Saturday, August 10, 2019

Feser on simplicity

Ed Feser has responded to a critique of divine simplicity. It's useful to see how a Thomist  of his caliber fields objections:

Take the latter point first. Though its critics often treat the notion of divine simplicity as an unimportant curiosity, there are good reasons why the Church Fathers, the medieval Doctors, and two ecclesiastical councils regarded it as essential to orthodoxy. For one thing, it is a consequence of God’s ultimacy. For anything composed of parts is ontologically posterior to those parts, and can exist only if something causes the parts to be combined. Hence if God were composed of parts, there would have to be something ontologically prior to him and something which combines those parts, thereby causing him to exist. But there is nothing ontologically prior to or more ultimate than God, and nothing that causes him. To be the uncaused cause of everything other than himself is just part of what it is to be God. Hence God cannot be composed of parts but must be absolutely simple.[1]

i) One problem with his argument is the equivocal notion of "parts". The paradigm-cases involve physical composite objects. It then extrapolates from paradigm-cases in creation  to God. 

But what about abstract objects. Timeless, spaceless objects like the Mandelbrot set, possible worlds, and the number Pi. These are certainly complex in a certain respect. Does that mean they're made of parts?

And even if we say they're made of parts, does that mean the "parts" of the Mandelbrot set, possible worlds, or the number Pi are ontologically prior to the whole? What sense does it make to say, in reference to complex abstract objects, that the whole is prior to the parts or the parts are prior to the whole? Do the parts "cause" the whole–or vice versa? Is that even meaningful in this context?

(Admittedly, some metaphysicians deny the existence of abstract objects, but I'm just using that as an illustration. It can work at a conceptual level whether or not you're a modal realist.)

ii) Underlying his argument seems to be a monadic prejudice about the ultimacy of the one over the many. That ultimate reality is simple or strictly unitary. But what if ultimate reality is irreducible complex? It seems to play on the crude intuition that big things are made of smaller things, so as you go down the scale there are ever fewer constituents until, at rock bottom, there's just one thing left. 

For another thing, divine simplicity safeguards God’s uniqueness. Where there is a distinction between a thing and its nature or essence, then that thing will not necessarily be unique. For example, there is a distinction between a given particular triangle and triangularity as a common nature or essence. Given that distinction, something other than that particular triangle might share that same nature or essence, so that there can be more than one triangle.  By contrast, according to the doctrine of divine simplicity, there is no distinction between God and his nature or essence. God just is his nature, so that it is not something that he could have in common with another thing. And if there cannot be anything else that has the divine nature, then there cannot even in principle be more than one God. In this way and others, divine simplicity protects monotheism.[2]

i) Which invites the familiar objection that divine simplicity is unitarian rather than Trinitarian. 

ii) Another problem is that Thomstic simplicity isn't a single claim but a bundle of claims. We might agree that God is identical his nature. God is not property instance of a generic nature. But that's a different claim that saying each divine attribute is identical with every other attribute.  

iii) In addition, there's a standard distinction between God's communicable and incommunicable attributes. Or the medieval exemplarist tradition, where the divine nature is imitable. 

Take divine ideas. Suppose possible worlds are divine ideas. But these allow for finite exemplifications. God instantiates his concept of the world in space and time. 

These are the reasons why defenders of divine simplicity sometimes go so far as to argue that to deny the doctrine entails atheism. For if being an uncaused cause and being absolutely unique entail simplicity, then to deny that there is anything that is simple or non-composite is implicitly to deny that there is an absolutely unique uncaused cause. And since to be God just is to be an absolutely unique uncaused cause, to deny divine simplicity is therefore implicitly to deny the existence of God.

The stakes in this debate are therefore much higher than Mullins lets on, and for a theist to refute the doctrine of divine simplicity would require more than merely raising objections of the kind Mullins does. It would require explaining how such objections could avoid inadvertently refuting theism itself.

Because Thomistic simplistic is a bundle of claims, it may be more plausible in some respects than others. 

In any event, Mullins’ objections, I have claimed, do not succeed. Consider first the claim that divine simplicity is incompatible with divine freedom. The objection begins by noting that the doctrine of divine simplicity holds that God is identical with his attributes.  God doesn’t merely have omnipotence but rather is his omnipotence; he doesn’t merely have omniscience but is his omniscience; and so on. But what is true of God’s attributes is true of his acts as well.  God just is his act of creating the world, for example. Now, since God exists necessarily, it follows that if God just is his omnipotence, then his omnipotence exists necessarily. But the same thing would have to be true of his acts, such as his act of creating the world. That act would exist necessarily as well. 

But in that case, the objection concludes, we have the result that the world exists of necessity.  And that is problematic in two ways. First, it seems obviously false, because though the world does in fact exist, it is hard to doubt that it could in principle have failed to exist. Second, the result seems to entail that God does not act freely. He had to create the world. But that is contrary to standard Christian doctrine, which holds that God freely created the world insofar as he could have opted instead not to create it. Similarly, it is Christian doctrine that God could have refrained from imparting grace to us, and this too is incompatible with the result that God’s acts are as necessary as he is.

Now, one way to see what is wrong with this objection is in terms of a distinction drawn by the Thomist philosopher Barry Miller.[3] Suppose Socrates grows a beard. That involves the acquisition by Socrates of a real property. But suppose Socrates becomes shorter than Plato, not because of any change in Socrates himself, but rather only because Plato has grown taller.  That involves the acquisition by Socrates of a mere “Cambridge property” rather than a real property.  The real change, and thus the acquisition of a real property, is in Plato rather than in Socrates.[4]Now, what the doctrine of divine simplicity claims – contrary to what Mullins supposes (in what he labels premise (9) of his argument) – is, not that all of God’s properties are identical and thus are as necessary as he is, but rather that all of his real properties are. He can have Cambridge properties that are merely contingent. And his having created the world is among these contingent Cambridge properties. That the world comes into being does not entail the acquisition of a real property by God, any more than Socrates’s becoming shorter than Plato by virtue of Plato’s becoming taller entails the acquisition of a real property by Socrates. When we keep this distinction in mind, we can see that divine simplicity does not have the implication that the creation of the world was necessary, and thus does not have the implication that God is not free.[v]

i) I prefer to say (contingent) relations rather than contingent properties.

ii) I first encountered the notion of Cambridge changes in Peter Geach. It's a legitimate distinction as far as it goes.

iii) Feser's response seems to assume the very issue in dispute: the question at issue is whether God can have contingent relations, given Thomistic simplicity. For Feser to respond by distinguishing between "real properties" and "contingent properties" takes for granted the bone of contention. So how does that actually show, rather than presume, that simplicity is compatible with divine freedom? 

iv) Moreover, the issue as I see it isn't whether God undergoes change. The point, rather, is that creation is an act of the divine will. If God's will is identical to God's essence, and God's essence is necessary, then isn't whatever God wills necessitated by his essence? 

Mullins also accuses advocates of divine simplicity of making “cheap” appeals to mystery in defending the doctrine, and in particular of resorting to “mysterious language” in explicating it. But his attack is aimed at a caricature, and he misses the point about language that the defender of divine simplicity is trying to make.

Now, there is nothing at all “mysterious” about analogical usage. On the contrary, it is an extremely familiar feature of ordinary language. Moreover, it is a kind of usage that is unavoidable when we try to describe things that are remote from everyday experience – as we do, for example, in physics. Consider general relativity, which describes the structure of time and space at the largest scale, and quantum mechanics, which describes the structure of matter at the smallest scale. Here we have to find some way to talk about physical realities that are very unlike the kind of physical things with which we deal in everyday life, which is the context in which our terms for physical things have their natural home. Hence we have to rely very heavily on analogy.

So, for example, general relativity speaks of “curved space.” This is very odd, when you think about it. The way we ordinarily use the term “curved” is to apply it to the shapes of the things that occupy space. For example, we speak of the curvature of the surface of a ball, which you would observe as you held it up before your eyes. Relativity requires us to apply the concept to the space through which the ball and other objects move, and this is a kind of curvature you cannot observe. Quantum mechanics requires us to speak of certain entities as exhibiting both particle-like and wave-like properties. This too is very odd. Being a “particle” in the usual sense of the term rules out being a “wave” in the usual sense, and vice versa.

Analogical language can be legitimate if you're able to isolate and identify what the comparison shares in common. To take his own example, "spacial curvature" is metaphorical in Relativity. That's used for popular expositions. But unless you can translate the metaphor into a literal proposition, the comparison is opaque. Indeed, it merely posits an analogy. But that can only be justified if it's possible to show the point of analogy. Since every analogy has an element of disanology, we need to demonstrate where the disanalogy leaves off and the analogy begins.  

Because we have to stretch the meaning so far, the usage is somewhat mysterious. But there is no “cheap” mystery here, because the theoretical reasons that have pushed us in the direction of stretching language in this way have paid for this usage, as it were. 

i) Up to a point, appealing to mystery is legitimate in theology and science alike. But it can't be sheer mystery, because sheer mystery is unintelligible. Sheer mystery fails to affirm or deny anything. It represents nothing. Divine simplicity can't be so radically unlike anything in human experience or imagination that we have no idea what the category stands for. 

ii) In addition, it's not as if Thomistic simplicity is a revealed truth which we are entitled to take on faith. Rather, Thomistic simplicity is a philosophical construct. As such, it must be able to withstand rational scrutiny. Appeal to mystery is evasive and illegitimate when it shields philosophical theology from rational scrutiny. 


  1. Quite right to question the assumption that plurality entails composition. Feser needs to do better.

    Feser’s rejoinder to Mullins misses the point. Divine willing is not a Cambridge property by the very fact that it is the deity willing. So relations those things that are willed have to God because they are willed by God is not a gloss on divine willing. It is a gloss on objects willed. So he just sidesteps the argument. This is why his account is dismissive. If God’s act of will is his essence, the modality will be transitive and so the world will be necessary. Other Thomists like Kretzmann have noted and conceded the point.

    The appeal to analogical predication won’t work here either. The reason is simple and Thomas himself knew it. Analogical predication is a consequence of simplicity and so simplicity has to be demonstrated prior to an appeal to analogical predication is made to defend it, otherwise it is circular. This is why Thomas puts simplicity prior to analogical predication in the Summa Theologia. This is a standard point in the literature and I am surprised that Feser fell prey to it.

    What is more, what Thomas has in mind is very different than mere ordinary language. It is precisely because God is simple that the mode of signification for any human language is skewed in reference to God. And this is so because all human language for Thomas is garnered from sensation, a la of the Aristotelian sort, and so all terms entail composition in their mode of signification. So his use of everyday language in the sciences just doesn’t wash.
    And when defenders of Simplicity such as Dolezal appeal to mystery in the face of the objection from divine freedom, yeah, that is a cheap appeal because we are not getting what we promised, namely a demonstration of conceptual compatability.

  2. ADS becomes more incomprehensible every time I read about it.

  3. Lloyd Gerson has an interesting essay in "Aristotle on Method" which contrasts Aristotle's and Plotinus' view on simplicity. There is a reason that Plotinus located Aristotle's 'thought thinking thought' not in the One, but in the Nous, intellection, says, Pltonius implies a kind of composition inapplicable to the One.

    Stump has interesting things to say on simplicity, say, in her "Aquinas" but it's far from clear that it works, as a critique of her attempt at reconciling it with Divine freedom (in "Metaphysics and God: Essays in Honor of Elenore Stump"). She herself remarks that being, to Augustine, involves order, genus, mode including God. And her idea that God can vary across possible worlds - both seem to involve plurality of a sort in God.

    Now, is there a reason we can know as to why there are only three perosns in the Godhead and not multiple instantiations of the Godhead, multiple Gods, if we reject absolute Divine simplicity?

  4. I saw the Theopolis Institute's discussion on Divine Simplcity. What are some other good critical interactions with the notion of Divine Simplicity?

    1. For what it's worth, I plan to do two more posts on the topic.