grazer philosophische studien (2016) 1-31 brill.com/gps Ontological Trivialism? How to Meinong a Carnap-Quine Seyed N. Mousavian University of Gothenburg, Sweden, Institute for Research in Fundamental Sciences (ipm), Iran [email protected] Abstract How hard is it to answer an ontological question? Ontological trivialism, (ot), inspired by Carnap’s internal-external distinction among “questions of existence”, replies “very easy.” According to (ot), almost every ontologically disputed entity trivially exists. (ot) has been defended by many, including Schiffer (1996; 2003; 2006) and Schaffer (2009). In this paper, I will take issue with (ot). After introducing the view in the context of Carnap-Quine dispute and presenting two arguments for it, I will discuss Hofweber’s (2005; 2007) argument against (ot) and explain why it fails. Next, I will introduce a modified version of ontological trivialism, i.e. negative ontological trivialism, (not), defended by Hofweber (2005), according to which some ontologically disputed enti- ties, e.g. properties, (almost) trivially do not exist. I will show that (not) fails too. Then I will outline a Meinongian answer to the original question, namely, ‘How hard is it to answer an ontological question?’ The Carnapian intuition of the triviality of internal questions can be saved by the Meinongian proposal that quantification and reference are not ontologically committing and the Quinean intuition of the legitimacy of inter- esting ontological questions can be respected by the Meinongian distinction between being and so-being. Keywords ontological commitment – meta-ontology – ontology – Carnap – Quine – Meinong 1 The Puzzle Let’s begin with the following meta-ontological question: © koninklijke brill nv, leiden, 2016 | doi 10.1163/18756735-09303004 0002785279.INDD 1 301935 6/23/2016 5:52:16 AM 2 doi 10.1163/18756735-09303004 | Mousavian O: How hard is it to answer an ontological question? (An ontological question being a question of existence) Thomas Hofweber (2005) argues that (O) is puzzling. The puzzle is that “there seem to be two contrary but equally good answers to the question” (Hofweber 2005, 257): very hard and very easy. Very hard, if one takes the question to be legitimate and, for example, like Quine (1948/2012) tries to read the ontology of the world off of our best overall theory of it as the values of the bound vari- ables of the theory. This project is “at least as hard as coming up with our best overall theory of the world” (Hofweber 2005, 258). Very easy, if one buys the so- called ‘triviality arguments’ and thus accepts ‘ontological trivialism’, (ot). (ot) implies that ontological questions are legitimate and trivial. More particularly, ot: Almost1 every2 ontologically disputed entity trivially (or obviously) exists.3 The debate over (O) goes back, at least, to Carnap (1956/2012)-Quine (1948/2012) dispute. Carnap, roughly speaking, classifies ontological questions as either 1 The modifier ‘almost’ is introduced in order to take into account the “rubbish bin of the non- existent” (see Schaffer 2009, 354), in which very few entities are thrown (such as perhaps contradictory entities). 2 Here ‘every’ should not be read as having ontological commitment, otherwise (ot) would be trivially true. Note that the debate over (ot) is about its non-vacuous truth or falsity. My formulation of (ot) presupposes that the sentence “there are ontologically disputed entities” has a true reading with no ontological commitment to the ontologically disputed entities. One way to accomplish this is to take the quantifier ‘every’ in (ot) to be ontologically neutral (then it can reach ontologically disputed entities even if they do not exist) and put the onto- logical commitment into the predicate ‘exists’. Alternatively, one may appeal to type-token apparatus. Accordingly, (ot) may be reformulated as ot′: “For almost every ontologically dis- puted type of entity, there trivially exists an entity of that type.” Then, even if ‘every’ in (ot′) is ontologically committing, it only brings ontological commitment to the types of entities, not to their tokens that are subject to ontological debate. I will follow the first strategy, for the reasons discussed in §9. 3 Jonathan Schaffer (2009) is explicit on such a formulation: “the contemporary existence de- bates are trivial, in that the entities in question obviously do exist” (Schaffer 2009, 357, origi- nal emphasis). Hofweber (2005) formulates the view similarly: “[A]n answer to the question whether or not there are properties, propositions or numbers follows immediately from the most uncontroversial premises imaginable” (p. 258, emphases are mine). Stephen Schiffer’s (1996, 2003, 2006) formulation of ontological trivialism in terms of “pleonastic Platonism” characterizes numbers and properties, among other things, as “language-created language- independent entities” that trivially exist as “something-from-nothing” entailments suggest. grazer philosophische studien (2016) 1-31 0002785279.INDD 2 301935 6/23/2016 5:52:16 AM Ontological Trivialism? | doi 10.1163/18756735-09303004 3 trivial or epistemically illegitimate, not both. Given a conceptual framework and a kind of entities, Carnap’s internal questions, namely, “the questions of the existence of certain entities of the new kind within the framework” (Car- nap 1956/2012, 17), are trivial and legitimate. Carnap thinks that his internal questions are trivial since they are not metaphysically committing. This is so, in turn, because such questions employ a non-metaphysical conception of reality (ibid). Carnap’s external questions, namely, “questions concerning the existence or reality of the system of entities as a whole” (ibid), are nontrivial but illegitimate. They are illegitimate theoretical questions since they are not theoretical questions in the first place. They are not theoretical since they are questions about the acceptance or denial of conceptual frameworks (or lan- guage structures). Hence, according to Carnap’s meta-ontology, no interesting ontological question is both nontrivial and legitimate. Quine (1948), in contrast, characterizes most ontological questions as nei- ther trivial nor illegitimate: They are not trivial since ontological questions need to be answered according to a theory and a theory is committed to the existence of those entities that the bound variables of the canonical formula- tion of the theory range over. Neither finding the best theory nor its canonical formulation is trivial.4 Moreover, most ontological questions are not illegiti- mate, from Quine’s (1951/1963) point of view. This is so, in turn, because Quine (1951/1961) does not acknowledge the internal-external distinction as charac- terized by reference to Carnapian conceptual frameworks. Otherwise put, for Quine, most ontological questions are nontrivial, in fact hard, and legitimate. (ot), in some sense, stands in between. (ot) describes most ontological questions as trivial and legitimate: (ot) is similar to Carnap’s view in so far as it takes many ontological questions to be ‘trivial’ but differs from that in taking such questions to be ontologically committing. As some recent formulations of (ot) emphasize, e.g. Stephen Schiffer (1996, 2003) and Jonathan Schaffer (2009), numbers, properties, propositions and fictional characters all obvious- ly exist.5 (ot) is similar to Quine’s view in so far as it takes many ontological questions to be legitimate but differs from that in treating them as ‘trivial.’ In this paper, I will take issue with (ot) from a Meinongian point of view. After the introduction, in §2, two sample arguments in defense of (ot) will be introduced. In §3, I will discuss Hofweber’s (2000, 2005, 2007) argument against (ot). In §4, I will explain why it fails. In §5, I will address an objec- tion to my criticism of Hofweber’s argument against (ot). I will then, in §6, 4 See Schaffer (2009) for further discussion. 5 Schiffer’s view makes some distinctions between properties and propositions, on the one hand, and fictional characters and artifacts, on the other hand. See note 7 below. grazer philosophische studien (2016) 1-31 0002785279.INDD 3 301935 6/23/2016 5:52:16 AM 4 doi 10.1163/18756735-09303004 | Mousavian introduce a modified version of trivialism, i.e. negative ontological trivialism (not), advocated by Hofweber (2005). Subsequently, in §7, I will argue that (not) fails too. In §8, I address an objection to my argument against (not). Finally, in §9, I will outline a Meinongian answer to (O), namely, ‘How hard is it to answer an ontological question?’ The intuition behind the Carnapian meta- ontology can straightforwardly be captured by the Meinongian proposal that quantification and singular reference are not ontologically committing. The intuition behind the Quinean meta-ontology can be saved by rejecting (ot). Ontological questions are legitimate but not trivial, given a distinct existence predicate in a Meinongian language. 2 Triviality Arguments: Schiffer and Schaffer As far as I can see, (ot) has never been directly argued for. Instead, the view has been backed by various forms of, what I call, ‘triviality arguments.’ For ex- ample, concerning the existence of numbers, Schaffer (2009, 357) offers the following triviality argument: Start with the debate over numbers. Here, without further ado, is a proof of the existence of numbers: [arg 1:] 1. There are prime numbers. 2. Therefore there are numbers. 1 is a mathematical truism. It commands Moorean certainty, as being more credible