Review of Oppy (2018)

A shadow of criticism has followed ontological arguments for almost a thousand years and counting. Irrespectively, the arguments continue to intrigue philosophical thought, and no decline is in sight. In particular, modal versions of the argument formulated by Hartshorne, Lewis, Plantinga and Gödel in the 1960’s and 70’s have helped to dispel the widely held suspicion that a simple logical blunder lies behind ontological arguments. As a result, recent discussions have shifted from assessing the argument’s validity towards its soundness and dialectical efficacy. This requires engaging with the philosophical issues inevitably raised by the argument, such as questions about the nature of concepts and arguments, existence and possibility. These have since stood at the forefront of the debate.

The concerns united by reference to ontological arguments form the subject matter of a recently published volume edited by Graham Oppy, himself one of the most prolific authors on the topic in the past 25 years. His informative introductory essay underlines important differences between the arguments commonly called “ontological.” Oppy suggests abandoning the search for unity suggested by the description “the ontological argument.” Instead, the commonalities should be viewed genealogically: “What is distinctive of ontological arguments is that their formulation has the right kind of connection to Anselm’s argument” (p. 11). Hence, fruitful engagement with and criticism of ontological arguments proceeds by cases.

This sets the tone for the volume’s first group of articles which are devoted to defenders and critics of the ontological argument, namely: Anselm, Aquinas, Descartes, Leibniz, Kant, Hegel, Gödel, Lewis, Plantinga, and Tichý. A second group of three essays dealing with overarching systematic issues surrounding the preceding arguments complements the volume. Here we find treatments of the relation between conceivability and possibility, the “fallacy” of begging the question, and the relation between existence, characterization and modality. Overall, the volume provides readers with informative up-to-date discussions of ontological arguments of scholarly value by senior researchers in the field. (With the notable exception of M. Inwood’s article on Hegel which lacks engagement with the literature on the subject.) At the same time, the essays are written in an accessible manner, rendering the volume suitable as an accompaniment to graduate-level courses on the subject. Due to limitations in space, I will refrain from summarizing and discussing all the contributions. For that purpose, Oppy’s introduction (pp. 2–5) is well suited. Instead, I will focus on three contributions I found particularly worth discussing.

The majority of ontological arguments treated in the volume—Anselm’s, Leibniz’s, Gödel’s, Plantinga’s—are shown to be deductive in nature by their interpreters. A noticeable rift opens up between them and Descartes’ argument, according to Lawrence Nolan’s interpretation. His article represents an important scholarly contribution because it virtually reverses the standard deductive reading, and plausibly so.¹ Developing a suggestion hinted at by M. Gueroult and J. Barnes, Nolan interprets Descartes’ so-called ontological argument as “the report of an intuition in the sense of a non-discursive, self-validating, intellectual apprehension” (p. 54). The aims Descartes pursues with the argument are persuasive rather than argumentative: all he points to serves the purpose of getting the meditator to have the relevant intuitive insight.

A strength of Nolan’s reading is that it allows us to make good sense of passages (e.g. Med. V, AT VII 68–69 and Princ. I., §15), where Descartes clearly glosses the cognition of God as an intuitive insight; these have always been difficult to accommodate within deductive interpretations of the argument. Moreover, Nolan convincingly shows that his reading coheres with Descartes’ skepticism towards a formal-deductive understanding of reasoning voiced in the Rules as well as the other philosophical doctrines he adheres to (pp. 57–65). However, the intuitive reading of the argument has to confront the following difficulty: what about the passages where Descartes overtly argues in a deductive manner?

Nolan uses two principal interpretive moves to provide a coherent picture in these cases (pp. 54, 66–71). First, he convincingly shows that the overtly argumentative passages, commonly taken to be Descartes’ argument, are best read as rebuttals of possible criticisms. They allow the meditator’s intuition not to be distracted by an unjustified conception, e.g. by understanding the distinction between essence and existence as a real rather than merely rational distinction. Second, he argues that, for historical reasons, Descartes aimed to present his philosophy in a manner adjusted to the scholarly discourse of his day, which put great emphasis on the syllogistic demonstrability of God’s existence.

While the latter may be correct as a matter of historical fact, Nolan’s line of interpretation may be bolstered by a more penetrating understanding of the relation between intuition and deduction. It is Descartes’ view that deduction is necessary in case one does not have clear and distinct, intuitive insight at one’s disposal (p. 61). Although this legitimizes ascribing priority to intuitive over deductive insight, it does not imply a merely historical explanation of the occurrence of deduction. Rather, one might—in line with Descartes’ general manner of proceeding in the Meditationes—explain the deductive arguments as necessary steps towards intuitive insights. It helps to take into consideration what the condition for distinctly perceiving a given content is: being able to tell it apart from others (Princ. I., §45). If that is the case, then the arguments delivered to fend off criticisms are not merely negative or persuasive, but positively contribute to the distinctness of the meditator’s perception and thus to its intuitiveness. This should not be understood as a criticism, but as additional support for Nolan’s reading. In my view, Nolan’s essay represents a lasting contribution to our understanding of Descartes. Moreover, it points to a version of the ontological argument that might merit systematic development in the light of recent advancements in the epistemology of intuition.²

Other than authors who defend the ontological argument, the volume features some of its most important critics in Aquinas, Kant, and Lewis. Among these, L. Pasternack’s perceptive and well-informed article on Kant is one of the best discussions currently available. It sets the record straight on the nature of Kant’s case against the ontological argument. Contrary to popular wisdom, the latter extends well beyond the familiar line “existence is not a real predicate”. Pasternack distinguishes two main strands of criticism within Kant’s argumentation in the Critique of Pure Reason, the first of which targets an analytic, the second a synthetic reading of the judgment “God exists” (p. 102). Kant argues that an ontological argument insisting on the analytic reading of the statement is dialectically flawed, i.e. does not add up to an argument at all (pp. 104, 106)³, while a synthetic reading rests on the thesis that existence is a “real predicate” which Kant disputes (pp. 106–115).

The soundness of the second part of Kant’s criticism rests on an argument against existence being a real predicate, which Pasternack deems inconclusive. His rendering is as follows: if existence were a property of objects, then a concept specifying all the properties of the object, but lacking existence as a mark, would fail to fully articulate the object in question, leading to a “mismatch” between concepts and their objects. This argument is unconvincing for two related reasons. First, it leaves open why this mismatch should be deemed problematic according to Kant, which needs to be established for the argument to be sound. Pasternack appears to agree on this point, which leads to the second weakness of the argument: it is susceptible to the “obvious rebuttal” Pasternack puts forward. Basically, it consists in making the mismatch disappear by allowing existence to be a mark of concepts (p. 114).

However, a more convincing reading of Kant’s argument is possible. Immediately after the passage Pasternack quotes in support of his reading, Kant writes: “Even if I think in a thing every reality except one, then the missing reality does not get added when I say the thing exists, but it exists encumbered with just the same defect as I have thought in it; otherwise something other than what I thought would exist” (A600/B628). It emerges from this sentence that the alleged “mismatch”, i.e. a concept’s not fully capturing all the properties its instances exhibit, is not what is at issue, at least as far as Kant perceives matters. On the contrary, his point concerns instantiation, or the relation between concepts and objects, in general. This addresses the first weakness of Kant’s case as interpreted by Pasternack. But what is Kant’s point then?

Kant argues that a concept’s instantiation does not correspond to any addition of properties to it; rather, a concept’s instantiation amounts to the object’s having just the properties the concept specifies. A plausible way of construing this claim is: A concept’s content consists in the conditions an object has to meet in order to count as an instance of it. Kant can be understood as showing that this view cannot be upheld if one understands existence as a property. The reasoning can be understood as follows. If existence were a property of objects and being an instance of a concept is to exist, then a concept’s instances would consequently have to bear the property “existence”. If “falling under a concept” consists in an object’s conforming to the conditions set by the concept and existence is one of these conditions, then existence would have to be a mark of the concept. But this would render some existence-judgments analytic—a view Kant takes himself to have refuted at this point in the discussion. If one grants this, it follows that existence is not a mark of any concept. But if it still holds that instances of concepts exist, then a concept’s instantiation consists in an object’s conforming to the conditions set and exhibiting the property “existence” additionally. As the latter is not part of a concept’s content, this content’s identity therefore cannot consist in a specification of what it takes to be its instance, no matter how completely or incompletely it captures an object’s properties.

Kant therefore does not argue that a mismatch between concepts and objects is problematic as such, but that a specification of a concept’s content in terms of conditions instances have to meet is impossible given that one accepts the following three theses:

1. a concept is individuated by the conditions on objects to count as instances of it;

2. existence is a property of objects;

3. existence-judgments are synthetic.

Kant’s point therefore is: the view that existence is a property is indicative of a misunderstanding of what concepts and “falling under a concept” are. Pasternack’s rebuttal misses the mark in relation to this issue, for accepting existence as property and conceptual content either leads to the implausible view that all existence-judgments are analytic or, if they remain synthetic, precludes a conception of concepts as specifying the conditions of what it is to fall under them.

Alongside “the usual suspects”, Pavel Tichý’s work on the ontological argument makes an unexpected appearance in the volume. As is convincingly shown by G. Oddie’s essay, Tichý offers one of the most penetrating and revealing interpretations of Anselm’s Proslogion III, i.e. the passage serving as inspiration for what is known as “the modal ontological argument”. Tichý delivers a logically valid reconstruction of Anselm’s argument as well as an unfamiliar axiological criticism of its soundness.

The reading rests on Tichý’s ontology, fundamental to which is the distinction between “two entirely different types of entity” (p. 199): individuals (such as Donald Trump) and offices (such as “the President of the U.S.A”). Intuitively, offices are either occupied by an individual or not, where occupancy is to be understood as a property of the office rather than the individual. Formally, offices are partial functions mapping world-time pairs to individuals which are undefined when the office goes unoccupied. What an office is—its essence—is given as a set of conditions called requisites which have to be borne by occupants to count as such (p. 203).

Within this framework, the modal ontological argument aims to derive the necessary occupation of “the divine office” (p. 205), which Tichý interprets as “that individual office such that no individual office is greater than it” (p. 206). Anselm’s formula, thus understood, singles out a second-order office, that is, an office occupied by a first-order office rather than an individual. Glossing over the details of the reconstruction, the Proslogion III argument derives necessary existence as a requisite of the greatest office, yielding the conclusion that the divine office is necessarily occupied. This yields a “valid” argument according to Oddie (p. 209).

Compared to the standard modal ontological argument known from the writings of Harthshorne and Plantinga, Tichý’s interpretation of the argument has one key advantage. The standard version treats existence in all possible worlds as an essential property of God and derives God’s existence from His/Her possible existence plus S5. The argument is often criticized for begging the question because the premise that God’s existence is possible cannot be substantiated in a non-circular fashion. G. Priest’s offers one way of putting the difficulty (p. 265).⁴ According to the premises of the argument, the following two entailments hold: (1) God’s actual existence follows from His/Her possible existence; (2) God’s actual existence entails His/Her possible existence. “God exists” and “possibly, God exists” are therefore equivalent according to the argument’s premises. Hence, presupposing the possibility of God’s existence is question-begging insofar as it is equivalent to presupposing God’s existence. By contrast, Tichý’s reconstruction derives the necessary occupancy of the divine office via an axiological premise, namely: necessarily occupied offices are always greater than ones which are not (p. 208). The truth of this premise can be assessed independently, and hence God’s necessary existence gets established in a more satisfactory way.

However, this premise also renders the argument unsound according to Tichý. The claim that necessarily occupied offices are always greater than ones which are not is subject to counterexamples, one of which is: the office “the discoverer of the incompleteness of arithmetic” is contingently occupied, whereas the office “the pick of the morally most depraved”, where “pick” refers to a choice function to be applied in case of a tie, is necessarily occupied. Yet, the former is plausibly “greater” than the latter (p. 212). Therefore, Oppy concludes with Tichý, the argument rests on an implausible axiology of existence. Attempts at weakening the relevant requisite (e.g. either being God or else the pick of the morally best) will, while ending up necessarily occupied, fail to prove the existence of God at all world-times, for God is not merely the relatively morally best being, but the absolutely best (pp. 212–213).

Oddie’s simultaneously fascinating and accessible discussion of Tichý’s reconstruction will hopefully lead to the recognition of what strikes me as the most convincing version of a Proslogion III-style modal ontological argument. Further discussion may delve deeper into the axiological questions raised by Tichý. As is always the case with criticisms resting on counterexamples, they may show that, but not explain why, some thesis is false. What principled reason against Anselm’s axiology can be given?