star Wars Episode IV: A New Hope. To simplify things a bit, I’ll just pretend it had the longer title when it was released and use ‘Star Wars’ to refer to the Star Wars canon.

[3] When it was released it was titled The Empire Strikes Back, which was later changed to Star Wars Episode V: The Empire Strikes Back. To simplify things a bit, I’ll just pretend it had the longer title when it was released and use ‘Star Wars’ to refer to the Star Wars canon.

[4] This way of presenting the problem of truth in serial fiction follows Caplan (2014: 69).

[5] Thanks here to Ben Caplan for discussion. For further discussion of the relationship between authorial intent and meaning (hence truth) in fiction, see (e.g.) Carrol (1992), Levinson (1996), Livingston (2005), Davies (2006), and below.

[6] Even if you doubt that the relationship is strictly part to whole, Hope clearly bears some part-like relationship to the Star Wars canon.

[7] Caplan (2014) attributes this sort of reply to Travis Kendall and Shieva Kleinschmidt. Proponents of temporal eternalism include Levinson (1980), Jackman (1999), and Stoneham (2003). A related view in the metaphysics of time is eternalism, according to which purely past objects, such as dinosaurs, are just as real as purely future objects, such as Martian outposts, which are in turn just as real as present objects, such as you and me. Proponents of eternalism include Heller (1990), Mellor (1998), and Sider (2001). The claim that future facts affect past or present content is crucial to this reply, and it is not licensed by eternalism alone. Furthermore, eternalism may be dispensible if there can be future facts without eternalism being true. But what is indispensible to the reply is that there are future facts and they affect past content.

[8] Some might think even this is too much to ask for. Just as we may not know what a statue represents if we only see a proper part of it, we may not know what is true according to a serial fiction after having only been exposed to some but not all of it. I think the spatial case and the temporal case are disanalogous, however. It is plausible that, in 2014, we knew that, according to AMC’s The Walking Dead, Tyrese was alive. Now we know that, according to AMC’s The Walking Dead, Tyrese is not alive. To preview my own take on serial fiction, serial fictions change over time in a way that is fundamentally different from how a statue “changes” across space.

Another reply is due to an anonymous referee. Maybe retconning cases are skeptical scenarios, and Lucas is not in a position to know, but we think he is because we are ignoring the relevant skeptical scenario. This is an interesting suggestion. I doubt, however, that we think he is in a position to know what the story is so far because we are ignoring a skeptical scenario. Even when we explicitly attend to the possibility of retconning, our confidence that he is in a position to know what the story is so far remains. Even if it is possible for Star Wars to be retconned so that it turns out that, all along, Vader was a pineapple, it still seems that Lucas is now in a position to know that, according to the Star Wars story so far, Vader is not a pineapple.

A further suggestion due to the same anonymous referee goes as follows. Maybe retconning scenarios are like fake barn cases, and so are skeptical scenarios. So maybe Lucas can only know what the story is so far if he is not in fact in a retconning scenario. But a number of epistemologists (and ordinary folk) deny that fake barn scenarios really are skeptical scenarios. (See Gendler and John Hawthorne (2005), Sosa (2007), and Colaco, Buckwalter, Stich, and Machery (2014). And if they’re right, there is no special reason for thinking that the future facts concerning whether retconning will occur are sufficient to now rob Lucas of knowledge of what the story is so far.

[9] Thanks here to David Doerksen. It’s unobjectionable that sometimes what happens later can be responsible for the truth of an earlier claim. Predictions, for instance, are true only if what is predicted later turns out to be the case. And if that’s “backward causation”, it’s surely of an unobjectionable sort. (See Nolan and Jenkins (2008).) But what happens with Hope on the temporal eternalist view is more insidious: something that seemingly had what it took to be false according to Hope in 1977 is instead true because of what happened in 1980. By contrast, predictions don’t generally have what it takes to be false at the time they are made. Thanks to Lewis Powell for discussion.

[10] McGonigal (2013) raises some similar objections to this sort of reply, which he calls “Extreme Realism”.

[11] Cameron’s view can be seen as rejecting (1) since he thinks it is neither true nor false that, in 1977, according to Hope, Vader is Luke’s father. (Well, neither true nor false according to Hope-.) Caplan (2014) argues that this is a mistake since it misinterprets Hope. After all, Obi-Wan tells Luke that Vader betrayed and murdered his father. I consider Caplan’s criticism of Cameron to be sound. So I will instead consider a view that is related to Cameron’s view but which Cameron does not accept. This view is just like Cameron’s except that this view endorses (1).

[12] Recall that we are considering a version of Cameron’s view that endorses (1).

[13] McGonigal (2013) raises a similar objection.

[14] Versions of these sorts of objections can be found in Cappelen and Lepore (2005) and elsewhere.

[15] McGonigal (2013) doesn’t go so far as to explicitly endorse relativism as the solution to the problem of serial fiction, but he does argue that it fares better than some competing accounts.

[16] On David Kaplan’s account they also include times. See Kaplan (1989).

[17] For an extended criticism of relativism, see Cappelen and Hawthorne (2009).

[18] This sort of objection is due to Cappelen and Hawthorne (2009). I am considering which propositions are true with respect to Hamill’s circumstance rather than what Hamill thinks or says in his circumstance. This is to forestall the objection that having Hamill think or say that Vader is Luke’s father would shift Hamill’s circumstance to one in which Empire is relevant.

[19] It’s plausible to suppose that characters are not really functions from contexts to contents. See Braun (1995). As Braun observes, Kaplan’s informal remarks on character in Kaplan (1989: 505) are at odds with his formal remarks in Kaplan (1989: 548, 551-2), and, as Braun (1995) argues, it is plausible that no extensional function can adequately capture the notion of Kaplanian character.

[20] Caplan notes he is indebted to an anonymous referee and to Jeff Speaks (p.c.) for raising the problem.

[21] More carefully, it seems that 1947-Ford is inconsistent with 2015-Ford, and, so, with Ford’s Life.

[22] For general defenses of the claim that an artwork can change over time, see McFee (1980), McFee (1992), and Rohrbaugh (2003). For some sympathy for the claim, see Kaplan (2011). David Sanson suggests that we can think of the serial fiction on this view as an abstract “box”. Which propositions are contained in the box change over time. Less heuristically, I am inclined to the view that the “box” is a formal, rather than material, part of the fiction, in roughly the sense of Koslicki (2008). Obviously, the proposal leaves a lot of unanswered questions. I am trying to be ecumenical while still gesturing in the direction of the solution I prefer. McGonigal (2013) briefly considers a somewhat related view to the one on offer here.

[23] Let my “t-matter” be the matter I have at t. I do not think I am the sum of my “t-matters”. Rather, on the view I prefer, I am identical to or coincident with my matter plus my form (roughly in the sense of Koslicki (2008)). Both my matter and my form are different at different times (or I am coincident with different form/matter hylomorphs at different times) and facts about my form or matter at a time plausibly explain or partly explain why certain propositions are true of me at certain times. Thanks to Ben Caplan and Kathrin Koslicki for discussion.

[24] This might seem to just relocate the problem of change: instead of asking how one thing—Ford—can have apparently incompatible properties, we can now ask how one thing—Ford—can be constituted by different matter at different times. I take it that on the standard, endurantist, coincidentalist solution to the problem of change, this is just a brute fact about Ford. He can be constituted by different matter at different times in virtue of being the sort of thing he is.

[25] Well, Ford might be an extraordinary object. But not in the relevant sense. Though I’ve sketched a bit of the solution to the problem of change I prefer, I won’t pretend that other theories of change over time could not be extended to account for the changing matter of serial fictions.

[26] Thanks here to an anonymous referee.

[27] One important area of computer science, automata theory, studies abstract machines that change their internal state upon receiving certain inputs. And in library and information sciences, bibliographic objects, including documents, are distinguished in terms of different “abstraction levels”: item, manifestation, expression, and work. It seems that the working assumption in the field is that documents are abstract objects despite being capable of being created and altered. See Huitfeldt, Vitali, and Peroni (2012).

[28] The musical works I have in mind include songs by Shellac (http://www.touchandgorecords.com/bands/band.php?id=22). The allegedly abstract that seem to change are not confined to socially constructed entities. Biological species have hallmarks of the abstract but seem capable of undergoing change as well.

[29] It is important to emphasize that insofar as the problem of change is, at bottom, a genuinely metaphysical problem, the solution we end up endorsing to the problem of change and extend to the problem of truth in serial fiction will not, in the end, collapse into some form of contextualism or relativism, since those are, at bottom, linguistic views.

[30] David Sanson (p.c.) has suggested one important disanalogy: Russellian propositions about fictional characters, if they exist, seem unable to undergo intrinsic change in the way that Russellian propositions involving you or me might. That’s because, in turn, it is implausible that fictional characters, if they exist, are capable of undergoing intrinsic change in the way you or I might. I don’t think this disanalogy undermines the main point, however. For further discussion, see (e.g.) Braun (2005) and Braun (2015).