Revisiting the ‘Wrong Kind of Object’ Problem

Any uniform semantic treatment of fictional names (e.g., ‘Frodo’) across parafictional statements (e.g., ‘In The Lord of the Rings, Frodo was born in the Shire’) and metafictional statements (e.g., ‘Frodo was invented by Tolkien’) runs into a variation of the ‘wrong kind of object’ problem. The problem arises when an analysis of one of these statements inappropriately attributes a property to an object. For example, it would be problematic if an analysis implied that flesh and blood individuals are invented by someone, and similarly problematic if an analysis implied that abstract objects are born in a certain region. Abstract object theory has provided a solution to this conundrum by distinguishing two modes of predication: encoding and exemplifying. Recently Klauk has argued that the problem reappears for the analysis of explicit parafictional statements in this theory. In this paper we formalize the objection and show that one can distinguish three issues in connection with the ‘wrong kind of object’ problem. We then address them in turn. ∗Copyright c © 2020, by Merel Semeijn and Edward N. Zalta. This paper is forthcoming in Organon F. The authors would like to thank the audience at PhiLang2019, Emar Maier and two anonymous Organon F reviewers for their valuable feedback. This research is supported by the Dutch Research Council (NWO), Vidi Grant 276-80-004 (Emar Maier). Merel Semeijn and Edward N. Zalta 2 1 The ‘wrong kind of object’ problem The semantics of statements about fictions requires a distinction between four different types of statements that feature fictional names (i.e., names of fictional entities). Consider the following four statements about Frodo: (1) Frodo had a very trying time that afternoon. (2) Frodo was born in the Shire. (3) In/According to The Lord of the Rings, Frodo was born in the Shire. (4) Frodo was invented by Tolkien. We shall categorize, and subsequently refer to these statements, using the following terminology: • (1) is a fictional statement, i.e., it is part of a fictional narrative (e.g., (1) is a quote from Tolkien’s The Lord of the Rings). Such statements are usually understood as neither true nor false but rather as fictional truth-makers; the act of storytelling determines what is true in the fictional story. • (2) and (3) are parafictional statements, i.e., statements about what is true in some fiction but which are not part of the original storytelling. These statements are usually analysed as being true or false (depending on the content of the fictional story) and can be ‘explicit’ like (3) or ‘implicit’ like (2) (depending on whether the prefix ‘In/According to fiction/story s’ is overt). • (4) is a metafictional statement, i.e., a statement about a fictional entity as a fictional entity that can be true or false. Any uniform semantic treatment of fictional names across these different types of statements runs into a variation of ‘the problem of the wrong kind of object’. If we adopt a realist approach and assume that the name ‘Frodo’ refers uniformly to an abstract object (e.g., Zalta (1983; 1988a) or Inwagen (1977)), we run into difficulties with the interpretation of (1), (2) 1Here we follow Recanati’s (2018) terminology. Note that a sentence like (1) can also function as an implicit parafictional statement if used in a discussion about the content of The Lord of the Rings. 2We adopt this term from Klauk (2014) although he uses it only to refer to the realist variant of the problem. 3 Revisiting the ‘Wrong Kind of Object’ Problem and (3); abstract objects are not the right kind of things to have trying times or be born in certain regions. On the other hand, if we adopt an antirealist approach and take the name ‘Frodo’ to refer uniformly to a flesh and blood individual in a set of counterfactual or pretense worlds (e.g., Lewis (1978), Walton (1990), or Maier (2017)), we run into difficulties with the interpretation of (4); flesh and blood individuals are not the right kind of things to be invented. In this paper, we investigate fictional names by using the theory of abstract objects (hereafter object theory) as a theoretical framework (Zalta 1983; 1988a). We use this framework to solve the realist version of the ‘wrong kind of object’ problem. Before delving into this debate we shall examine what’s required of a uniform semantic analysis of fictional names in light of the debate on ‘mixed’ discourse (section 2). We then turn to the solution that object theory offers to the problem of the wrong kind of object. This solution centers around a distinction between two modes of predication: encoding and exemplifying (section 3). We discuss a challenge to this solution that has recently been posed by Klauk (2014) according to which the problem persists for explicit parafictional statements (section 4). We offer a formalisation of this objection in object theory and show that it gives rise to three separate issues that need to be addressed. We explain how object theory addresses those issues by pointing out that the story operator creates a hyperintensional environment akin to that created by propositional attitude reports (section 5). We conclude with some insights that emerge about entailment relations among fictional, parafictional, and metafictional statements (section 6). 2 A uniform semantics for fictional names In addition to the distinction between fictional, parafictional, and metafictional statements, discussions about the analysis of fiction have also introduced a distinction between discourse that is ‘internal’ to the fiction and discourse that is ‘external’ to the fiction. The distinction attempts to separate (a) discourse that describes the content of a fiction from a perspective within the fiction (e.g., talking about Frodo as a flesh and blood individual) and (b) discourse that describes the content of a fiction from a perspective outside the fiction (e.g., talking about Frodo as a fictional character). We’ll see below, however, that some sentences constitute ‘mixed discourse’ in that they combine internal and external forms of Merel Semeijn and Edward N. Zalta 4 discourse. These sentences gives rise to the question of whether a uniform semantic analysis of internal and external discourse is possible. And this question can be posed for our initial distinction among sentences: is there a uniform semantic analysis of fictional, parafictional, and metafictional statements? In what follows, we shall examine some arguments in favour of a uniform semantic analysis across internal and external discourse. Second, we will reformulate the desideratum of a uniform semantic treatment of fictional names. As it turns out, the literature on mixed discourse only establishes a need for a uniform analysis across parafictional and metafictional statements. It is not clear whether this desideratum extends to a uniform analysis across fictional, parafictional and metafictional discourse.