For convenient reference, I quote the relatively formalized version of Robinson's argument that he gave in the course of replying to Phil Blosser's rebuttal of his original argument:
1. If God is absolutely simple (P), then his act of will to create is identical with his essence (R)
2. If God’s act of will to create is identical with his essence (R), thenh is act of will to create is necessary. (Q)
3. If God is absolutely simple (P), then his act of will to create is necessary. (Q) (From 1,2 by Hypothetical Syllogism)
4. God is absolutely simple. (Premise S)
5. Therefore, God’s act of will to create is necessary (R). (From 3,4 byModus Ponens)
Support for (2) is given by the following argument:
6. If God’s essence is had by him necessarily, then if anything is identical with his essence it is necessary.
7. God’s essence is had by him necessarily. (Premise)
8. Therefore, anything identical with his essence is necessary. (From 6, 7 [MP])
(7) I take to be uncontroversial and by that I mean that any Christian should agree with it on its face. (6) can be supported by Liebniz’s Law: (x) (y) [(x = y), then (P) (Px, ≡ Py)]. For any x and any y, if x is identical to y, then if x has a property P then y must have that same property P and vice versa.
Now the main argument, i.e. (1)-(5), is logically valid. It is also a legitimately ad hominem argument in that it's designed to demonstrate the mutual incompatibility of two doctrines Aquinas held. But (1) and (2) are ambiguous: the senses of ‘essence’ and 'identical with...' are either not Aquinas' senses or, if construed in line with Aquinas, are not those which Robinson needs for his argument. In support of that claim, I shall take the difficulties with Robinson's argument in the order most convenient for exposition. Once that is done, Blosser's defense of Aquinas—which is essentially that of Aquinas himself—will emerge as apposite albeit as insufficiently explanatory.
As a preliminary, it is important to get clear about what the 'absolute' in 'absolute divine simplicity' means for Aquinas. For there is a sense of 'absolute' that is very easy to slip into using, but on which Aquinas would have rejected (1)'s antecedent—if he had ever taken that sense seriously to begin with. Thus, if there is no differentiation whatsoever in God, and if Aquinas believed as much, then Aquinas' God would be indistinguishable from the "One" of Proclus. But no professing, orthodox Christian takes or could take God to be absolutely simple in that sense; indeed, no Christian such as Aquinas would deny that there are relations ad intra within God. If God is a living, loving God, as we all affirm, then even granted that all his internal relations are eo ipso reflexive, they are not merely notional. Leave aside the thorny question how the divine nature can be thought to exhibit such relations in abstraction from the interrelations of the Three Divine Persons—e.g., as in God’s eternally knowing the plurality of the divine ideas as ways in which his being could be participated. We can only take one problem at a time. The fact remains that, just by professing the doctrine of the Trinity, Aquinas was fully committed to holding that there are real relations within God and thus some form of differentiation in God. Whatever Aquinas meant by 'absolute' divine simplicity, then, he was very unlikely to have intended it in a sense that would be incompatible on its face with a dogmatic affirmation he made in keeping with Tradition. He just wasn’t clueless enough to make a mistake like that. So if (1) ascribes to Aquinas an essentially pagan sense of ADS, then (1) is true only because it is a material conditional whose antecedent is false. That of course is not what Robinson is after at all. But Aquinas' commitment to affirming some differentiation and plurality in God is important for understanding, if only by contrast, what he means by ADS and how that is relevant to assessing (1) and (2).
Let's start with the obvious: for Aquinas, there is no sort of metaphysical "composition" in God. In ST Ia Q3, he argues that God is not composed of: (i) matter and form; (ii) nature and subjectum; (iii) essence and existence; (iv) genus and difference; (v) substance and accident. Thus God is absolutely simple not in the sense of being wholly undifferentiated ad intra, but in the sense that he is not composed of parts in any of the ways Aquinas thought it possible in principle for something to be so composed. The negative formulation is essential for Aquinas inasmuch as he insisted that we know of God not what he is—i.e., his nature—but what he is not. (Citing John of Damascus, Aquinas premises as much in ST Ia Q2 A2—before discussing the divine attributes or even proving God’s existence.) We know the divine nature not so much by affirming perfections of him, which we can and should do, but by subtracting from the conceptual content of those affirmations whatever is distinctive of the perfections' finite instances. Only then may we affirm the (infinite) perfections as contained "supereminently" and mutually “identically” in God. That of course enshrouds God in a kind of conceptual haze—at least on the level of natural as distinct from revealed theology. Indeed, Aquinas vigorously denied that God is one kind of being among others, one that just happens to be immeasurably more perfect than the others. God is Being in the fundamental and fullest sense: Ipsum esse subsistens. In natural theology, we can only explain "by way of negation and remotion" what that means. Such a method isn’t terribly illuminating: we can say that God thus exhibits such-and-such perfections, but we cannot explain how. Yet given the subject matter, that result is surely appropriate; anything more would be hubris. It explains why Aquinas formulated ADS as he did. Granted (i)-(v) and given what Aquinas took to be the range of possible candidates for composition, there just aren't any ways left for God to be composed (Article 7). For Aquinas, then, ADS is a primarily negative claim.
Now on such an account of the absoluteness of DS, there is no reason to believe that Aquinas would accept (1) and (2) in senses that would be useful for Robinson's argument. Of course one might argue, as Robinson implicitly does, that (1) follows from affirming (iii)—the identity of God's existence with God's essence—along with another characteristic thesis of Aquinas. To sum it up:
(9) God's existence is identical with his actuality [by Aquinas' doctrine of God as actus purus](10) God's actuality is identical with his "act of will to create" [by ADS]
(11) God's existence is identical with his essence [from (iii) above, a conjunctive part of ADS]
(12) God's existence is identical with his act of willing to create (MP on 9,10)
(13) Ergo, God's act of will to create is identical with his essence (MP on 11, 12)
That argument too is logically valid, and I believe Aquinas would have accepted it. But that is useless for Robinson's purpose because the concepts of essence and identity that Robinson invokes throughout his arguments, as distinct from the terms themselves, are not those of Aquinas.
Regarding essence (or nature; in this context the terms are interchangeable for Aquinas), consider Aquinas’ prior argument that God is the same as his nature:
…in things not composed of matter and form, in which individualization is not due to individual matter--that is to say, to "this" matter--the very forms being individualized of themselves--it is necessary the forms themselves should be subsisting "supposita." Therefore "suppositum" and nature in them are identified. Since God then is not composed of matter and form, He must be His own Godhead, His own Life, and whatever else is thus predicated of Him. (ST Ia Q3 A3 resp).
Recall that, in Aquinas’ metaphysics, even angels are identical with their natures because they are purely spiritual “forms” not individuated by matter and thus are, each of them, sui generis (better: sui speciei). But he never thought that a true, complete description of the nature that is a given angel would tell one everything that angel ever did or would do. He believed that all angels have done at least one thing with libertarian freedom, and thus that not everything they do is deducible in principle from what they are by metaphysical necessity. For Aquinas, therefore, the claim that such-and-such is identical with its nature does not mean, as a general proposition, that everything truly predicable of it is an essential property of it and thus is “had necessarily” by it. Nor is the latter even logically equivalent to the former. Without risking too much anachronism, we may formalize his actual claim thus:
For any x and any sortal (not: general) predicate K, if Kx and x is not individuated by matter, then for any y, if Ky and y is not individuated by matter, then x is the same K as y.
Therefore and in particular:
For any x and y, if x is God and y is God, then x is the same God as y.
That is what relevantly follows from Aquinas' denial that God is composed of himself and his nature or essence—the properly negative form of the claim that God is “identical with” his nature. It’s no different from the case of angels.
Note that the concept of identity I’m ascribing to Aquinas here is relative not absolute identity. (Peter Geach has been the strongest and most original contemporary advocate of RI; see especially his Reference and Generality, 3rd edition, 1980).The thesis that there is such a thing as absolute identity is far from indisputable; one bit of evidence that it is incompatible with Aquinas’ thought is precisely the sort of argument that Robinson gives. (Try to explicate Aquinas’ Trinitarian theology with absolute identity. The thing just can’t be done.) What accordingly undermines Robinson’s argument—as an attempt, that is, to saddle Aquinas with a consequence Aquinas would have denied—is that it construes God’s “essence or nature” for Aquinas along the same lines as those of Alvin Plantinga (Does God Have a Nature?, Marquette U.P., 1980) and Christopher Hughes (On a Complex Theory of a Simple God, Cornell UP, 1989): as a collection of properties that, on ADS, seems to become a single necessary property with which God is falsely, perhaps even incoherently, said by Aquinas to be identical in the sense of absolute identity. That is why Robinson feels free to invoke Leibniz’s Law (LL) in support of (6), which he in turn adduces as the first premise of a sub-argument [(6)-(8)] in support of (2). LL’s antecedent employs absolute not relative identity as though the former were itself an unequivocal and uncontroversial notion. But Robinson’s invocation of LL is anachronistic and thus question-begging, just as I think the broader arguments of Plantinga and Hughes are for that and other reasons. Hence there is no reason to believe that Aquinas would have accepted either (1) or (2) in senses that would give Robinson’s argument any purchase.
Well then: how are we to construe Aquinas’ claim that God’s “existence” (i.e., esse, or act of being) is identical with his essence or nature? Well, God’s nature is pure to-be: unqualified esse. Unlike angels or any creature, he doesn’t receive being from anything else: He is Being Itself, so that whatever else there is gets its existence, as well as its kind of existence, from him. Such is God’s aseity, which rules out any sort of composition inasmuch as composition entails a composer other than the composed and at least one constituent that is not the composed (ST IA Q3 A7). Therefore, God is not “composed” of his essence and his existence. He doesn’t “have” existence as though there were something to him that isn’t his existence but which could receive it and thus enter into composition with it. But once again, that does not entail that whatever God does is an essential property of him and thus is done by absolute as distinct from hypothetical necessity. What it does entail is that God’s “to-be,” in the unqualified and thus incomprehensible sense of his actus purus, is the same God as God’s nature, in the sense of ‘nature’ expounded above: a subsisting “absolute form” (ST IA Q3 A3). That God cannot be composed of his nature and his esse, so understood, is simply a logical consequence of his aseity. But that is perfectly compatible with the divine nature, again so understood, being libertarian-free, at least to some extent.
To sum up. Two exegetical results, taken together, absolve Aquinas of the charge of inconsistency: the construal of God’s nature or essence not as a collection of apparently distinct but really identical “properties,” but as subsisting, absolute “form” of which God’s “properties” are merely notional expressions; and the rejection of absolute in favor of relative identity. Thus, the relevant identity relations entailed by ADS should be framed as: God is the same God as God's nature, which is the same God as God’s existence, which is the same God as God’s actus purus. That is why, even though God wills whatever he wills only eternally in one act of will, it remains logically open to Aquinas to claim that some of what God wills is necessary only ex suppositione not by necessity of nature. There is differentiation in God, and some of it consists in his (logically) contingent activity, which befits him but is not necessitated by anything other than his free, rationally befitting choice to manifest and communicate his goodness ad extra.
Blosser is basically correct to invoke that defense and to characterize Robinson’s conception of the divine nature as some sort of “engine of absolute necessities.” Of course anything with a nature is, by that very fact, such that certain things hold necessarily of it; yet for the reasons I’ve given, it does not thereby follow that whatever is identical with its nature does all that it does by necessity of nature. The only problem with Blosser’s critique is that he failed to explain why that is so for Aquinas and thus what was wrong with Robinson’s conception as an exegesis of Aquinas. I have filled that lack.
Apparently, Robinson’s larger purpose in attacking ADS, as exemplified in Aquinas, is to contrast the Latin tradition in natural theology unfavorably with the Orthodox, as exemplified in Gregory Palamas’ distinction between the divine essence and the divine energies. Not having read Palamas myself, I have nothing useful to say on my own account about his theology. But if what Robinson says about Palamas and the essence-energies distinction is reliable, then I see no reason to hold that Palamas’ natural theology is fundamentally incompatible with the Latin tradition about ADS. All we need is a more sensitive and less anachronistic reading of the Latin tradition.