Existence: Essays in Ontology

Existence: Essays in Ontology 1. Introduction This wonderful collection of most of van Inwagen’s recent essays on topics in fundamental ontology is certainly to be welcomed.1 Many of the essays are focused on articulating and arguing for van Inwagen’s preferred meta-ontology, which he calls neo-Quineanism. In addition to these essays, Existence also contains essays on the eliminability of variables, the status of fictional entities, the nature of ontological categories, the nature of properties, mereology and change over time, the comparative merits of constituent and relational ontologies, a probing discussion of metaphysical issues in the philosophy of mind, as well as an interview with van Inwagen that describes his intellectual development and his assessment of his philosophical contributions. Each essay is written in van Inwagen’s characteristically charming and impeccable style. I am grateful that this collection exists in whatever sense of ‘exists’ or manner of existence is appropriate to apply. As noted above, many of the book's essays defend all or some of the theses that constitute neo-Quineanism. These theses are as follows: (1) Being is not an activity. (2) Being is the same as existence. (3) Existence is univocal. (4) The single sense of being is adequately captured by the existential quantifier of formal logic. The fifth thesis is a conjunction of methodological commitments about how to conduct ontological disputes in light of 1–4. Roughly, and briefly, participants in ontological disputes should examine the ontological implications of whatever they wish to affirm, and these ontological implications should be rendered using the first-order quantifiers and variables. Existence is exceptionally rich, and regrettably I can discuss only a small piece of it. Specifically, I’ll focus on van Inwagen’s defence of thesis 3. In Section 2, I’ll try to clarify what thesis 3 expresses. In Section 3, I’ll discuss how van Inwagen reconciles thesis 3 with the claim that all physical objects are mereological simples or living beings, but nonetheless ordinary people can truly say, ‘There are chairs’. In Section 4, I’ll examine one of van Inwagen’s arguments for thesis 3. In Section 5, I will discuss a version of ‘anti-realism about ontology’ that is not refuted by van Inwagen’s arguments for 3. 2. What does thesis 3 express? In several of the essays in Existence, van Inwagen appears to defend both the claim that the expression ‘exists’ is univocal and the claim that existence is univocal. But while I understand what van Inwagen means when he claims that ‘exists’ is univocal – he means that ‘exists’ has exactly one meaning, that this expression is not semantically ambiguous – I’m not sure what he means by ‘existence is univocal’. What sort of thing does van Inwagen take existence to be? Is it, for example, a property? And is it the sort of thing that by van Inwagen’s own lights can be aptly said to be univocal? If existence is univocal, is ‘univocal’ univocal? Or is existence univocal in a different sense from the sense in which ‘exists’ is univocal? One interpretative possibility is that all van Inwagen means when he utters ‘existence is univocal’ is that ‘exists’ is univocal. On this interpretation, we have two merely stylistic variants of the same semantic thesis. This interpretation is somewhat supported by the casual way in which van Inwagen alternates between saying that ‘exists’ is univocal and existence is univocal.2 However, there is another possible interpretation: when van Inwagen claims that existence is univocal, the opposing view that he targets is the view that there are modes, ways, or kinds of existence. That existence comes in modes, ways, or kinds is both historically important and worthy of critique, and hence it would be unsurprising if it were among van Inwagen’s targets. Moreover, in the context of defending van Inwagen’s preferred meta-ontology, van Inwagen (50–58) criticizes Heidegger’s philosophy of being, which includes the claim that there are modes of being, and Meinong’s distinction between subsistence and existence (60, fn. 16). So it is not implausible that, when van Inwagen utters ‘existence is univocal’, he is denying that there are modes, ways, or kinds of existence. Suppose this interpretation is correct. Then thesis 3 conflates two claims that are not obviously equivalent. Ryle denied that ‘exists’ is univocal but he also did not believe that there are modes or kinds of existence. Contrary to van Inwagen’s claim (63), Ryle does not argue that existence is equivocal; perhaps he would have said that it is a category mistake to say that existence rather than ‘exists’ is equivocal.3 Regardless, Ryle’s focus is on ‘exists’ rather than on existence. Ryle denied that there are modes or kinds of existence partly because he does not believe that there is a general sense of ‘exists’ that applies to anything that the various senses of ‘exists’ apply to. But note also that one could accept the metaphysical claim that there are modes of existence while claiming both that no ordinary language expression has any of these modes as its meaning (or semantic value more generally), and that ‘exists’ has as its sole meaning (or semantic value) the ‘general’ way of existing that everything has, regardless of the kind of thing it is.4 (Here are two analogies to consider. Consider a language that has a univocal predicate for being coloured although it contains no predicates for particular colours; consider a language in which a kind term like ‘jade’ is univocal although jade is itself something like a disjunctive kind.) If thesis 3 does conflate two non-equivalent theses, then they should be disentangled and we should assess which if any of van Inwagen’s targets are refuted by his arguments for thesis 3. In what follows, I will treat 3 as if it only expressed the semantic thesis about ‘exists’. 3. Ontology and meta-ontology Thesis 3 initially seems to conflict with van Inwagen’s own metaphysical practice. When doing metaphysics, van Inwagen denies that there are chairs.5 But van Inwagen also claims that ordinary assertions of ‘there are chairs’ are literally true but compatible with what van Inwagen asserts when he says ‘there are no chairs’.6 The first essay of Existence, titled, ‘Introduction: Inside and Outside the Ontology Room’, defends this claim. If there were two meanings of ‘there is’, the ordinary person’s assertion and van Inwagen’s assertion could both be true. Thesis 3 rules this out. However, even if ‘there is’ is not ambiguous but there are other possible meanings this expression could have had, we could see how the ordinary person’s assertion and van Inwagen’s could both be true: perhaps van Inwagen has adopted a new (possibly better from a metaphysical perspective) meaning for ‘there is’ and is using it when stating his metaphysical theses. But van Inwagen (37–38) denies this as well: not only are ontological expressions such as ‘there is’, ‘exists’, ‘being’, univocal, but also these expressions could not have had different meanings than what they in fact have. I wondered at this point how van Inwagen individuates expressions; if an expression essentially has the meaning it in fact has, ontological expressions could not have had different meanings. (If expressions don’t essentially have their actual meanings, why couldn’t ‘there is’ have had a different meaning? The answer seems to be that there is no alternative possible meaning; more on this soon.) But nonetheless we could still reconcile the ordinary person’s assertion with van Inwagen’s assertion: perhaps contrary to appearances, only the ordinary person uses ‘there is’, while van Inwagen uses an expression that looks, sounds, and functions syntactically and inferentially like the ordinary ‘there is’ while differing in meaning. (But, as before, perhaps the meaning van Inwagen has adopted is a metaphysically better meaning to adopt.) Provided, of course, that there is another meaning possible for ‘there is’, which van Inwagen denies. We will return to his reason for this denial in Section 4. Whether ontological disputes should be conducted in ordinary language or in a technical language has been the subject of much recent debate. For example, Sider (2009) considers framing ontological disputes in what he calls ‘Ontologese’; the quantifier-expression used in Ontologese is stipulated as having the most metaphysically fundamental meaning an expression of this sort can have. Van Inwagen (1) also suggests the adoption of a technical language for conducting ontological disputes that he calls ‘Tarskian’. In Tarskian, sentences like ‘chairs exist’ have the logical form taught to us by first-order logic. But Tarskian does not contain a quantifier expression that differs in meaning from ‘exists’ in ordinary English, and so in this respect Tarskian is not Ontologese. When do we speak Tarskian? Van Inwagen (3–5) distinguishes two contexts in which one might utter the sentence ‘there are chairs’. One of these contexts is the context of ordinary life; another of these contexts van Inwagen dubs ‘the ontology room’. ‘There are chairs’ expresses one proposition – a true one – in the context of ordinary life, but it expresses a different proposition – one that is false – in the ontology room. I take it that van Inwagen’s position is that ‘there are chairs’ is a univocal sentence, provided that ‘chairs’ is univocal, yet the sentence as a whole is context sensitive. The key thing is that somehow this shift in context brings about a shift in the logical form of proposition expressed: in the ontology room, ‘chairs exist’ wears its logical form on its face, but the form it has in the context of ordinary life is not the form of ‘There is an x such that x is a chair’.7 I don’t see how the context of an utterance could affect the logical form of a sentence in the way van Inwagen needs. I do accept that what sentence has been expressed in a given context is sensitive to features of that context. Van Inwagen (7, fn. 12) mentions the string of English words, ‘Landing planes can be dangerous’, and I accept a position (that he also mentions) that this string corresponds to two sentences, one of which contains a phrase denoting an action-type (‘landing planes’) while the other contains a phrase equivalent to ‘planes that are landing’. Context can settle which sentence is used in this context; but once context does this, it doesn’t further settle the logical form of the sentence expressed. But I doubt that the string of English words ‘there are chairs’ corresponds to two distinct sentences. Moreover, I suspect van Inwagen’s position conflicts with what seems to me a plausible principle about context-sensitivity: CS: If a complex sentence is context sensitive, it has some context-sensitive proper part that explains or grounds the context sensitivity of the sentence.8 If CS is correct, then van Inwagen should tell us which part of existence avowals or denials determines the context sensitivity of the avowal or denial. Is it the quantifier expression or the remaining material that is adjoined to it? It is widely accepted that ‘there is’, ‘some’, and other locutions like these are context-sensitive expressions. But the standard way of understanding how they are context sensitive does not fit well with van Inwagen’s picture. On the standard picture, these expressions correspond to domains of entities, and they are context sensitive in that a given expression corresponds to a different domain in different contexts. In the context of a party, I can truthfully say, ‘There aren’t any bottles of beer except what are in the cooler outside’ even though there are many bottles of beer all over the world. In the context of the party, we are talking about fewer things than in other contexts. This familiar phenomenon has been noted and exploited by ontologists with views opposed to van Inwagen’s. The mereological universalist thinks that, whenever there are some things, there is a whole composed of those things. This profligate ontology seems contrary to common sense. Many mereological universalists have attempted to reconcile their affirmation of sentences like ‘There is a whole made out of Kris and the moon’ with the ordinary person’s affirmation of ‘There is nothing made out of Kris and the moon’. According to the standard story, what the universalist says in one context is true but so is said in ordinary contexts, because the domain of quantification invoked in the latter are proper subdomains of the domain of quantification invoked in the former context, and these subdomains do not contain the fusion of Kris and the moon.9 So both the mereological universalist and van Inwagen distinguish between a context in which ontological assertions are made and ordinary contexts, and by doing so, both the mereological universalist and van Inwagen hope to reconcile their ontological claims with what initially appeared to be common-sensical denials of them. But prima facie the universalist is in a better position than van Inwagen. First, the mereological universalist explicitly identifies the context sensitivity of the quantifier expressions as the source of the context sensitivity of sentences that avowal existence and thereby respect CS, whereas van Inwagen is (as far as I can tell) silent on the source of the context-sensitivity. Second, the mereological universalist appeals only to a well-understood phenomenon, the positing of which is antecedently motivated by considerations that have little to do with fundamental metaphysical disputes, whereas we do not yet have a theory of how it is that avowals of existence could be context-sensitive in the way that van Inwagen claims, and the positing of this context-sensitivity seems motivated largely by the need to reconcile van Inwagen’s metaphysics with the assertions made by ordinary people. (Obviously, van Inwagen cannot appeal to the familiar fact that we often quantify restrictedly, since, unlike the mereological universalist, van Inwagen believes in far fewer kinds of things than the ordinary person prima facie appears to believe in.) Van Inwagen has offered a theory of how positive existential sentences uttered in ordinary contexts might be true, even given his ontology. But what about negative existential sentences? Consider the sentence ‘Nothing is made of that cat and that dog’ uttered in an ordinary context. In that context, the sentence expresses a truth. As noted earlier, the universalist explains how it expresses a truth: the quantifier is restricted. What is van Inwagen’s explanation? Consider a log cabin. According to van Inwagen (11–14), in an ordinary context the sentence ‘there is log cabin made out of those logs’ expresses a proposition that is necessarily equivalent to (but not identical with) the proposition expressed in the ontology room by ‘there are simples arranged log-cabin-wise’. By parity of reasoning, in an ordinary context the sentence ‘nothing is made of that cat and that dog’ should express a proposition that is necessarily equivalent with the proposition expressed in the ontology room by ‘no simples are arranged that-cat-and-that-dog-wise’. But that latter proposition is false, since there are simples in just that arrangement. So the ordinary disavowal should express a false proposition as well. Why focus on ordinary common-sensical utterances, rather than utterances made in the context of scientific inquiry? Suppose an astrophysicist says, ‘Some celestial object is causing this strange pattern of light’. Why think that this sentence doesn’t have the logical form it seems to have? The fact that the object causing this pattern is unlikely to be alive or a mereological simple doesn’t provide a reason to think this. Ordinary language didn’t develop for the purposes of theorizing about the physical world, but scientifically regimented languages did. In short, I don’t see why only the metaphysician speaks in Tarskian. The scientist who cares about precisely describing physical reality would be well-served by adopting a language with a relatively transparent logical structure. Why think that she hasn’t but the metaphysician has? But, if the scientist is speaking in Tarskian in these contexts, what she says is incompatible with van Inwagen’s preferred ontology – and this strikes me as a worse consequence than merely flying in the face of common sense about what there is. I remain doubtful whether a sentence like ‘Ted has more mass than Jason’ has a particularly complex logical form. There’s some funny business about how to handle the time at which Ted is more massive than Jason, but that’s about it. But if ‘Ted’ and ‘Jason’ refer to my car and my guitar respectively, the sentence is false in van Inwagen’s ontology room although true in the context of ordinary life. Are names context-sensitive? Is ‘has more mass than’ context-sensitive? It seems to me implausible to claim this, but it also seems to me that, given CS, van Inwagen might need to. 4. Van Inwagen’s arguments for thesis 3 Van Inwagen offers two arguments for thesis 3. The first is van Inwagen’s (41–43 and 61–63) argument from number. Roughly, it is this. Existence is closely related to number: ‘Fs exists’ is true if and only if the number of Fs is greater than 0. Numerals, however, are univocal, as is ‘the number of’. These expressions are univocal if and only if ontological expressions are univocal. So ontological expressions are univocal. Elsewhere, I’ve examined whether this argument shows that there are no modes of being.10 And we’ll further discuss it in Section 5. Accordingly, I’ll focus on the second argument here. Let’s call this second argument the argument against expansion; it appears on pages 69–70. The target of this argument is Putnam’s (2004) claim that, although (perhaps) we cannot say that arbitrary fusions of things exist in the ordinary sense of ‘exist’, we can introduce what he calls a ‘conventionally extended’ sense of ‘exist’ on which we can. If there are alternative possible meanings for ‘exist’, there are the sort of extended meanings that Putnam mentions; but if there no such extended meanings, there are no alternatives to what we actually mean with our ontological expressions. Van Inwagen (69–70) offers the following straightforward argument: in order to introduce an extended sense of ‘there is’ to range over new entities (such as arbitrary fusions) one must antecedently believe that there are these entities. But, if one antecedently believes that there are these entities, one is already quantifying over them, and hence the putatively extended meaning for the quantifier is simply the original meaning the quantifier already had. As van Inwagen (69–70) nicely puts it, ‘Extending the meaning of a term so that the term will apply to objects beyond those it already applies to is precisely analogous to extending a geographical boundary: you can extend a geographical boundary to encompass new territory only if that territory is already there.’ I will offer two responses to the argument against expansion. The first response is that Putnam could concede that one can redraw ‘ontological boundaries’ only against a pre-existing belief that there are things for the extended meaning to apply to, but deny that this belief must be in the ‘objects that would be in the domain of the new quantifier’. Let me explain. Suppose for the sake of illustration we take the semantic value of the existential quantifier to be a higher-order property of properties; call this property ‘E’. Consider now the property of being a cat, the property of being a dog, and the property of being the fusion of a cat and a dog. The property of being a cat and the property of being a dog instantiate E, but the property of being the fusion of a cat and a dog does not. But there is another property, E*, such that any property that instantiates E instantiates E* but some properties, among them being the fusion of a cat and a dog, instantiate E* without instantiating E. E* is the putatively possible extended meaning for our ontological expressions. In order to invoke E* rather than E, perhaps one must have an antecedent belief in something like properties, but one needn’t have an antecedent belief in things that are fusions of cats and dogs. The second response is that no antecedent belief in objects of any kind is necessary for adopting an alternative meaning for ontological expressions. Rather, what is necessary is a change in the pattern of use of ontological expressions, and this change in use might be unnoticed by members of the linguistic community in question. The pattern of use of an expression consists in sincere and literal utterances of sentences containing that expression, as well as dispositions to sincerely and literally utter sentences that contain it. One of the factors that determines the semantic values of our expression is their pattern of use; if the pattern of use changes significantly, the semantic value of that expression might change as well. Van Inwagen (9) says of himself that he is enough of a Wittgensteinian to think that it is not possible for very much of what we ordinarily say to turn out false. But what if the folk change what they ordinarily say? What if sentences like ‘there are wholes made of cats of dogs’ were sincerely, literally, and frequently used in a wide range of ordinary situations? I suspect that, if use were to change in this way, the proposition expressed by ‘there are wholes made of cats and dogs’ would not be the one this sentence currently expresses. I suspect that ontological expressions such as ‘there are’ would acquire a subtly different meaning, and this is why these sentences would express different propositions. Perhaps there is another explanation as to how these sentences would come to express different propositions in those circumstances; perhaps a systematic change in relevant features of their contexts of utterance somehow brings this about. I don’t know what this other explanation looks like, but I also don’t know how it works on van Inwagen’s own account. It seems to me that the argument against conventional expansion doesn’t succeed. Van Inwagen (42–44) offers a second reason in support of the argument against expansion. Consider the null individual, that putative entity that is a part of everything. There is no such entity, but if we can introduce ‘conventionally extended’ meanings for ontological expressions, there should be a possible meaning for ‘there is’ such that, when ‘there is’ has that meaning, ‘there is a null object’ expresses a truth.11 Van Inwagen thinks that there is no such meaning, and hence anyone who tried to speak a language like this would not even be playing with words, but rather with meaningless sounds, marks, or scratches. I don’t think van Inwagen is clearly right here, but I’ll concede that he is and investigate whether there is a principled way to accept some ‘conventionally extended meanings’ and reject others. A key idea of philosophers like Hirsch is that putatively different ontological theories are just different ways of expressing the same underlying facts about concrete reality. Perhaps the underlying facts about concrete reality that all adequate languages must ‘agree on’ are facts about the properties and relations of the mereologically simple entities. The null individual, were it to exist, would be a mereological simple. Accordingly, a language in which ‘there is a null object’ expresses a truth is not adequate, unlike, e.g., a language in which ‘every collection of simples has a fusion’ expresses a truth. 5. Contextualist anti-realism about ontology In this final section, I’ll briefly sketch a version of anti-realism about ontology that both captures the spirit of Putnam’s and Hirsch’s views but accepts thesis 3 and but accepts thesis 3. The key move is that, just as van Inwagen claims, sentences containing ontological expressions are context-sensitive. It seems to me that what is central to the spirit of Putnam’s and Hirsch’s views is that there is no unique context that is the context to use existential sentences in order to express ‘the ontological facts’. In short, there is no context that is the ontology room. But there might nonetheless be many ontology rooms – that is, many contexts in which one can use existential sentences to express true propositions that they would not ordinarily express. Putnam and Hirsch might cheerfully grant that there is the context that van Inwagen calls ‘the ontology room’, although the definite description is inapt, for there are also contexts in which ‘for any objects, there is a whole composed of those objects’ expresses a truth. The proponent of this view makes use of the same machinery van Inwagen makes use of; she traffics in other senses of ‘exists’ to no greater extent than van Inwagen himself. Moreover, this version of ‘anti-realism about ontology’ could embrace a model of why ontological expressions are context-sensitive: they are context-sensitive because they are sensitive to standards of application. Sensitivity to standards of application is a familiar phenomenon. Consider ‘tall’. ‘Hugh Jackman is tall’ expresses a truth in some contexts of utterance because the standards governing ‘tall’ in that context are fairly ordinary, but in the context of discussing basketball players, ‘Hugh Jackman is tall’ expresses a falsehood. The spirt of ‘anti-realism about ontology’ is that there are different standards that one could have for what it takes to be an object. On particularly lax standards, the goofy objects countenanced by the unrestricted mereologist have what (little) it takes, but on much stronger standards, we cannot even truthfully say ‘tables exist’. Note also that this version of ‘anti-realism about ontology’ has the following advantage over van Inwagen’s view: it appeals to a familiar phenomenon that is already known to be widespread and merely finds a further application for it. Note also, though, that one could embrace this kind of contextualist anti-realism without offering this particular mechanism for why ontological expressions are context sensitive. Can one accept this explanation for the context-sensitivity of ontological expressions and still accept thesis 4? Probably not. Thesis 4 tells us that the meaning of the existential quantifier is exhausted by the syntactic and inferential rules governing that expression; these are the things you grasp when you study a textbook of logic. But if ontological expressions are sensitive to standards, this fact is something you wouldn’t learn simply from a textbook on logic. So it would seem that there is more to the meaning of ontological expressions than what is represented by the existential quantifier? On this view, what is the relationship between ordinary existential expressions and the existential quantifier? I think not like the relation between ordinary conditionals and the material conditional; the latter is an expression with less content but nonetheless with sufficient content to be a meaningful expression. But if you like the view I have been sketching, you should regard the existential quantifier as a contentless symbolic device, the mastery of which aids you in understanding the structural features of genuinely meaningful ontological expressions. Van Inwagen’s arguments for thesis 3 are presented independently of the considerations in favour of the theses that succeed them, but perhaps what we have learned is that there are deeper connections between theses 3 and 4 than we antecedently thought.12 How fares this version of anti-realism against van Inwagen’s argument from numbers? Perhaps this argument works against Putnam and Hirsch’s views as they state them. But it seems less successful against the version of ‘anti-realism about ontology’ I’ve articulated. On this view, sentences containing numerals are also context sensitive, as they also seem to be according to van Inwagen. Imagine the following exchange at the physical plant at Western Washington University: ‘We need the exact count of chairs for the next shipment, so no screw-ups this time. How many chairs are in the truck?’ ‘Fifty-two.’ ‘What?’ ‘The number of chairs in the truck is fifty-two!’ Presumably that last sentence expresses a truth in the context in which it was uttered although it expresses a falsehood in van Inwagen’s ontology room. Finally, note that there could still be an interesting question about whether there are metaphysically privileged contexts – those that we ought to place ourselves in if we wish to describe ‘the structure of reality’. Perhaps Hirsch and Putnam will say there are none; perhaps Sider 2009 (and van Inwagen?) could say that there is exactly one; and perhaps a friend of modes of being could say that there is more than one. Provided that we can apply the notion of fundamentality/naturalness/structure to contexts of utterance rather than to the utterances themselves, many meta-ontological debates could be recapitulated.13 Let me wrap things up. It is the nature of these critical reviews that they primarily contain criticism, and this can obscure how much the reviewer admires the book he is reviewing. So let me reiterate that I have barely touched on the metaphysical depths that lie within this excellent book. Suffice it to say, I wholeheartedly recommend that metaphysicians study Existence carefully.14 Footnotes 1 Existence: Essays in Ontology. By Peter van Inwagen. Cambridge University Press, 2014. viii + 262 pp. £50.00 cloth, £18.99 paper. 2 For example, observe that the official statement of thesis 3 on page 60 is ‘Existence is univocal’ but the very first sentence is about philosophers’ views on whether ‘exists’ has different meanings. 3 Perhaps van Inwagen’s claim that Ryle thinks that existence is equivocal is some evidence that all van Inwagen means by ‘existence is equivocal’ is that ‘exists’ is equivocal. 4 In McDaniel (2009, 2010a, b), I have argued for the independence of the semantic question of whether ontological expressions are polysemous from the metaphysical question of whether there are modes of being. I accept the latter while remaining neutral on the former. 5 See Van Inwagen (1990). 6 Let me be clear upfront that I intend to be talking about the meanings of sentences expressed in various contexts, and not what speakers might mean by these sentences. Trying to unambiguously talk about sentence-meaning rather than speaker-meaning has proved to be exhausting, and hopefully will prove to be unnecessary now that I have warned you about my focus. 7 See van Inwagen (11–2, fn. 21). 8 A complex sentence is one composed of more than one meaningful word or phrase. Here, I set aside possible examples of context-sensitive sentences that consist in a single word. An example: ‘There!’. For a recent defense of CS, see Dorr (2014), who calls CS “constituency”. Thanks to Mike Rieppel for discussion. 9 See, e.g. Lewis (1986: 211–13) and Hudson (2001: 105–12). 10 See McDaniel (2009) and Turner (2010) for further reflections. 11 I agree with van Inwagen that the null object does not exist. Part of his case against the null object is that we know that nothing is a common part of Bas van Fraassen and Hilary Putnam (44–45). Friends of constituent ontologies, according to which van Fraassen and Putnam have properties as constituents, might disagree, provided that these properties are universals. Constituent ontologies are the target of the tenth essay in this collection. 12 Another avenue to explore: in light of theses 3 and 4, what van Inwagen should say about mass quantification, which also is not captured by the ordinary count-quantifier of the first-order logic. 13 If we go this route, probably the better thing to say is that fundamentality/naturalness/structure apply to utterances-in-a-context. 14 I thank Ross Cameron, Joshua Spencer, Jason Turner, Janice Dowell, Mike Rieppel and Robbie Williams for helpful comments on earlier drafts. References Dorr C. 2014. Transparency and the Context-Sensitivity of Attitude Reports. In Empty Representations , ed. García-Carpintero M., Genoveva M.. Oxford: Oxford University Press. Hudson H. 2001. A Materialist Metaphysics of the Human Person , Ithaca (NY): Cornell University Press. Lewis D. 1986. On the Plurality of Worlds , Basil: Blackwell Publishing. McDaniel K. 2009. Ways of being. In Metametaphysics: New Essays on the Foundations of Ontology , ed. Chalmers D., Manley D., Wasserman R.. Oxford: Oxford University Press. McDaniel K. 2010a. A return to the analogy of being. Philosophy and Phenomenological Research  81: 688– 717. Google Scholar CrossRef Search ADS   McDaniel K. 2010b. Being and almost nothingness. Nous  44: 628– 49. Google Scholar CrossRef Search ADS   Putnam H. 2004. Ethics Without Ontology . Cambridge (MA): Harvard University Press. Sider T. 2009. Ontological realism. In Metametaphysics: New Essays on the Foundations of Ontology , ed. Chalmers D., Manley D., Wasserman R.. Oxford: Oxford University Press. Turner J. 2010. Ontological pluralism. Journal of Philosophy  107: 5– 34. Google Scholar CrossRef Search ADS   Van Inwagen P. 1990. Material Beings . Ithaca (NY): Cornell University Press. © The Author 2017. Published by Oxford University Press on behalf of The Analysis Trust. All rights reserved. For Permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Analysis Oxford University Press

Existence: Essays in Ontology

Loading next page...
 
/lp/ou_press/existence-essays-in-ontology-sHFwLGbGyF
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of The Analysis Trust. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
0003-2638
eISSN
1467-8284
D.O.I.
10.1093/analys/anx063
Publisher site
See Article on Publisher Site

Abstract

1. Introduction This wonderful collection of most of van Inwagen’s recent essays on topics in fundamental ontology is certainly to be welcomed.1 Many of the essays are focused on articulating and arguing for van Inwagen’s preferred meta-ontology, which he calls neo-Quineanism. In addition to these essays, Existence also contains essays on the eliminability of variables, the status of fictional entities, the nature of ontological categories, the nature of properties, mereology and change over time, the comparative merits of constituent and relational ontologies, a probing discussion of metaphysical issues in the philosophy of mind, as well as an interview with van Inwagen that describes his intellectual development and his assessment of his philosophical contributions. Each essay is written in van Inwagen’s characteristically charming and impeccable style. I am grateful that this collection exists in whatever sense of ‘exists’ or manner of existence is appropriate to apply. As noted above, many of the book's essays defend all or some of the theses that constitute neo-Quineanism. These theses are as follows: (1) Being is not an activity. (2) Being is the same as existence. (3) Existence is univocal. (4) The single sense of being is adequately captured by the existential quantifier of formal logic. The fifth thesis is a conjunction of methodological commitments about how to conduct ontological disputes in light of 1–4. Roughly, and briefly, participants in ontological disputes should examine the ontological implications of whatever they wish to affirm, and these ontological implications should be rendered using the first-order quantifiers and variables. Existence is exceptionally rich, and regrettably I can discuss only a small piece of it. Specifically, I’ll focus on van Inwagen’s defence of thesis 3. In Section 2, I’ll try to clarify what thesis 3 expresses. In Section 3, I’ll discuss how van Inwagen reconciles thesis 3 with the claim that all physical objects are mereological simples or living beings, but nonetheless ordinary people can truly say, ‘There are chairs’. In Section 4, I’ll examine one of van Inwagen’s arguments for thesis 3. In Section 5, I will discuss a version of ‘anti-realism about ontology’ that is not refuted by van Inwagen’s arguments for 3. 2. What does thesis 3 express? In several of the essays in Existence, van Inwagen appears to defend both the claim that the expression ‘exists’ is univocal and the claim that existence is univocal. But while I understand what van Inwagen means when he claims that ‘exists’ is univocal – he means that ‘exists’ has exactly one meaning, that this expression is not semantically ambiguous – I’m not sure what he means by ‘existence is univocal’. What sort of thing does van Inwagen take existence to be? Is it, for example, a property? And is it the sort of thing that by van Inwagen’s own lights can be aptly said to be univocal? If existence is univocal, is ‘univocal’ univocal? Or is existence univocal in a different sense from the sense in which ‘exists’ is univocal? One interpretative possibility is that all van Inwagen means when he utters ‘existence is univocal’ is that ‘exists’ is univocal. On this interpretation, we have two merely stylistic variants of the same semantic thesis. This interpretation is somewhat supported by the casual way in which van Inwagen alternates between saying that ‘exists’ is univocal and existence is univocal.2 However, there is another possible interpretation: when van Inwagen claims that existence is univocal, the opposing view that he targets is the view that there are modes, ways, or kinds of existence. That existence comes in modes, ways, or kinds is both historically important and worthy of critique, and hence it would be unsurprising if it were among van Inwagen’s targets. Moreover, in the context of defending van Inwagen’s preferred meta-ontology, van Inwagen (50–58) criticizes Heidegger’s philosophy of being, which includes the claim that there are modes of being, and Meinong’s distinction between subsistence and existence (60, fn. 16). So it is not implausible that, when van Inwagen utters ‘existence is univocal’, he is denying that there are modes, ways, or kinds of existence. Suppose this interpretation is correct. Then thesis 3 conflates two claims that are not obviously equivalent. Ryle denied that ‘exists’ is univocal but he also did not believe that there are modes or kinds of existence. Contrary to van Inwagen’s claim (63), Ryle does not argue that existence is equivocal; perhaps he would have said that it is a category mistake to say that existence rather than ‘exists’ is equivocal.3 Regardless, Ryle’s focus is on ‘exists’ rather than on existence. Ryle denied that there are modes or kinds of existence partly because he does not believe that there is a general sense of ‘exists’ that applies to anything that the various senses of ‘exists’ apply to. But note also that one could accept the metaphysical claim that there are modes of existence while claiming both that no ordinary language expression has any of these modes as its meaning (or semantic value more generally), and that ‘exists’ has as its sole meaning (or semantic value) the ‘general’ way of existing that everything has, regardless of the kind of thing it is.4 (Here are two analogies to consider. Consider a language that has a univocal predicate for being coloured although it contains no predicates for particular colours; consider a language in which a kind term like ‘jade’ is univocal although jade is itself something like a disjunctive kind.) If thesis 3 does conflate two non-equivalent theses, then they should be disentangled and we should assess which if any of van Inwagen’s targets are refuted by his arguments for thesis 3. In what follows, I will treat 3 as if it only expressed the semantic thesis about ‘exists’. 3. Ontology and meta-ontology Thesis 3 initially seems to conflict with van Inwagen’s own metaphysical practice. When doing metaphysics, van Inwagen denies that there are chairs.5 But van Inwagen also claims that ordinary assertions of ‘there are chairs’ are literally true but compatible with what van Inwagen asserts when he says ‘there are no chairs’.6 The first essay of Existence, titled, ‘Introduction: Inside and Outside the Ontology Room’, defends this claim. If there were two meanings of ‘there is’, the ordinary person’s assertion and van Inwagen’s assertion could both be true. Thesis 3 rules this out. However, even if ‘there is’ is not ambiguous but there are other possible meanings this expression could have had, we could see how the ordinary person’s assertion and van Inwagen’s could both be true: perhaps van Inwagen has adopted a new (possibly better from a metaphysical perspective) meaning for ‘there is’ and is using it when stating his metaphysical theses. But van Inwagen (37–38) denies this as well: not only are ontological expressions such as ‘there is’, ‘exists’, ‘being’, univocal, but also these expressions could not have had different meanings than what they in fact have. I wondered at this point how van Inwagen individuates expressions; if an expression essentially has the meaning it in fact has, ontological expressions could not have had different meanings. (If expressions don’t essentially have their actual meanings, why couldn’t ‘there is’ have had a different meaning? The answer seems to be that there is no alternative possible meaning; more on this soon.) But nonetheless we could still reconcile the ordinary person’s assertion with van Inwagen’s assertion: perhaps contrary to appearances, only the ordinary person uses ‘there is’, while van Inwagen uses an expression that looks, sounds, and functions syntactically and inferentially like the ordinary ‘there is’ while differing in meaning. (But, as before, perhaps the meaning van Inwagen has adopted is a metaphysically better meaning to adopt.) Provided, of course, that there is another meaning possible for ‘there is’, which van Inwagen denies. We will return to his reason for this denial in Section 4. Whether ontological disputes should be conducted in ordinary language or in a technical language has been the subject of much recent debate. For example, Sider (2009) considers framing ontological disputes in what he calls ‘Ontologese’; the quantifier-expression used in Ontologese is stipulated as having the most metaphysically fundamental meaning an expression of this sort can have. Van Inwagen (1) also suggests the adoption of a technical language for conducting ontological disputes that he calls ‘Tarskian’. In Tarskian, sentences like ‘chairs exist’ have the logical form taught to us by first-order logic. But Tarskian does not contain a quantifier expression that differs in meaning from ‘exists’ in ordinary English, and so in this respect Tarskian is not Ontologese. When do we speak Tarskian? Van Inwagen (3–5) distinguishes two contexts in which one might utter the sentence ‘there are chairs’. One of these contexts is the context of ordinary life; another of these contexts van Inwagen dubs ‘the ontology room’. ‘There are chairs’ expresses one proposition – a true one – in the context of ordinary life, but it expresses a different proposition – one that is false – in the ontology room. I take it that van Inwagen’s position is that ‘there are chairs’ is a univocal sentence, provided that ‘chairs’ is univocal, yet the sentence as a whole is context sensitive. The key thing is that somehow this shift in context brings about a shift in the logical form of proposition expressed: in the ontology room, ‘chairs exist’ wears its logical form on its face, but the form it has in the context of ordinary life is not the form of ‘There is an x such that x is a chair’.7 I don’t see how the context of an utterance could affect the logical form of a sentence in the way van Inwagen needs. I do accept that what sentence has been expressed in a given context is sensitive to features of that context. Van Inwagen (7, fn. 12) mentions the string of English words, ‘Landing planes can be dangerous’, and I accept a position (that he also mentions) that this string corresponds to two sentences, one of which contains a phrase denoting an action-type (‘landing planes’) while the other contains a phrase equivalent to ‘planes that are landing’. Context can settle which sentence is used in this context; but once context does this, it doesn’t further settle the logical form of the sentence expressed. But I doubt that the string of English words ‘there are chairs’ corresponds to two distinct sentences. Moreover, I suspect van Inwagen’s position conflicts with what seems to me a plausible principle about context-sensitivity: CS: If a complex sentence is context sensitive, it has some context-sensitive proper part that explains or grounds the context sensitivity of the sentence.8 If CS is correct, then van Inwagen should tell us which part of existence avowals or denials determines the context sensitivity of the avowal or denial. Is it the quantifier expression or the remaining material that is adjoined to it? It is widely accepted that ‘there is’, ‘some’, and other locutions like these are context-sensitive expressions. But the standard way of understanding how they are context sensitive does not fit well with van Inwagen’s picture. On the standard picture, these expressions correspond to domains of entities, and they are context sensitive in that a given expression corresponds to a different domain in different contexts. In the context of a party, I can truthfully say, ‘There aren’t any bottles of beer except what are in the cooler outside’ even though there are many bottles of beer all over the world. In the context of the party, we are talking about fewer things than in other contexts. This familiar phenomenon has been noted and exploited by ontologists with views opposed to van Inwagen’s. The mereological universalist thinks that, whenever there are some things, there is a whole composed of those things. This profligate ontology seems contrary to common sense. Many mereological universalists have attempted to reconcile their affirmation of sentences like ‘There is a whole made out of Kris and the moon’ with the ordinary person’s affirmation of ‘There is nothing made out of Kris and the moon’. According to the standard story, what the universalist says in one context is true but so is said in ordinary contexts, because the domain of quantification invoked in the latter are proper subdomains of the domain of quantification invoked in the former context, and these subdomains do not contain the fusion of Kris and the moon.9 So both the mereological universalist and van Inwagen distinguish between a context in which ontological assertions are made and ordinary contexts, and by doing so, both the mereological universalist and van Inwagen hope to reconcile their ontological claims with what initially appeared to be common-sensical denials of them. But prima facie the universalist is in a better position than van Inwagen. First, the mereological universalist explicitly identifies the context sensitivity of the quantifier expressions as the source of the context sensitivity of sentences that avowal existence and thereby respect CS, whereas van Inwagen is (as far as I can tell) silent on the source of the context-sensitivity. Second, the mereological universalist appeals only to a well-understood phenomenon, the positing of which is antecedently motivated by considerations that have little to do with fundamental metaphysical disputes, whereas we do not yet have a theory of how it is that avowals of existence could be context-sensitive in the way that van Inwagen claims, and the positing of this context-sensitivity seems motivated largely by the need to reconcile van Inwagen’s metaphysics with the assertions made by ordinary people. (Obviously, van Inwagen cannot appeal to the familiar fact that we often quantify restrictedly, since, unlike the mereological universalist, van Inwagen believes in far fewer kinds of things than the ordinary person prima facie appears to believe in.) Van Inwagen has offered a theory of how positive existential sentences uttered in ordinary contexts might be true, even given his ontology. But what about negative existential sentences? Consider the sentence ‘Nothing is made of that cat and that dog’ uttered in an ordinary context. In that context, the sentence expresses a truth. As noted earlier, the universalist explains how it expresses a truth: the quantifier is restricted. What is van Inwagen’s explanation? Consider a log cabin. According to van Inwagen (11–14), in an ordinary context the sentence ‘there is log cabin made out of those logs’ expresses a proposition that is necessarily equivalent to (but not identical with) the proposition expressed in the ontology room by ‘there are simples arranged log-cabin-wise’. By parity of reasoning, in an ordinary context the sentence ‘nothing is made of that cat and that dog’ should express a proposition that is necessarily equivalent with the proposition expressed in the ontology room by ‘no simples are arranged that-cat-and-that-dog-wise’. But that latter proposition is false, since there are simples in just that arrangement. So the ordinary disavowal should express a false proposition as well. Why focus on ordinary common-sensical utterances, rather than utterances made in the context of scientific inquiry? Suppose an astrophysicist says, ‘Some celestial object is causing this strange pattern of light’. Why think that this sentence doesn’t have the logical form it seems to have? The fact that the object causing this pattern is unlikely to be alive or a mereological simple doesn’t provide a reason to think this. Ordinary language didn’t develop for the purposes of theorizing about the physical world, but scientifically regimented languages did. In short, I don’t see why only the metaphysician speaks in Tarskian. The scientist who cares about precisely describing physical reality would be well-served by adopting a language with a relatively transparent logical structure. Why think that she hasn’t but the metaphysician has? But, if the scientist is speaking in Tarskian in these contexts, what she says is incompatible with van Inwagen’s preferred ontology – and this strikes me as a worse consequence than merely flying in the face of common sense about what there is. I remain doubtful whether a sentence like ‘Ted has more mass than Jason’ has a particularly complex logical form. There’s some funny business about how to handle the time at which Ted is more massive than Jason, but that’s about it. But if ‘Ted’ and ‘Jason’ refer to my car and my guitar respectively, the sentence is false in van Inwagen’s ontology room although true in the context of ordinary life. Are names context-sensitive? Is ‘has more mass than’ context-sensitive? It seems to me implausible to claim this, but it also seems to me that, given CS, van Inwagen might need to. 4. Van Inwagen’s arguments for thesis 3 Van Inwagen offers two arguments for thesis 3. The first is van Inwagen’s (41–43 and 61–63) argument from number. Roughly, it is this. Existence is closely related to number: ‘Fs exists’ is true if and only if the number of Fs is greater than 0. Numerals, however, are univocal, as is ‘the number of’. These expressions are univocal if and only if ontological expressions are univocal. So ontological expressions are univocal. Elsewhere, I’ve examined whether this argument shows that there are no modes of being.10 And we’ll further discuss it in Section 5. Accordingly, I’ll focus on the second argument here. Let’s call this second argument the argument against expansion; it appears on pages 69–70. The target of this argument is Putnam’s (2004) claim that, although (perhaps) we cannot say that arbitrary fusions of things exist in the ordinary sense of ‘exist’, we can introduce what he calls a ‘conventionally extended’ sense of ‘exist’ on which we can. If there are alternative possible meanings for ‘exist’, there are the sort of extended meanings that Putnam mentions; but if there no such extended meanings, there are no alternatives to what we actually mean with our ontological expressions. Van Inwagen (69–70) offers the following straightforward argument: in order to introduce an extended sense of ‘there is’ to range over new entities (such as arbitrary fusions) one must antecedently believe that there are these entities. But, if one antecedently believes that there are these entities, one is already quantifying over them, and hence the putatively extended meaning for the quantifier is simply the original meaning the quantifier already had. As van Inwagen (69–70) nicely puts it, ‘Extending the meaning of a term so that the term will apply to objects beyond those it already applies to is precisely analogous to extending a geographical boundary: you can extend a geographical boundary to encompass new territory only if that territory is already there.’ I will offer two responses to the argument against expansion. The first response is that Putnam could concede that one can redraw ‘ontological boundaries’ only against a pre-existing belief that there are things for the extended meaning to apply to, but deny that this belief must be in the ‘objects that would be in the domain of the new quantifier’. Let me explain. Suppose for the sake of illustration we take the semantic value of the existential quantifier to be a higher-order property of properties; call this property ‘E’. Consider now the property of being a cat, the property of being a dog, and the property of being the fusion of a cat and a dog. The property of being a cat and the property of being a dog instantiate E, but the property of being the fusion of a cat and a dog does not. But there is another property, E*, such that any property that instantiates E instantiates E* but some properties, among them being the fusion of a cat and a dog, instantiate E* without instantiating E. E* is the putatively possible extended meaning for our ontological expressions. In order to invoke E* rather than E, perhaps one must have an antecedent belief in something like properties, but one needn’t have an antecedent belief in things that are fusions of cats and dogs. The second response is that no antecedent belief in objects of any kind is necessary for adopting an alternative meaning for ontological expressions. Rather, what is necessary is a change in the pattern of use of ontological expressions, and this change in use might be unnoticed by members of the linguistic community in question. The pattern of use of an expression consists in sincere and literal utterances of sentences containing that expression, as well as dispositions to sincerely and literally utter sentences that contain it. One of the factors that determines the semantic values of our expression is their pattern of use; if the pattern of use changes significantly, the semantic value of that expression might change as well. Van Inwagen (9) says of himself that he is enough of a Wittgensteinian to think that it is not possible for very much of what we ordinarily say to turn out false. But what if the folk change what they ordinarily say? What if sentences like ‘there are wholes made of cats of dogs’ were sincerely, literally, and frequently used in a wide range of ordinary situations? I suspect that, if use were to change in this way, the proposition expressed by ‘there are wholes made of cats and dogs’ would not be the one this sentence currently expresses. I suspect that ontological expressions such as ‘there are’ would acquire a subtly different meaning, and this is why these sentences would express different propositions. Perhaps there is another explanation as to how these sentences would come to express different propositions in those circumstances; perhaps a systematic change in relevant features of their contexts of utterance somehow brings this about. I don’t know what this other explanation looks like, but I also don’t know how it works on van Inwagen’s own account. It seems to me that the argument against conventional expansion doesn’t succeed. Van Inwagen (42–44) offers a second reason in support of the argument against expansion. Consider the null individual, that putative entity that is a part of everything. There is no such entity, but if we can introduce ‘conventionally extended’ meanings for ontological expressions, there should be a possible meaning for ‘there is’ such that, when ‘there is’ has that meaning, ‘there is a null object’ expresses a truth.11 Van Inwagen thinks that there is no such meaning, and hence anyone who tried to speak a language like this would not even be playing with words, but rather with meaningless sounds, marks, or scratches. I don’t think van Inwagen is clearly right here, but I’ll concede that he is and investigate whether there is a principled way to accept some ‘conventionally extended meanings’ and reject others. A key idea of philosophers like Hirsch is that putatively different ontological theories are just different ways of expressing the same underlying facts about concrete reality. Perhaps the underlying facts about concrete reality that all adequate languages must ‘agree on’ are facts about the properties and relations of the mereologically simple entities. The null individual, were it to exist, would be a mereological simple. Accordingly, a language in which ‘there is a null object’ expresses a truth is not adequate, unlike, e.g., a language in which ‘every collection of simples has a fusion’ expresses a truth. 5. Contextualist anti-realism about ontology In this final section, I’ll briefly sketch a version of anti-realism about ontology that both captures the spirit of Putnam’s and Hirsch’s views but accepts thesis 3 and but accepts thesis 3. The key move is that, just as van Inwagen claims, sentences containing ontological expressions are context-sensitive. It seems to me that what is central to the spirit of Putnam’s and Hirsch’s views is that there is no unique context that is the context to use existential sentences in order to express ‘the ontological facts’. In short, there is no context that is the ontology room. But there might nonetheless be many ontology rooms – that is, many contexts in which one can use existential sentences to express true propositions that they would not ordinarily express. Putnam and Hirsch might cheerfully grant that there is the context that van Inwagen calls ‘the ontology room’, although the definite description is inapt, for there are also contexts in which ‘for any objects, there is a whole composed of those objects’ expresses a truth. The proponent of this view makes use of the same machinery van Inwagen makes use of; she traffics in other senses of ‘exists’ to no greater extent than van Inwagen himself. Moreover, this version of ‘anti-realism about ontology’ could embrace a model of why ontological expressions are context-sensitive: they are context-sensitive because they are sensitive to standards of application. Sensitivity to standards of application is a familiar phenomenon. Consider ‘tall’. ‘Hugh Jackman is tall’ expresses a truth in some contexts of utterance because the standards governing ‘tall’ in that context are fairly ordinary, but in the context of discussing basketball players, ‘Hugh Jackman is tall’ expresses a falsehood. The spirt of ‘anti-realism about ontology’ is that there are different standards that one could have for what it takes to be an object. On particularly lax standards, the goofy objects countenanced by the unrestricted mereologist have what (little) it takes, but on much stronger standards, we cannot even truthfully say ‘tables exist’. Note also that this version of ‘anti-realism about ontology’ has the following advantage over van Inwagen’s view: it appeals to a familiar phenomenon that is already known to be widespread and merely finds a further application for it. Note also, though, that one could embrace this kind of contextualist anti-realism without offering this particular mechanism for why ontological expressions are context sensitive. Can one accept this explanation for the context-sensitivity of ontological expressions and still accept thesis 4? Probably not. Thesis 4 tells us that the meaning of the existential quantifier is exhausted by the syntactic and inferential rules governing that expression; these are the things you grasp when you study a textbook of logic. But if ontological expressions are sensitive to standards, this fact is something you wouldn’t learn simply from a textbook on logic. So it would seem that there is more to the meaning of ontological expressions than what is represented by the existential quantifier? On this view, what is the relationship between ordinary existential expressions and the existential quantifier? I think not like the relation between ordinary conditionals and the material conditional; the latter is an expression with less content but nonetheless with sufficient content to be a meaningful expression. But if you like the view I have been sketching, you should regard the existential quantifier as a contentless symbolic device, the mastery of which aids you in understanding the structural features of genuinely meaningful ontological expressions. Van Inwagen’s arguments for thesis 3 are presented independently of the considerations in favour of the theses that succeed them, but perhaps what we have learned is that there are deeper connections between theses 3 and 4 than we antecedently thought.12 How fares this version of anti-realism against van Inwagen’s argument from numbers? Perhaps this argument works against Putnam and Hirsch’s views as they state them. But it seems less successful against the version of ‘anti-realism about ontology’ I’ve articulated. On this view, sentences containing numerals are also context sensitive, as they also seem to be according to van Inwagen. Imagine the following exchange at the physical plant at Western Washington University: ‘We need the exact count of chairs for the next shipment, so no screw-ups this time. How many chairs are in the truck?’ ‘Fifty-two.’ ‘What?’ ‘The number of chairs in the truck is fifty-two!’ Presumably that last sentence expresses a truth in the context in which it was uttered although it expresses a falsehood in van Inwagen’s ontology room. Finally, note that there could still be an interesting question about whether there are metaphysically privileged contexts – those that we ought to place ourselves in if we wish to describe ‘the structure of reality’. Perhaps Hirsch and Putnam will say there are none; perhaps Sider 2009 (and van Inwagen?) could say that there is exactly one; and perhaps a friend of modes of being could say that there is more than one. Provided that we can apply the notion of fundamentality/naturalness/structure to contexts of utterance rather than to the utterances themselves, many meta-ontological debates could be recapitulated.13 Let me wrap things up. It is the nature of these critical reviews that they primarily contain criticism, and this can obscure how much the reviewer admires the book he is reviewing. So let me reiterate that I have barely touched on the metaphysical depths that lie within this excellent book. Suffice it to say, I wholeheartedly recommend that metaphysicians study Existence carefully.14 Footnotes 1 Existence: Essays in Ontology. By Peter van Inwagen. Cambridge University Press, 2014. viii + 262 pp. £50.00 cloth, £18.99 paper. 2 For example, observe that the official statement of thesis 3 on page 60 is ‘Existence is univocal’ but the very first sentence is about philosophers’ views on whether ‘exists’ has different meanings. 3 Perhaps van Inwagen’s claim that Ryle thinks that existence is equivocal is some evidence that all van Inwagen means by ‘existence is equivocal’ is that ‘exists’ is equivocal. 4 In McDaniel (2009, 2010a, b), I have argued for the independence of the semantic question of whether ontological expressions are polysemous from the metaphysical question of whether there are modes of being. I accept the latter while remaining neutral on the former. 5 See Van Inwagen (1990). 6 Let me be clear upfront that I intend to be talking about the meanings of sentences expressed in various contexts, and not what speakers might mean by these sentences. Trying to unambiguously talk about sentence-meaning rather than speaker-meaning has proved to be exhausting, and hopefully will prove to be unnecessary now that I have warned you about my focus. 7 See van Inwagen (11–2, fn. 21). 8 A complex sentence is one composed of more than one meaningful word or phrase. Here, I set aside possible examples of context-sensitive sentences that consist in a single word. An example: ‘There!’. For a recent defense of CS, see Dorr (2014), who calls CS “constituency”. Thanks to Mike Rieppel for discussion. 9 See, e.g. Lewis (1986: 211–13) and Hudson (2001: 105–12). 10 See McDaniel (2009) and Turner (2010) for further reflections. 11 I agree with van Inwagen that the null object does not exist. Part of his case against the null object is that we know that nothing is a common part of Bas van Fraassen and Hilary Putnam (44–45). Friends of constituent ontologies, according to which van Fraassen and Putnam have properties as constituents, might disagree, provided that these properties are universals. Constituent ontologies are the target of the tenth essay in this collection. 12 Another avenue to explore: in light of theses 3 and 4, what van Inwagen should say about mass quantification, which also is not captured by the ordinary count-quantifier of the first-order logic. 13 If we go this route, probably the better thing to say is that fundamentality/naturalness/structure apply to utterances-in-a-context. 14 I thank Ross Cameron, Joshua Spencer, Jason Turner, Janice Dowell, Mike Rieppel and Robbie Williams for helpful comments on earlier drafts. References Dorr C. 2014. Transparency and the Context-Sensitivity of Attitude Reports. In Empty Representations , ed. García-Carpintero M., Genoveva M.. Oxford: Oxford University Press. Hudson H. 2001. A Materialist Metaphysics of the Human Person , Ithaca (NY): Cornell University Press. Lewis D. 1986. On the Plurality of Worlds , Basil: Blackwell Publishing. McDaniel K. 2009. Ways of being. In Metametaphysics: New Essays on the Foundations of Ontology , ed. Chalmers D., Manley D., Wasserman R.. Oxford: Oxford University Press. McDaniel K. 2010a. A return to the analogy of being. Philosophy and Phenomenological Research  81: 688– 717. Google Scholar CrossRef Search ADS   McDaniel K. 2010b. Being and almost nothingness. Nous  44: 628– 49. Google Scholar CrossRef Search ADS   Putnam H. 2004. Ethics Without Ontology . Cambridge (MA): Harvard University Press. Sider T. 2009. Ontological realism. In Metametaphysics: New Essays on the Foundations of Ontology , ed. Chalmers D., Manley D., Wasserman R.. Oxford: Oxford University Press. Turner J. 2010. Ontological pluralism. Journal of Philosophy  107: 5– 34. Google Scholar CrossRef Search ADS   Van Inwagen P. 1990. Material Beings . Ithaca (NY): Cornell University Press. © The Author 2017. Published by Oxford University Press on behalf of The Analysis Trust. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Journal

AnalysisOxford University Press

Published: Jan 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off