Robert Merrihew Adams on Brian Leftow’s God and Necessity
Like their medieval scholastic predecessors, the best known Continental Rationalist philosophers of the seventeenth century (Descartes, Spinoza, Malebranche, and Leibniz) held that God exists necessarily, and that all possibilities along with all necessary truths about anything actual or possible are grounded in God. They were not all in agreement, however, as to how those possibilities and necessities are grounded in God. The contrast between the views of Descartes (1596-1650) and Leibniz (1646-1716) may serve as background for approaching what I take to be the most thoroughly articulated and argued study of the relation between God and necessity (and possibility) that our own time has produced.
Descartes’s main question about possibilities and necessities is what makes the “eternal truths,” as he calls them, true. His answer is voluntarist to an extent that is unusual in Western philosophical and theological thought. He held that the eternal truths are voluntarily caused or legislated by God. He claimed that though they are in fact necessary, God could have made them false, declaring in The Philosophical Writings of Descartes (vol. III (1991), 358-9):
I do not think that we should ever say of anything that it cannot be brought about by God. For since every basis of truth and goodness depends on his omnipotence, I would not dare to say that God cannot bring it about that … 1 and 2 are not 3. I merely say that he has given me such a mind that I cannot conceive … a sum of 1 and 2 which is not 3.
Descartes’s view on this subject found some support in the seventeenth century but has never been prevalent in modern philosophical theology. It is worth noting, however, that in the twentieth century a voluntarism that looks similar in some ways, and comparably extreme, was asserted (without reference to Descartes) by the great (and self-consciously not philosophical) theologian Karl Barth in his Church Dogmatics (vol. II, 533-42).
Among Descartes’s early modern opponents on this point was Leibniz. In his view, possibilities and necessities are grounded not in God’s will but in God’s understanding. Moreover, the main question about possibilities and necessities to which Leibniz offers a theological answer is not what makes them true but what grounds their being. He holds that “if there is a reality in the Essences or possibilities, or indeed in the eternal truths, that reality must be founded in something existing and Actual.” So how can there be truths so absolutely necessary that they could and would be true even if no contingent being existed? Leibniz’s answer is that the reality of possibilities and of the eternal truths “must be founded … in the Existence of the necessary Being, in which being possible is sufficient for being Actual.” Specifically, he holds that “the Understanding of God is the region of eternal truths, or of the ideas on which they depend” — that they have ontological standing, or being, insofar as they are understood by God and represented in the divine understanding. This is for Leibniz the basis of an argument for the existence of God “from the reality of eternal truths” — that is, from the assumption that there are truths so necessary that they could and would be true no matter what, even if nothing contingent existed (Monadology, §§43-45). He treats this, in effect, as an inference to the best explanation of the reality of facts of possibility and necessity, where the explanation is metaphysical, and turns on a constitutive rather than a causal relationship.
Fifty years ago, such arguments were easily dismissed as of merely antiquarian interest, since almost all philosophers and many philosophical theologians were convinced that the existence of God could not be absolutely necessary. Since the 1960s, such certainties have been dissolving in the analytical mainstream of Anglophone philosophy, as assumptions about what necessity must or could be have come to seem problematic. In this philosophical context there is more room for theories which involve, as Brian Leftow’s does, the hypothesis of a necessarily existing deity.
In his magnum opus, God and Necessity, Leftow stakes out a position that is intermediate between those of Descartes and Leibniz, having points of agreement with each of them. It is the voluntarist aspect of the position, however, that gets the most emphasis and the most vigorous argument in the book — perhaps because Leftow realistically expects it to be the target of the most opposition. The tightrope on which he seeks to balance his position is a distinction between what he calls “secular” and “non-secular” truths and possibilities. There is not room here to present a full account of Leftow’s developed distinction; but, roughly speaking, a secular sentence is one that if true would state no truth about God, or about divine attributes or events, though it might imply one. Secular sentences are about actual or possible creatures of God, or their (secular) attributes and interactions. Leftow’s position is theologically voluntarist about possibilities and necessities expressed by secular sentences — almost all of them — but not in general about those expressed by non-secular sentences, which he sees as determined by God’s nature (Leftow, 248-50, 275-76).
As regards God’s existence, in particular, Leftow holds that “His nature makes Him necessary — His choice cannot alter this. Thus … God exists necessarily.” Similarly, as regards pure logic and pure mathematics, Leftow agrees with Leibniz that their ontological grounds are to be found in God’s thought — God’s actual thinking. And in these cases “God’s nature dictated the result. He would have so thought no matter what.”
Leftow has relatively little to say about logic and mathematics and their ontological grounding. His “main concern,” he writes, “is a theory of secular truths” — a theological theory about them and the status they may have as possible or even necessary (Leftow, 251). Secular possibilities and necessities, Leftow insists, do not follow from, and are not determined or guaranteed by, God’s nature. Thus he opposes what he calls “deity theories,” which view “secular” possibilities and necessities as grounded in deity, where by “deity” is meant the nature or individual essence of God. Leibniz’s theory of the reality of eternal truths is a prime example of the “deity theories” that Leftow opposes. Contrary to them he holds that secular attributes enter logical space by God’s thinking them up — though the divine essence allowed God not to think them up and thus to omit them from logical space. Since God thought up secular attributes, on Leftow’s view, almost all propositions involving them acquire their status as necessary, contingent, or impossible by a voluntary divine decision that was not predetermined by the divine essence.
The following is a simplified (but I hope not weakened) form of Leftow’s “main argument against deity theories” (Leftow, 209 ff.). The “necessary truth about creatures alone” that he proposes as an example is “water = H2O.” Analytical metaphysicians of possibility and necessity have in fact discussed this example, and most regard it as a necessary truth. Issues could be raised about the role of the word “water” in the example, however, which would lead us away from our theological topic. We can avoid those issues by substituting the closely related identity “H2O = H2O.” That will not seriously affect character of Leftow’s argument. He will not claim that a negation of “H2O = H2O” could be true; but he argues that “H2O = H2O” would be untrue “were H2O not possible nor so much as impossible, but instead just not a denizen of logical space.”
How could that be? Well, on Leftow’s theory, nothing in God’s nature requires that God think about H2O, and if God had never thought up H2O and therefore had never created it, H2O would not be a denizen of logical space, and the proposition “H2O = H2O” would not exist and therefore would not be true. On a deity theory such as Leibniz’s, however, it follows from the nature of God, and is therefore necessary, that if God exists, God does think of H2O and H2O is a denizen of logical space, as an object of God’s thought, and of course “H2O = H2O” (or “If H2O exists, then H2O = H2O”) is a necessary truth. And this is just what Leftow objects to, for in implying that necessarily, if God exists, then H2O is a denizen of logical space, the deity theory also implies that necessarily, if H2O were not a denizen of logical space, then God would not exist. “Surely,” Leftow protests, “deleting a chemical property from logical space should not affect whether God exists.” It would be better, he concludes, to avoid “deity theories” of the grounding of “secular” possibilities and necessities (by which of course he does not mean avoiding all theological theories on the topic).
One may respond, of course (as Leftow recognizes), that according to a deity theory these implications really pose no threat at all to God’s existence. For if the deity theory is right, both God’s existence and God’s understanding of H2O, and the truth of “H2O = H2O” as represented in God’s understanding, are absolutely necessary, as reasons included in God’s essence ensure that there is no possible alternative to them. Leftow grants, in effect, that the conditional proposition “necessarily, if H2O were not a denizen of logical space, then God would not exist” is not just counter-factual but counter-possible (having an antecedent clause that could not possibly be true), if a deity theory is right. He claims, however, that it would be wrong to dismiss it as not worth worrying about, because it is also counterintuitive. He argues that “deity is the property having which makes God divine. Intuitively, facts about water do not help make God divine,” and are thus “irrelevant to the job the property deity does.” For what is relevant to God’s possession of such intuitively divine attributes as omniscience and omnipotence is not what secular possibilities there are in logical space, but what God knows, and can do, about whatever secular possibilities are there (Leftow, 240).
The deity theorist’s best response to this argument, perhaps, is that “making God divine” is too narrowly religious a description of “the job” belonging to deity as a property, if we understand deity (as Leftow does) as the essence of God. More comprehensively, one might think, “the job” that God’s essence does includes, metaphysically, being the nature by reason of which God necessarily exists as the ground of being, the being on which all facts of every sort depend. If the job of God’s essence is as large as that, is it really counterintuitive to suppose that secular possibilities such as that of H2O are too petty or too profane to be provided for in the divine essence? Moreover, a deity theory like Leibniz’s account of “the reality of eternal truths” may be appealing precisely to philosophers who find it unintuitive to suppose that there is any way in which being such a thing as H2O actually is could have failed to be at least possible.
Harder to shrug off for Leibnizian deity theorists may be Leftow’s efforts to problematize their assumption that the divine essence could contain all the materials needed for representing or thinking of all the secular possibilities. How, Leftow asks, could those possibilities be represented in the divine mind? He focuses on a view, found in Thomas Aquinas and other medieval theologians, according to which the properties that possible creatures would have are imitations (imperfect, of course) of the divine nature, so that God is able to know those creaturely properties by knowing how his own essence is capable of being imitated (Summa Theologiae, Part I, qu. 15, article 2.). The imitation relation of creatures to God is part of Leibniz’s philosophy also, though not part of his account of possibilities as ontologically grounded in God’s understanding. Leftow argues that while the imitability of the divine nature might in this way give God very general concepts of created substance or created mind as such, it does not offer a promising way for God to conceive of such creatures in complete detail, nor to conceive of material creatures at all (Leftow, 153 ff.). Leftow, rightly, does not treat this as a decisive objection to deity theories. The great obstacle to accepting it as a decisive objection is the assumption of most theists that God’s cognitive powers are infinitely greater than ours. That assumption suggests that the divine nature might indeed enable God to have certain ideas that we have, even if we have not managed to see how that would be possible. Nevertheless, if (as I agree) deity theorists (including Aquinas and Leibniz) have not provided a convincing development of an imitability theory, or any other theory, of how the divine essence might enable God to understand all possible creatures, that is a weakness in their theories.
Can Leftow do better? He does offer a theory of how God has concepts of possible and actual creatures. It is “a causal theory of divine content” — a reductive theory, according to which concepts in God’s mind are reducible to God’s possession of certain powers. He holds that “the ontology of God’s mental content need invoke only powers of and causings by God and divine mental events.” For example, “God’s thinking that Fido is a dog is that part of His life that would in conjunction with certain other divine mental events bring it about that Fido is a dog” (Leftow, 312-15). Both the appeal and the implausibility of this striking view, I think, are rooted in the same point: that the divine concept of the producible object is to get its content from God’s knowledge of possible divine mental events that would produce the object in actual existence, rather than the other way around. The attraction of the view is that it promises a way of avoiding problems that have been found in other accounts, including resemblance accounts, of mental representation. The implausibility is that the view does not clearly provide for God to know more about what it is for Fido to be a dog than that it is whatever certain divine powers and mental events would cause.
To put more pressure on Leftow’s view, consider the thought that a responsible creator of the actual world had better have known more about what it is like for Fido to feel pain than just how that occurrence would be causally related to powers and mental events in the creator. Leftow does not disagree. Such considerations lead him to entertain the view that “God can have some quality of experience whence … He can fully conceive of pain.” But how is that consistent with his apparently sweeping claim that “the divine mental event which is God’s thinking that P is simply that part of the divine life that has or would have effects suitable to that content”? He does suggest that God might be able to have the full conception of pain “merely from His understanding of His power to have” a relevant quality of divine experience. Where the question is how God can know what it is like to feel pain, however, no appeal to supposed knowledge of divine powers to cause or to have an experience (as distinct from actually having a relevant experience) has to my mind much more explanatory force than the simple “somehow or other” which is as available to the deity theorist as to Leftow (Leftow, cf. 286, 314).
God and Necessity is a monumental work on its topic, rich in theoretical alternatives critically discussed — many more of them than could be discussed here. I found it extremely stimulating and rewarding. Its likeliest readers are professional philosophers. It presents an intricate tissue of analytical arguments, some of them formulated partly in the symbolism of modern formal logic. It is hard reading, perhaps harder than it needed to be. But it is definitely a book to be reckoned with.
Also Recommended from MRB: