Peter van Inwagen has been influential in metaphysics for many years, and Existence: Essays in Ontology contains a range of his recent metaphysical papers sharing the connection of arguing for interesting and provocative ontological claims. All but two of the essays (the introductory essay and ‘Alston on Ontological Commitment’) are previously published, but their collection into one volume is justified by their collectively displaying van Inwagen's sophisticated approach to ontology. A collection of thirteen essays covering, amongst other things, metaontology, fictional entities, properties, categories, and mental causation (although notably not his notorious ‘organicist’ ontologyFootnote ¹ ) cannot be summarised piecemeal in this space, but fortunately van Inwagen's unifying approach affords the opportunity to bring the themes together. Thus I will reconstruct as much as possible of van Inwagen's position around the core principles informing the book.

Van Inwagen self-identifies as a Quinean, following many since the mid-20^th century. He differs from most, though, in giving a detailed account of what he takes that to mean. ‘Being, Existence, and Ontological Commitment’ articulates and defends five theses jointly comprising a supposedly Quinean metaontology:

1. (1) ‘Being is not an activity’ (54).

2. (2) ‘Being is the same as existence’ (58).

3. (3) ‘Existence is univocal’ (60).

4. (4) ‘The single sense of being or existence is adequately captured by the existential quantifier of formal logic’ (71).

5. (5) The Quinean criterion of ontological commitment (79-86).

These theses continually inform van Inwagen's approach. (2) and (3) are particularly important in ‘Existence, Ontological Commitment, and Fictional Entities’, where they guide his account of fictional entities. After finding that repudiating fictional entities would be unattractive van Inwagen notes the persuasive thought that fictional entities might exist in a special sense, or something-less-than-exist. This Meinongian move falters because of his conviction that there is no such separate kind of existence, leading him to a theory on which fictional characters exist but their apparent ‘apparatus of predication’ (106) is ambiguous between two distinct relations: having and holding. Having is the ordinary relation between an entity and its properties, while holding is an unclear relation between a fictional character and what others might describe as the properties that it has according to the fiction. Van Inwagen concedes an inability to define ‘holding’, but insists that it ‘makes sense if fictional discourse makes sense’ (108).

(4) plays a role in ‘Alston on Ontological Commitment’, which engages with William Alston's well-known critique of thesis (5). On van Inwagen's treatment the critique applies the paradox of analysis to what Alston calls ‘ontological reductions’. Briefly, supposing an ontological reduction from

to

we have an apparent problem. For Alston, the first sentence is obviously existential, so if the second sentence is an adequate translation then it commits us no less than the first to holes' existence. On the other hand, if it isn't an adequate translation, it can't help us say what we wanted to say so we achieve nothing. After indicating some technical problems in Alston's formulation of the problem van Inwagen goes on to defend the Quinean criterion. The second statement does not commit us to holes by virtue of its relation to the first: the pressing question is whether the obvious translation of the first or the second into quantifier-variable form is appropriate. If we (metaphysicians) endorse the obvious translation of the first, we are committed to holes if we accept the first statement. If the obvious translation of the first is inappropriate, the second statement may provide a clearer account of what one is committed to by virtue of uttering the first statement. One seeks such paraphrases not to eliminate ontological commitments but to ‘remove merely apparent ontological commitments’ (148), a project which does not aspire to the sentences' synonymy (148).

Important to determining appropriateness for van Inwagen is the distinction explained in the introductory essay between propositions expressed by sentences inside and outside ‘the ontology room’ (1), which he considers a context of utterance separating metaphysicians from ordinary-language considerations. While ‘outside propositions’ are the folk's stock-in-trade, ‘only metaphysicians . . . have ever considered – ever entertained, ever grasped, ever held before their minds – the inside proposition’ (6). This aims to insulate metaphysicians from the accusation of failure to respect language-use without exposing them to the claim that they are disputing issues of mere choice: these two attacks (respectively van Fraassen's and Putnam's) are answered using essentially the ontology room machinery in ‘The New Antimetaphysicians’.

(5), the Quinean criterion of ontological commitment, plays an absolutely central role in the collection. The difficulty of avoiding quantification over fictional characters drives the account in ‘Existence, Ontological Commitment, and Fictional Entities’, the pitfalls of attempting to eliminate quantification over abstract objects lead van Inwagen to begrudgingly accept the existence of properties in ‘A Theory of Properties’, and the conviction that causal relations and explanations are sufficient to accommodate useful causal talk lies behind his (rather quick) dismissal of events in ‘Causation and the Mental’. In each case the strategy is fairly stable: select a subject matter, examine the kind of talk required by that subject matter, test translations with different apparent commitments, and choose that which best satisfies criteria like parsimony, generality, and plausibility.

Each paper proceeds carefully through its subject matter, often reaching surprising results from a few simple claims – with the possible exception of ‘Causation and the Mental’, which is hard to follow. However problems arise most readily when the reader tries, as is natural, to reconcile claims made between chapters. The only paper expressly written for this volume is the introduction (‘Alston on Ontological Commitment’, though previously unpublished, is a revision of a symposium paper given in Alston's memory), which clarifies some of the work, but aside from this there is only the minimal cohesion imparted by phrases of the type ⌜ see chapter n ⌝. While expected, this only postpones van Inwagen's responsibility to answer pressing questions about how his claims sit together as a whole.

One possible tension is caused by the principle that ‘The following is true of everything: for every property, it has either that property or its negation.’ (151) This is an assumption in several arguments in the book. However, when van Inwagen compares his theory of fictional characters to those of Amie Thomasson and Nicholas Wolterstorff an advantage cited for his own position is that, being less specific, it commits him to answering fewer awkward questions. One might worry that by van Inwagen's own standards, this is anything but an advantage: as he says elsewhere, ‘if what you have said about X leaves it an open question whether certain specifiable members of X have the property F, you have not said enough about X’ (216). For a specific example of saying too little, van Inwagen claims only that fictional characters belong within the category abstract object (108), declining to state whether they constitute a category of some level or are a class within a more inclusive category.

Another significant tension arises, this time concerning van Inwagen's overall approach, between ‘Being, Existence, and Ontological Commitment’ and ‘What is an Ontological Category?’. He describes the metaontology summarised above as ‘essentially Quine's’ (53), itself a hedging of the claim in the chapter's forerunner that he ‘would be willing to call it an exposition of Quine's metaontology’,Footnote ² but certain divisions introduced in ‘What is an Ontological Category?’ encourage the question: how Quinean is van Inwagen's position? Therein he identifies three conceptions of ontology implicit within the analytical tradition:

• Bergmann's conception: ‘ontology is the study of the ontological structure of objects’ (183).

• The bare Quinean conception: ‘it is the business of the practitioners of ontology to produce and defend answers to [the question “What is there?” that] strike them and their peers as “general” and “abstract” and “systematic” ’ (184).

• The categorical conception: ‘it is the business of ontology to [answer the question “What is there?”] in terms of a specification of the ontological categories’ (185).

Van Inwagen labels the second the ‘bare Quinean’ conception because he thinks Quine would find it an ‘attractive formulation of the nature of ontology’ (185, n. 4) though he stops short of directly branding anyone (including Quine) as a bare Quinean. Van Inwagen is not a bare Quinean, however: he subscribes to what I have labelled the ‘categorical conception’. For van Inwagen it is definitional of an ontological category that it be a natural class (‘natural’ in a primitive sense, ‘class’ in the sense of class-as-many) that comprises a substantial chunk of what there is (193–5). By van Inwagen's definition of metaontology, as comprising the questions ‘What does “there is” mean?’ and ‘What methods should be employed in the investigation of what there is?’, it is unclear whether the categorical conception counts as part of van Inwagen's metaontology: he seems to think not (200), while the wider metaontological literature suggests the contrary. Now that metaontology has transcended the boundaries set by van Inwagen's coinage of the term,Footnote ³ I am inclined to say that the masses overrule the founder, meaning that in at least one significant respect – its project of specifying objective ontological categories – van Inwagen's metaontology looks un-Quinean. Is this a merely classificatory matter? It may turn out to be so. I think, however, that it encourages deep concern about whether the five theses of the Quinean metaontology are suited to uncovering the supposed ‘non-conventional’ (186) distinctions required by the categorical conception.

A final worry, to which I can only gesture here, regards van Inwagen's repudiation of inference to the best explanation in metaphysics. In ‘Relational vs. Constituent Ontologies’, the reliance of the latter kind – ontologies on which objects have ‘ontological structure’ (204) – on this method is one of van Inwagen's reasons to reject it, as he accepts van Fraassen's critique of IBE, but he claims that relational ontologies (on which objects have only mereological structure) make no such appeal: ‘I should never want to say that the fact that greenness was a property of both the apple and the book explained the fact that they were both green’ (215). He claims instead that we are committed to properties as a consequence of our beliefs. But for anyone engaged in an interpretative project (a prerequisite for translation), especially considering the aforementioned distinction between inside and outside propositions, it is hard to see how appeal to IBE can be avoided across the board. After all, it is key to van Inwagen's position that the truth of ‘When a cold poker becomes hot, that is a change’ does not entail that there are changes – changes being, for him, events (245) – because he does not believe that such things exist. In that case, an account of causation like that sketched in ‘Causation and the Mental’ looks as if it ought to explain how those beliefs of van Inwagen's are consistent with one another. If not, how are interpretations of causal talk restricted? Van Inwagen exhorts the ontologist to ‘tie all your terms of art to ordinary language by some sort of thread that can be followed’ (217) but without the requirement to select the best available explanation of our discourse, this is no constraint. Unfortunately the intended scope of this rule of thumb is left unclear, but it is unlikely that he intends it to be so loose as to require only that an account of Fs be an account of Fs.

As indicated earlier, Existence is a collection of essays brought together from their original, disparate, contexts, which makes the drawing of these tensions far easier. The essays were not written to stand together, so explanations are missing that the author might have felt obliged to give in a monograph. Such difficulties, alongside the excessive boldness of some of his principles, are likely to be the recourse for those who, like myself, are not converted to van Inwagen's ‘rather extreme ideas about ontology’ (238). However, to borrow van Inwagen's reaction to Quine's attempt to defend nominalism, ‘[i]ts value lies in its contributions to metaontology, not in its contributions to ontology’ (136). Van Inwagen's careful explorations of various aspects of ontology serve to demonstrate more clearly the theoretical underpinnings of a certain species of ontological realism, and this is useful for both its defenders and its critics.