Identifying Nothing: Anti-Realist Strategies for the Identity of Fictional Characters

Abstract

According to fictional anti-realism, fictional characters should be excluded from the ontological inventory. Even though ficta are not assumed to be genuine entities, some issues concerning their identity seem to be genuine ones. Anti-realist philosophers may adopt three different strategies in order to deal with them: the Negation Strategy (i.e., such problems are not genuine ones), the Translation Strategy (i.e., such problems should be translated in terms of ficta-surrogates, genuine entities that replace ficta), and the Simulation Strategy (i.e., such problems should be handled within the pretense that ficta are genuine entities). In this paper, I shall argue in favor of the Translation Strategy as it shows some analytical advantages over its rivals, especially in treating the interplay between identity issues about ficta and ordinary narrative/interpretive practices.

1. Introduction

Anti-realism about ficta is the ontological view according to which there are no such entities as fictional characters1. Typically, anti-realist philosophers are willing to replace them with some further entities (ficta-surrogates, as they may be called) that would play the same explanatory role as fictional characters while being metaphysically more palatable. For instance, if there are no ficta, hence there is no such entity as Miss Marple, something else should play the role of contributing to the truth of the following sentence:

(1) Miss Marple was created by Agatha Christie.

For the sake of simplicity, I shall distinguish two main approaches to fictional anti-realism. The first one goes semantically: since the proper name “Miss Marple” has no reference in reality, either it has some other kind of semantic value or its contribution to the sentences in which it occurs must be explained on a cognitive/pragmatic basis2. The second approach appeals to make-believe: whatever people say about fictional characters is uttered within some sort of pretense according to which there are such entities as fictional characters3.

Both the semantic and the pretense-theoretic accounts have to face identity issues about those very entities that they would like to exclude from the ontological inventory. For instance, even if there is no Miss Marple or Jessica Fletcher, it seems reasonable to wonder whether the following sentence is true or false:

(2) Miss Marple is the same as Jessica Fletcher.

Anyone who is acquainted with both Agatha Christie’s stories and the television series Murder, She Wrote should have noticed an array of significant analogies between the two characters. Indeed, Jessica Fletcher was initially conceived of as an American version of Christie’s fictional detective. An explicit identification was not allowed due to some copyright troubles; still, some hints remain: the title of the show, which clearly recalls Murder, She Said, a 1961 film based on Christie’s stories, and the choice of Angela Lansbury, who had played Miss Marple in the 1980 film The Mirror Crack’d, for the main role. On this basis, one may want to argue that Miss Marple and Jessica Fletcher are ultimately the same character. Still, one may also want to argue the opposite on different grounds—e.g., in virtue of their showing too many dissimilarities with respect to their characterization (among others, Fletcher is a novelist while Marple is not). Be that as it may, the identity riddle raised by (2) seems to be a genuine one, even though Marple and Fletcher are not genuine entities.

It is far from obvious how anti-realist philosophers should deal with sentences such as (2). In what follows, I shall present three different strategies available for fictional anti-realism. While the first and the third ones are quite ready-to-use and hence have been widely applied, the second strategy seems to be less practical insofar as it requires more preliminary work. Still, as I shall argue in a while, it shows some analytic advantages over its rivals. Generally speaking, the aim of this article may be stated as follows: if fictional anti-realism were true, which theoretical perspective should be favored in order to deal with identity statements that purport to concern ficta?

The Negation Strategy is a very natural move. Since there are no fictional entities, the problem of their identity cannot be a genuine one. After all, it is not possible to distinguish or identify things that do not even exist: no identity without entity. Unsurprisingly, this strategy is well-represented both in the semantic and in the pretense-theoretic fields:

“[…] about questions of identity of fictional entities, there is no fact of the matter when they are identical. This is so because there is no actual individual entity that bears the identity relationship to itself”4.

“One may well dismiss the matter [of identifying a character of one work with a character of another work] on the ground that the questions are silly”5.

Despite its intuitive appeal, on second thoughts, the Negation Strategy is likely to leave us with a strong sense of dissatisfaction. Even the anti-realist wants to concede that (2) is perfectly intelligible: it must have a precise meaning and a definite truth value. In fact, she presumably wants to maintain that (2) does not concern ficta but ficta-surrogates instead—perhaps games of make-believe, according to the pretense theorist, or some sort of cognitive/descriptive/propositional contents according to other strands of anti-realism.

The Translation Strategy follows this idea: since fictional characters have been excluded from the ontological inventory in favor of ficta-surrogates, identity issues (purportedly) about the former should be reformulated in terms of the latter. Such a move requires a huge amount of translation work. As far as I know, no systematic attempt along this path has ever been made.

The Simulation Strategy seems to offer a more practical alternative: while trying to fix the truth value of (2), we just make as if Jessica Fletcher and Miss Marple were genuine entities. Let us take a brief look at a couple of examples from the philosophical literature:

“[…] in situations in which fiction is being discussed, and we are speaking as though the characters, events, etc., are real, our sentences are to be understood as prefixed by ‘According to NR, …’ [where NR stands for Naive Realism]”6.

“[…] since I deny the existence of fictional characters, I take [their identity conditions] to hold not in reality but rather within the scope of the pretense that there are such things as fictional characters”7.

As a matter of fact, such a strategy particularly fits within a pretense-theoretic framework. Nevertheless, it is crucial to point out that it is not necessary to endorse a fictionalist stance towards ficta in order to make use of the Simulation Strategy: an anti-realist may decide to tackle identity issues about fictional characters within the pretense that there are such entities, without believing or assuming that any talk about fictional characters is to be understood as produced and consumed within such a pretense.

To this end, it would be sufficient to articulate the Simulation Strategy as a three-step move. First, let us call (BR) our Best Realist Theory and (BA) our Best Anti-Realist Theory. Secondly, let us assume that any sentence in the language of (BR) has an appropriate translation into the language of (BA). Finally, at this point, let us stipulate that we shall keep talking in the language of (BR) just because it is easier to handle. By assumption, were we interested in a metaphysically accurate way of speaking, a proper translation would always be available. Now, such a strategy can be worked out in detail.

2. The Simulation Strategy

Before sketching a minimal realist account of fictional characters, I shall attempt to corroborate our assumption by showing a plausibly exhaustive translation schema into the language of fictional anti-realism. To begin with, it should be noted that sentences such as (1) and (2) have at least two different readings. From an internal perspective, with respect to the fictional worlds depicted by Christie’s stories and by the television series Murder, She Wrote, (1) and (2) are indisputably false. Within Christie’s stories, it is not the case that Miss Marple was created by Agatha Christie; actually, there is no Agatha Christie inside the fictional world depicted by Agatha Christie through her stories8. Rather, the flesh-and-blood Miss Marple was presumably created by her parents (or by God, maybe, depending on one’s religious beliefs). Within the relevant stories, (2) is false as well: Jessica Fletcher, as a human being, is clearly not the same as Miss Marple as a human being. While Jessica Fletcher the person is distinct from Miss Marple the person, though, one may wonder whether Jessica Fletcher the character is distinct from Miss Marple the character.

Indeed, from an external perspective, with respect to the fictional worlds depicted by Christie’s stories and by the television series Murder, She Wrote, the truth value of (2) is disputable—as we noticed above—and (1) is undoubtedly true: it really was Agatha Christie who created the fictional character known as Miss Marple. Following Bonomi (2008), I shall call paratextual the internal-perspective reading of fictional sentences such as (1) and (2), and I shall call metatextual the external-perspective reading of such sentences. I shall also assume that any fictional discourse is either paratextual or metatextual in this sense9.

It is now feasible to outline a minimal anti-realist account of fictional characters, which shall play the role of our (BA) by furnishing a translation schema for each of these two kinds of fictional discourse. Let us base (BA) on the intuition that, whenever we talk about Miss Marple, we are not referring to an exotic entity that inhabits some mysterious Platonic or Meinongian realm, for there is no such entity. Rather, we are referring to a certain plurality of Marple-depiction tokens, e.g., the verbal descriptions contained in Christie’s novels and short stories (strictly speaking, all the spatio-temporal descriptions contained in all the physical or mental copies of Christie’s novels and short stories). Moreover, let us assume that, at least in typical cases, when I assert something about Miss Marple, I am referring to a canonical plurality of Marple-depictions, i.e., the Marple-depictions compliant with a canonical narrative tradition (the canon), where “compliant with” means “coherent and complete with respect to the canonical narrative tradition”10. On these grounds, I suggest paraphrasing the paratextual reading of

(3) Miss Marple is a detective.

as follows:

(3*) All Marple-depictions (compliant with the canon) are detective-depictions.

In the same vein, I suggest paraphrasing the metatextual reading of (1) as follows:

(1*) Marple-depictions are (collectively) such that Agatha Christie was the first human being who produced any of them.

We are now equipped with a general schema for each kind of fictional discourse:

(PT) all N-depictions are F-depictions.

(MT) the N-depictions are F.

Where N is typically a proper name and F is a unary predicate. The (PT) schema can be easily adapted to cover n-ary predications as well: for instance, a binary predication would have the form “All N-depictions and M-depictions couples are R-depictions”.

Let us recall our assumption: any sentence in the language of (BR) has an appropriate translation into the language of (BA). Our translation schemas (PT) and (MT) corroborate the assumption. Therefore, we can finally take the third step of the Simulation Strategy, i.e., we shall keep talking as if the fictional characters Miss Marple and Jessica Fletcher really existed. In other words, we shall address identity matters within the (BR) pretense. What does (BR) look like, though? We need to know that before proceeding any further.

There are many different realist theories about ficta on the philosophical market11. All of them have something in common when it comes to the identity of fictional characters, namely a tendency to identify or distinguish ficta on the basis of their characterization within the relevant stories. In this respect, realist accounts seem to just develop a commonsensical intuition about fictional stories, which is presumably taken for granted among both common readers and experienced literary scholars: i.e., that the way in which character x and character y are characterized by their author/s is relevant when it comes to deciding whether x and y are the same character or not. Hence, we may sketch a minimal realist account of fictional characters, which shall play the role of our (BR), by furnishing the following sufficient and necessary conditions for the identity of ficta:

(ISN) necessarily, x and y are the same fictional character if and only if x and y are ascribed exactly the same properties within the relevant stories.

Although equipped with a strong sense of plausibility, such a property-based criterion leads us to a pair of unwanted consequences that we need to take into account.

First, there is a problem about indiscernibility: if (ISN) were true, indiscernible fictional characters would count as a single character. Such an outcome sounds far less plausible than (ISN) itself. In The Lord of the Rings, for instance, J.R.R. Tolkien depicts Sauron’s army as composed of hundreds of orcs who are not discernible from one another. While reading the books, we just come to know some of their generic features as orcs, but we are not able to distinguish each one of them from the others by virtue of some individually specific properties. Still, in our minds, we certainly depict Sauron’s army as composed of distinct individuals—and so did Peter Jackson while shooting the films based on Tolkien’s novels. Similarly, a story may concern a couple of indiscernible twins, i.e., such that no individuating feature is specified within the story, and we would still imagine them as two distinct individuals. Undoubtedly, within the relevant stories, there are hundreds of distinct (although indistinguishable) orcs and two distinct (although indistinguishable) twins. Yet, according to (ISN) within reality there is just one single fictional character corresponding to Sauron’s army and just one single fictional character corresponding to the couple of twins.

Secondly, there is a problem about fragility: if (ISN) were true, fictional characters would be extremely fragile entities. It would be sufficient to lose one property, even an accidental or secondary one, for a fictum to lose its identity altogether. We may distinguish two kinds of metaphysical fragility. To begin with, Miss Marple would be temporally fragile. Let us imagine that Stephen King decides to write a sequel story to Christie’s novel Sleeping Murder: Miss Marple’s Last Case. According to King’s story, Miss Marple moves from England to Cabot Cove, Maine—the fictional town where Jessica Fletcher’s investigations usually take place. We are supposed to imagine that it is that very Miss Marple, the one we are so familiar with, who moves to Cabot Cove, according to King’s story. Nevertheless, it follows from (ISN) that King’s Miss Marple is not the same fictional character as Christie’s Miss Marple: within the relevant stories, the former moves to Cabot Cove while the latter does not12. Moreover, Miss Marple would be modally fragile as well. It seems true that Agatha Christie could have written crime stories very similar to the actual ones except for one detail: Miss Marple lives in St. John Mead rather than in St. Mary Mead. This certainly looks like a possible scenario. Still, it follows from (ISN) that Miss Marple could not have lived in St. John Mead rather than in St. Mary Mead. That possible Miss Marple would not be the same character as the actual one13.

Perhaps (ISN) can be adjusted to neutralize at least the latter of the two problems that affect (BR). At first glance, it seems reasonable to assume that some properties are ascribed to Miss Marple essentially while others only accidentally. For instance, it may be presumed that being a human person (within the stories) is essential to Miss Marple, while living in St. Mary Mead or moving to Cabot Cove are not. Accordingly, (ISN) may be amended as follows:

(ISN*) necessarily, x and y are the same fictional character if and only if x and y are ascribed exactly the same essential properties within the relevant stories.

It would follow from (ISN*) that King’s Miss Marple is the same as Christie’s Miss Marple and that Miss Marple could have lived in St John Mead rather than in St Mary Mead.

Unfortunately, identifying an uncontroversially essential property of some fictional character is quite an insidious task. A plausible candidate for Miss Marple would be the following: a human being gifted with extraordinary observation and reasoning skills, who solves plenty of murder cases. This really looks like a property that Miss Marple could not ever fail to possess without ceasing to be herself/itself. However, if it were sufficient to possess this property in order to be Miss Marple, then it would follow that Sherlock Holmes is the same character as Miss Marple. (ISN*) should then be amended as follows:

(IN) necessarily, x and y are the same fictional character only if x and y are ascribed exactly the same essential properties within the relevant stories.

(IN) is much weaker than its predecessors, for it only provides us with necessary conditions for the identity of fictional characters: in order to be Miss Marple, it is necessary to be ascribed, within the relevant stories, the property of being a human person gifted with extraordinary observation and reasoning skills, who solves plenty of murder cases.

Is it actually true, though? Let us imagine a future scenario. Tomorrow, among the unpublished works of Agatha Christie, a previously unnoticed manuscript will be found. Its very title is quite telling: Awaking Singer: Miss Marple’s True Last Case. Christie’s manuscript tells the story of an opera singer, Miss Marple, who suddenly awakes from a very long sleep, throughout which she has dreamt of being a person gifted with extraordinary observation and reasoning skills, who solved plenty of murder cases. The news is astonishing! Miss Marple is not a detective, albeit an amateur one, for she has just been dreaming all the time14.

There may be even trickier cases. Let us imagine a science fiction story about Miss Marple written by Margaret Atwood. Its protagonist is a teenager who lives in New York in the XXI Century. She has a good taste for vintage shows and really enjoys watching Murder, She Wrote on television, but she would be totally unable to solve murder cases herself. One day, while sniffing around the science lab at school, she runs into a mysterious time machine through which people can access alternative past scenarios. Then, clearly with some surprise, our teenager finds herself in the shoes of an elderly amateur detective who lives in St. Mary Mead, England, during the first half of the XX Century and solves plenty of murder cases thanks to her extraordinary observation and reasoning skills. Apparently, according to Atwood’s story, Miss Marple comes from the future. What would we say with regards to such deviant cases? It seems true that Atwood borrowed a fictional character from Christie and elaborated on it in order to create a new story, significantly divergent from the canon. In addition, it also seems true that, according to the canonical stories, Miss Marple does not come from the future. In other words, the following sentences are seemingly both true:

(4) The protagonist of Atwood’s story is the character created by Christie.

(5) The protagonist of Christie’s stories is not the character that features in Atwood’s story.

In order to properly engage with Atwood’s story and fully enjoy it, we need to take its protagonist as the very same character created by Christie. This is quite a good reason to value (4) as true. On the contrary, if we took the protagonist of Christie’s stories as a teenager who came from the future by means of a mysterious time machine, we would significantly alter our engagement with Christie’s stories, let alone our appreciation of them. This is a very good reason to value (5) as true. We are then left with a bizarre consequence: x is the same character as y, while y is not the same character as x. There seems to be a failure of symmetricity with regards to the identity relation between fictional characters15.

It is worth recalling that we are still reasoning in the language of (BR), i.e., within the pretense that fictional characters are genuine entities that literally exemplify properties and entertain relations between them. As we just observed, within (BR) there are some troubles about identity across different stories16. At least three explanations can be offered by the Simulation Strategy. First: fictional characters are bizarre entities that violate the logical laws of identity. This option is not authentically problematic from the point of view of the anti-realist. For we are just pretending that fictional characters are bizarre entities that violate the logical laws of identity: this is what the Simulation Strategy is all about. The second option is that the relation at issue is not strict numerical identity but some sort of weaker sameness relation that is not symmetric, so it may be the case that x is the same as y while y is not the same as x. As we shall try to show in a while, this choice is quite promising, but it would be better developed after a change or strategy. The third option is that a Geachian notion of relative identity may be employed. One may say, for instance, that Atwood’s Miss Marple is the same character as Christie’s Miss Marple with respect to Atwood’s deviant story, and yet they are distinct characters with respect to Christie’s canonical stories.

In sum, while trying to build (BR) on a very robust intuition, i.e., that the identity of ficta depends on their characterization within the relevant stories, we came into trouble. There are a couple of ways out. We may abandon the property-based identity criterion and search for a different one, which would have nothing to do, either directly or indirectly, with the features ascribed to characters in their fictional worlds. Such a move seems gratuitously counterintuitive to me. Unless provided with a strong independent reason, one should not reject the idea that our attempts to identify fictional characters ultimately rely upon the creative practices of those human beings who made them up in the first place. Rather, it may be asked whether it is strict numerical identity that we are concerned with when we wonder about the sameness relation between Miss Marple and Jessica Fletcher, or Christie’s and Atwood’s Miss Marple. It is worth exploring this question in further detail. In what follows, I shall try to show that the Translation Strategy better suits such an exploration.

3. Back to the Translation Strategy

As the Simulation Strategy for fictional anti-realism requires, we have tackled some identity issues about ficta within the realist pretense that fictional characters are genuine entities, either objects or properties, and that one can genuinely wonder whether an identity relation holds between a character x and a character y. As we just saw, though, ficta can behave in quite a bizarre way. In particular, among other strange phenomena, there seems to be a failure of symmetricity in some cases. Now I want to argue that, in order to properly analyze such cases from an anti-realist point of view, the Translation Strategy is more effective. For it shall allow us to treat Miss Marple and Sherlock Holmes not as two individuals, or two properties, but as two pluralities or collections of individuals: the Marple-depictions and the Holmes-depictions (compliant with certain narrative traditions).

Let us finally step out of the pretense and consider (BA) again. Agatha Christie wrote several novels and short stories about the same fictional woman, Miss Marple. By doing so, she produced hundreds of fictional depictions that, taken together, constitute a canon that is still widespread among passionate readers all over the world. Within this narrative tradition, it is feasible to single out a plurality of canonical Marple-depictions, i.e., the plurality composed of all and only those Marple-depictions which are coherent and complete with respect to the canon. Therefore, any member of such a plurality is an amateur-detective-depiction, an elderly-unmarried-woman-depiction, a living-in-St-Mary-Mead-depiction, and so on and so forth, and none of them is, e.g., a green-spooky-Martian depiction. We shall refer to this canonical plurality by means of a plural definite description: the Marple_c-depictions. Some decades after Agatha Christie’s creative activities, Margaret Atwood decided to write a science fiction sequel to Miss Marple’s popular stories. By doing so, she produced several fictional depictions that, taken together, constitute a derivative and somehow deviant narrative tradition. Within the latter, it is feasible to single out a plurality of deviant Marple-depictions, i.e., the plurality composed of all and only those Marple-depictions that are coherent and complete with respect to Atwood’s deviant tradition. Therefore, any member of such a plurality is a coming from the future-depiction, a teenage-girl-depiction, an elderly unmarried-woman-depiction17, and so on and so forth, and none of them is, e.g., a green-spooky-Martian-depiction. Again, we can introduce a plural definite description in order to refer to this deviant plurality: the Marple_d-depictions.

I have just outlined a plausible anti-realist picture of what it literally means for an author to ‘create’ a character and what literally is a fictional character. As the Translation Strategy requires, we should now translate (4) and (5), i.e., two sentences that purportedly concern fictional characters, into the metaphysically transparent language of anti-realism. Namely, we should reformulate (4) and (5) in terms of our ficta-surrogates. Before translation, though, a further step is needed.

Typically, the higher the level of popularity of a fictional character is, the bigger the number of derivative traditions will be. Think of Mickey Mouse, for instance. At the very beginning, he/it was created by Walt Disney, who ascribed some specific features to him/it; hence, a corresponding plurality of Mickey-Mouse_c-depictions arose18. Later on, the successful fate of this character led many different countries in the world to adopt him/it and give rise to specific ‘national’ depictive traditions, which are linked together to some extent but also quite independent from each other. Hence, many corresponding pluralities of Mickey Mouse-depictions arose—a French one, a Mexican one, an Italian one, and so on. In virtue of the articulated causal chain that ties them all, they can be collected together into a huge single plurality of Mickey-Mouse depictions. Presumably, it is this generic plurality that we have in mind when we posit questions about Mickey Mouse. Is Mickey Mouse the same as Donald Duck? Of course, he/it is not, and we do not need to resort to the original/canonical Mickey Mouse in order to settle this question.

Analogously, there is a general plurality of Marple-depictions (i.e., the Marple_g depictions) to which both the Marple_c-depictions and the Marple_d-depictions belong in virtue of their being linked together within the same articulated causal chain19. When we ask whether Atwood’s Miss Marple is the same character created by Christie, we presumably have in mind the general character of Miss Marple, i.e., the plurality of Marple_g-depictions. In other words, we are wondering if Atwood’s depictions belong to the general depictive tradition initiated by Christie. Thus, (4) can be translated into

(4*) The Marple_d-depictions are part of the Marple_g-depictions.

Such a parthood relation is to be understood as follows: any Marple_d-depiction is one of the Marple_g-depictions. On the other hand, when we wonder if the protagonist of Christie’s stories belongs to Atwood’s derivative tradition, the answer is clearly negative. Indeed, (5) can be translated into

(5*) It is not the case that the Marple_c-depictions are part of the Marple_d-depictions.

Again, this is to say: it is not the case that any Marple_c depiction is one of the Marple_d-depictions. Both (4*) and (5*) seem to be true. Figuratively, it is useful to conceive of such pluralities as sets. The set of the Marple_d-depictions is included in the general set of the Marple_g-depictions (but the opposite does not hold); on the other hand, the set of the Marple_c-depictions is not included in the set of the Marple_d-depictions—in fact, they are disjoint sets. Moreover, the set of the Marple_c-depictions would be nothing but the difference between the general set and all the derivative sets included in it20.

This analysis in terms of ficta-surrogates is aimed at explaining the apparent failure of symmetricity detected above. We do not need to revise our notion of strict numerical identity when it comes to fictional characters; it is sufficient to replace it with a non-symmetric relation, i.e., inclusion between pluralities of fictional depictions. Interestingly, within a general narrative tradition, symmetricity seems to fail anytime a change of genre occurs. Atwood, e.g., develops Christie’s original tradition by switching to a different literary genre: sci-fi crime fiction instead of realistic crime fiction. Parody is also a case in point. In the 1988 film Without a Clue, directed by Thom Eberhardt, Sherlock Holmes turns out to be an actor hired by John Watson to pretend to be a brilliant detective who solves plenty of murder cases. In reality, Dr. Watson is the clever guy. Role reversal is a typical device of humorous fiction. Eberhardt’s film develops the original tradition initiated by Arthur C. Doyle in a deviant way, through a change of genre: from realistic crime fiction to humorous crime fiction. Generally speaking, any expansion of a narrative universe may generate a failure of symmetricity of this sort.

The Translation Strategy seems to be suitable not only for ordinary narrative practices such as narrative expansions but also for ordinary interpretive practices such as the one illustrated by the following case21. In Franz Kafka’s masterpiece The Metamorphosis, it is told the story of Gregor Samsa, a salesman who wakes up one morning only to find himself transformed into a horrible, gigantic vermin. The narrator does not say anything explicit about what kind of vermin readers are supposed to imagine. Nevertheless, there are some hints here and there throughout the fictional text produced by virtue of Kafka’s creative activities. Now, an interpretive conundrum occurs. In fact, on the basis of exactly the same textual evidence, Vladimir Nabokov and Primo Levi have argued that Gregor Samsa was transformed into a beetle22, while the majority of literary critics usually take for granted that Gregor Samsa was transformed into a cockroach. Hence, both these sentences are true:

(6) Gregor Samsa is a cockroach according to Nabokov and Levi.

(7) Gregor Samsa is a beetle according to the majority of literary critics.

At first glance, one may suspect that different characters correspond to different interpretive acts and wonder whether the following sentence is true or false:

(8) Gregor Samsa the cockroach is the same as Gregor Samsa the beetle.

Although the ‘two’ Samsa are characterized differently, commonsense would presumably suggest to us that (8) is true, and I do not see any good reason to reject it in such cases: there is just one single fictional character, Gregor Samsa, who is interpreted as being transformed into a beetle (within the relevant story) according to a certain interpretive act, and is also interpreted as being transformed into a cockroach (within the relevant story) according to a different, yet equally legitimate, interpretive act.

Here is Kafka’s case analyzed from the point of view of a fictional anti-realist who wants to apply the Translation Strategy to it. Let us first consider the plurality of canonical depictions of Gregor Samsa, which arise from Kafka’s novella, i.e., the Samsa_c-depictions. Indisputably, it is true of any Samsa_c-depiction that it is a vermin-depiction23. Also, it is neither true that any Samsa_c-depiction is a beetle-depiction nor that any Samsa_c-depiction is a cockroach-depiction: the exact kind of vermin was just left undetermined by the author. Readers can only rely upon some textual hints that appear to be compatible with both the beetle and the cockroach rendering of what is going on in this Kafkian disturbing world24. Therefore, we may translate (6) and (7) as follows:

(6*) the Samsa_c-depictions are interpreted as beetle depictions by Nabokov and Levi.

(7*) the Samsa_c-depictions are interpreted as cockroach depictions by the majority of literary critics.

(6*) and (7*) are both true, as expected. According to Nabokov and Levi, Gregor Samsa is depicted by Kafka as transformed into a beetle. On the other hand, according to the majority of literary critics, Gregor Samsa is depicted by Kafka as transformed into a cockroach. Still, Nabokov, Levi, and most of the critics are all talking about the same plurality of fictional depictions: the canonical one generated by Kafka’s creative activities. Thus, (8) can be translated into the language of anti-realism as the true sentence

(8*) the Samsa_c-depictions interpreted as beetle-depictions are the same plurality as the Samsa_c-depictions interpreted as cockroach-depictions.

4. Conclusions

It may be observed that the Simulation Strategy would be as effective as the Translation Strategy, and perhaps even more straightforward, in treating ordinary interpretive practices, such as the one exemplified by Kafka’s case illustrated above. Within the pretense of (BR), one would just say that there is a single fictional character, Gregor Samsa, who is interpreted as being transformed into a beetle by some readers and as being transformed into a cockroach by others. Nevertheless, our main purpose here is to show that the Translation Strategy is equipped with some refined tools particularly useful in treating more troublesome cases of narrative expansion, which may elicit bizarre phenomena. In addition, such a strategy can be easily extended to cover interpretive practices as well. Virtually, by means of (BA), the anti-realist should be able to translate and analyze any problematic sentence that purports to concern fictional characters as genuine entities. The drawback of such a method is easy to find; it makes language far more complicated than it would be within the (BR) pretense. Not an unaffordable cost for the anti-realist, though, for she should be content enough to complicate the semantics in order to keep the ontology as simple as possible.