Book review: GARCÍA-CARPINTERO, Manuel. Relatar lo Ocurrido como Invención: Una Introducción a la Filosofía de la Ficción Contemporánea (Madrid: Ediciones Cátedra, 2016).

Gustavo de Azevedo About the author
GARCÍA-CARPINTERO, Manuel. . Relatar lo Ocurrido como Invención: Una Introducción a la Filosofía de la Ficción Contemporánea (Madrid: Ediciones Cátedra, 2016). ISBN: 978-376-3593-4


The book reviewed here advances a fregean theory of reference able to stand up to Kripekean objections to descriptivism. It also claims that fictions are an invitation to imagine situations or pretending assertions and, in spite of it, fictions are objects of knowledge too, since they can refer to reality and we refer to fictional objects. In this review I present a summary of García-Carpintero's ideas and outline some objections to them.

Fiction; Fregean theory of reference; Pretense theory

Had we made a list of the most philosophically problematic entities, fictional objects would surely have been among them. Many discussions on Metaphysics and the Philosophy of Language, Mind and Mathematics have at some points involved such entities. Currently there are three main families of theories about fictional objects, namely, possibilism, neo-meinongianism and creationism. Notwithstanding the subtitle of the book, Relatar lo Ocurrido como Invención goes further than a mere introduction to the Philosophy of Fiction. In this book García-Carpintero puts foward his own set of theses about the theme. Moreover, it is not only concerned with the epistemology of fiction, as one would assume from the title. All fundamental problems involving fiction1 1 Including the three families of theories mentioned before. are addressed in detail there, from the most basic to the most complex topics until Chapter 5 which is, like the epilogue, written in the form of a literary essay. In the first two chapters, he introduces the most important speech acts theories and applies them to fictional speech (or writing) acts. In Chapter 3, there is a discussion over fictional worlds qua possible worlds. After that, in Chapter 4, the lengthiest one, he states how two ways of reference occur, namely, from a non-fictional world to a fictional one and vice-versa.

Firstly, with the intent of establishing a background thesis of speech acts, García-Carpintero makes an analysis of Austin’s, Grice and Strawson's, and Williamson's ideas about this point. Austin’s theory appears in order to distinguish the illocutionary force from the propositional content of an utterance, which can be found in every occurrence of a speech act, although the first semantic aspect is essentially related to institutions, conventions or common practice, while the second constitutes the truth conditions for enunciative utterances. That is, every illocutionary act expresses a proposition, even if the same proposition may be expressed by sentences with different illocutionary force. Similarly, acts with the same illocutionary force may represent different propositions.

Grice and Strawson claim that illocutionary acts are not fundamentally connected to conventions, but to communicative intentions instead. Such a conception, according to García-Carpintero, is imbued in a psychological analysis of linguistic actions that are constituted by two intentions: primary and procedural. The first intention concerns a propositional attitude, whilst the second looks for the satisfaction of the first intention by an inference reached via the recognition of the primary intention. Besides that, García-Carpintero explains the Gricean theory of conversational implicatures with the goal of pointing out that there is more to be considered in an utterance than its literal meaning.

None of the conceptions explored in the first two sections of the first chapter (the ones I have talked about in the last two paragraphs) are adopted in the book. The background thesis we should have in mind to read the subsequent chapters is based on the following rule: (KT) "Asserting that p is correct if and only if, she who asserts that p puts her audience in a position to know p" (p. 42). This is the knowledge transmission rule, which, in contrast to a psychological-descriptive conception, is normative. García-Carpintero, following a theory adduced by Williamson, argues that neither conventions nor intentions are sufficient to provide a correct account of assertive acts. The former is not sufficient because asserting is not a convention, although the use of certain expressions in assertions is conventional. And the latter is also not sufficient because there are utterances which one asserts without the intention of doing such.

Section 2.1 gives an explanation of Searle's ideas about assertion in fiction. Two main points are made: (i) speech acts in fiction are pretended; (ii) there are two types of norms ruling speech acts, which are vertical and horizontal. The first relates language to the world and the second involves extralinguistic conventions. Linguistic acts in fiction are ruled by the second type. Section 2.2 is dedicated to defend the claim that imagination is a de se propositional act which is immune to error through misidentification. This kind of propositional act is essential to the construction of fiction (here recognized as pretense). The idea introduced in Section 2.2 is further developed in Section 2.3 through the analysis of Walton’s and Currie’s theses. Walton proposes that sentences in fiction are pretended and must be analyzed within the scope of a fictional operator, since the propositions expressed by these sentences cannot be true simpliciter. Currie defends the claim that a fictional work is characterized by the author's communicative intention that some proposition should be imagined by the audience. That is to say, every fiction author concocts her work with the intention of leading the audience on an imaginative process of propositions by the means of recognizing this very same intention.

Chapter 3 is dedicated to an analysis and criticism of a conception due to Lewis, and states a normative proposal about the fictional discourse. After explaining which idea of possible worlds2 2 Parafctional truths. he is adopting, García-Carpintero exposes two interpretations of the Lewisian theory of fiction. Both interpretations assume that assertions contained in fiction are pretended. Nonetheless, the first restricts the accessibility relation to the actual world, while the second does not have such a restriction. Here are the two interpretations: AN1 - "A sentence of the form Ff (P) is true if and only if P is true at the closest possible world to the actual world where the pretended assertions which constitute f are known by the pretended narrator" (p. 79); AN2 - "A sentence of the form Ff (P) is true if and only if, if w is a possible world in which the propositions which constitutes the context of f are true, P is true at the closest possible world to w in which the pretended assertions which constitutes f are known by the pretended narrator"(p. 81). F is a fictional operator, f is an indicator of the fiction we are talking about, and P is a proposition.

Section 3.2 discusses three main objections to Lewis' theory. (1) It cannot deal with impossible propositions in fiction since possible worlds must be consistent. For instance, there are, intentionally or not, stories that contain contradictory propositions. According to García-Carpintero's interpretation of Lewis, each contradictory proposition is true in one fiction but not their conjunction, which calls into question the intentions of some authors; (2) Some fictional truths are derived from the conventions related to the literary genre (e.g., comedy, drama etc.) and not directly from the standard meaning of the sentences expressed in the text. For example, a movie character may have scopophobia, while the actor who plays that character has been filmed all the time; (3) Some of the narrator's assertions are not true in the story and some truths in the story are not asserted by its narrator. There are cases in which the narrator tries to deceive the audience by non-trustworthy assertions. And there are also truths that the audience should infer from the narrator's assertions. This is because some parafictional truths are not true in worlds closest to the one in which the narrator knows what she is saying.

García-Carpintero explains his own ideas Sections 3.4 and 3.5. The normative analysis does not exclude the relevance of communicative intentions since, after all, these intentions determine that the assertion norms are not applied to fiction, but to the norms of fictionalizing (make-fiction), which make an invitation to imagine certain propositional contents3 3 Propositions are structures composed of properties and objects. . More precisely, the rule adopted by García-Carpintero is: "(F) Fictionalizing p is correct if, and only if, one puts her expected audience in a position of imagining p and p deserves to be imagined by such audience." (p. 89). Insofar as we can, and in fact do, imagine contradictions, this rule accommodates contradictory stories. It also deals well with truths derived from genre conventions. If the screenplay invites us to imagine a scopophobic character, it does not matter if the actor representing the character was in front of a lot of cameras. Thus, the truth of a sentence of the form Ff (P) depends on the author’s success in attracting the expected audience to imagine P. Otherwise, a story like Kafka's The Metamorphosis could be criticized in the wrong way. When he wrote that a man turns into an insect, he did not imply that there is a possible world where a human being can become an insect4 4 Supposedly all human beings are necessarily human beings. , this is just an invitation to imagine such transformation.

Another problem concerning the Lewisian idea is that the storytelling act is also part of the story, but if a fiction asks us to imagine a world without rational beings, this would be impossible, since the narrator herself is a rational being telling the story. Hence, F deals better with such a story, since the fiction asks us to consider as true the fictional proposition that there are not rational beings in this world. That is, if we adopt F, we do not need to consider such a story to be inconsistent, withal, it does not exclude the real author from fictional analysis. García-Carpintero claims that we should take into account the author's intentions, even if the author’s context does not determine the content of the story. That is to say, the details of the author's biography must not be counted as part of the fictional content, even though they determine which speech acts were used. Thus, only data from the author’s work are relevant to its interpretation, although the author’s communicative intentions are relevant to determine the content of the work.

Section 4.1 exposes the classical quarrel involving descriptivist and Millian theories of proper names. (In Section 4.2 García-Carpintero manifests preference for a version of the former.) Nothing new concerning the well-known Fregean theory of reference is presented on this section, but the characterization of this theory is precise and more than sufficient for the proposals of the book. The same comment should be made about the Millian theory, which includes Kripke's modal argument against descriptivism. This argument, according to García-Carpintero, cannot show that there are no descriptive elements in proper names, although it shows that proper names are rigid designators, while definite descriptions usually are not. He argues that demonstratives and indexicals have descriptive contents, even though they are rigid designators. Therefore, no necessary relation between rigid designation and descriptive content can be established. At this point, García-Carpintero starts a long explanation about indexicals.

Indexical terms have two kinds of meanings. One relates to their rule of use and the other to each use of them. For example, on the one hand, the indexical "I" has a rule of use, namely, it refers to the person who makes the utterance. On the other hand, each time "I" is used, it refers to a different person if used by different speakers. So indexicals have a common and a specific meaning for all and each occurrence, respectively. García-Carpintero's suggestion is to extend this analysis of indexicals to other referential terms, but the challenge is to maintain that there is a descriptive content associated to these terms without sustaining that they are synonymous (referential terms and its associated descriptive content). From this point on, he makes use of Stalnaker's pragmatic presuppositions theory in order to meet this challenge. Inasmuch as descriptive content is part of a presupposed proposition, it is not part of the asserted proposition. This derives from the claim that some presuppositions, instead of being part of semantic content, are pragmatically presupposed. He addresses Kripke's modal argument with the very same conception which avoids treating definite descriptions as part of the asserted content of referential terms. If definite descriptions are, as adduced by García-Carpintero, only part of presupposed content, then claiming that they are not rigid designators does not dismiss definite descriptions as part of the content of referential terms.

In Section 4.3 García-Carpintero develops some consequences of a fregean theory of reference regarding fictions. He advocates that names of fictional characters are disguised descriptions. Moreover, such terms do not work as rigid designators, since they denote different objects in different possible worlds compatible with fictional truths of the work of fiction. This supposedly occurs because fictional objects are incomplete and there are many possible worlds compatible with all fictional truths in a fictional work.

Also, García-Carpintero advocates the thesis that we should put a fictional operator in front of any fictional truth. That is, "Sherlock Holmes is a detective" is false, while "Fsh (Sherlock Holmes is a detective)" is true (sh indicates one of Sherlock Holmes stories). Thus, if one asserts that Sherlock Holmes exists, she is saying something false, while it is true that Fsh (Sherlock Holmes exists). Appealing to fictional operators avoids commitment to the existence of fictional objects. There are three problems for this strategy: (i) Ironic sentences inside a fictional work cannot have the correct treatment; (ii) Transfictional sentences whose truth are independent from fiction like "Sherlock Holmes is more famous than any real detective", cannot be read with the fictional operator and remain true; (iii) As a consequence of (ii), the two readings problem remains (against simplicity). In spite of the mentioned thesis being simpler than neomeinongians theories, that can hardly be seen as a real advantage since neomeinongians theories make use of extraneous distinction between properties or property relations and have no concern with simplicity.

Section 5.1 addresses problems (i) and (ii). García-Carpintero5 5 (iii) is not addressed. It is a consequence of García-Carpintero's thesis choice. states that, if we utter the sentence appearing in (ii) as a metaphor, we were doing nothing but proposing to imagine Sherlock Holmes as an actual object of reference. After all, it is possible to assert the proposition expressed by the sentence appearing in quotation marks at (ii) without referring to any fictional object, viz., we could assert it via a paraphrase which would only contain assertions about fictional works. However, such an answer does not help with problem (i) because ironic statements inside fiction cannot be seen as a metaphoric use of the singular reference apparatus. Hence, the appeal must fall back over presuppositions. Wherefore, the main theses presented in this book are:

  1. Speech acts of fiction construction are pretended, and García-Carpintero calls these speech acts "presuppositions generators". These presuppositions generators produce propositions whereby we imagine what is true in a fiction;

  2. Such propositions must be interpreted with a fictional operator Ff , where f indicates the work of fiction in analysis;

  3. Fictional discourse does not commit one to the existence of with fictional objects.

Why does that sort of pretending speech acts matter? Section 5.2 sketches an answer, which is better developed in the epilogue. Fictions are not only able to trigger emotions, they also provide knowledge, which may be propositional, experiential (by putting yourself in someone’s shoes), and practical (know-how), although García-Carpintero is only concerned with propositional knowledge is this book. He argues that the differences between fiction and reality at the ontological and illocutionary levels do not imply that we cannot acquire knowledge from fiction, since there are many fictions that talk about reality. That is to say, some fictional work contains true propositions about reality, even though fictions are not under the aegis of truth.6 6 This lack of compromising with truth is the illocutive difference. The ontological difference is obvious. However, there is an epistemic disparity between fiction and reality. An author of fiction intends to bring about imaginative acts about propositions. On the other hand, someone who writes a non-fictional piece, e.g., a scientific paper, must have the intention to sustain her ideas. Although he considers this a strong objection, García-Carpintero (p. 186) thinks that we can acquire knowledge from fiction in a similar way than we acquire knowledge from testimony7. If we were authorized to import propositions about the real world into fiction, we could, likewise, be authorized to infer propositions about the real world from fiction. In the conclusion of Section 5.2, García-Carpintero defends the view that, besides being a source of knowledge, fiction matters because it triggers emotions in the following way: being engaged in a fictional game is pretending that some propositions are true, therefore the emotions felt from fiction are equally pretended, i.e., the fear we feel when watching a movie is, actually, fearf, although these two reactions (fear and fearf) remain indistinguishable regarding their phenomenal features.

Lastly, Section 5.3 is dedicated to the fictional aspect of visual arts and music. Pictorial representations express propositions as well as linguistic representations. If an appropriate pictorial representation of p is made, she who perceives such representation will imagine herself as seeing that p. So, differently from linguistic representation, the pictorial one is about what a spectator sees. In other words, the truths are the product of de se fictional propositions about what someone sees from a representation. Abstract painting is not a hindrance to the thesis above. It is produced with the same intention to cause some visual imaginative experience on a spectator as non-abstract painting is also produced. How about music? A musical oeuvre generates fictional propositions. Thus, a musical piece expresses fictional truths via melodic, harmonic and rhythmic structures listened by an appropriate spectator. This spectator can feel fictional emotions provided by her imaginative act similarly to the fictional literature engagement.

Relatar lo Ocurrido como Invención is probably the most complete introductory book ever written about philosophy of fiction. Nevertheless, its main theses, (1), (2) and (3), would need further explanation. It seems to this reviewer that the answers to problems (ii) and (iii) regarding (2) are unsatisfactory and (i) is not even addressed. Thesis (1) and (2) seem to subsume each other, which jeopardizes the alleged simplicity of García-Carpintero's theory. The same goes for the distinction between two kinds of emotions. Why postulate entities like fearf if it is numerically identical to fear?. It is also hard to be convinced by the combination of the thesis that we should not be committed to the existence of fictional entities with the thesis that some metafictive sentences are true. The book is full of very interesting examples, but sometimes some of them seem to be meant rather as exercises of literary criticism than as support for the arguments. The claim that some sentences containing fictional names must be seen inside the scope of a fictional operator, while others must be paraphrased (in order to avoid ontological commitments), seems a bit arbitrary. Nonetheless, Carpintero's theory has many advantages over artifactualists and neo-meinongian theories. Had he renounced the use of the fictional operator in favor of paraphrases,7 7 In spite of claiming fictional objects are artefacts (p. 168), the rest of his argumentation shows another direction. he could still avoid the ontological commitment via the claim that fiction is an invitation to imagine propositions. It could be hard to obtain these paraphrases, but the strategy would preserve his nice version of descriptivism. Moreover, this would not preclude that sentences inside fiction refer to reality nor that a fictional work can be a source of knowledge. Summarazing, in the view of this reviewer, the book has two merits, namely, exposing the most recent debate about fiction and exhibiting a fair fregean theory over fictional objects, which makes it a mandatory reading for those interest on the philosophy of fiction.


  • GARCÍA-CARPINTERO, Manuel. Relatar lo Ocurrido como Invención: Una Introducción a la Filosofía de la Ficción Contemporánea. Madrid: Ediciones Cátedra, 2016.

  • 8
    Article info CDD: 101

  • 1
    Including the three families of theories mentioned before.

  • 2
    Parafctional truths.

  • 3
    Propositions are structures composed of properties and objects.

  • 4
    Supposedly all human beings are necessarily human beings.

  • 5
    (iii) is not addressed. It is a consequence of García-Carpintero's thesis choice.

  • 6
    This lack of compromising with truth is the illocutive difference. The ontological difference is obvious.

  • 7
    In spite of claiming fictional objects are artefacts (p. 168), the rest of his argumentation shows another direction.

Publication Dates

  • Publication in this collection
    Jul-Sep 2017


  • Received
    06 Aug 2017
  • Reviewed
    03 Oct 2017
  • Accepted
    04 Oct 2017
UNICAMP - Universidade Estadual de Campinas, Centro de Lógica, Epistemologia e História da Ciência Rua Sérgio Buarque de Holanda, 251, 13083-859 Campinas-SP, Tel: (55 19) 3521 6523, Fax: (55 19) 3289 3269 - Campinas - SP - Brazil