5. Rival Ultimates

Chapter 4 tested the metaphysical rivals: positions that contest the machinery of Time, Space, and Substance without proposing a rival ground beneath it. This chapter turns to rival ultimates — accounts that do propose one. Each names an ultimate offered as more fundamental than structure, asked to occupy the ground the tradition reserved for God.

The classical-theistic constructions are the most fully developed and take up most of the chapter, but they are engaged as one family of rival ultimates rather than as a category apart. Each is taken in its strongest available form and pressed on one question: can an ultimate be specified as prior to the structural conditions without merely relocating the roles those conditions play? The inquiry distinguishes three verdicts. Some candidates fail because their proposed grounding relation has no determinate content on the book’s analysis. Some retain the structural roles under another vocabulary. Others remain coherent but carry a grounding or explanatory burden the chapter cannot close by contradiction. The bar is the strongest form the literature has yet produced; where a more powerful version exists or later emerges, the argument is answerable to it, not exempt from it.

The chapter works through three families of these, each gathered under its own cluster. The first are the theistic arguments that reach for a divine ground without presupposing structure: the cosmological and foundation arguments, the arguments from design and fine-tuning, the ontological and modal arguments, and the moral, pragmatic, and transcendental ones (§5.1). The second are the doctrines of the divine nature built to keep that ground ultimate without regress, simple without composition, and independent of the structure it grounds: divine simplicity and the via negativa (§5.2). The third are the accounts of the ultimate that are not classically theistic but make the same bid against structure: Spinozist pantheism, Neoplatonism, process theology, Madhyamaka emptiness (§5.3). §5.4 draws together what the engagements across these families reveal.

A note on scope: this chapter presses contemporary Christian-analytic theistic counter-constructions because that is where the literature has been most fully developed: Rasmussen, Plantinga, Pruss, Swinburne, Craig, Leftow, Stump, Dolezal. The structural pressures collected here are not in-house Christian difficulties. The same moves are made across the Islamic falsafa tradition (the medieval Arabic philosophical inheritance of Greek metaphysics): Avicenna’s contingency argument, Al-Ghazali’s Tahāfut (the Incoherence of the Philosophers) and the Kalām cosmological tradition the contemporary argument inherits. They are made too across the Jewish philosophical tradition (Maimonides on divine simplicity and negative theology). Where a parallel premodern development is doing the same conceptual work, it is registered in the relevant section. The argument is not against any one tradition but examines a structural pattern that recurs across traditions. It does not treat every use of divine, ultimate, or apophatic language as one undifferentiated target; each engagement is evaluated at the resolution and with the grounding commitments its own defenders specify.

A note on disaggregation: Aquinas’s Five Ways are not answered in a single section because the structural analysis does not treat them as a single argument. Each Way targets a different structural condition: motion and efficient causation (§6.7.2), contingency (Chapters 2–3 and 4), degrees of perfection (§§6.2–6.6), teleology (Chapter 7). Each is taken up where that condition is derived, so the Five Ways are met across the book rather than in one place.

5.1 Theistic Arguments for a Divine Ground

This first cluster gathers the theistic arguments that reach for a divine ground without presupposing structure: the cosmological and foundation arguments (§5.1.1–§5.1.4), the arguments from design and fine-tuning (§5.1.5–§5.1.6), the ontological and modal arguments (§5.1.7–§5.1.8), and the transcendental, moral, and pragmatic ones (§5.1.9–§5.1.11). Each reaches past the world toward a God offered as prior to and more fundamental than time, space, and substance.

Each is pressed on a single question: can it be carried through without leaning on the structural roles it was meant to stand free of? The recurring finding is often relocation rather than refutation. A ground that acts, knows, creates, sustains, or explains must at least bear directed dependence, relational differentiation, and real grounding content. Whether those roles entail the richer types of temporal succession, spatial extension, and causal persistence depends on the specific form of the ground and must be argued rather than assumed. A candidate that posits an essence-prior or transcendent ground may remain a live metaphysical proposal; the chapter asks whether it has supplied an intelligible grounding relation or merely named one.

5.1.1 Rasmussen’s Foundation Argument

Among contemporary arguments for a divine ground, Rasmussen’s foundation argument¹ presses the case on the narrowest and most rationalist basis: no appeal to revelation, scriptural authority, or parochial religious commitment, only the demand that things have explanations. It belongs to a lineage running from Avicenna’s contingency argument (burhān al-ṣiddīqīn) in the Kitāb al-Shifā,² through Leibniz’s Monadology³ and Clarke’s demonstrative cosmology, to its contemporary reconstructions in Pruss⁴ and Koons.⁵ Because it stakes everything on explanatory principles the present account also accepts, it earns engagement at full strength before any reply.

The argument proceeds in three moves.

1. The Foundation. Every contingent fact has a sufficient reason (the PSR), so the regress of reasons terminates in a being whose existence is not contingent on anything further, a Supreme Foundation.

2. Maximal greatness. Any limit on this Foundation’s nature would itself be a contingent fact, and explaining the limit would require something beyond the Foundation, which contradicts its ultimacy, so the Foundation must possess every positive attribute to the highest degree.

3. Personhood. Consciousness is among the positive attributes, so the Foundation is personal.

The result is classical theism reached by explanatory first principles alone: a maximally great necessary being, argued for without a premise the naturalist has already refused. What makes the argument difficult to resist is its rationalist honesty: it does not retreat into mystery; it runs the Principle of Sufficient Reason (the PSR: every contingent fact has a sufficient reason for being so rather than otherwise) without flinching and lets the derivation land where it lands.

What the argument gets right is granted from the start: explanatory regress does not terminate in a brute fact. Where it goes wrong is the identity of the terminus. The terminus is located by §2.7 in the structural roles themselves: T, S, and Φ (the three structural conditions, borne by Time, Space, and Substance) as the conditions under which anything is coherently a something-rather-than-nothing. Rasmussen accepts the PSR and rejects that terminus: the roles, he would say, are facts, facts require reasons, and the reason must be something outside them that can ground them without itself requiring grounding. The dispute is not about whether the PSR generates pressure; both accounts agree that it does. It is about where that pressure comes to rest. If the structural roles are themselves the conditions of coherent fact-hood, they do not stand in line with other facts awaiting a reason; they are what allows anything to be a fact waiting for one. The argument reaches its Foundation by three lines, and each, carefully followed, recovers the very structure it means to displace: the Argument from Limits, the Mind-Grounding Isomorphism, and the direct appeal to the PSR.

First. The argument from limits runs straight off the PSR: any limit on the Foundation’s nature would be a contingent fact requiring further explanation, so the Foundation must possess every positive attribute without restriction. Each restriction would point beyond the Foundation. Maximal greatness follows from the PSR itself, requiring no independent theological assumption, which is why it lands even on readers suspicious of theism.

The move risks treating every determination as a limiting defect. At role-resolution, a proposed Foundation must at least be determinate, distinguishable from what it grounds, and stand in a real asymmetric dependence relation to it. None of that yet proves clock-time, spatial extension, or ordinary causal persistence in an atemporal ground. It does show that maximality cannot be obtained simply by subtracting every form of order, difference, and bearing; a wholly indeterminate posit supplies no intelligible grounding content. The Argument from Limits mistakes the conditions for existence for restrictions on a hypothetical limit-free existent, and concludes by describing ∅ under a greatness-theoretic heading.

A theist will object that this describes ∅, not the infinite plenitude classical theism intends: a Foundation lacking every limitation is maximally full, not empty. But the objection trades on an equivocation the reply can refuse. The limit at issue is not privation, some perfection the Foundation lacks, but determination: to be anything at all, even infinite fullness, is to be this and not that, full and not empty, being and not nonbeing. The principle descends from Spinoza’s determinatio negatio est (determination is negation, Letter 50), generalized by Hegel into the omnis determinatio est negatio since quoted in his name, and pressed here to its theological conclusion. Determination is not a defect to be maximized away; it is what being a something costs, and a candidate that sheds it entirely does not rise above the roles but falls beneath them into ∅.

So the appeal to infinite fullness faces a fork: either it keeps that determinacy, and it is a determinate, therefore structured, ground, the floor already conceded; or it renounces determinacy, and the plenitude is ∅ under a superlative. Answering that the Foundation merely cannot be empty, since emptiness is no perfection to lack, guards it against privations it never faced and leaves untouched the determinacy it cannot shed. A bolder theist welcomes the second horn: the God of the via negativa is “beyond being,” so naming it ∅ merely fails to rise past determinate categories. But “beyond being” either marks something determinate, this and not that, and falls back into the first horn, or it marks no determination at all, which §2.7.1 establishes is not eminence beyond being but ∅, nothing under the honorific “beyond.”

Second. The mind-grounding isomorphism extrapolates from the way finite minds ground propositional content to the way the Foundation grounds existence itself: if minds carry meaning across cognitive acts (entertaining truths, sustaining explanatory relations, grounding one content in another), a maximally unrestricted version of that same relation is the best candidate for grounding reality. So Mind is the most general category of grounding, its range of configurations exceeding what structural mediation could confine.

Range of configurations is idle if the capacity to host configurations is missing. Any mind that entertains a proposition must carry that proposition across its articulation (Time), distinguish it from neighboring propositions (Space), and persist as the same mind across the act of entertaining (Substance). A mind that performs no such act is not a mind; it is the word mind with no referent. A classical theist will protest that this pictures only a finite, discursive mind: the divine intellect entertains nothing across an interval but grounds all truths timelessly and without succession, in one simple act or in nothing beyond its own nature. Grant the disclaimer entirely; it changes nothing, because grounding of any kind, timeless grounding included, rests on a minimal floor. An intellect that grounds truths without performing any temporal act must still stand to what it grounds in an asymmetric order of prior and posterior (T as grounding-order), remain distinct from those truths rather than collapsing into them (S as relational distinction), and retain its identity as their ground across the relation (Φ as identity-retention). A timeless, non-discursive intellect that grounds anything has thereby conceded the floor; one that meets none of these conditions grounds nothing at all. Timelessness lifts the mind above succession, not above structure. Rasmussen attributes to his Foundation everything a mind does (entertaining truths, grounding reasons, sustaining explanatory relations), which means the Foundation is already operating on the structural conditions the argument is supposed to explain. Run carefully, the Principle does not elevate Mind above T, S, and Φ. It shows that Mind is where T, S, and Φ operate most densely, which makes it a high-order instance of structure, not an alternative to it. The wider reach of mental grounding is not its escape from the roles. It is the roles running richer.

Third. The direct appeal to the PSR skips theistic content altogether and invokes the PSR unadorned: the structural roles are facts, facts require reasons, and the reason must be a necessary being that grounds them without itself requiring a ground. This version is difficult to answer without appearing to abandon rationalism, because refusing the PSR’s demand at the structural level looks like allowing a brute fact at the foundation.

The PSR has no purchase against the structural Conditions if the proposed contrast class is simply their total absence: §2.7 argues that ∅ is not a coherent alternative state. That result blocks a demand for a cause of the Conditions construed as an ordinary contingent fact. It does not, by itself, eliminate an essence-prior proposal that takes the Conditions to depend on a primitive metaphysical ground; such a proposal incurs a further explanatory burden rather than being answered merely by the absence of ∅. The present reply is that Rasmussen has not yet shown why a necessary structured ground requires an additional personal Foundation, not that every possible grounding gloss is already contradictory.

The Foundation, moreover, inherits a structural burden. Rasmussen can press that the relevant grounding or explanation need not be causal or temporal: the Foundation may ground without producing, explain without sustaining-through-time, or source without performing a causal act across an interval. The reply should therefore remain at role-resolution. Any intelligible source/ground relation must at least distinguish source from grounded, order the dependence asymmetrically, and obtain as a real explanatory or grounding fact rather than as a bare label. The book identifies these as the S-, T-, and Φ-roles in their minimal forms: relational distinction, asymmetric grounding-order, and real bearing or identity-retention. This does not by itself prove that an atemporal Foundation instantiates clock-time, spatial extension, or causal persistence; it establishes that the Foundation is not wholly structureless.

Analogical predication is the classical theist’s natural recourse at exactly this point: God grounds not univocally but analogically, so the floor is granted only in an eminent, non-creaturely mode. It changes nothing here. Either the analogical relation has determinate content, in which case it carries the ordering, distinction, and identity-retention the floor names, and a structured ground has been conceded under an analogical description; or it has none, in which case the relation specifies ∅ (§2.7.1). There is no third term between structure and ∅ for the analogical move to occupy. The full pressure on the classical via negativa runs at §5.2.2 and §5.4; the reply here needs only the floor, and the analogical move cannot slip it.

Reading the roles at this minimal level is not a retreat from the fuller T, S, and Φ the book defends elsewhere, nor an equivocation that proves a thin conclusion and asserts a thick one. The minimal reading fixes only the floor any grounding relation must meet (relational distinction, identity-retention, asymmetric ordering), and the reply needs only that floor: no candidate ground, however the theist refines it, drops beneath it. Whether the ground’s temporality, extension, and substantiality also take their richer everyday type is a further question the reply here neither needs nor settles. The richer type-upgrade is secured only where the separate change-realizer arguments apply; an atemporal ground stands outside that conclusion by hypothesis, so the reply against it rests on the minimal role floor alone. That is enough: a ground that refuses even the floor — asymmetric ordering, relational distinction, identity-retention — specifies ∅ (§2.7.1), while a ground that grants the floor has thereby conceded a structured ground, whatever it declines at the level of everyday type. There is no loophole by which a ground might satisfy the minimal conditions while escaping the roles themselves: that the floor (asymmetric ordering, relational distinction, identity-retention) simply is T, S, and Φ read at role level is fixed upstream, by the role-definitions of §1.4 and the Structural Requirement (§2.4), not asserted afresh here nor borrowed from §5.4. Condition-Individuation (§2.6.11) secures separately only that there is no fourth axis, and the remaining disagreement, whether a structured ground is aptly named God, is the one §5.4 takes up.

The classical admission that God cannot make a square circle was always a concession that coherence binds the Foundation. The present argument names what coherence consists in: the description of what T, S, and Φ together require, read at the level of relational distinction, identity-retention, and asymmetric ordering rather than specifically causal-temporal production. A Foundation bound by coherence is a Foundation operating within structure, not outside it.

The abductive variant on these arguments reaches the same terminus by inference rather than demonstration: best-explanation reasoning rather than deductive PSR application, as in Koons.⁶ The reply is parallel. A posit must be able to do the explanatory work attributed to it; doing that work requires determinate ordering, differentiation, and bearing at role-resolution. An abductively posited Foundation therefore carries structural commitments. Whether those commitments reduce the Foundation to the book’s structure or remain features of a distinct personal ground is the live abductive comparison, not something the role-floor alone decides.

Two objections press at this site. The selection-among-possibilities challenge asks: if Nature is the necessary structural foundation, what selects which contingent configuration it realizes? The question is internal to Nature: which configuration obtains within the necessary structure, not a question about Nature itself. The modal PSR as applied to contingents within Nature (§4.1.3) answers the explanatory pressure without an external intentional selector, while the absence of a coherent contrast class for Nature’s non-instantiation (§2.7) blocks projecting the question back onto the Conditions; the level-distinction at §2.8.4 fixes why, since contingency is a property of configurations within Nature, not of the necessary structural ground. The aggregation challenge asks: if T, S, and Φ are each necessary, why are they not three independent necessary foundations that aggregate to one, the structural form of monotheism? Because T∧S∧Φ is not a conjunction of three independently necessary grounds but one structural fact under three logical-categorical readings. The Triunity of Nature (§2.6.9) and Condition-Individuation (§2.6.11) establish that instantiation of any one Condition analytically requires the others, so there is no coherent specification of T, S, or Φ as a foundation prior to their conjunction. Rasmussen’s aggregation move, which correctly collapses multiple independent necessary foundations to one, does not reach T∧S∧Φ, because the Triconditional identity forecloses the independence it requires.

What converges. Each of the three lines to Rasmussen’s Foundation recovers the structure it set out to replace: the Argument from Limits mistakes the structural conditions for restrictions on a hypothetical limit-free existent and ends up describing ∅ under a greatness-theoretic heading; the Mind-Grounding Isomorphism attributes to the Foundation everything a mind does (entertaining truths, grounding reasons, sustaining explanatory relations), so the Foundation is already operating on the structural conditions it was meant to explain; and the direct PSR line has no purchase where the proposed contrast class is ∅, which fails before it has been described; the essence-prior gloss that survives carries a further explanatory burden rather than an answer, leaving a Foundation bound by coherence as a Foundation operating within structure.

Rasmussen is right that explanatory regress must terminate somewhere, and right about nearly everything the terminus must be. The one divergence is its identity: not a necessary being standing over the roles, but the roles themselves, which are necessary because nothing is a something without them. What classical theism has been pointing at under the name Foundation is what Chapter 1 and §2 have been proving under the name T ∧ S ∧ Φ. For neighboring contemporary positions in the same methodological cluster, the engagement runs at §3.4 and §3.5 rather than here: Pruss and Rasmussen,⁷ and Pizzi.⁸ The cluster reaches □∃x x; the five-domain convergence reaches □(T∧S∧Φ); the latter closes candidates the former leaves open. Pizzi⁹ is the closest in formal method, and explicitly leaves the identity of the necessary existent open in his §7, a stop-point this argument crosses. Divine simplicity is taken up next, at §5.2.1: whether the necessary ground can be identical with its attributes in a way that exempts it from the structural conditions this section has established apply to any ground capable of the grounding work. Rasmussen’s personalist last stage (the move from necessary foundation to intentional selector, and from intentional selector to personal being) is engaged directly at Chapter 6: §6.4.1 (Awareness), §6.4.2 (Knowledge), and §6.7.1 (Power) test whether awareness, knowledge, and agency are genuine additions to the necessary structural ground or configurations it realizes. The lever Rasmussen would press hardest is the book’s own Recursive-Depth → Phenomenal-Interiority arc (Chapter 8): the place where the framework most invites the personalist conclusion, since it is the book’s own account of how interiority arises. That recursive structural depth gives rise to phenomenal interiority is not a concession that the ground is antecedently personal, but the explanation of how persons arise within the structure rather than standing over it. Two precision notes on scope govern the divergence. The book’s necessity claim is de dicto with respect to the structural role (necessarily, any ground that discharges the modal-grounding function is T∧S∧Φ as a co-instantiated conjunction), not the de re claim that this particular token instantiation is necessarily this token. The modal-collapse risk (whether necessary identity collapses contingent configurations of Nature into necessities) is foreclosed by the level-distinction §2.8.4 establishes: necessity attaches to the Conditions, not to the configurations they permit to vary.

5.1.2 Cosmological Argument

The First Cause demand is natural theology’s oldest cosmological move and its most systematic: if T, S, and Φ are necessary, what caused them to be? The demand presses the account of §§2.1–2.6 to either produce a cause outside the structural order or concede that necessity has a source the account has not named. Three claims carry the reply. First, the Kalām traversal argument is a spatial metaphor applied to a successional order, and Oppy’s deflation of Hilbert’s-Hotel absurdities¹⁰ shows the metaphor does not transfer. Second, the contingency-of-the-whole move conflates the contingency of individual events with the contingency of the order they occur in and illicitly exports the former to the latter. Third, the PSR appeal, pressed as a demand for something rather than ∅, presupposes a contrast class whose incoherence §2.7 has already established; pressed as the Big Conjunctive Contingent Fact, it demands a necessary explainer outside Nature that the PSR alone does not deliver, as §4.1.3 shows.

The three forms sort into two wings the reply treats differently: the necessity-seeking wing (Leibniz, Koons, Pruss) is answered by being grounded, since □(T∧S∧Φ) already supplies the necessary terminus it reaches for; the causation-seeking wing (the Kalām and its kin) is answered by being dissolved, since a boundary is a feature of the order, not an event outside it.

The argument concedes at the outset that the Leibnizian tradition is correct on a narrower point: causal series within the structural conditions do admit, and often require, further explanation. The claim pressed here is weaker than the target that tradition is built to refute. The order through which causal series run is not itself a series. Causal vocabulary applies to what occurs within the order, not to the order itself. A framework that makes causation possible is not among the candidate effects causation explains.

The cosmological tradition offers a resilient natural-theological resistance to the sufficiency of structural necessity. Its core commitment is simple: something must have started things. The universe is a vast succession of causes and effects extending backward in time; either that succession is infinite or it has a first member. An infinite past, the objector contends, is incoherent: actual infinities cannot be traversed, and we could never have arrived at the present if infinitely many prior moments had to elapse first. A finite past therefore requires an originating cause, and such a cause cannot itself lie among the events it initiates. The argument’s classical roots run through the Islamic kalām tradition (medieval speculative theology that pressed dialectical-rational defenses of revealed doctrine): Al-Kindi’s On First Philosophy¹¹ gives the earliest systematic case for the impossibility of an infinite past, and Al-Ghazali’s Tahāfut al-Falāsifa¹² develops the mature premodern version that the contemporary recovery directly invokes.¹³ The point has been refined by its contemporary inheritors into three distinct argument forms: the Kalām appeal to traversal, the Koonsian appeal to mereological contingency, and the Pruss appeal to the Principle of Sufficient Reason.¹⁴ Each converges on the same conclusion: a necessary being outside the causal order, identified with God. The present account lands next to the Leibnizian wing of this tradition — □(T∧S∧Φ) and their necessary ground being adjacent endpoints — and diverges from it only on the line taken and the commitments incurred along the way.

The challenge bears directly on §§2.1–2.6. T, as defended there, is a necessarily instantiated structural role, derived without PSR machinery and without any external originating cause. The First Cause objector grants the necessity and demands a cause for it: the account has explained why Time must obtain, not what started it. Three forms of the move take up the objection; the narrower point that individual causal series within the order admit explanation stands untouched.

First. The Kalām traversal argument, regimented, runs:

1. Causal principle. Whatever begins to exist has a cause.

2. Finite past. The universe began to exist: an actual infinite sequence of past events cannot be traversed, for if infinitely many moments had to elapse before now, now could never arrive, so the past is finite.

3. Conclusion. The universe therefore has a cause; and a cause not itself in time must be a timeless agent, which is God.

The traversal argument is a spatial metaphor dressed as a structural claim. ‘Traverse’ treats the past as a finished road Now had to travel down, a road whose length must be completed before the traveler arrives. But in a successional order, Now is not a traveler reaching a destination. It is the current coordinate of the order’s state. You do not need to ‘finish’ infinity to be at a specific point within it, any more than you need to visit every point on an infinite plane to be at coordinate (0,0). Oppy¹⁵ formalizes the point: the absurdities of Hilbert’s Hotel require the simultaneous rearrangement of discrete objects, a spatial and substantial paradox that does not map onto temporal succession. You cannot shift past moments aside to make room for the present. The mathematical tools Craig invokes to generate absurdity do not apply to the domain he applies them to.

Craig¹⁶ has pressed a generalized version of the point in reply: the absurdities follow from the actuality of the infinite itself, not from any rearrangement procedure; Hilbert’s Hotel is a vivid illustration of a more general impossibility rather than its only form. Granting actual-infinite-absurdity in general (the cases involving completed infinite collections accessed as totalities), the question is whether a temporal past counts as that kind of completed-collection-accessed-as-a-totality. It does not. On a tensed or passage-centered theory, the past need not be treated as a completed collection accessed from outside; on a tenseless block theory, the past can form a completed ordered totality. The traversal objection therefore depends partly on the theory of time and cannot be dismissed by one model alone. The tensed/tenseless question is the standing pivot of the Kalām dialectic, but the account need not adjudicate it here: on the composition defended at §7.2.2, tenseless order and tensed passage are not rival readings of Temporality but the two features it jointly requires, and the reply holds on each. Taken as ordered dimension, the traversal worry does not arise: there is no formation to complete, only an ordered manifold at whose interior Now sits. Taken as directed passage, grant the finitist the point in full — successive addition, if it establishes anything, establishes that the past is finite; that much is conceded below, since the account is compatible with a finite-past cosmology. A finite past is a chronometric boundary within T, the earliest coordinate the passage has reached from within, not a moment-prior-to-T from which T departed. So the theological inference fails on both features alike: what it needs is not a boundary but a cause outside the order, and the boundary is a feature of the order (§7.2.2). The Kalām’s own conclusion about finitude is granted; only the step from finitude to an external cause is refused. Craig’s stronger reply keeps open the question whether a beginningless temporal order can be actually infinite without anyone traversing it from an external vantage. Oppy’s response weakens the traversal metaphor; it does not by itself settle every argument against an actual infinite past.

Craig’s empirical prong is separable from the conceptual one. The Borde–Guth–Vilenkin theorem¹⁷ is standardly cited as proving the universe had a past boundary, and Craig treats it as empirical confirmation of Kalām’s second premise. Two points cut the weight. First, BGV establishes only that geodesics with positive average Hubble expansion are past-incomplete, not that the universe as such began. Aguirre and Gratton¹⁸ and Carroll¹⁹ describe eternal cosmologies in which past-incompleteness of inflating geodesics is compatible with an eternal ambient spacetime. Craig and Sinclair’s Blackwell Companion reply²⁰ presses that Aguirre–Gratton requires a quiescent low-entropy past boundary that itself functions as a beginning, and that Carroll’s eternal cosmologies require thermodynamic time reversal at a past boundary that operates likewise; the 2014 Greer–Heard Forum exchange with Carroll restates the point.

Granted: the quiescent boundary, where one is required, is a beginning in the chronometric and thermodynamic senses. It is not a beginning in the categorial sense the Kalām requires. A past boundary is the first coordinate of T along the dimension (the earliest position the passage has reached from within), not a moment-prior-to-T from which T departed. The chronometric beginning is a feature within T (§§2.4.6, 7.2.2); the categorial beginning, which the Kalām needs and the cosmological model does not deliver, is foreclosed by Parasitic Negation on T (§§2.1.5, 5.1.6).

Second, even if the universe has a finite past boundary in its own time-coordinate, that boundary is a feature within T, S, and Φ: the first chronometric coordinate of the successional order, not an event caused by anything outside it. A boundary of the order does not transport the order’s explanation outside the order. The argument concedes the empirical finding and denies the theological inference it is supposed to underwrite.

A newer wave of anti-infinite-past work has migrated away from actual-infinity absurdities toward paradox-based causal finitism. Koons²¹ presses Grim Reaper cases (scenarios in which an infinite staggered sequence of potential killers generates contradiction) as evidence that infinite causal sequences are metaphysically impossible even when mathematically coherent. Pruss’s Infinity, Causation, and Paradox²² generalizes the point into a systematic case. These arguments are not touched by Oppy’s Hilbert deflation, because they do not turn on Hilbert-style rearrangement; they target the supposition that infinitely many prior events could each causally produce the next. The present reply does not contest causal finitism in general (the account is compatible with a finite-past cosmology) but denies that causal finitism, even if sound, establishes what Kalām requires. A finitely-extended past is still a past within T; its first chronometric coordinate is the earliest position the passage has reached from within, not an event caused by something outside the order it would initiate. The Kalām framework presents this as a forced choice between (i) the first moment of T has no cause (denying Premise 1) and (ii) the cause is not within T (and so outside T). The present account takes a third position. The present account treats the first chronometric coordinate as a boundary of the temporal order rather than automatically as an effect of something outside it. The Kalām defender can contest that category assignment and argue that the universe’s beginning is an event falling under the causal principle. The dispute therefore turns on the scope of causation, not on a contradiction generated merely by asking for a cause. The categorial-vs-chronometric distinction is developed directly at §5.1.6: what is foreclosed is a moment-prior-to-T from which T departed; what is admitted is a chronometric boundary as the first coordinate of T itself. Grim Reaper considerations constrain the shape of causal series; they do not relocate causation to the order’s exterior.

A further reply is available, though the categorial response does not require it. Grim Reaper cases idealize location, propagation, and mechanism in order to isolate a causal-structural paradox. The idealization cannot be rejected merely because it omits ordinary physical realization; the relevant question is whether the causal stipulations are jointly satisfiable and whether their contradiction generalizes from the constructed series to every beginningless causal past. Physicality is a substance carrying causal power across T-succession through S-extension (§1.4, §2.4.11); a putative cause that occupies no region and transmits nothing across one is not a faint or limiting-case cause but no cause at all. The dilemma this poses to the causal-finitist deployment is exact. Either the reapers are genuine causes — located in S, carrying Φ, propagating at finite speed — in which case the impossibility the converging series generates is located in Spatiality and Physicality rather than in infinity or in causation as such: infinitely many physical signals converging on a single instant drive their spatial separation toward zero and their energy density without bound, a merely physical impossibility that establishes nothing about a metaphysically necessary First Cause. Or the reapers are the idealized spaceless triggers the paradox in fact stipulates, in which case what it proves impossible was never a causal sequence on the present account, and ‘infinite causal sequences are impossible’ mislabels a result about an unsatisfiable non-causal stipulation. A beginningless past is untouched on either horn: it is not a staggered swarm of located conditional agents racing to act inside a shrinking window, but the past-directed order of T itself, whose earliest coordinate is a boundary rather than an effect. The Grim Reaper thus constrains, at most, the arrangement of embodied causes; it does not reach the successional order those causes run within, which is the only thing the Kalām needed it to reach.

This verdict converges with the critical literature rather than standing alone. The unsatisfiable-pair diagnosis — Shackel’s demonstration that Benardete constructions merely encode two jointly unsatisfiable conditions,²³ pressed against causal finitism by Schmid and Malpass²⁴ — locates the contradiction in the stipulation rather than in the infinity, and observes that purely non-causal Benardete variants (in which each reaper’s action brutely correlates with its predecessors’ rather than being caused by them) generate the paradox with no causal chain for finitism to sever. That non-causal variant is decisive here: if a swarm carrying no causal power at all yields the same contradiction, the impossibility was never a fact about causation reaching infinitely back. The present account agrees on the diagnosis and adds only that the past-directed order of T is not a Benardete series to begin with, so nothing in the family bears on whether that order has a first coordinate.

Loke²⁵ presses a further refinement: the only coherent transition from a timeless cause to a temporal effect is free agential initiation, so the first cause must be personal. The refinement inherits the structural reply already given. A ‘timeless cause’ stands outside T by hypothesis, and standing outside T is standing outside the conditions under which causing has content; a free agent who initiates temporality must persist as the same agent across the initiation (Substance), distinguish its pre- and post-initiation states (Space), and stand in a before-and-after relation to the effect it initiates (Time). The personal-agent refinement does not escape the structural conditions; it concentrates them. Loke’s argument is a contribution to the personalist question deferred to §5.1.4, not a distinct way past the present response.

Morriston presses two further objections the account can absorb. The first is inductive: the causal principle that anchors the Kalām, that whatever begins to exist has a cause, draws its support only from cases of rearrangement within the order, so its warrant does not extend to an absolute beginning of the order itself.²⁶ The account does not lean on this objection, since its own reply is categorial rather than inductive, but the objection reinforces from a second direction that the causal idiom is at home within the order and strains when lifted outside it. The second objection is sharper against the personalist fix: a timeless agent’s producing a first effect is itself a transition from not-creating to creating, and a transition is a succession, so the agential initiation Loke and Craig invoke reintroduces the very order it was meant to precede.²⁷ That is the structural reply already given, reached from the opposite side: the transition presupposes Time, Space, and Substance rather than delivering them from outside.

A corollary follows if the theist absorbs Morriston’s point rather than resisting it. The standard absorption denies the transition outright: God’s willing of the world is timeless and immutable, so there was never a state of not-creating for the first effect to succeed. The absorption then leaves one exit from necessity: a willing immutable in fact yet able to have been otherwise, the necessity of supposition Aquinas distinguishes from absolute necessity.²⁸ The exit keeps the world contingent, but at the price of a modal contrast (a possible circumstance in which the changeless willing is absent) that nothing in the timeless picture grounds; whether that contrast survives divine simplicity is the modal-collapse question pressed at §4.4.9 and engaged at §5.2.1. The theist who declines the exit, identifying the timeless willing with the necessary divine essence so that no modal distinction remains between what God is and what God wills, delivers the present account’s conclusion in theological dress: there is no vantage, temporal, volitional, or modal, at which God exists and the world does not, and obtaining in every circumstance is necessity under its ordinary modal name. Those premises yield a necessary natural order directly: □(T∧S∧Φ) with a divine middleman attached. And the middleman contributes nothing to the necessity itself, since it is the modal closure of the willing, not any further feature of the agent, that closes off every circumstance lacking a world. What remains in dispute on that horn is not whether the order is necessary but whether its necessity needs an owner, and that is the personalist question already deferred to §5.1.4.

Second. The contingency-of-the-whole move grants, with Koons,²⁹ that an infinite past is possible. The series still consists entirely of contingent beings, events that did not have to happen. A series of contingent beings, even an infinite one, is itself contingent. The contingency of the whole demands a necessary cause outside the series. That cause is God.

The move trades on two senses of contingent. Individual events may be contingent even if the structural possibility of an order is necessary. It does not follow that this particular total history is necessary. Koons’s argument gets traction if the Big Conjunctive Fact or total series is itself a contingent fact requiring explanation, whether or not it is an additional concrete object over and above its members. But a succession is not a thing over and above what succeeds; it is just the parts in the order they take. And that order obtains within a necessary framework whose conditions are what make succession possible at all. Member-level contingency is real. Whole-level contingency, treated as a further fact requiring a further ground, is a misattribution: a property that belongs to what sits inside the framework has been exported to the framework holding them. Hume’s Dialogues IX³⁰ and Paul Edwards³¹ diagnosed the underlying slip as a fallacy of composition, treating a property of the parts (contingency) as a property of the whole, and the diagnosis is sound and worth crediting. The reply does not, however, rest on the Hume–Edwards explanatory principle that explaining each member thereby explains the conjunction, the very inference Pruss’s contingency argument is built to reject. Its load-bearing move is the one already made: whole-level contingency is a misattribution because the order obtains within a necessary framework, and the demand for an external ground of the series-as-a-whole has no coherent target once ∅ and a necessitating Nature are both foreclosed (§4.1.3). Koons’s mereological premise is precisely what is supposed to rescue the argument from the composition charge; denying that the whole is an entity over and above its members withholds that premise without leaning on the Hume–Edwards inference. What Koons locates at the top of his derivation (necessary ground) is already present at the bottom, as T, S, and Φ. The accounts share a destination. This one arrives without the mereological premise that the series is an entity additional to its elements, a premise Koons himself flags³² as load-bearing and contestable.

Third. The PSR appeal presses, with Pruss,³³ that every contingent fact requires a sufficient reason. The entire sequence of contingent events is itself a contingent fact: it could have gone differently in countless ways. Its sufficient reason must be a being whose existence is not itself contingent, and whose nature explains why this history rather than another obtains. That being is God.

The PSR question splits into two the reply must keep apart. Read as ‘why an ordered history rather than none at all?’, it presupposes ∅ as the contrast, and ∅ is no contrast: §2.7.2 closes it on independent modal and temporal grounds, and sufficient reason bottoms out where alternatives run out, which is at the roles that make alternation possible. Read as ‘why this configuration rather than another?’ — Pruss’s Big Conjunctive Contingent Fact, the conjunction of all contingent configurations of Nature — the alternatives are genuine and the PSR does real work: rival configurations compete within the structural conditions, and each contingent fact keeps its contingent explanation there.

What does not follow is the further step, that the conjunction as a whole requires a necessary explainer outside Nature. The book denies that this demand automatically requires an explainer outside Nature. But the alternatives are not exhausted by ∅ and a Nature that necessitates every configuration: a theist may posit a necessary personal ground whose contingent intention explains one history without necessitating it. That proposal must face the grounding, freedom, and modal-collapse burdens developed in §§5.1.1–5.2.1 rather than being eliminated by the PSR split alone. The dissolution does not run through the Hume–Edwards principle that explaining each conjunct explains the conjunction, the inference Pruss’s argument is built to reject; it denies that the external demand has an automatic target, as §4.1.3 develops in full.

Hume anticipates the move in Dialogues IX:³⁴ Cleanthes’ reply to Demea is that if a necessary being must be posited at all, the material universe itself is a candidate, and the theist offers no non-theological ground for disqualifying it. The present argument completes Cleanthes’ gesture. The ‘necessary material universe’ Cleanthes floated was not specified enough to bear the weight; □(T∧S∧Φ) specifies it. The Humean cosmological option the Dialogues left undeveloped is the one the structural account takes up. Pruss’s own defense of the PSR against Humean attack³⁵ is uncontested here; the present argument simply does not conscript the principle into service outside the domain where it has purchase. The standard critical literature presses harder, rejecting the principle rather than confining it: Rowe³⁶ argues that the PSR is not rationally compelling and can be denied without irrationality, and Mackie³⁷ presses both the causal principle and the demand for an explanation of the totality as unmotivated once every member has been explained. The reply here needs neither rejection. It grants the PSR full force within the structural conditions and withholds only its application to them, a narrower and less contestable claim than the outright refusal Rowe and Mackie mount.

A kindred deflationary move presses from the opposite side: the call for a first cause is a pseudo-question, Rundle³⁸ and Grünbaum³⁹ argue; ‘why is there something rather than nothing’ has no grip because no coherent answer-space exists. The role account agrees with half of this and rejects the other half. It agrees that ‘what caused the structural order?’ terminates, because the idiom of causation applies within that order rather than to it. It rejects the dismissal of the further question ‘why is the order necessary rather than absent?’, because that one does have an answer: §§2.1–2.8 give it. The cosmological demand is not empty; it is a question with a structural answer where theism has given a theological one. Rundle and Grünbaum close the inquiry without resolving it; the present argument resolves it without closing it.

Kant frames the deepest version of the critique. The cosmological argument extends causality beyond the field of possible experience, where alone the concept has warrant, and so misapplies a category whose jurisdiction is the world of appearances (Critique of Pure Reason);⁴⁰ Kant presses further that the argument covertly relies on the ontological proof it claims to do without. The First Antinomy sharpens the impasse (KrV):⁴¹ equally compelling cases can be built for a first beginning in time and against one, which for Kant shows the question has outrun the conditions under which causal reasoning stays legitimate. The account agrees that causation does not apply to the order as a whole, and for a recognizably Kantian reason: the idiom belongs to what occurs within the field, not to the field itself. It parts from Kant on the status of the conclusion. His verdict is epistemic, that reason cannot settle the question and must fall silent; the verdict here is metaphysical, that the question is malformed, because the order is necessary (§§2.1–2.6) and its necessity has a structural ground rather than a causal one. Where Kant reaches a limit of knowledge, the account reaches a categorial boundary.

One objection cuts across all three forms and is the family’s most resilient. Schopenhauer⁴² named it the Taxicab objection: the Principle of Sufficient Reason is hired to carry the argument as far as a first cause and then waved off once the destination is reached, the terminus exempted from the very demand that powered the regress. Oppy⁴³ presses the contemporary form as a symmetry: whatever explanatory privilege the theist claims for the divine terminus, whether necessity, self-explanation, or brute factuality, the naturalist can claim with equal right for the universe or its fundamental laws, so the demand for a cause of the order underdetermines the theological answer. The objection lands against any account that stops by fiat. It does not land here. The account exempts nothing: it grants the structural order no standing waiver from a demand that otherwise applies, but denies that the demand was ever well-formed against the order, since causation relates elements within the order and the order is not one of its elements. The symmetry Oppy exploits is real, and it is conceded rather than resisted, because it is not met by making the necessary ground simpler or more unified than a physical candidate. It is met by locating necessity in T, S, and Φ themselves (§§2.1–2.6), which are neither a divine terminus nor a rival physical one but the conditions any candidate terminus, divine or physical, would have to instantiate. The taxicab is never hired, so there is no fare to dodge.

A note on division of labor. Several adjacent theistic moves belong to neighboring sections and are deferred there rather than redeployed here. The personalist extension (why think the necessary ground is a mind rather than a set of structural roles?) receives its dedicated engagement in §5.1.4 (Swinburne’s simplicity argument) and §5.1.1 (Rasmussen’s Mind-Grounding Isomorphism). Divine simplicity as the classical counter-ground runs through §5.2.1. The Thomistic per se causal series and the Aquinas/Feser/Kerr pure-actuality response are handled in §§5.1.3 and 5.2.1. The Boethian eternal present is §4.2.1. The worry that structural necessity propagates to Spinozan modal collapse is §4.4.9. The PSR as a standalone rationalist demand (distinct from its deployment here as a first-cause lever) receives its fullest engagement in §4.1.3. The present section confines itself to the specific demand that the order require a cause; the surrounding terrain is covered in those sections.

A final consideration turns the objection back on the conditions of its own statement, a move called retorsion: the call for a first cause presupposes what it is calling on to be caused. The asking unfolds across moments, the words occupy relational positions, and the thought persists across its articulation and transmission: an objection to the necessity of T, S, and Φ raised within T, S, and Φ. The obvious reply is that enacting the structure while demanding its cause shows only the structure’s actuality in the world where the demand is issued, not its necessity. Granted, but modal necessity does not need to be closed here; §§2.1–2.6 did that. This section defends the sufficiency of that closure against calls for an additional explanatory step. What the retorsion shows is that every idiom in which the call can be issued (temporal succession, spatial articulation, substantial persistence) is parasitic on the very order the call purports to require a cause for. Causation explains moves inside the field. It cannot explain the field: not because the field is mysterious, but because ‘cause’ is a relation between elements of the field, and the field is not one of its own elements.

What converges. Each of the three forms concedes the very order it means to extend: the Kalām traversal treats the past as a finished road Now had to travel down, a spatial metaphor that does not map onto successional order, where Oppy’s Hilbert deflation blocks the rearrangement absurdities and any finite past boundary is a feature within T rather than an event outside it; the contingency-of-the-whole move concedes a necessary framework and then misassigns member-level contingency to it, committing the fallacy of composition Hume and Edwards diagnosed; and the PSR appeal either presupposes ∅ as a contrast the structure forecloses or, in its Big-Conjunctive-Fact form, demands an external necessary explainer the PSR split alone does not deliver (§4.1.3), leaving the personal-ground variant to the burdens of §§5.1.1–5.2.1, with sufficient reason bottoming out where alternatives run out.

The narrower point (that individual causal sequences admit further explanation) stands; the argument does not contest it. What the argument contests is the step beyond: the assumption that the order through which sequences run is itself a sequence, that the language of causation lifts from elements to their ground, and that the call for a first cause can be satisfied by anything other than what makes calls, causes, and sequences possible in the first place. The regress objector runs on Nature while asking what Nature rests on. It does not rest on anything; it is what makes resting on possible at all. With the First Cause demand answered, §5.1.3 turns to the mechanism classical theism offers the First-Cause picture: the per se series of simultaneous causation. Swinburne’s probabilistic recovery of the inference — theism pressed as a simpler hypothesis than its rivals — is §5.1.4.

5.1.3 Simultaneous Causation

The order of cause-before-effect is not the only order causation admits. A hand moving a stick that presses a stone illustrates concurrent dependence: while the contact and applied force obtain, the stone’s acceleration is produced without a temporal gap between the applied force and its effect. Releasing the hand need not stop all motion instantaneously, because inertia remains. Simultaneity is a causal structure too, and Aquinas sees in it the template he needs: an eternal God could simultaneously ground every moment of a temporal world without entering the time of what it grounds. The per se causal series runs through a stack of concurrent dependencies rather than a sequence of prior ones. Cause at the top, effect at the bottom, no temporal interval between them. Remove the top and the whole column falls in the same moment. If the picture works, the stack can be finite upward while terminating in an atemporal source, and the atemporal cause at the summit never needs to enter the time of the stones it moves.

Simultaneous causation is the mechanism that lets classical theism have both the atemporal God and the temporal world. The tradition developed the doctrine to resolve a specific problem: how an atemporal, unchanging being could cause a temporal, changing world without either being changed by the causing or leaving a gap in the causal order. Aquinas (ST I.2.3) deploys simultaneous causation in the First Way, the per se series that stops at an Unmoved Mover. Aquinas’s De Aeternitate Mundi addresses explicitly the coexistence of eternal cause and temporal effect. Avicenna’s continuous-creation doctrine in the Kitāb al-Shifā develops the parallel position in the Islamic tradition: God is the necessary, atemporal cause whose causing of contingent existents is itself eternal even though the existents are temporally extended.⁴⁴ The continuous-creation model is not a separate escape but the same atemporal-sustaining picture in Islamic dress, and it meets the reply given below to the occasionalist and conservationist kindred: a continuously creating cause must still individuate what it sustains (S), order it (T), and carry it across (Φ). Kant’s Critique of Pure Reason refines the concept in the Second Analogy, distinguishing logical from temporal priority and illustrating with the ball resting on the pillow. Stump and Kretzmann’s “Eternity”⁴⁵ proposes ET-simultaneity as a relation between the eternal present and each temporal moment. Craig’s God, Time, and Eternity⁴⁶ develops the contemporary version closest to biblical-theism commitments: God timeless without creation, temporal with creation. Against the doctrine: Wolterstorff’s “God Everlasting”⁴⁷ presses the case for divine temporality against Boethius and Aquinas. Padgett’s God, Eternity, and the Nature of Time⁴⁸ defends a relative-timelessness alternative. Mullins’s The End of the Timeless God⁴⁹ presses the most sustained contemporary argument that timelessness is internally incoherent. Morriston⁵⁰ has pressed the simultaneous-causation move specifically for two decades: if the cause is eternal, the effect must be equally eternal. The present reply sits closest to the Mullins–Wolterstorff wing in holding that timeless grounding does not cohere, and adds what that wing does not specify: the failure is not primarily biblical or doctrinal but structural; the logical priority the doctrine requires is itself a structured relation, and the structure is what was supposed to be escaped.

Applied here, the move offers the theist a path to concede the structure’s dominion over the created order while preserving God as external ground. The argument holds that any ground of the structure must itself instantiate the roles. The simultaneous-causation theorist counter-claims that grounding need not be temporal, spatial, or substantial: it can be purely logical, a relation of eternal dependence that places the effect within T, S, Φ (the three structural Conditions, borne by Time, Space, and Substance) without placing the cause there. On this picture, the demand that grounds instantiate what they ground fails at the central claim: God grounds the structure atemporally, the structure grounds creatures temporally, and the argument’s reach stops one level short of fundamentality.

The move bears directly on the Cartesian Certainty for change of §2.2, the Intermediate Necessity of §2.3, the identification in §2.4 of the modal constraint with T, S, and Φ, and the Triconditional account in §2.6 of the Conditions as Nature. Simultaneous causation is the specific mechanism by which classical theism tries to exempt God from the structural conditions while accepting their application to creatures. If the exemption holds, a changeless ground is possible after all, provided the ground relates to change by logical rather than temporal priority. Three forms of the move take up the objection: the per se causal series, Kant’s logical/temporal distinction, and Craig’s modal refinement. The replies below argue that each discovers temporal asymmetry inside the logical priority it was supposed to replace.

First. The per se causal series, regimented, runs:

1. Two series. A per accidens series (events in time) differs from a per se series (grounding relations that hold at every moment).

2. The First Way. Aquinas’s First Way holds that a per se series of causal dependencies requires no temporal extension: it is a relation of ontological priority rather than temporal sequence.

3. Conclusion. So God can be the per se cause of the world without being in time: the world depends on God at every moment, and the dependence relation is not itself a temporal event.

If C is a sufficient cause for E, whenever C obtains E obtains. The will of the Maximally Great Being (MGB) is held to be eternal, changeless, and sufficient. An eternal sufficient cause entails an eternal effect. Morriston⁵¹ presses the point: if the cause is eternal, the effect must be equally eternal; if the effect is finite, a simultaneous cause must be equally finite. (The Stump–Kretzmann ET-simultaneity reply to this eternal-cause/eternal-effect point is engaged at §4.2.1.) The per se series resolves nothing, because the structural question reappears within it: how can an eternal, changeless ground produce a finite, changing effect without the ground itself being modally or structurally differentiated between its producing and its non-producing state? The doctrine gestures at logical priority while silently importing the temporal directionality it was supposed to avoid. If temporal priority is denied, the causal relation must still supply asymmetric dependence; otherwise the remaining relation is structurally indistinguishable from simultaneous coexistence, two things obtaining together without one producing the other. The issue is not whether every cause must come earlier in clock-time. The issue is whether “cause” has determinate content without ordered priority, distinguishable relata, and a stable dependence relation. Those are the structural roles in non-clock form. That identification is not stipulated here. The minimal floor was established independently at §5.1.1, showing that even a timeless grounding intellect must exhibit asymmetric grounding-order, relational distinction, and identity-retention across the relation in order to count as a ground at all. Asymmetric ontological priority is not being relabeled temporal by fiat; it is being held to the floor that any grounding relation must clear, and that floor is T, S, and Φ read at their most minimal. The simultaneous-causation theorist cannot keep the priority while refusing the floor, because the priority is precisely what the floor articulates.

The deeper problem cuts across all three forms: whatever the priority relation is called, simultaneous causation turns the contingency of creation into a dilemma. If God is necessarily creator or sustainer, and the sustaining relation is sufficient, then the sustained order follows necessarily from God. A necessary sufficient sustainer does not leave the sustained order contingent unless the sustaining relation is itself contingent. But if the sustaining relation is contingent, creatorhood is not necessary in God; and if the sustaining relation is necessary, the particular order God sustains is necessary with God. It will not do to answer that Nature is necessary in any case: the necessity of the structural conditions is affirmed here (□(T∧S∧Φ)), and §4.4.9 shows why necessary structure does not necessitate any particular configuration within it. That reply is available for the structure and unavailable for a contingent created world. A necessary complete sufficient cause would necessitate its effect. The classical theist therefore distinguishes God’s necessary existence from a contingent free willing or from the contingent obtaining of the creation relation. Whether that distinction is coherent under timelessness and divine simplicity is the modal-collapse question; simultaneous causation alone does not settle it. The doctrine alone has not preserved a free creator beyond Nature; it has purchased the atemporal ground at the price of leaving in dispute the contingency it was meant to protect.

A second pressure compounds the first. Creation, considered not as a clock-time event but as a dependence relation, has structural requirements of its own. A creating relation has an asymmetric order: source and recipient are not interchangeable, and the dependence runs from ground to grounded rather than the reverse. It has a distinction: creator and creature must be distinguishable relata rather than one undifferentiated item. It has relation-stability: the recipient must be something whose being is held by the source, not a label that vanishes as soon as the relation is named. Those are the structural roles in non-clock form. Strip T from the relation and there is no directed dependence, only co-obtainment. Strip S from it and there is no creator/creature distinction. Strip Φ from it and there is no recipient whose being is sustained or produced. Eternal creation avoids a first temporal moment only by shifting the issue from temporal beginning to asymmetric dependence; it does not remove the structure that makes dependence intelligible. There is no third option in which creating obtains without order, distinction, and a bearer of the relation: that description names the absence of a creating relation, not a special kind of one.

Feser’s defense of the per-se series through the act/potency distinction merits direct engagement at this point. The per-se causal series is, on Feser’s reading in Aquinas⁵² and Five Proofs,⁵³ not merely a sequence of concurrent dependencies but an analysis of what it is for one thing to be the actualizer of another: the cause supplies actuality that the effect has only potentially, in the very instant of causation. The causal series terminates in actus purus (pure act without potency) because any cause that retains potency is itself in need of actualization.

The structural reply does not contest the act/potency analysis on its own terms; it shows that T, S, and Φ are what the analysis describes. Actualization supplies asymmetric dependence and distinction between actualizer and actualized, directly yielding thin order and difference roles. Thomistic concurrent actualization need not be a temporal transition from an earlier potential state, and dependence need not by itself be spatial extension. The richer T, S, and Φ types require the separate realizer arguments. Feser’s act/potency machinery, taken on its own terms, is an analysis of what asymmetric causal dependence consists in, and the structural roles are what that analysis bottoms out in. The conclusion is the same as the main reply: the per-se series, whether described in the language of actualization or in the language of asymmetric dependence, presupposes the structural conditions at the terminating cause rather than exempting it from them.

The engagement with the actus purus form of simplicity and with the analogy-of-being machinery that underwrites it lives at §5.2.1; the structural question raised there and here is the same: whether a fully-actualized being can bear the causal roles the doctrine requires without bearing the structural conditions those roles presuppose.

Second. Kant’s logical/temporal distinction pushes the case further. A cause and its effect can be simultaneous in time (the ball resting on a pillow creates the indentation at the moment it rests there) while the cause is logically prior: the indentation depends on the ball, not vice versa. Classical theism uses this template for God’s relation to the world.

Kant’s distinction is real and defeats the move rather than rescuing it. Logical priority without temporal priority requires a structural relation that holds between simultaneous states, a relation of asymmetric dependence between distinct elements within a single moment. Such a relation requires distinguishable relata, asymmetric dependence, and determinate obtaining at role-resolution. Logical priority within one moment is not thereby temporal asymmetry, and relation-stability is not automatically ordinary persistence. The ball-and-pillow analogy illustrates this vividly: the indentation depends on the ball because mass, resistance, and gravity are structurally active between them. The analogy works because ball and pillow are both in time, both in space, both substantial. Remove the structural roles and the analogy collapses: there is no ball and no pillow, only two labels with nothing between them. Kant’s distinction describes the structure carrying the logical priority, not an escape from it. Without what the pillow supplies, the ball has nothing to indent, and the asymmetric-dependence relation the analogy illustrates has no content to carry.

Third. Craig’s modal refinement goes furthest. God exists timelessly without creation and temporally with creation; the transition is not an event in God but a relational change: God stands in new relations to creatures without changing intrinsically. Simultaneous causation on this reading is what timeless grounding looks like at the temporal interface.

The refinement concedes the point while renaming the concession. If God stands in new relations to creatures with creation, then God’s relational profile changes from no-relation-to-creatures to standing-in-relation-to-creatures. A change in relational profile is a change, and a change requires before-and-after, a prior state differing from a subsequent state.

Here the strict-timelessist will invoke the distinction between real and merely Cambridge change: God’s coming to stand in relations to creatures is, on this reply, an extrinsic Cambridge change (as Socrates comes to be shorter than a growing Theaetetus without himself altering), not a real change in God, so timelessness survives untouched. The distinction is genuine; the Mullins–Padgett objection is that it does not cover this case. Craig’s view is pressured by the claim that becoming temporal changes the subject’s mode of existence rather than merely adding an external predicate. Craig denies this and treats the change as relational. The Mullins–Padgett objection holds that the Cambridge-change distinction does not cover this case: becoming temporal changes the subject’s mode of existence rather than merely adding an external predicate, and the move from real to Cambridge change requires defense rather than mere assertion. Cambridge change is the correct description of “shorter than”; it is the wrong description of “timeless, then temporal.” (The kindred ET-simultaneity reply, which relates the eternal present to each temporal moment without either term changing, is engaged at §4.2.1.) Timelessness-without-creation plus temporality-with-creation requires God to occupy two distinct modal-relational states and a transition between them. That transition is itself a structural event. Craig has not avoided making God temporal at the moment of creation. He has made God’s atemporality conditional on not-creating and God’s temporality conditional on creating, a structural description of a being whose modal status varies with its relational context. The result is not a being outside the structure, but a being whose structural position depends on which relational role it occupies.

Mullins⁵⁴ pushes the critique further: if God’s atemporality is contingent on God’s non-relations, then God’s atemporality is not a necessary feature of the divine nature but a contingent fact about the divine’s relational profile, and classical theism was supposed to deny exactly that. Padgett⁵⁵ presses from a different direction: the transition Craig posits (from timeless-without-creation to temporal-with-creation) is not coherent unless God was already in some temporal frame within which a transition could occur, which is why Padgett opts for relative-timelessness from the start rather than trying to combine strict atemporality with a creative act.

A kindred variant is Malebranchean occasionalism: God is the sole genuine cause, and what appear as natural causal relations are divinely ordained regularities holding between occasions rather than mechanisms. The reply is the same in thinner vocabulary. For God to occasion distinct events, the creaturely occasions must be differentiated and ordered, and the decree must supply determinate explanatory bearing. Occasionalists may construe the divine intention as one timeless decree rather than a persisting mental episode, so the move secures the minimal roles while leaving the loaded divine types contested. Occasionalism has not eliminated the structural conditions; it has relocated them into the divine willing itself. Another kindred: the conservationist account,⁵⁶ on which God continuously sustains creatures in existence without directly causing their operations. The sustaining relation is a constant causal contact that persists across time, and persistence across time is the admission the doctrine was trying to avoid.

The changeless-ground objection is §4.2.1. The simplicity variant, which grounds the atemporal-cause picture in actus purus, is §5.2.1. The probabilistic-simplicity variant is §5.1.4. The PSR-demand variant is §4.1.3. The Modal Collapse worry that necessity propagates from ground to grounded is §4.4.9.

What converges. Three forms of the exemption, each discovering temporal asymmetry inside the logical priority it was supposed to replace: the per se causal series must either supply asymmetric dependence, conceding the structural floor at its terminating cause, or strip directionality and lapse into simultaneous coexistence, and Morriston’s point stands that an eternal sufficient cause entails an eternal effect; Kant’s logical/temporal distinction requires asymmetric dependence between distinct elements within a single moment, where distinction, persistence, and directionality are T, S, and Φ at the single-moment level and the ball-and-pillow analogy works only because ball and pillow are both in the structure; and Craig’s modal refinement concedes the point while renaming it, since a change in relational profile is a change and Mullins’s critique cuts through, atemporality contingent on non-relation being no necessary feature of the divine nature.

Plato set the problem in the Timaeus: time is the moving image of eternity. The formulation gave Western metaphysics its longest-running project: to describe a relation between the eternal and the temporal that did not collapse one into the other. Augustine inherited the project; Boethius articulated eternity as the complete and simultaneous possession of illimitable life; Aquinas formalized simultaneous causation as the mechanism; Kant sharpened the logical-temporal distinction; Craig, Stump, and Kretzmann refined the apparatus across the last half-century. What simultaneous causation loses across the refinements is the still point: the eternal ground from which the world was supposed to hang. Two and a half millennia of work on the eternity/time interface converge on the same structural conditions the doctrine was designed to hold at arm’s length: simultaneity, relational priority, and donation of being are all structured relations, and the structure is what the doctrine was attempting to transcend.

5.1.4 Argument from Simplicity

Richard Swinburne develops a mathematically precise case for theism in the analytic tradition. The Coherence of Theism⁵⁷ establishes that the concept of God is internally consistent. The Existence of God⁵⁸ argues that, given the evidence of the natural order, consciousness, religious experience, and the moral phenomena, the posterior probability of theism (its probability after the evidence is taken into account, in the technical Bayesian sense) exceeds one-half. Simplicity as Evidence of Truth,⁵⁹ his Aquinas Lecture at Marquette, articulates the underlying epistemology: simpler hypotheses have higher prior probability as a matter of methodological necessity, not sentimental preference. The argument stands apart from the Five Ways not as a refinement but as a genre change. Swinburne does not try to demonstrate God exists; he calculates the odds. And on the metric he specifies, the odds favor a single infinite mind.

The Swinburnean case concedes more ground than any other contemporary theist move. Swinburne does not deny the structural roles are present, does not deny they admit coherent formulation, does not deny the argument for their necessity has some force. He argues rather that a divine mind is a simpler, more probable explanation for why the structure obtains at all. The move is a sharp one: probabilistic rather than structural, comparative rather than exemptive. Swinburne⁶⁰ argues that a single infinite mind, a being with a single, infinite degree of each positive property (omnipotence, omniscience, perfect goodness), constitutes a simpler hypothesis than the complex, multi-component physical universe. By Occam’s razor, the simpler hypothesis has higher prior probability as an explanatory ground. The physical universe, with its vast array of particular constants, particles, and forces, is maximally complex by comparison. The theist hypothesis wins on comparative parsimony.

Applied to the argument, the move is elegant. Swinburne does not deny T, S, and Φ are present; he argues they are better explained by positing a simpler posterior ground than by accepting them as brute or self-grounding. If the structural proof is valid and its premises are true, comparative simplicity cannot lower its conclusion’s probability. Swinburne’s case instead pressures the premises or explanatory sufficiency of the structural account: it argues that, given contested background assumptions and evidence, theism is the better overall explanation. The move requires a simplicity metric that operates across ontological levels and favors infinite-mind posits over structural-role posits. Secure the metric and the argument is defeated by comparative probability rather than refutation.

The contemporary literature on simplicity runs thick, and the picture is less favorable to Swinburne than his standing in the field might suggest. On simplicity as a truth-indicator: Sober’s Ockham’s Razors⁶¹ gives the most sustained philosophical analysis of simplicity’s role in scientific inference, and concludes that simplicity functions only relative to background assumptions, never as an unconditional criterion. Huemer⁶² presses the deflation further: the standard rationales for parsimony hold in scientific settings but fail to transfer, leaving ontological simplicity without evidential value in typical philosophical contexts. Against Swinburne: Mackie’s The Miracle of Theism⁶³ presses the Bayesian prior-assignment problem most trenchantly: prior probabilities are stipulated, not derived, and Swinburne’s stipulations do the theological work before the calculus runs. Philipse’s God in the Age of Science?⁶⁴ presses the scientific-methodology angle: Swinburne’s criterion is alien to the actual practice of inference in physics. Oppy’s Arguing about Gods⁶⁵ catalogs the failures of the probabilistic arguments across their range. Dawkins’s The God Delusion⁶⁶ offers the rhetorical version of the deflation (a God who continuously monitors and controls every particle in the universe cannot be simple) whose analytic substance was already present in Mackie and later refined by Philipse. The present reply sits closest to Sober on simplicity’s background-relative operation and to Philipse on the scientific-methodology point, and adds what neither alone specifies: the background relative to which Swinburne’s simplicity claim would have to be evaluated is T, S, and Φ themselves, which forecloses the comparison in principle.

The argument turns against the necessity claim of §2.5 and the exhaustion claim of §2.6. If simplicity favors theism over structural necessity, the modal conclusion of the argument is defeated by comparative probability even if its logic is sound. The reply runs on two levels, and the order matters. The load-bearing level concerns the candidate, not the comparison. Whatever wins the parsimony contest is offered as the ground of the structural order, and being a ground of any kind requires the minimal floor established at §5.1.1: asymmetric grounding-order, relational distinction from what is grounded, and identity-retention across the grounding relation, which is T∧S∧Φ read at the structural level. A single infinite mind that grounds the roles thereby instantiates them; it cannot be the simpler thing the roles reduce to, because it already presupposes them. This level says nothing about who performs the comparison, so it is untouched by the rejoinder that the book’s own case for the roles is equally an exercise of embedded reasoning. The second level concerns the comparison itself, is genuinely weaker, and is placed second for that reason: comparative parsimony ranks candidates within a fixed background of alternatives, and it has no purchase on a background weighed against its own absence, because §2.7 has already shown there is no coherent contrast class in which the roles fail to obtain. With no articulable alternative on the menu, there is nothing for the razor to weigh the roles against. This is not the claim that the roles are exempt from rational support; the transcendental case of §§2.1–2.6 supports them. It is the narrower claim that comparative parsimony is the wrong instrument, since it requires a slate of rivals the roles do not sit on. Three forms of the move take up the objection: one-mind parsimony, cross-level simplicity, and Bayesian refinement.

First. The one-mind parsimony move holds that an infinite mind has a single unified essence (one being with one set of infinite properties) while the present account posits three distinct roles each requiring separate grounding. Occam’s Razor, the objector presses, favors the one over the three: infinity in a single subject is simpler than multiplicity across structural roles.

The objection misapplies Occam’s Razor by confusing ontological levels. The razor governs explanatory entities within a background of structural possibilities: it operates by comparing candidate explanations of contingent facts against a stable background against which simplicity judgments have meaning. It is coherent to ask whether electrons or fields are the simpler posit for explaining electromagnetic phenomena, because both are candidates within structure and the simplicity comparison has a stable background. It is not coherent to apply the same criterion to structure itself: structure is not a candidate explanation within a background of structural possibilities; it is the background.

That structure is background rather than candidate is not asserted but earned at §2.7, which shows there is no coherent contrast class in which the roles are absent; with no such alternative, the roles are not one posit among rivals but the condition any posit occupies. The one-versus-three arithmetic the objection runs never gets started either: §5.1.1 shows T, S, and Φ are not three independent posits but one structural fact under three categorial readings, so there is no count of three for a count of one to beat.

Actual simplicity comparisons are structured cognitive acts, but that act-side fact does not prevent them from evaluating hypotheses about the structure of reality. The stronger reply must concern the candidate and the metric: whether the divine-ground hypothesis already incurs structural commitments, and whether any background-neutral simplicity measure favors it. Occam’s razor was made to cut entities within a shared background. It cannot cut the background itself.

Second. Cross-level simplicity raises the stakes. Simplicity applies across ontological levels and not merely within them, the objector presses: a single infinite mind is simpler than the structural conjunction even as an ultimate explanation, and the comparison evaluates whether structure or the mind is the better ultimate posit rather than presupposing structure as its background.

The cross-level move raises the altitude of the comparison but not the standing of its candidate. To rank the mind against structure as ultimate posits is still to offer the mind as the ultimate ground, and a ground of any kind carries the §5.1.1 floor: asymmetric grounding-order, relational distinction from what it grounds, identity-retention across the relation. There is no altitude at which a grounding mind sheds those and remains a ground. So the cross-level comparison never reaches a vantage where structure is one optional posit among others; it relocates the contest to a level where the theistic candidate has already instantiated the structure it was meant to displace. Swinburne’s argument does not offer an alternative to structure; it describes a configuration of it. The evaluator-side point reinforces this without bearing the weight: any such comparison is performed by a structurally embedded reasoner, with no unstructured vantage from which to weigh the conditions of weighing. But the decisive fault lies in the candidate, not in the one comparing.

Third. The Bayesian refinement turns the comparison probabilistic. The prior probability of the theistic hypothesis, given its internal simplicity and explanatory scope, exceeds the prior probability of a necessary structural architecture; updating on the evidence of a contingent-looking, complex physical universe, the posterior favors the infinite-mind hypothesis. The inference is not deductive but probabilistic, and the probability calculus is on the theist’s side.

The Bayesian refinement’s real weakness is not that priors are computed by embedded reasoners; pressed that way the point would prove too much, sweeping in the book’s own reasoning along with Swinburne’s. The weakness is that the prior assigned to “the theistic hypothesis” is already doing the theological work before the calculus runs, as the comparison-class problem below makes plain.

Set the prior-assignment problem aside and the candidate-side reply still binds: for an infinite mind to be a coherent posit, that mind must stand in determinate cognitive and grounding relations to what it knows and what it grounds. Its omniscience covers what-is-the-case, and what-is-the-case has structural content: distinguishable objects, temporally articulated states of affairs, persisting carriers of becoming. The mind’s mode of cognition need not be discursive (running through premises and conclusions in succession, the way a finite intellect does) for the point to bind. Even non-discursive cognition, where the divine intellect grasps all things in a single timeless act rather than thinking them through step by step, cognizes content that is structurally articulated; and the grounding relation a mind bears to what it grounds is itself asymmetric, sustained, and obtaining between distinguishable relata. Aquinas’s intuitive cognition, the Boethian tota simul, and the Stump–Kretzmann ET-simultaneity are engaged at length in §4.2.1, which shows each variant requires structural conditions through the relations it bears to what it knows and what it grounds even where it denies them in the mode of cognition. T, S, and Φ are required by what the mind cognizes and what it grounds, not by how it cognizes. The non-discursive exit narrows the line through the mode of divine cognition; it does not eliminate the structural conditions on its content or on the grounding relation that connects the mind to what it grounds. A mind so related to its contents and to what it grounds carries more articulated commitments than the bare role-floor. That does not yield a theory-neutral numerical simplicity ranking, because Swinburne counts properties and laws under a different metric. It does undercut any claim that the divine hypothesis is obviously simpler before the metric and background assumptions are defended. The classical rejoinder that the divine mind has no composition at all, and so is not a configuration of anything, is the actus purus line; it is taken up at §5.2.1, where the strict-simplicity claim must still meet this floor, since even a metaphysically simple mind stands in an asymmetric, distinct, identity-retaining grounding relation to what it grounds. The reply here needs only that floor and leaves the actus purus case to its own site.

The infinite mind is not a cheaper ontology. It is a configuration that presupposes the full structural architecture at maximum richness, and simplicity comparisons conducted at that level run on the conditions they are trying to weigh.

One underlying premise of Swinburne’s cumulative case deserves direct engagement in this section. Swinburne holds that personal explanation (explanation in terms of an agent’s intentions) is an irreducibly distinct form of explanation, not reducible to structural or scientific explanation, and at least as fundamental as structural explanation.⁶⁷ The premise carries weight independent of the simplicity comparison: even granting the structural-conditions reply above, if personal explanation is irreducibly distinct from structural explanation, then an irreducibly personal ground remains a live explanatory terminus. The book’s reply concedes the irreducibility claim while denying the entailment. That personal explanation is irreducible to the vocabulary of physics is true: intentions are not fully translatable into particle velocities or field configurations. Ordinary personal explanation involves an agent, an intention directed toward a distinguishable outcome, and continuity between intending and acting. A timeless theist will deny that intention is formed across time. The role-level point survives — content, direction, and explanatory bearing remain — but the loaded temporal and persistence types cannot be inferred from personal explanation alone. The irreducibility of personal explanation to physics is compatible with T, S, and Φ being necessary conditions for personal explanation to occur. Swinburne’s canonical examples (the agent who freely decides, who selects among possibilities, whose intentions cause outcomes) each instantiate the structural conditions in the act of being a personal explanation. The irreducibility claim survives; what it does not establish is that the irreducible personal level is free of T, S, and Φ or prior to them. A personal explanation that instantiates T, S, and Φ in its own performance is not an alternative ground to the structural conditions but a high-order configuration of them.

A separate line of pressure runs through the comparison class itself. Swinburne ranks one infinite mind against the complex physical universe and concludes the mind is simpler. The class, presented as theism versus its rivals, is gerrymandered against the strongest alternative theisms. Process theism⁶⁸ posits one dipolar deity whose abstract pole carries fewer essentially distinct attributes than the classical theist mind. Panentheism⁶⁹ posits one substance with internal differentiation rather than a divine substance externally related to a created one, reducing the number of fundamental ontological categories. Open theism⁷⁰ drops the strong omniscience condition that does heavy work in the Swinburnean parameter count. On Swinburne’s own metric (fewer essentially distinct properties scoring as simpler) several sibling theisms score equal or better than the Swinburnean theism his prior assignments single out, and the calculus does not adjudicate among them without theological commitments that precede it. The simplicity case, pressed honestly through its full class of rivals, narrows to a contest among theisms rather than a contest between theism and structural necessity. Mackie’s⁷¹ prior-assignment objection sharpens here: the priors Swinburne assigns to “the theistic hypothesis” elide which theism, and the elision does the disputed theological work before the Bayesian calculus runs.

A kindred move drops the Bayesian apparatus for an aesthetic-simplicity appeal: simpler ultimate posits offered as intrinsically more credible than complex ones, independent of probability calculus. Huemer⁷² has already cut the move off from within the parsimony literature, finding the aesthetic rationale delivers no evidential value in ontological disputes; the structural reply explains the residual pull. Aesthetic simplicity is an assessment made by embedded reasoners, structurally configured to find the simpler posit more satisfying because structural cognition naturally compresses. The aesthetic judgment is not a view from outside structure; it is a feature of how structure evaluates itself. Another kindred: Dawkins’s⁷³ inverse parsimony argument (a mind capable of monitoring every particle cannot be simple) arrives at an adjacent conclusion by a different way. The present reply sharpens it: not only is the mind not simple in the information-theoretic sense, but its simplicity is uncomputable without the structural background, and the structural background is what the simplicity claim was meant to explain away.

A note on direction. The reply does not depend on which direction the parsimony comparison is run. Oppy⁷⁴ develops the naturalistic counterpart most thoroughly: where Swinburne argues theism wins on parsimony, Oppy argues naturalism wins on the same metric, one ontological category, no additional minds, a necessary-or-brute natural terminus reached without theological posits. The present account is stronger than Oppy’s at the structural level. Any appeal to parsimony presupposes the necessary standards against which the evaluation runs, and those standards are T∧S∧Φ themselves. The within-structure-versus-of-structure distinction that blocks Swinburne’s pro-theism parsimony blocks Oppy’s pro-naturalism parsimony with equal force: parsimony does not adjudicate between structure and an alternative to structure in either direction, because parsimony requires the conditions it would have to evaluate. The reply runs at a second layer as well. The candidates being compared are themselves causes (Swinburne’s mind that grounds structure, Oppy’s structure that grounds phenomena), and being-a-cause requires causally structured temporal succession (T), which by the Triconditional Cross-Entailment (§2.6.7) requires also relational extension (S) and a persistence medium across which influence is transmitted (Φ). Neither candidate can be specified as a cause without the full conjunction T∧S∧Φ. Parsimony’s evaluative standards and the candidates it was meant to evaluate both run on T∧S∧Φ; the reply blocks the move at both layers.

The strict-simplicity variant, which presses parsimony through actus purus rather than probability, is §5.2.1. The simultaneous-causation variant is §5.1.3. The foundation-argument variant is §5.1.1. Rasmussen’s stage-by-stage personalist construction (the move from necessary foundation through selection-among-possibilities to personal being) receives its primary reply at §5.1.1 and is carried to Chapter 6 (§6.4.1 Awareness, §6.4.2 Knowledge, §6.7.1 Power) rather than being developed here; this section engages Swinburne’s probabilistic simplicity case specifically. The PSR-demand variant is §4.1.3. Swinburne’s case is cumulative by design (simplicity sets the prior, and fine-tuning, the existence of consciousness, and the moral phenomena are aggregated toward a posterior exceeding one-half), so dismantling the simplicity wing in isolation would leave the aggregate standing if the other weights did. Each component must therefore be assessed at its own site (fine-tuning at §5.1.5, the PSR at §4.1.3, consciousness and moral awareness elsewhere). Weakening one component lowers the cumulative case, but it does not establish that every remaining evidential weight is zero. The §5 replies are reckoned together at §5.4.

What converges. Three forms of the simplicity move, each running on the architecture it was meant to outrun: Occam’s Razor is a within-structure evaluative tool that compares candidates against a stable structural background, and applied to structure itself the razor has no background to cut against; cross-level comparison never reaches a vantage where structure is one optional posit among others, since a grounding mind at any altitude carries the asymmetric order, distinction, and identity-retention of the §5.1.1 floor; and the Bayesian refinement imports the problem it was meant to solve, since the prior assigned to “the theistic hypothesis” does the theological work before the calculus runs and an infinite mind capable of grounding is a configuration at maximum structural richness, not an obviously simpler posit.

Swinburne’s work is the kind of work only a theist who takes secular philosophy with complete seriousness could produce: at every juncture he accepts the standards of the field and argues the theistic hypothesis wins on those standards. The precision is what lets the reply be precise too. The simpler hypothesis the objector recommends is a configuration of structure, not an alternative to it. Comparative parsimony, pressed against the architecture of comparison itself, collapses into a comparison the architecture had to underwrite to be performed. Hume gives the move its final form in Dialogues concerning Natural Religion (posthumous):⁷⁵ if a necessary being must be posited at all, the material universe itself is a candidate, and no non-theological consideration disqualifies it. Cleanthes’ suggestion was underspecified in 1779. The present argument supplies the specification: □(T∧S∧Φ). The necessary being Cleanthes gestured at is the structure Swinburne’s simplicity comparison was trying to outrun.

5.1.5 Fine-Tuning Argument

Precision is the fine-tuning argument’s wedge. Several fundamental parameters appear to lie in ranges compatible with complex structure that are narrow relative to commonly discussed comparison ranges. The often-cited 10¹²⁰ figure for the cosmological constant primarily expresses the discrepancy between observed vacuum energy and a naive Planck-scale estimate, not straightforwardly the width of a life-permitting interval. The tradition reads the broader sensitivity evidence as evidence of selection and the selection as evidence of a selector. The resulting inference is a serious empirical challenge to the necessity claim for T, S, and Φ (the three structural conditions, borne by Time, Space, and Substance) defended in §§2.1–2.6. The fine-tuning argument is a single inference (from narrow-window precision of the constants to selection) with two extensions the theist presses when the core inference meets resistance. The reply follows this shape rather than treating the three versions as independent objections.

Against the core inference, two lines apply pressure: McGrew, McGrew, and Vestrup⁷⁶ challenge the choice of measure, while Carter⁷⁷ and Bostrom⁷⁸ show that observer selection must be included in the likelihood model. Against the multiverse extension, the reply asks whether the higher-order mechanism and its parameters receive independent support or merely move the explanatory burden. Against Swinburne’s Bayesian cumulative case,⁷⁹ the issue is whether theism and its rivals supply justified comparative likelihoods for embodied observers of our kind.

Beneath all three forms lies a prior question: what alternative configurations and comparison class the probability model is licensed to range over. And beyond the bare inference lies a stronger reading (that the universe is not merely life-permitting but life-intended) which separates cleanly from the core argument and fails on its own terms.

The reply takes the prior question first, then the three forms, then the teleological gloss. The observation behind the argument is not in dispute. The constants of this universe do sit within narrow ranges compatible with complex structure. The account contests the inference from the observation to selection, not the observation itself.

The fine-tuning tradition is the empirical wing of natural theology’s design inheritance. Where Paley’s watchmaker reasons from biological complexity to a designer, the post-Paley inheritors (;⁸⁰;⁸¹;⁸² and⁸³ at selective points) reason from the values of the fundamental physical constants to the same conclusion. Substantial changes in the effective cosmological constant relative to the observed value can prevent galaxies and other complex structures from forming; the exact life-permitting range depends on the model and on which parameters are varied together. Vary the strong force coupling by a few percent and stable atomic nuclei do not form. The quantitative sensitivities have been catalogued with professional care. What is contested is whether precision of this kind, against the background of no further specification of the alternatives, is genuinely evidence of selection. The naturalist reply has two heads. Anthropic reasoning (;⁸⁴)⁸⁵ notes that observers can only observe observer-permitting conditions. This supplies a selection effect, not by itself a complete explanation of why any observer-permitting universe exists; its force depends on the cosmological ensemble and observation-selection model adopted. The probabilistic critique⁸⁶ argues that the likelihood calculations the theist runs presuppose a measure over an unbounded parameter space for which no canonical measure exists. The present account absorbs both replies and adds a third: the precision is not a feature of a configuration picked out from among rivals but a feature of what a configuration within T, S, and Φ must look like to be a configuration at all.

The challenge bears on §2.4 most directly. T and S, defended there as necessary structural conditions, admit no variation: not because the values of physical constants are fixed at a particular numerical setting, but because the roles T and S discharge are logically prior to anything a numerical setting could register. Fine-tuning compares this configuration to alternative configurations. The comparison is coherent only at a level where ‘configuration’ has content, which is within the structural order. What follows presses the point that the argument’s demand for a tuner applies the idiom of selection one level above where it has purchase. One question stands upstream of the entire dialectic and is taken first; three forms of the move then take up the objection at the level of the inference itself.

The prior question. Probability is downstream of possibility. Before assigning a physical chance distribution, one must establish a genuine space of alternative outcomes. A Bayesian confirmation argument can instead use epistemic probabilities over rival hypotheses, but it must still justify its comparison class, priors, and likelihoods. Mere ignorance about whether the constants are necessary does not by itself provide a canonical distribution over their values. This question sits upstream of the normalizability problem: McGrew, McGrew, and Vestrup⁸⁷ show that even granting a space of alternatives, no canonical measure over it exists; the prior question asks whether the space has been shown to exist at all.

The proponent has a practiced escape. Probability, they note, need not wait on metaphysics; rational agents assign credences to propositions whose modal status is fixed but unknown all the time. Goldbach’s conjecture is either necessarily true or necessarily false, yet a mathematician may rationally hold it to 0.99 on inductive grounds; so (the move runs) one need not first prove the alternative constants metaphysically possible in order to spread a credence across them and call the actual values surprising. The escape shows that a Bayesian design argument can use epistemic probabilities without positing an objective chance process that sampled the constants. Rational credences can be evidentially responsive even when the proposition is necessarily true or false. The remaining burden is comparative: the proponent must justify the likelihoods and priors under theism and its rivals, rather than infer an ontic selector directly from an objective frequency distribution. Epistemic probability can support an abductive conclusion, but only through defensible comparative likelihoods; it cannot be treated as if it revealed a literal sampling process over constants.

A second escape runs the other way: perhaps the alternatives only seem possible. Kripke⁸⁸ shows that some truths are necessary but knowable only a posteriori (that water is H₂O, that Hesperus is Phosphorus) so that what presents itself as a contingent setting can be a necessary one wearing the appearance of contingency. If the constants are necessary a posteriori, the apparent space of alternatives is an illusion of contingency, and the fine-tuning improbability never gets started. The present account does not lean on this claim, and notes the standing counter: Chalmers⁸⁹ argues from modal rationalism that ideal conceivability does track genuine possibility, so the appearance of contingency cannot be dismissed so cheaply. The dialectic is unsettled, and the account does not rest on settling it. It rests only on the asymmetry the prior question fixes: the burden lies with the proponent to establish that the alternatives are possible in the sense the measure requires, and that burden has not been discharged. And the account’s own positive footing is insulated from the dispute in any case. What it asserts of complex structure (that a structure-permitting, life-permitting configuration is always available to necessary T, S, and Φ) is a necessary possibility, not a claim that the constants could not have differed; a claim of that form stands whether or not ideal conceivability tracks genuine possibility, since conceiving the constants otherwise leaves a necessary possibility exactly where it was. The dispute the proponent would need to win to revive the inference is therefore one the account can decline from either side.

One discipline must be observed in pressing this. The claim is not that the numerical values of the constants are necessary: that is a separate and stronger thesis, argued elsewhere and not needed here. The claim is that the structural conditions T, S, and Φ are necessary, while the contingency of the particular numerical settings within them is unestablished. The argument requires that contingency as a premise and has not earned it. The conclusion is not ‘the values are necessary’ but ‘the values have not been shown to be contingent in the sense the inference needs’, and an inference that presupposes an unestablished premise does not transmit its conclusion.

First. The precision argument, in its contemporary form, runs from likelihood ratios rather than intuition. Collins and Barnes do not rely on intuition; they rely on comparative probabilities. If life is extraordinarily rare across the space of possible constants under naturalism, and non-rare under theism, then the narrow-window observation confirms theism. The form of the argument is Bayesian: P(tuning | theism) / P(tuning | naturalism) is large, and observing tuning updates credence toward theism.

The calculation requires a probability measure over the space of possible constants. McGrew, McGrew, and Vestrup⁹⁰ press a serious measure problem: an unbounded parameter space admits no uniform normalizable measure, and the ratio of life-permitting to life-forbidding regions is undefined without additional constraints. Defenders answer with restricted ranges, scale-sensitive priors, or direct likelihood comparisons, so the objection imposes a burden rather than refuting every Bayesian formulation. ‘One part in 10¹²⁰’ presupposes a flat measure over an interval that extends to infinity in both directions; a log-uniform measure would yield a different ratio; a measure weighted by physical plausibility another; and the argument offers no principled reason to prefer any of them. The sensitivity of the conclusion to the measure chosen is as large as the purported fine-tuning itself. Stenger’s The Fallacy of Fine-Tuning⁹¹ observes that many of the ‘fine-tuned’ constants are only fine-tuned relative to a particular parameterization; under different but empirically equivalent parameterizations, the apparent precision dissolves. Carroll⁹² notes that whether the tuning is ‘fine’ depends on what set of possible universes one is comparing the actual one to, a set the theist has not specified.

The most developed reply on the theist’s side is Collins’s,⁹³ which concedes that the full parameter space admits no canonical measure and restricts the comparison to an epistemically illuminated range: the interval of values across which the life-permitting effect of variation can actually be assessed, within which, Collins argues, a normalizable measure is available and the life-permitting sub-region is vanishingly small.

The restriction relocates the problem rather than resolving it. An epistemically illuminated range is bounded by what inquirers can currently model, not by what is metaphysically possible, so a measure defined over it is a measure over the reach of present physics, and confidence drawn from it is confidence about that reach, not about a space of ways the constants could have been. This is the objective-epistemic conflation the prior question already isolates, reappearing one level in: Collins secures a normalizable measure by substituting an epistemic boundary for a modal one. And even granting the illuminated range as a genuine space of alternatives, the measure over it is still unfixed, since a uniform and a log-uniform measure on the same finite interval deliver different ratios and the argument supplies no principled reason to prefer either. The coarse-tuning observation that the life-permitting band is narrow on any reasonable measure survives the restriction, but it establishes only that the band is narrow relative to the illuminated range, which is a fact about the range our physics can survey, not evidence that the constants were drawn from a wider space and selected from it.

Anthropic reasoning makes a complementary point that does not duplicate the measure problem. Carter⁹⁴ and Bostrom⁹⁵ argue that observers can only find themselves in observation-permitting conditions: not because observation causes the conditions, but because the selection effect is built into what it is to be an observer making an observation at all. Were the constants otherwise, no one would be measuring them. The present account extends the anthropic point rather than replacing it: T, S, and Φ are not conditions that happen to be observer-permitting; they are conditions that make observation possible in the first place. An observation that these conditions obtain is made from within them. That fact requires an observation-selection correction, but it does not by itself show that the observation has no evidential bearing beyond the conditions.

Second. The multiverse-insufficiency extension presses the design inference up one level. Conceding everything naturalist cosmology asks for (a universe-generating mechanism, an ensemble of universes, a stochastic process sampling parameter space) the theist (;⁹⁶)⁹⁷ argues that the fine-tuning demand reappears at the meta-level rather than dissolving. Any universe-generating mechanism must have its own structure: laws governing what universes it produces, a range of constants it can assign, a probability measure over outcomes. Those higher-order parameters are themselves either contingent (so the fine-tuning demand reappears at the meta-level) or necessary (so the theist concedes the destination of the present account). The regress of designers, the theist argues, does not terminate in the multiverse; it terminates in whatever fixes the multiverse.

Two replies. Hacking’s inverse-gambler objection warns that observing one apparently unusual outcome does not automatically support many unobserved trials. Contemporary multiverse arguments often add independent cosmology and an observation-selection rule, so the objection constrains naive inferences rather than every multiverse explanation.

The sharper reply is available within the modal framework defended in §§2.1–2.6 rather than from the fine-tuning dialectic’s neutral empirical register; conditional on that framework, S5 preserves the necessity of any genuinely necessary higher-order parameter. It does not identify the parameter or explain its high-resolution content. Necessary meta-parameters would be invariant features of the ensemble theory, but T, S, and Φ are too coarse to determine their specific values or probability law. The structural roles constrain any genuine universe-generating realizer; they do not identify every higher-order parameter or meta-tuner with the triad. Readers unconvinced by the prior modal framework can rely on the McGrew normalizability point alone: that closure is framework-neutral, and the S5 closure is a further sharpening available only on top of it.

Third. The Bayesian cumulative case, pursued most carefully by Swinburne,⁹⁸ takes the argument in a more restrained probabilistic direction, retreating from entailment to confirmation. Swinburne does not claim the fine-tuning observation entails a designer; he claims it raises the probability of theism relative to naturalism by a non-trivial factor, and that when combined with other probabilistic considerations, the cumulative case for theism becomes compelling: simplicity (the prior-from-simplicity wing of Swinburne’s case, engaged at §5.1.4), the existence of consciousness, and the phenomenon of moral awareness. Fine-tuning is one weight on a larger scale.

Monton⁹⁹ isolates the structural problem with the fine-tuning weight specifically. If God is omnipotent, God can instantiate conscious beings in any logically possible configuration. The present configuration is not therefore necessary for the existence of conscious beings. But if it is not necessary for the existence of conscious beings, the theist cannot explain why God selected this one rather than another except by appeal to divine preference, which does no Bayesian work: preference is compatible with any observation and discriminates between no hypotheses. The probability P(this configuration | theism) is as free as divine preference allows; which is to say, unanchored; which is to say, unable to contribute a determinate likelihood ratio.

A sharper version presses on the theist’s own commitments rather than on Bayesian bookkeeping. God is held to be a conscious being not in this universe. If the constants of this universe are necessary for the existence of conscious beings, no conscious beings exist outside this universe, and God does not exist as characterized. If these constants are not necessary for consciousness simpliciter, the fine-tuning argument must be stated more narrowly in terms of embodied observers or life of our general kind. That limits the conclusion but does not make the observation evidentially null; comparative likelihoods may still differ over how readily rival hypotheses produce such observers. The fine-tuning premise concerns embodied life or observers generated by physical laws, whereas God is proposed as an unembodied mind. That is not a simple equivocation, but it narrows the evidence: the likelihood comparison must explain why theism predicts a physically life-permitting cosmos of this sort rather than merely the existence of consciousness somewhere.

The teleological gloss. The inference examined so far concludes only that the constants were selected; it is silent on what for. A stronger reading is almost always present in the tradition and rarely separated out: that the universe is not merely life-permitting but life-intended, tuned for life, with life as the purpose the tuning serves. This gloss is separable from the bare inference, and it fails on its own terms, independently of whether the bare inference succeeds. It is also the reading that carries the argument’s rhetorical force, so its collapse matters.

Any-tuning. The gloss assumes that this configuration is specially suited to conscious life: that life is what these constants are for. But that only configurations near ours could host conscious beings is precisely what the argument has not shown. For all that is known, most configurations might support life of some kind; our inability to imagine it reflects the limits of our physics, not the limits of the possible.¹⁰⁰ Worse for the gloss, the theist’s own ontology already contains conscious beings that hold none of this universe’s constants fixed. God is a conscious being with no physical parameters at all. Angels and demons, on the traditional commitment, are conscious without bodies. The saved in heaven and the damned in hell are conscious persons in conditions that are not this universe’s physics, on the doctrine, not even this universe. If consciousness can obtain in God, in heaven, and in hell, then this configuration is not necessary for consciousness on the theist’s own accounting, and the ‘tuned for conscious life’ gloss contradicts the metaphysics it is meant to serve. The example is not a hostile construction; it is drawn from the proponent’s own creed.

The reversal. Scoped to the gloss rather than to the bare inference, the fine-tuning observation runs the other way. If God is omnipotent, the space of means by which conscious life could be realized is enormous: God could sustain life in any configuration, or with none, or by direct fiat. A universe that must be threaded through a needle’s eye of constants to permit life is therefore a strange choice for a being to whom all means are open: it is exactly the expedient a maker without options would be forced into, not one a maker with every option would prefer. The likelihood reversal makes this precise:¹⁰¹ once observability is taken as background, learning that the universe is naturalistically life-permitting cannot raise the probability of a supernatural tuner, and on the for-life gloss specifically it lowers it, since an omnipotent intender had no need of a delicately balanced mechanism at all. The positive thesis follows: life that needs no apparatus to be created, yet is given a vast and precarious one, is evidence against the apparatus being for it.

Function. The sharpest form of the objection turns on how teleology is individuated. A thing made for a purpose is identified as such by its serving that purpose: the test of ‘made for X’ is ‘succeeds at X.’ An omnipotent maker cannot fail at function: whatever such a maker intends, obtains. So if the universe were made for life, it would be unmistakably, abundantly, securely hospitable to life, because nothing could have constrained the maker to do less. What is observed instead is a cosmos in which life clings to a vanishing margin, sustains itself against constant threat, and is extinguished with ease. A product that barely and precariously discharges its supposed function is, by the only test teleology affords, not made for that function. This wing is immune to the ‘precious because rare’ reply: rarity is evidence of difficulty, and difficulty is exactly what omnipotence does not face.

Scale. Three observations reinforce the functional point. Qualitatively, life does not find a hospitable world; it engineers itself against a hostile one. Survival is a labor of building shelter, regulating temperature, and defending against an environment that is not arranged for the organism; cities are monuments to how much must be done to make a patch of the world habitable. The relation runs opposite to the gloss: life is tuned to its environment by adaptation, not the environment to life by design. Quantitatively, the life-permitting region is a vanishing measure of a cosmos overwhelmingly lethal to life: near-absolute vacuum, lethal radiation, temperatures at both extremes, durations and distances on scales that dwarf anything life occupies.[1] Structurally, the narrowing recurs one level down inside that vanishing region: the fraction that is habitable is not itself one hospitable environment but a partition of mutually exclusive niches, where crossing from one into another is lethal rather than merely difficult, an ocean drowns what land life requires, open air suffocates what ocean life requires, and each further subdivision (pressure band, temperature range, salinity, atmospheric composition) forecloses what its neighbors permit. There is no environment hospitable to life as such, only a fractal series of disjoint environments each hostile to whatever is not already fitted to it. The fracturing recurs even inside a single generalist species. The Bajau of maritime Southeast Asia carry a genetic adaptation, an enlarged spleen serving as an oxygen reservoir, that lets them free-dive to depths and durations unadapted humans cannot match, while Tibetan and Andean highlanders meet the same challenge of high-altitude hypoxia by two different physiological routes: lower hemoglobin concentration in the one population, higher in the other. Ethiopian highlanders answer the same hypoxic pressure a third way, holding hemoglobin at sea-level norms while still reaching oxygen saturation levels neither of the other two groups matches by their own route. Inuit and Yup’ik populations carry a variant of the gene CPT1A that lets a near-zero-carbohydrate, fat-heavy diet fuel the body without the metabolic cost that same diet would impose on someone without it. The Haenyeo free-divers of Jeju evolved their own route to breath-hold diving in cold water, independent of the Bajau’s. None of this is a claim of lethal incompatibility of the ocean-and-land kind; it is a claim about differential fit. Even the species most often cited as nature’s generalist exception has already sorted into sub-populations locally optimized for extremes the rest of the species only tolerates. A universe that is for life is, on inspection, almost entirely against it.

What unifies these wings is omnipotence. Each dysteleological observation invites the reply that the hostility is instrumentally necessary: that a universe capable of producing life must be large, old, and governed by laws that are lethal elsewhere. That reply binds a constrained maker, not an omnipotent one. The moment the maker is granted omnipotence, every appeal to instrumental necessity lapses, and the hostility, scarcity, and precariousness become unforced features of the product, which is to say, evidence about its purpose. This is the cosmological specialization of the evidential structure Draper¹⁰² presses for pain and pleasure: the distribution of good and ill in the world is more expected on indifference than on benevolent intention, and the gap is evidence.

Three replies are available to the gloss, and each fails for a stated reason. The instrumental-necessity reply binds the process, not the agent, and an omnipotent agent is not bound by the process. The minimalist-masterpiece reply (that a great maker might sign a single, spare, improbable work) does no Bayesian work, since ‘a great maker would do this’ is compatible with any observation whatever, and it collides directly with the suffering the spare work contains. The soul-making reply (that hardship is the point) concedes that the universe is not made for life’s flourishing but for its testing, which is a different purpose, post-hoc with respect to the fine-tuning data, and which reactivates the problem of evil rather than answering it.

The Fall. A last rescue holds that the tuning was once benign and was broken: that the world’s hostility postdates a fall, an original sin, a derangement of nature. The rescue faces a trilemma and forfeits its setting. Which conditions changed? If the physical constants changed, the claim is a physical hypothesis with no physical evidence, and one that runs against the very constancy the fine-tuning argument relies on. If only the biological or moral order changed, the cosmic hostility (vacuum, radiation, scale) was there from the start and is unexplained. If the change is purely spiritual, it does no work against the physical dysteleology at issue. And the rescue forfeits the project: an inference licensed only by an ancient text is no longer natural theology, which was the genre the fine-tuning argument entered to occupy. The collective form of the doctrine compounds the cost. A penalty visited on all conscious beings for the act of two, through a headship none of the penalized consented to or could have refused, is a moral structure an omnipotent maker had every means to avoid, who could, for instance, have created each pair on its own world to stand or fall for itself, with no representation imposed. That the collective arrangement was chosen when non-collective ones were available is not a necessity to be excused but a choice to be assessed, and it is assessed under the problem of evil, engaged in its own right elsewhere.

The eliminativist wing of the fine-tuning debate presses harder than the present argument does. Stenger¹⁰³ argues that the appearance of fine-tuning is a statistical artifact: that under a full accounting of the possible parameterizations, no remarkable precision remains to be explained. (The Ikeda–Jefferys reversal, a distinct claim, is engaged above against the for-life gloss rather than against the bare observation.) The present account does not commit to the eliminativist’s stronger claim. Some sensitivity is real; the question is what the sensitivity is of. The eliminativist answers: of nothing in particular, once the parameterization is normalized. The present account answers: of the configuration to the conditions prior to configuration. The two answers converge on rejecting the design inference and diverge on whether any feature of the observation survives the rejection. The present account preserves the observation and denies the inference; the eliminativist denies both. Where the eliminativist says the theist is chasing nothing, the present account says the theist is chasing the structure.

Probability rides on modality. The three forms share a structure worth stating directly: every probability is conditional on the modal frame that fixes its sample space, so probability questions are asked within a structure, never of it. That cuts against the tuner from two sides. Tuning requires a standard to tune toward, and the standard admits no tuner: if it is contingent, fine-tuning collapses into preference-matching and a regress of standards opens; if it is necessary, there is nothing to tune, since what holds necessarily holds in virtue of its necessity, not by being selected for. And the precision of any probability assignment tracks the resolution of the modal map beneath it, so a coarse map that under-distinguishes necessity from contingency inflates apparent contingency and, with it, apparent improbability. The narrow-window number is calibrated against what, at the structural level, is not itself a probability-bearer; precision against an ill-formed structure is fool’s-precision, and the argument presents that precision as evidence about the very structure it was conditioned on.

A note on division of labor. The adjacent constructions that extend the fine-tuning debate are covered in their own sections. The Platonist reading of physical laws (whether the constants are prescriptive features of a separate domain or descriptive features of structural necessity) is §4.2.4. The Wigner problem of mathematical applicability runs close to fine-tuning and is engaged in §4.3.6. The multiverse taken as full-blown modal realism is §4.4.1. Swinburne’s broader cumulative case for theism, of which fine-tuning is one component, is engaged on its simplicity wing in §5.1.4. The cumulative structure does not rescue the fine-tuning weight: Swinburne’s method aggregates several individually-weak considerations toward a posterior above one-half, but each faces the same structural-presupposition reply and its own comparative burdens wherever it is pressed (fine-tuning here, simplicity at §5.1.4, the PSR at §4.1.3), so the contribution of this weight must be reassessed before the cumulative posterior can be claimed. The §5 replies are reckoned together at §5.4. The problem of evil (reactivated above by the for-life gloss and by the appeal to a fall) is engaged in its own right in its own treatment, to which the present section defers. The present section confines itself to the direct inference from narrow-window observation to a tuner, and to the for-life gloss laid on top of it; the companion moves are handled where they belong.

What converges. A prior question, three forms, and a separable gloss converge on a qualified verdict. Fine-tuning can enter a Bayesian case, but only after its comparison class, priors, measure, and observation-selection effects are defended. The multiverse extension can move rather than remove the explanatory burden, yet necessary higher-order parameters are not thereby identical with T, S, and Φ. The cumulative case must show that theism predicts embodied observers in a cosmos like this more strongly than its rivals; the existence of an unembodied divine mind neither proves nor disproves that likelihood claim. Finally, the for-life gloss, separable from the bare inference, fails on its own terms: on the theist’s own ontology consciousness obtains without this configuration, an omnipotent maker cannot fail at the function the gloss assigns, and the world’s overwhelming hostility is evidence against its being for life rather than for it.

A final consideration: the argument for a tuner is delivered within the configuration it calls tuned. The precision of the constants is measured across moments, in laboratories configured in space, using instruments that persist through measurement; the inference from precision to selection is run by minds that instantiate the very structural roles whose contingency the argument is trying to leverage. The observation ‘these constants are precisely set’ is a report from within a world where the constants obtain; the inference ‘therefore a setter’ imports the idiom of selection from domains where selection has content (domains where unselected options also obtain) into a domain where no such options are available to compare against. The actual measurement and argument are structured acts, but a classical theist denies that divine selection is a temporal engineering process using pre-existing space and material. Retorsion establishes the structure of inquiry, not the absence of an external selector. The stronger objections remain the comparative-likelihood, measure, and grounding burdens developed above. The precision is a feature observed from within one structured configuration. Whether it is also evidence of selection remains a comparative explanatory question rather than something retorsion settles.

[1]: The scale disparity, quantified. The observable universe spans roughly 93 billion light-years and roughly 13.8 billion years; a single known planet has hosted life for about 4 billion of those years, and reflective life occupies a fraction of a percent of that span. The precondition compounds the disparity: the mammalian radiation that eventually produced humans followed directly from the Cretaceous–Paleogene extinction roughly 66 million years ago, in which an asteroid impact eliminated the non-avian dinosaurs and most other species then living; the ecological niches that opened were what primates, and later humans, expanded into. A universe built for a species that arrives this late, by way of a mass extinction, and occupies this little of it, is a strange one to have spent 13.8 billion years preparing, most of it filled with sterile voids.

5.1.6 Intelligent Design

The intelligent design argument is the biological-scale version of the design inference engaged at cosmological scale in §5.1.5. Its proponents, Behe,¹⁰⁴ Dembski,¹⁰⁵ Meyer,¹⁰⁶ and Plantinga¹⁰⁷ at selective points, argue that biological complexity exhibits features (irreducible complexity, specified complexity, information content) that no unguided process can produce, and that the inference to a designer is the best explanation of those features. The Discovery Institute has carried the position into legal and pedagogical fora, and the Kitzmiller v. Dover (2005) decision rendered its legal status in the United States, but the underlying philosophical structure is what concerns the present argument.

ID is not simply fine-tuning at biological scale. Its core commitment is the forward design inference from specified or irreducible biological complexity to intelligent agency. Some religious deployments add special creation, an initially perfect order, or Fall-based explanations of dysteleology, but these are not commitments of ID as such and should be treated as optional extensions rather than the defining package. The Fall-based extension is what makes the religiously extended package self-sealing in a way fine-tuning is not. Fine-tuning claims forward tuning of constants and faces straightforward probabilistic and structural rebuttals. ID has a built-in absorber for every poor-design rebuttal: the recurrent laryngeal nerve, the human spine, the ichneumon wasp, the long catalogue of biological inefficiencies and cruelties[^5-1-6-dysteleology-catalogue]: these are read not as evidence against design but as evidence of fall from design. The argument flips dysteleology into confirmation. The present reply engages the core inference and both extensions, not just the core.

First. Against the design inference, the present reply draws on the four probability–modality couplings developed at §5.1.5, named here for ease of reference: the standard-side dilemma (a tuner needs a standard, contingent or necessary, and both options foreclose the tuning), the consciousness self-application (the designer cannot be exempted from the constraints that motivated the inference), the modal-structure mask (improbability claims at the structural level are modal questions disguised as probability questions), and the resolution bound (modal-map coarseness inflates apparent contingency). Where §5.1.5 reads “constants” the argument reads “biological configurations.” Where it reads “tuning,” it reads “designing.” The transfer is not automatic, because the two scales do not pose the improbability in the same key. Cosmological fine-tuning is a claim about the modal status of the constants, whether they could have been otherwise. Biological improbability is pressed as something nearer a first-order probability over a space of chemical and genetic configurations, and that is where the argument’s intuitive force actually sits.

The biological case cannot be closed merely by transferring the cosmological objections. It turns on empirical models of mutation, selection, drift, developmental constraint, and the causal history of particular systems. Modal and probability questions still matter, but the decisive comparison is whether evolutionary mechanisms or design better explain the evidence under defensible likelihoods.

A further pressure bears on the inference from within its own founding image. Paley’s¹⁰⁸ watch is recognized as designed only against a backdrop that is not: the artifact stands out because the heath it lies on does not bear the same marks. The design inference is therefore contrastive at its root, reading “designed” off a difference between the worked object and the unworked ground. The theistic use of the inference then extends the verdict to everything, the watch and the heath alike, since the same designer is said to have made the ground the watch rests on no less than the watch. But that extension consumes the contrast that licensed the first step: if the ground is designed too, there is no unworked backdrop left against which the watch’s design could register, and “designed” stops marking any detectable difference. The inference cannot both earn its conclusion locally (this object, unlike its surroundings, is designed) and generalize that conclusion globally (the surroundings are designed as well) without dissolving the very contrast that gave the local reading its content. Universalized design is indistinguishable from no design at all: not because nothing is designed, but because the predicate has lost the difference it was introduced to track.

The analogy that anchors the inference also marks where it fails. A watch cannot produce another watch, and it is that incapacity, not the watch’s intricacy alone, that makes an external maker feel mandatory: nothing about a watch could bring a second watch about. Living things have exactly the capacity the artifact lacks. They reproduce, vary, and are filtered, so the inference cannot be carried from the watch to the organism without importing the watch’s sterility as a silent premise, which is to assume rather than show that life, like a watch, could not have come about unaided. The point can be put as a controlled comparison. A kennel holds a hundred dogs. In the first case a breeder selects two of them whose cross yields an animal well suited to a particular hunt: the mating is intended and the result is designed. In the second case the same two dogs are simply the last pair left together and breed as dogs do, with no one intending anything: the litter is identical to the breeder’s. Set the two animals side by side. Nothing in either records which history produced it. One was designed and one was not, and the designed one carries no mark of the fact. What the comparison demonstrates is that agency is the sole variable differentiating the two cases; all product-level features are held constant. That result grounds a general claim about the domain the design inference operates in. Design is a prospective process: the designer models a function and selects for it. Selection is a retrospective process: the environment tests against function and retains what fits. Both are function-tracking processes; they differ only in whether the specification precedes or follows the production. For any functional product a designer can yield — any product whose distinguishing features consist in fitness to a function — a sufficiently calibrated selective filter operating on that same function can yield the same, because the features the designer specified-for are precisely the features selection retains. Functional similarity alone may underdetermine whether a trait was designed or selected, because different histories can converge on similar products. But causal histories can leave traces in genealogy, constraint, redundancy, and error-patterns. The design inference must therefore compare full historical models rather than assume either that function uniquely reveals agency or that products can never contain evidence of their origin.

This closes the appeal to specified complexity. A specification is a function-profile: the set of functional properties an object must instantiate to serve its purpose. Selection is differential reproduction based on fitness to the environment. Fitness to the environment is a function-profile. Selection can cumulatively retain variants that function better in a given environment, but it is not literally intentional specification: fitness landscapes emerge from organism–environment relations rather than from a prior semantic target. Dembski’s design filter claims to detect specifications that are improbable given chance but probable given design; they are equally probable given cumulative selection operating on the same function-specification — which is the core of the cumulative-selection reply Sober develops.¹⁰⁹ A specification reliably hit by selection is not a signature of design; it is the predictable output of any process, intentional or not, that tracks function over sufficient iterations. The inference therefore requires a premise that rules out functional-equivalent selection, and that premise cannot be established by examining the product — only by examining the causal history the product-level evidence is being used to infer.

The empirical critique of ID’s specific apparatus, Behe’s irreducible complexity, Dembski’s design filter, Meyer’s information argument, lives in evolutionary biology and is not the structural argument’s burden; that work has been done there, notably by Sober.¹¹⁰ One strand of Behe’s case is not purely empirical, however. The Edge of Evolution claim (that specified biological information is a quantity the unguided process can generate only within narrow limits) is a thesis about what generates structural information, not only about what the biology shows; the structural reply to it, how the immanent operators of change, composition, and constraint generate form rather than merely presuppose it, is developed where the operators are (Chapter 7), and the information-theoretic features are taken up at §4.3. The couplings at §5.1.5 and the comparative-history burden above bear on the design inference here, however that generative question is settled. The present argument shows what the inference must establish before it runs; the biological literature contests the empirical premises themselves.

Second. A historical Fall narrative posits an earlier condition within temporal history from which later corruption proceeds. It need not posit a moment outside T or a categorial origin of time. The relevant objections are empirical and theological — whether there was such an initial condition, what changed, and whether it explains the evidence — not Parasitic Negation alone. There is no moment-prior-to-T from which the first change departed, no temporal position at which “perfect creation” was instantiated as an initial condition outside the order it would initiate (§§2.6.6, 7.2.2, 7.6). The quantitative question of how far back along the dimension the passage runs (whether the past is chronometrically finite or infinite) is a separate question within T, engaged at §5.1.2 against the Kalām’s traversal and BGV arguments. The fall narrative in its privileged-t₀ form needs the categorial t₀; the Kalām presses the quantitative question. The two are dispatched at different sections by different machinery. Parasitic Negation applied to T (§§2.1.5, 2.6.6) is the move ID’s privileged-t₀ commitment cannot survive: any beginning would already instantiate the order it claims to begin. The question “when did time begin?” admits two readings. The quantitative reading (how far back does succession run) has a numerical answer within whatever cycle one is asking from. The categorial reading (what moment yielded T) is malformed: a moment is what T provides, not what produces T. A moment of perfect creation followed by decay is structurally describable as an event within time. Its difficulty is that ID evidence does not establish such a moment and the biological and cosmological records may count against particular versions of it. The categorial closure dispatches the fall narrative’s t₀; it does not dispatch the Kalām’s quantitative question, which is the work of §5.1.2.

Third. The asymmetric decay vector, the third commitment, reads observed imperfection as evidence of fall from initial perfection. The reading conflates two distinct phenomena: entropy increase (a structural feature of T running) and moral or teleological decline (a normative feature attributed to history). The conflation is what gives the fall claim its empirical footing: every ad hoc construction, vestige, inefficiency, and cruelty in the biological record is reabsorbed as evidence of a perfect prior state from which the present has departed. Entropy increase should not be identified without argument either with moral decline or with Temporality itself. Thermodynamic arrows depend on physical laws and boundary conditions; a Fall theology that links them to corruption owes an additional empirical and doctrinal account. Reading entropy as moral-teleological decline imports a normative direction the present account locates nowhere. T running is not T falling. The fall reading also lacks a historical anchor. It requires a moment at which a first human moral agent acted in a way that initiated decay, but roughly half a billion years of metazoan suffering predate any organism that could be such an agent, and contemporary population genetics rules out descent of Homo sapiens from a single founding pair within the span the lineage permits.¹¹¹ Plantinga’s O Felix Culpa¹¹² relocates the fall into a privileged theological moment whose empirical anchorage is set aside; the move preserves the theological structure at the cost of giving up exactly the historical fall the original argument needs. Without the entropy-as-fall reading, and without a historical fall to anchor it, the dysteleology cases revert to what they were before being absorbed by the theodicy: evidence against the design inference, not evidence for it.

A note on scope. The fall claim also makes a directional empirical prediction (that observed history runs away from an initial perfection), and there are aggregate gradients in human social life and in religious demand that appear to run the other way. Those gradients are contested empirical terrain, and the present section does not rest on them: the fall narrative is already foreclosed structurally, by the categorial-t₀ argument and the entropy misreading above, whatever the long-run social data ultimately show. The substantive case that these gradients positively confirm the structural account is made where that work belongs (§8.12.4 on the moral arc and §8.14.4 on declining religiosity) and is left to stand or fall there on its own evidence.

What closes. The core inference and its two extensions dispatch separately, and the package does not survive their separation. The design inference carries the couplings at §5.1.5 and the comparative-history burden developed above. The privileged origin moment is foreclosed by the present argument that T has no exterior or prior origin moment, a categorial closure that leaves the quantitative question of a finite chronometric past to §5.1.2 and Chapter 7. The asymmetric decay vector misreads entropy as moral decline. A Fall-based absorber can make a religious design theory resistant to dysteleological evidence, but it is an optional theological auxiliary rather than part of ID’s core. It incurs empirical and explanatory burdens concerning the initial condition, mechanism, and scope of corruption. Without the absorber, the design inference at biological scale inherits the burdens of the design inference at cosmological scale, which §5.1.5 leaves undischarged.

[1]: The catalogue extends to reproduction itself. The American College of Obstetricians and Gynecologists estimates that as many as one in four recognized pregnancies ends in miscarriage (ACOG estimate, cited in Deshpande and Kucirka, StatPearls, 2023), and a substantial further share of fertilized eggs are lost before implantation is ever detected. A designer credited with intending the birth of every child inherits, on the same accounting, a design in which a large fraction of attempts at that outcome fail before term.

5.1.7 Modal Ontological Argument

Plantinga’s modal ontological argument¹¹³ runs as follows.

1. A being is maximally great just in case it is omniscient, omnipotent, and morally perfect in every possible world, so that maximal greatness includes existence in every possible world.

1. Possibility premise. Possibly, a maximally great being is instantiated: ◇□G.

2. S5. Possible necessity yields necessity: from ◇□G, □G.

3. Conclusion. A maximally great being exists in every possible world, the actual one included: □G, and so G.

The crux is the possibility premise: is maximal greatness coherent?

The argument has antecedents and successors that sharpen what is being claimed. Hartshorne¹¹⁴ develops an earlier modal version of the argument from Anselmian materials, and Adams¹¹⁵ defends the broader thesis that the modal-ontological strategy survives the Kantian objection that existence is not a predicate. Plantinga’s own assessment in The Nature of Necessity is more modest than the argument’s later reception: the argument shows that belief in a maximally great being is rational, not that disbelief is irrational. The objection developed below establishes a stronger result. It does not merely show that disbelief is rational alongside belief; it shows that the classical version of the argument cannot succeed even at its more modest target.

The argument’s modal core relies on the same fixed-domain S5 framework the present book operates within. The present account does not contest the transfer rule; it contests whether the classical possibility premise supplies a determinate candidate within the global domain of real possibility.

Try to conceive of an atemporal, aspatial, immaterial being. The conception fails not because the being is alien or unlike anything in experience but because the description subtracts from its object the conditions of statehood itself: temporal location, spatial-relational differentiation, substantial bearing. The collapse of the conceiving act (succession of imaginative acts, distinction of the imagined being from background, sustained representational substance, each withdrawn by the very content entertained) is the symptom of that subtraction rather than the engine of the argument; the load is carried by what the description withdraws from its object, made exact in the conditions-for-statehood argument below. The general result has already been pressed at §3.3: structural absence cannot be coherently conceived, because every attempt to represent it instantiates what it denies, and the asymmetry between conceivability-to-possibility and inconceivability-to-impossibility favors the downward direction when the inconceivability is of this structural kind. The maximally great being is the case where structural absence is dressed in classical-theistic predicates. What follows works the same point through Plantinga’s machinery directly.

Plantinga’s possibility premise concerns maximal excellence — omniscience, omnipotence, and moral perfection — instantiated in every world. Atemporality, aspatiality, immateriality, and strong divine simplicity are additional classical-theist doctrines, not premises of the modal ontological argument itself. Coherent possibility, in the relevant sense, requires that the conjunction’s content be specifiable in a way that does not contradict itself or contradict the conditions on which content is possible. If a defender adds the strongest negative classical attributes, the book can press whether a personal being so characterized retains determinate cognitive, volitional, and grounding roles. That pressure does not show that Plantinga’s thinner maximal-greatness predicate is contentless; it challenges one theological development of the candidate.

A defender may attempt the Plantingian reply: vehicle and content come apart. One can conceive a colorless object using a colored display; conceiving an atemporal being does not require the thought itself to be atemporal. The reply works for single properties: color, mass, charge, shape, all can be absent without the object collapsing. T, S, and Φ fail the test categorically. An object lacking temporal location is not an object existing timelessly but an object of which no moment can make ‘this exists’ true; an object lacking spatial-relational differentiation is not an object with unusual geometry but an object with no interval separating it from anything else; an object lacking Substance is not an object existing for zero duration but an object with no being through which predication could run.

T, S, and Φ are not properties within a state of affairs; they are conditions for statehood. The colorless-object analogy does not carry here, not because the maximally great being is missing more properties, but because it is missing the conditions under which properties could be instantiated at all.

The claim is scoped to concrete existents. Abstract objects (numbers, propositions, the possible worlds Plantinga’s own abstract-worlds actualism reifies) bear properties without instantiating T, S, or Φ, and nothing here denies it; the status of abstracta is treated in its own right at §4.2.4. But the maximally great being is not an abstract object. It is a concrete person who knows, wills, and acts, and those are the very acts that require succession, distinction, and a persisting subject. The Platonist’s atemporal existents are no refuge for it, and neither is a third category: the classical theist may reply that God is neither a structural concretum nor an abstractum but sui generis, pure act on the divine-simplicity model, and so exempt from the dilemma. The reply relocates the burden rather than discharging it. A being that knows, wills, and acts has already been given a personal profile whose exercise requires succession, distinction, and a persisting subject, whatever category label is invoked to secure the exemption; the constitutive case that a bearer of those acts must instantiate T, S, and Φ is pressed against the ipsum esse construal directly at §6.1.

Because these are conditions for statehood and not features within a state, the description does not specify an object that then proves impossible (an impossibility would make the possibility claim about it false, true in no world) but fails to specify a possibility-candidate at all.

The diagnosis should not be placed automatically before truth-value. Standard semantic treatments can assign falsehood, presupposition failure, or undefinedness to defective descriptions. Here the safer verdict is dialectical: the defender owes a coherent account of how the added classical attributes combine with personal agency; semantic emptiness does not follow merely from their negations. Where standard modal semantics assigns falsity to an impossible proposition, the fully classical candidate leaves no determinate proposition for any world to evaluate: the presupposition on which the description’s content depends is exactly what the description withdraws. The possibility premise, read with the full classical conjunction, is not false but empty: it carries no modal content for S5 to operate on.

Possibility-Premise Emptiness · R13 · Result of the Conceivability Asymmetry

The possibility premise of Plantinga’s modal ontological argument carries no modal content: the conjunction of classical attributes specifies ∅ under personal description, and S5 cannot transmit necessity from an empty input. The argument is therefore not unsound but inapplicable. The local Chapter 5 verdict rests on the §3.3 conceivability-asymmetry result and the §5 machinery already established; the Result form anticipates Supernatural Preclusion (§6.1) via Horn 1 of the Trilemma Closure, where the same result is restated under Supernatural Preclusion once that result is in place downstream. (§3.3, §6.1)

The standard objection remains central: S5 transfers the possibility premise but does not support it. The structural critique can add that some classical enrichments of the candidate create tensions with personal agency. It should not claim that Plantinga’s possibility premise carries no modal content simply because those additional doctrines are contested. What is described by the conjunction of the classical attributes is not a genuine possibility-candidate, because what is described is not a possible configuration but the absence of the conditions for any configuration.

The present objection should also be distinguished from the symmetry objection developed by Oppy¹¹⁶ and Sobel.¹¹⁷ The symmetry objection observes that the same modal apparatus generates a parody: possibly there is no maximally great being, therefore necessarily there is no maximally great being. The objection’s force is that the argument and its parody share a form, so neither tells us which possibility actually obtains. The present objection does not rely on parody. It denies that the classical maximally great being is in the possibility space at all. The parody objection is consistent with the possibility premise being meaningful; the present objection denies that the possibility premise carries the kind of modal content that being true or false would require. Where the symmetry objection leaves the modal-ontological project a stalemate, this objection denies the project its starting move.

The S5 inference is intact. The modal framework is intact. What fails is the input. A modus ponens (the inference from P and ‘if P then Q’ to Q) transmits a conclusion only from a premise that carries modal content; a possibility premise that specifies no possibility-candidate carries none, so nothing is transmitted, regardless of how impeccable the inference rule. The failure is not that the premise is false (a false premise would still be a premise) but that no modal content is present for truth or falsity to attach to.

A second front is available, independent of the foregoing rebuttal. Van Hamme¹¹⁸ argues that modal logic alone does not supply a substantive account of a necessary personal God. This should be stated precisely. S5 can transfer necessity to any determinate proposition in the global domain W; it does not restrict its scope to structural vocabulary. What it cannot do by itself is establish that personhood-attributing predicates — omniscience, omnipotence, moral perfection — form a coherent necessary package or that they apply to the candidate described. The modal-ontological argument therefore needs more than the transfer rule: it needs an independently supported, determinate possibility premise for a personal being.

This front constrains the present account too, harmlessly: the book does not derive the personal or the axiological from S5 alone. Chapter 6 must ground its claims about value, agency, and greatness in the structural and dialectical arguments specific to them. The two rebuttals therefore do different work. The first challenges the classical candidate’s content; the second blocks the inference from a bare modal rule to the full personal-theistic conclusion. Together they show that S5 neither repairs an empty specification nor substitutes for an argument about the attributes.

The formal derivation. A more rigorous version of the modal ontological argument does not assert the possibility premise but derives it. Gödel’s ontological proof (his axiomatization of positive properties, sharpened in Anderson’s emendation against modal collapse and analyzed in Hájek’s formal treatment) lays down that every property is positive or its negation is, that positive properties are closed under entailment, and that necessary existence is positive, and from these proves as a theorem that a God-like being, one bearing every positive property essentially, is possibly instantiated; S5 then delivers necessary existence. Deriving the possibility premise looks to answer the present objection at its root: the premise is no longer left unsupported but established by proof.

It does not answer it. Gödelian derivations shift the burden to the positive-property axioms, their consistency, and the definition of God-likeness. Those axioms do not by themselves encode atemporality, aspatiality, or immateriality, so the earlier structural critique cannot simply be transferred to them. Formal derivability still does not show that the axioms provide the intended theological interpretation. Formal derivability manufactures no more modal content than the axioms supply: the proof is valid, and whether its input carries the intended theological content is exactly what the axioms leave unsecured.

The personhood front closes the same gap from the other side: Gödel’s positive properties include the personhood-attributing predicates, omniscience and omnipotence and moral perfection, that fall outside the structural vocabulary any modal logic can certify, so even granting the axioms their domain the apparatus cannot necessitate the personal. The Gödelian formalization raises the rigor of the derivation; it leaves untouched the status of what is derived.

A respondent may revise maximal greatness to drop the classical attributes, making the maximally great being temporal, extended, and embedded. The revision changes the subject in two respects. First, the present argument does not dispute such a being’s possibility, only the inference from possibility to necessity by S5: the inference depends on the being’s maximal greatness across worlds, which the revision has redefined and which now requires independent argument. Second, the revised being instantiates T, S, and Φ rather than grounding them, and a being so embedded depends on structure for its existence. The aseity the modal-ontological tradition was specifically trying to secure (necessary existence as a feature of the being rather than of its conditions) is the property §6.1 and §6.10 (Horn 3) establish the embedded deity cannot have. The revised argument may go through to a different conclusion, but the conclusion is no longer the conclusion the classical modal-ontological project was constructed to deliver.

What is established. S5 validates the transfer from possible necessary existence to necessary existence, but it does not establish the possibility premise or the coherence of the full theological package. Structural tensions arise when strong negative classical attributes are added to a personal agent; those tensions qualify the candidate without making Plantinga’s thinner premise semantically empty. The personhood front closes the same line from the inference side: modal logic cannot necessitate the personal even where the structural conditions are present. Revising the maximally great being to drop the classical attributes preserves the inference at the cost of what the classical project was specifically trying to secure: aseity as a feature of the being rather than of its conditions.

No rescue by changing the modal logic. A defender may retain the classical candidate while rejecting S5, so that ◇□G no longer transfers to □G. This is not self-undermining merely because it is stated in modal vocabulary; it is a rival account of real possibility and its global domain. But it is self-defeating for Plantinga’s specific derivation: the step from possible necessary existence to necessary existence is the inference that makes the modal ontological argument go. Rejecting S5 does not save that derivation; it replaces it with a different argument that needs its own modal semantics and its own defense of the possibility premise. The present account maintains the fixed-domain S5 analysis of §2.1. Granting it, the argument still needs a determinate classical candidate; rejecting it leaves the Plantingan transfer unavailable. Either way the original possibility line has not been secured.

A parallel epistemological line. Pascal’s wager is taken up at §5.1.11, where the temptation to run the same possibility-of-classical-necessary-being premise through a probability claim is closed off and the wager is returned to its own decision-theoretic ground.

Plantinga’s reserve: de re/de dicto and EAAN. Plantinga would press two further questions on the present account. First, the modal-scope question: □(T∧S∧Φ) can be read de dicto (necessarily, whatever constitutes the modal constraint on change instantiates the three roles) or de re (these specific structural occupants necessarily exist). The two readings differ on whether contingent states of Nature inherit necessity from Nature’s necessity. The de dicto reading is the intended scope: what is necessary is that the roles be instantiated, not that any specific instantiator is necessary. Contingent states of Nature remain contingent within a necessary structural framework (§2.8.4); the scope-limit argument at §2.5 (‘P4’s necessity attaches to F, not to every item within the field F makes possible’) is the relevant disambiguation; and the co-instantiation claim at §2.6.5 secures the de dicto reading without collapsing into de re necessity for contingent occupants. Second, Plantinga’s Evolutionary Argument Against Naturalism (Warrant and Proper Function¹¹⁹): the present argument is delivered by cognitive faculties selected for survival; whether such faculties can reliably track necessary structural conditions is the EAAN question. The structural account’s response (that any functioning cognitive faculty operates within T, S, and Φ by the transcendental conditions on operation itself) lives in Chapter 8.

5.1.8 Modal-Grounding Argument

Arguments of this kind reason toward a divine being. The modal-grounding argument reasons toward God as the explanation of modality itself: of why there are necessary truths and real possibilities at all, rather than nothing modal to speak of. Because it leans on the same modal facts the present account is built from, it is the most direct theistic competitor to the book’s own grounding of necessity, and it warrants engagement on its own terms rather than as a restatement of the ontological argument §5.1.7 handles.

The argument’s most developed form is Leftow’s. Historically, theist modal theories based all necessary truth on God’s nature; Leftow (God and Necessity¹²⁰) argues instead that necessary truths about possible creatures and kinds are fixed by God’s unconstrained imagination and choice, with the broadly logical modalities reduced to causal modalities seated in God and events in God’s mind. The distinctive move is a global economy argument: theists are committed to some theist modal theory, and the explanatory merits of a good one (its economy against the heavy ontologies of Lewisian modal realism and abstracta-laden ersatzism) are offered as an argument for God’s existence. Adjacent constructions press the same thought from other directions: Plantinga’s suggestion that propositions are best construed as divine thoughts, since there are far more of them than finite minds could entertain (“Two Dozen (or So) Theistic Arguments”¹²¹); and the theistic activism, or absolute creationism, of Morris and Menzel,¹²² on which abstracta (modal truths among them) depend for their existence on God’s intellectual activity. The shared structure is an inference to the best ground:

1. Grounding demand. Modal and necessary facts are real and cannot be brute: they require a ground.

2. Mind as ground. A mind is the best or only adequate ground for them.

3. Conclusion. A divine mind exists as the ground of modality.

Grant the demand in full. The argument is right that modal facts are not free-floating: that “possibly P” and “necessarily P” cannot be ungrounded surds, true with nothing about Reality to fix them. Grant, too, that a plurality of concrete non-actual worlds (Lewis), or a Platonic heaven of modal abstracta, is a heavy and contestable ontology. The demand for a ground for modality is legitimate, and the dissatisfaction with possible-worlds Platonism is one the present account shares rather than resists.

What the argument adds to that demand is a mind to satisfy it, and the structure satisfies it first. Reality’s fixed global domain of real possibility grounds modal status at §2.1: Modal Stability derives S5 from evaluation over one complete domain W, while the Bottleneck distinguishes that global modality from local reachability. Transition Closure is a narrower auxiliary result concerning onset from capacitylessness, not a third derivation of proposition-wise S5. Modal status is therefore not conferred by an intellect entertaining worlds; it is fixed by the global modal structure the account articulates. The explanatory work Leftow credits to God’s imagination and choice, and Morris and Menzel to God’s intellectual activity, is already discharged one level down, by the conditions any possibility presupposes. This bears directly on the global-economy form: the theist theory is advanced as the best ground precisely because its rivals are taken to be Lewisian realism and abstracta-Platonism. A structural account is a third option that grounds the modal structure without a mind and without a plurality of worlds, so at that level the inference from “best available ground” to God is blocked at the comparative step it depends on.

This front must be drawn at its real scope. What §2.1 grounds is the modal structure: that modal status is determinate, and determined by structure. The modal structure is schematic and existence-silent: it rigidifies what is possible and necessary without settling which contingent things exist or which specific creaturely natures populate the space of the possible. Leftow’s theory reaches further than the modal structure, aiming to explain why these creaturely kinds are the possible ones. The present account does not, on its own resources, fix that de re creaturely content; it fixes the modal structure, not the full roster of what the modal structure permits. A defender can therefore concede that the modal structure needs no mind and still press that the content of creaturely possibility does. Taken alone, this front narrows the argument’s reach; it does not close it.

The argument’s deeper trouble is independent of how much of modality the structural account claims to ground: the proposed ground cannot do the grounding without carrying a structural burden of its own. A classical divine intellect is typically described as atemporal, non-spatial, and immaterial. The relevant pressure is whether it can bear determinate modal content and stand in an asymmetric grounding relation to possibilities without instantiating at least thin order, difference, and bearing roles. The revised §5.1.7 verdict does not make the candidate semantically empty by definition; it leaves a substantive grounding burden. Actual finite imagining involves succession, distinction, and persistence, while the classical theorist proposes one timeless intellectual act. Even on that proposal, determinate contents and asymmetric dependence remain. This secures a minimal structural burden; whether it entails the loaded temporal, spatial, and physical types remains disputed.

A defender might relax the ground to an embedded, temporal mind, one that does instantiate the conditions. Then it is no longer the classical God whose aseity the theory was built to secure, and a mind so described instantiates Time, Space, and Substance rather than grounding them: it depends on the structure for its own operation, and so cannot be modality’s ultimate ground. The revision keeps a grounder by giving up the ground’s ultimacy.

The argument points at something real: modality is not explanatorily idle. The present account locates its modal structure in Reality’s fixed global possibility-domain and the structural Conditions later derived for worlds in that domain. The general result of §4.1.10 is narrower than an absolute sourcelessness thesis: no candidate whose own status is modal can confer the modal domain it already inhabits. A mind posited as an internal modal item therefore cannot ground W from outside; an essence-prior or transcendent mind remains a further metaphysical posit whose grounding credentials require independent argument.

What is established. The modal-grounding argument is right to refuse free-floating modality, and the structural account supplies a non-theistic ground for the global modal domain. That blocks the global-economy inference to God at the comparative step. A classical divine mind stripped of every structural role also fails to supply an intelligible active grounding relation on the present account. A defender may instead posit an essence-prior or transcendent ground, but then the proposal has moved beyond a modal-grounding explanation and owes an independent metaphysics of that ground. What the argument sees — the need for modal explanation — the structure supplies; what it adds — a specifically divine mind — remains unearned by the modal facts alone.

5.1.9 Transcendental Argument

The transcendental argument for God presses logical necessity as its strongest case. The Laws of Identity, Non-Contradiction, and Excluded Middle are absolute, universal, and not plausibly derivable from any contingent feature of the world; something must secure their status, and a purely structural account can seem too thin to carry that weight, so on this argument the securing ground is God. In its most ambitious form the claim widens past logic to intelligibility as such — the very possibility of making sense of anything — but logic is where it bites hardest, and that is where the engagement begins. Pressed for the relation that ground bears to logic, the argument divides into three genuine positions: classical presuppositionalism (;¹²³)¹²⁴ — the transcendental argument in the strict sense — places logic outside God, as externally grounded by a necessarily rational being; theistic conceptualism (;¹²⁵;¹²⁶;¹²⁷)¹²⁸ places logic inside God, as a structural feature of divine thinking; Cartesian voluntarism places logic downstream of God, as the free output of divine will. The three moves cover the principal theistic grounding strategies considered here (logic external to and grounded by God, logic internal to divine nature or thought, and logic dependent on divine will) and none secures a distinct theistic ground without unpaid burdens: voluntarism appears to presuppose the alternatives it would decree, presuppositionalism owes a non-circular account of its Author, and conceptualism converges on the structural conditions under a theistic description.

Regimented, the argument runs:

1. The laws of logic (Identity, Non-Contradiction, Excluded Middle) are absolute, universal, and necessary.

2. Their necessity cannot derive from any contingent feature of the world; something must secure it.

3. Only a necessarily rational being can secure necessary rationality.

4. Therefore a necessarily rational being, God, grounds logic.

The presuppositionalist move is, in its contemporary form, the most philosophically developed of the three. Van Til’s The Defense of the Faith¹²⁹ and Bahnsen’s Van Til’s Apologetic¹³⁰ stand at the head of the presuppositionalist tradition, which argues that any non-theistic worldview cannot account for the preconditions of intelligibility (logic, induction, morality) without borrowing from the theism it purports to reject. Morris and Menzel¹³¹ reframe the tradition with the view that has dominated the last three decades: theistic activism, on which God’s necessary thinking constitutes the realm of abstract objects, including logical truths. Plantinga’s Does God Have a Nature?¹³² had earlier pressed the companion question whether God stands subject to the logical laws or is identical to them, and settled on a version of the conceptualist reply. Leftow’s God and Necessity¹³³ is the most systematic contemporary treatment: defending a modal conceptualism on which the modal facts are grounded in God’s nature. Welty¹³⁴ develops a theistic conceptualist position that locates logic inside God rather than outside. Craig’s God Over All: Divine Aseity and the Challenge of Platonism¹³⁵ presses the same aseity motive to a different destination: surveying divine conceptualism among the realist options and finding it wanting, Craig concludes that abstract objects (propositions, properties, logical truths) are not entities of any kind, so the apparent Platonic realm is dissolved by anti-realist deflation rather than relocated into divine intellection. The voluntarist position runs through Descartes’s letters to Mersenne¹³⁶ and has few contemporary defenders, though Frankfurt¹³⁷ and Curley¹³⁸ give it sympathetic reconstructions. Against these stand anti-theistic-grounding positions: Sober’s¹³⁹ Humean line that logical truths require no grounding beyond their formal character; and Sher’s¹⁴⁰ logical realism, which treats logic as objective but not as requiring a rational ground.

The present account is closer to Sher’s realism than to any of the theistic positions, but the reason is specific: logic describes formal constraints on determinate contents and relations. Where a manifold is offered as a concrete change-realizing reality, the structural Conditions identify the roles it must realize. Logic alone does not derive the full T, S, and Φ types from any arbitrary manifold of relata.

The challenge bears directly on §2.6 and §2.7. T, S, and Φ are identified at §2.6 as the Triconditional Identity. Questions of the form “why is there something rather than nothing?” are dissolved at §2.7, which argues that ∅ (the Null State: the configuration in which none of T, S, or Φ is instantiated) is incoherent on multiple independent grounds. The divine-logic objection presses both: against the structural account, that logical necessity is a distinct domain requiring its own ground; against §2.7, that the positive necessity of the laws ruling out ∅ requires a ground of its own, and the ground must be a necessary rational being. The reply grants the necessity and universality of the core logical laws. T, S, and Φ can supply concrete instances of determinacy, difference, and ordered dependence, but logic is more general than any particular physical manifold. The structural account therefore offers an interpretation of logic’s application; it does not derive all logical truth from the loaded types alone.

The three moves share a prior assumption worth naming: each treats the laws of logic as prescriptions, edicts that hold because something with the authority to lay them down has done so, which is why a legislator seems required to account for their force. But the “law” in “laws of logic” is descriptive, not prescriptive. It issues no command that a manifold of distinct relata could obey or flout; it states what being such a manifold already comes to. A description needs something to describe, not someone to enact it, and once the laws are read this way the office of lawgiver the three moves compete to fill stands empty before any of them is assessed.

The distinction can be put as one of prepositions. What holds necessarily of God, that a given act is necessarily open to him or that his nature is what it is, is a law of God, not a law from God: it describes him rather than issuing from his authority, and he could not have prescribed it to himself, since to prescribe anything already presupposes the modal space in which the prescription would hold. A lawgiver who must stand under the laws in order to legislate them has not grounded them but exemplified them, and the appeal to divine authority regresses at the very point it was meant to terminate. Three forms of the move take up the objection.

First. The classical presuppositionalist treats logic as externally grounded. On Van Til and Bahnsen’s view, logical truths require something outside themselves to secure their necessity, and the only candidate is a necessarily rational, necessarily existent being whose nature underwrites the laws without being reducible to them. Every argument presupposes logical laws; every use of logical laws presupposes their ground; every ground of necessary rationality must itself be necessary; and the only necessary thing that can ground necessary rationality is a necessarily rational being. Call this the authoring relation: God authors logic by being the kind of being whose existence makes logical truth possible.

The authoring relation generates a modal dilemma at its own level. If the necessarily rational being (call it the Author) is the source of logical possibility, then the counterfactual “were the Author not to exist, logic would not obtain” must be true. But ‘were the Author not to exist’ is itself a logical possibility, a consistent hypothesis about what could fail to obtain, and its being a logical possibility presupposes the laws of logic holding. If the Author’s non-existence is logically possible, logic cannot depend straightforwardly on the Author’s actual existence. If it is logically impossible, the theist may say that logical necessity is grounded in the Author’s essence rather than externally governing it. That answer faces a circularity or explanatory-priority burden, but it is not formally inconsistent merely because logical predicates apply to their proposed ground.

The regress completes the point. If logic is grounded in the Author’s existence, what grounds the fact that the Author’s existence has the relevant logical character? The presuppositionalist’s candidate is the Author’s necessary nature, but “necessary nature” is itself a logical characterization, and the grounding runs in a circle. Bahnsen’s transcendental formulation (“the impossibility of the contrary”) concedes the point when pressed: “the impossibility of the contrary” is a logical modal claim, and the impossibility it invokes is the very kind of necessity the argument is trying to ground. The presuppositionalist account, pressed for substance, supplies a name for what underwrites logical necessity rather than an independent grounding of it.

Second. The contemporary conceptualist relocates logic inside the divine mind. On the conceptualist view (;¹⁴¹;¹⁴²;¹⁴³),¹⁴⁴ logical truths are not external to God and grounded by God; they are constitutive of God’s thinking. The Law of Identity is not a law God obeys and not a law God creates; it is a structural feature of how God necessarily thinks. Logic is God thinking. The view is a substantial improvement on the classical presuppositionalist version: it eliminates the authoring relation (which the previous move showed to be incoherent) and replaces it with an identity-style grounding that avoids the regress. Leftow’s God and Necessity develops the most careful version of this position, and it is genuinely resilient to the reply given to move one.

The concession is real. The reply is that the position succeeds only by locating structured content inside the divine mind. The point should not be put as though divine thinking must be discursive in the creaturely sense. The conceptualist explicitly denies that God reasons through logical truths one after another. The pressure is weaker and harder to evade: if logical truths are grounded in divine intellection, then divine intellection must contain determinate logical content. The Law of Identity requires determinacy of content; the Law of Non-Contradiction requires the distinction between a content and its negation; the Law of Excluded Middle requires an exhaustive contrast between a proposition and its negation. T does not enter because Excluded Middle is itself a temporal sequence. It enters through the asymmetric grounding and inferential roles the conceptualist needs: ground to grounded, premise to consequence, alternative-space to exhausted alternatives. These are not temporal episodes in a divine psychology. They are structural features of the content the conceptualist places in God.

Leftow¹⁴⁵ is explicit that the divine modal activity is not temporally successive and does not involve distinguishable mental states. But if the content of that activity is genuinely modal (if it fixes the space of possibilities and the laws governing it) it must still be determinate, internally articulated, and capable of grounding asymmetric entailment relations. That is structure. Plantinga¹⁴⁶ entertains exactly this worry under the heading “the bootstrapping objection”: if God’s nature grounds logic, and logic applies to God’s nature, then the grounding relation runs in a circle. The standard conceptualist reply distinguishes the explanatory priority of God’s nature from its logical priority, but the distinction requires logical apparatus to state, and the apparatus is what the account is trying to ground. The conceptualist account converges on the same structural conditions the argument names; the remaining disagreement is whether those conditions are best described as features of a divine mind or as features of structure as such.

Third. The Cartesian voluntarist makes logic the free output of divine will. On Descartes’s letters to Mersenne,¹⁴⁷ the eternal truths (including the laws of logic and mathematics) are established by God as a king establishes laws in his kingdom. God could have decreed otherwise; the fact that God decreed these laws rather than others is what makes them necessary; and the modal force of the laws traces entirely to the divine will. Frankfurt¹⁴⁸ reconstructs the view as genuinely coherent: a position on which God’s omnipotence extends to the logical as well as the natural order. Curley¹⁴⁹ gives a more cautious defense, reading the Cartesian view as a limit position rather than a settled doctrine.

Voluntarism faces the deepest difficulty of the three: the act of decree appears to presuppose what it was meant to create. A temporal decree would involve succession, distinction, and persistence. Cartesian voluntarists can instead construe God’s willing as eternal. The remaining objection is that free selection among alternative logics appears to presuppose determinate alternatives, identity, and consequence relations already available to the will; this yields a logical-priority problem without assuming a temporal divine psychology. The source of logic must already be inside logic to perform the act of sourcing. The objection is sharper than the parallel regress in move one, because voluntarism is committed to the contingency of the decreeing (God could have willed otherwise) and contingency is the mode of choice between alternatives. For there to be alternatives, there must be a space of distinct options; for a choice among them, an asymmetric dependence of outcome on selection; for the choice to hold, a chooser distinct from and stable across what is chosen. Voluntarism demands every structural condition the theory is trying to explain, in the very act of explaining it. This is why contemporary theism has largely abandoned the position: not because the motivation is wrong (the motivation is to protect divine omnipotence from external constraint) but because the machinery is self-defeating at a level even the conceptualist can see. Leftow¹⁵⁰ devotes an entire chapter to showing why voluntarism cannot be the theist’s fallback, precisely because it presupposes what it tries to ground. The present reply agrees with Leftow’s diagnosis and presses the point one level further: the conditions voluntarism presupposes are T, S, and Φ, and once they are secured, no further appeal to divine will is needed to underwrite logical necessity.

A kindred deflationary move presses from outside theism entirely. Sober¹⁵¹ argues, in a Humean spirit, that logical truths need no grounding at all: their status as truths is a matter of their formal character, not of any metaphysical relation to a ground. Sher’s Epistemic Friction¹⁵² develops a logical realism on which logic is objective but not grounded in any further entity; logical truths track formal features of reality the way mathematical truths track quantitative ones, and no rational author is required. The structural account agrees with Sober and Sher that no being is required to ground logic, and adds a constructive account of what does underwrite logical necessity: the structural roles. Where Sober stops at the formal character of logical truth and Sher stops at objective formal features of reality, the present account specifies which features those are and why they obtain necessarily. The convergence is on denying the theistic grounding; the divergence is on whether anything positive can be said about what logic describes. The structural account says yes, and says it is what the present book has been defending from §2.1 on.

Two adjacent objections are addressed elsewhere. Divine simplicity as the doctrine that God is identical to the divine attributes (bearing directly on the conceptualist move) is §5.2.1. Swinburne’s case that the necessary ground must be a mind rather than a set of roles is §5.1.4. The present section confines itself to the positive claim that logic requires a rational ground, and contests it on the three strategies available for what that ground could be.

What the three moves show. The three strategies receive different verdicts. Voluntarism faces the strongest internal difficulty, because an act of decree appears to presuppose the logical and modal alternatives it is meant to create. Presuppositionalism must explain how a necessarily rational Author grounds logic without merely describing that Author through logical predicates. The conceptualist view may coherently locate logical content in divine intellect, but then it becomes a rival metaphysical interpretation of the relation between logic and mind rather than a simple escape from structural constraints. The present account offers a non-theistic structural interpretation of logic; it does not establish by retorsion alone that no essence-prior or conceptualist account could be coherent. The remaining question is comparative: which account best explains the necessity and applicability of logic without adding an unearned grounding relation.

The widened claim. Pressed on logic, the presuppositionalist widens the target: the claim becomes not that logic requires a divine ground but that intelligibility itself does — that the very possibility of making sense of anything (logic, induction, the uniformity of nature, the fit between mind and world) presupposes God (;¹⁵³,¹⁵⁴ whose 1985 debate with Stein ran the demand for logic, induction, and moral standards together). The widening looks like a strengthening. It is a category error, and naming the category dissolves it.

Intelligibility is not a substance that something must supply; it is a relation, the fit that obtains when a tracker converges on a structure already there. A text is intelligible to a reader who has acquired its language and opaque to one who has not, though the text is unchanged: what varies is not a quantity of intelligibility resident in the ink but the match between an acquired competence and an antecedent structure. The relation has two terms, and each is already accounted for. The objective term is structure — that there is something with determinate identity, distinction, and exclusion to be tracked at all — and that is what T, S, and Φ secure, the burden this chapter has already discharged. The subjective term is a competence that converges on the structure, acquired the way any tracking is acquired, through exposure and adjustment. Neither term, and so not the relation they compose, requires a further being to supply it.

The presuppositionalist’s demand — what grounds intelligibility? — therefore mislocates its object. To ask what supplies the fit between a competence and a structure is to treat a relation as though it were a thing, on the model of asking what grants a key its fit to a lock: nothing grants it, the cut of the one matches the cut of the other, and the matching is not a third item in need of a donor. Once intelligibility is seen as the convergence of a tracker on a structure rather than a gift bestowed from above, the question that seemed to demand God as its answer turns out to have no place for an answerer.

The demand fails from the other side as well, once the relation is turned on the being the argument offers as its ground. To make sense of something is an achievement: a competence closing a gap, gathering a structure it did not already hold. The relation runs in a direction, from not-yet-tracking to tracking, and it presupposes a knower who lacked the structure before acquiring it. An omniscient being holds all structure already, so it has no gap to close and nothing it comes to track: nothing is intelligible to such a being, not through any deficiency but because making sense of names the work of a finite tracker meeting a structure it did not author. The argument has reached for the one being to whom the phenomenon it is built on cannot apply. The second reading of the phrase fares no better: if to make sense of the world were to make it, to author the structure rather than converge on it, that is voluntarism, and it collapses under the act of decree exactly as the third move showed. On the first reading there is nothing for an omniscient ground to make sense of; on the second the making already presupposes the structure it would produce. Intelligibility is left as nothing a divine being could supply.

The same dissolution handles the induction prong the presuppositionalist presses through Hume. That the future’s resemblance to the past cannot be justified without circularity is a real problem, but a divine guarantor does not solve it: the belief that a sustaining God will keep nature uniform rests on exactly the inductive habit under examination, so the guarantor relocates the circle rather than closing it. T, S, and Φ provide conditions under which ordered inquiry and stable comparison are possible, but they do not entail that the contingent laws or regularities of nature remain uniform. The Humean problem of induction therefore survives the bare structural floor; a divine guarantor may not solve it without circularity, but the triad alone does not solve it either. The direction of support then runs opposite to the one the argument needs: we hold nature uniform because we meet it as uniform, not because anything announces its uniformity to us, so the regularity counts as evidence that the structure is all there is rather than as a trace of a guarantor standing behind it.

The reliability rejoinder. Grant that intelligibility is a tracker converging on structure — what makes the tracking reliable, apt to latch onto what is there rather than to misfire? The demand repeats the category error one level in. Reliability is not a further substance the relation must be given; it is the success-condition that individuates something as tracking. Tracking admits degrees of reliability: a competence can correlate with a structure while misfiring in some conditions. Reliability therefore requires an account of error rates, calibration, and the processes that sustain successful representation; it is not guaranteed by the definition of tracking alone. At the structural floor, though, the demand thins: whatever competence tracks T, S, and Φ at all already operates within them, as a key that opens the lock is already cut to it. The fallible work of exposure and adjustment is just the process by which non-converging candidates fall away; what the word “competence” picks out is what survives. Nor does the evolutionary debunker’s wedge — selection tracks fitness, not truth — find purchase here, for the structure tracked is not a contingent regularity from which fitness could come apart but the precondition of there being distinct representational states to select among at all: any competence operating on distinct states already operates within T, S, and Φ. The wedge bites on contingent first-order content, never on the structural floor.

What converges. The section closes on two prongs. The first, developed through the three moves, is that the structure logic describes is necessary and no theistic strategy has shown it to require a rational author. The second is that intelligibility — the broadened target to which the argument retreats once the first prong holds — is a learned relation to that structure, not a substance any being must provide. The first removes the need for a ground; the second removes the need for a giver. Together they block a quick inference from logic or intelligibility to God. The comparative question remains whether a theistic, realist, conventionalist, or structural account best explains logical necessity, induction, and cognitive reliability without circularity.

A final consideration, offered as illustration rather than argument: the case that logic requires a rational ground is itself made in logic, and even the logician’s vocabulary is borrowed from the structure it evaluates. “Therefore” is there + fore, space and time fused; “necessarily” renders the Latin ne + cedere, not to withdraw, and withdrawal is a structured act: a succession of positions across a relational distance by something that persists. The etymology settles nothing; it only shows the structure surfacing in the very words built to describe it.

5.1.10 Moral Argument

The moral argument reasons from the reality of objective moral facts to God as their necessary ground. It has proved resilient precisely because it leans on a premise few want to deny (that some things are genuinely right or wrong, independent of what anyone believes or desires) and asks what could underwrite it. It must be engaged where it actually pushes, not as a restatement of the ontological argument §5.1.7 takes up: its force is axiological, not conceptual, and turns on the authority of obligation rather than the content of a concept.

The argument’s contemporary forms tie moral obligation specifically to a personal will. Adams¹⁵⁵ identifies moral goodness with resemblance to God and obligation with God’s commands; Evans¹⁵⁶ argues that obligations bear marks (authority, universal scope, inescapable bindingness) that only the commands of a personal, loving God adequately explain. Hare¹⁵⁷ presses the “moral gap” between the demand morality makes on us and our capacity to meet it, a gap divine assistance is needed to close. Craig¹⁵⁸ presses the argument in its most widely debated deductive form (if God does not exist, objective moral values and duties do not exist; objective moral values and duties do exist; therefore God exists), the popular syllogism that fixes the argument’s public profile, with all the philosophical weight resting on its first premise: that authority of the relevant kind requires a personal commander. Baggett and Walls¹⁵⁹ give the argument its strongest contemporary form, recasting it as an abduction rather than a deduction: theism is offered as the best explanation of the full moral landscape at once, its values, its obligations, its knowability, and the possibility of moral transformation, so that no single premise carries the inference and the cumulative fit is what is offered as evidence. Behind these stands Mavrodes’s¹⁶⁰ point that objective moral facts sit awkwardly, are “queer”, in a world with no more than nature in it; and even Mackie,¹⁶¹ no theist, allows that objective values, were there any, would fit more easily in a theistic universe than a naturalistic one. The shared structure is an inference to the best ground: moral facts are real and categorically binding; only a personal will supplies authority of that kind; therefore God.

Grant the realism in full. Moral facts are not projections, conventions, or sentiment got up as fact; that gratuitous cruelty is wrong is true, and true independently of any attitude taken toward it. Grant, too, that bindingness is the crux: an obligation that bound no one would not be an obligation, and a naturalism that reduces value to preference or fitness-tracking does struggle to recover the authority that genuine obligation carries. The argument’s animating worry is legitimate, and the present account shares it rather than dissolving it.

First. The ground the argument seeks is real but mislocated. What the argument adds to that worry is a personal will to supply the authority, and the dependence of morality on such a will is exactly what the book contests. The ground of moral bindingness is located at §8.12 in the structural constitution of persons and their standing toward one another, not in commands issued from outside that order; obligation is read off what persons are and how they stand to one another, not conferred by a will above them. If that identification holds, the argument’s central premise (that objective morality depends for its existence on God) is false, with the ground already supplied structurally and no further person required to issue it.

This front must be drawn at the register Chapter 8 actually licenses. The argument of §8.12 proceeds by structural identification, not strict derivation: it exhibits where moral bindingness is seated, but does not derive the full content of obligation the way Chapter 2 derives modal status. A determined moral realist can decline the identification (holding that authority of the relevant kind still requires a commanding will, however well morality maps onto the structure of persons) and so keep the dependence premise alive. Taken alone, this front shifts the burden; it does not discharge it.

Second. The commanding will the argument names cannot be had for free. The argument’s deeper trouble does not wait on Chapter 8. Grant, for the sake of the argument, that moral authority does require a personal ground: the ground the argument names still cannot supply it. The classical commanding will is typically described as atemporal, non-spatial, and immaterial. The revised §5.1.7 result does not make such a being semantically empty by definition. The pressure is whether a timeless divine nature or will can ground directed obligations to distinct persons while preserving determinate authority and content. Ordinary commands are temporal acts. Divine-command theorists may instead ground obligation in God’s eternal nature or standing will. That move avoids a temporal issuing event but still owes an account of the directed relation between commander, content, and obligated persons. Each is a condition of succession, relation, or persistence, and a ground precluded from time, space, and substance has not been shown able to supply any of them. Adams’s resembling goodness, Evans’s loving commander, Hare’s assisting God each presuppose a personal agent doing personal things: willing, relating, sustaining. Pressed to name the will that obligates, the moral argument points at a ground whose standing to command is exactly what remains unexplained.

A defender might relax the ground to a personal agent that does instantiate the conditions: temporal, related, present to those it binds. Then the proposal departs from strict classical timelessness. It may still claim ultimacy, but it must explain how an agent participating in temporal and relational moral life grounds rather than merely exemplifies moral authority. The revision keeps a commander by giving up the commander’s ultimacy.

The argument tracks something real. Moral facts are binding, and their bindingness is not self-standing: it is grounded. The mistake is the location: what grounds obligation is not a will behind the world but the structural constitution of persons within it (§8.12), the same inversion the book runs throughout, fixing on a genuine feature and mislocating its bearer. The doctrines this argument carries with it (divine-command theory as an account of where obligation is grounded, and Hare’s moral gap) are taken up in their own right in the Excursus, where the structural account is in hand to reconstruct what survives once the existence argument falls.

The abductive form needs separate handling, since it rests not on the single commanding act the second front dismantles but on a comparative claim: that theism explains the moral landscape better than any rival. The comparison fails at both ends. On the theist’s side, an explanation is only as good as the standing of its explanans, and the classical explanans enters the comparison owing the relation problem above: a ground precluded from time, space, and substance must account for the explanatory, sustaining, and commanding relations the abduction credits to it before its explanatory fit can be scored.

On the rival side, the inference is run against a strawman naturalism that reduces value to preference or fitness-tracking. The genuine rival is robust non-naturalist realism, worked out most fully by Wielenberg:¹⁶² objective moral facts as brute necessary truths standing in non-causal making relations to their base properties, requiring no commander and no reduction to preference. That account is the near-twin of the structural identification at §8.12, obligation seated in what persons are and how they stand to one another, and it blunts Mavrodes’s queerness charge: necessary moral facts are a serious rival to divine-command accounts, but their normativity is not explained merely by analogy with necessary structural conditions. The non-naturalist still owes an account of authoritative reasons and their relation to natural persons.

Against that rival the abductive inference does not clear its comparative step, for the same reason the modal-grounding argument does not clear its own (§5.1.8): a personal ground is named as the best explanation only because the field of rivals was drawn to exclude the structural one.

What is established. The moral argument is right that objective morality is real and binding, and right that its bindingness must be grounded. It is wrong about the ground. Where Chapter 8’s identification holds, obligation is seated in the structural constitution of persons rather than in a will above them (§8.12); and independently of that identification, the classical ground faces a substantive relation problem: its timeless nature or will must ground determinate obligations to distinct persons without relying on an unexplained command relation. What the argument sees (that morality is grounded) the structure supplies. What it added (that the ground is a divine will) names a being whose capacity to will and command is the very thing still owing an account.

5.1.11 Pascal’s Wager

Pascal’s Pensées¹⁶³ advances a prudential argument for theistic belief that has shaped the apologetic literature for three and a half centuries. The wager does not claim that the God of classical Christian theism exists, nor that the evidence weighs in favor of existence. It claims that, under uncertainty, belief is the rational bet: finite stakes are wagered against infinite expected return. Where the cosmological and ontological arguments engaged earlier in this cluster purport to establish a conclusion about what is the case, the wager purports only to settle a question about what to do under irresolvable uncertainty. Its appeal is that it appears to bypass the metaphysical disputes the other arguments inherit. The appearance is misleading: the wager does not bypass the metaphysics; it presupposes it, and what the presupposition invites and costs must be assessed on the wager’s own decision-theoretic terms.

Contemporary reconstructions sharpen the wager’s defended form. Jordan¹⁶⁴ develops the canonical analytic treatment, treating the wager as a many-gods-robust decision under genuine option in the William James¹⁶⁵ sense, and shows that the wager survives several classical objections when properly reconstructed: the principle of insufficient reason, the mixed-strategies dilemma developed by Hájek,¹⁶⁶ and the symmetric-payoff parodies. Lycan and Schlesinger¹⁶⁷ press the dominance argument: granted any non-zero credence in the classical Christian God, no available strategy dominates belief. The load-bearing premise across these reconstructions is the same: P(God exists) > 0, where the God in question is the necessary-being God of classical Christian theism whose existence underwrites the infinite expected payoff.

The modal smuggle. The wager’s prudential framing invites a modal reading it cannot pay for. Non-zero objective chance would require a genuine possibility, but Pascal’s wager ordinarily uses epistemic credence. A rational agent may assign non-zero credence to a proposition that is in fact impossible because its modal status is unknown. Thus P(God exists) > 0 does not by itself entail metaphysical ◇(God exists). The wager therefore need not assert a metaphysical possibility premise; it needs a defensible epistemic credence. The God on whose existence the wager’s infinite payoff depends (the God of classical Christian theism, omniscient and omnipotent and morally perfect across worlds) is a necessary being on the tradition’s own commitment (Anselm, Proslogion II; Aquinas, Summa Theologiae I.2.3; Plantinga¹⁶⁸). The classical conception forecloses a contingent reading: God exists is therefore equivalent to □(God exists), with no contingent witness available. If metaphysical possibility of a necessarily existing God were independently established, S5 would yield necessary existence. The wager’s epistemic credence does not establish that modal premise. The argument and its S5 defenses are carried at §2.1.5, and the parallel line through the modal ontological argument’s explicit possibility premise is closed at §5.1.7. The wager’s probability premise does not entail its metaphysical conclusion; it supports only a decision under uncertainty.

The wager resembles Plantinga’s argument only if epistemic probability is conflated with metaphysical possibility. Kept distinct, the wager is a decision-theoretic argument under uncertainty, not a disguised S5 proof. Plantinga explicitly uses a metaphysical possibility premise and S5. Pascal uses credence and expected utility. Conflating the two would be an error; the wager must be assessed on decision-theoretic rather than modal-transfer grounds.

The epistemic reading. A defender might press that the wager’s probability is epistemic, not metaphysical: Pascal is gambling under ignorance, not measuring credence over modal alternatives. This is the wager’s standard interpretation, not a retreat. Decision theory routinely operates with epistemic credences under uncertainty. The harder problems are different: assigning coherent probabilities across rival religions and payoff schemes, handling infinite utilities, accounting for mixed strategies and finite costs, and determining whether belief is voluntarily controllable. The epistemic reading therefore preserves the wager while exposing it to the standard decision-theoretic objections.

A further pressure. The classical-Christian-God specification is not optional for the wager: it is what underwrites the infinite payoff. The payoff structure requires that the wagered-for being be such that its existence settles eternity in the manner the wager presupposes: omnipotent enough to deliver infinite reward, omniscient enough to recognize the wagered belief, morally engaged enough to honor it. A merely contingent god, or a god whose attributes fall short of the classical specification, does not deliver the payoff structure the wager needs. The many-gods objection presses this from one side: too many candidates fit the prudential frame, and the calculation cannot select between them. The present objection presses it from the other side: only the classical necessary-being God delivers the payoff at all.

What closes. The wager does not establish God’s existence and should not be converted into an S5 derivation. As prudential reasoning, it depends on contestable epistemic credences, utility assignments, and a treatment of the many-gods problem. The classical-Christian specification supplies one payoff model but does not receive probability merely from being describable. The wager can remain a pragmatic recommendation for some agents under some credal states; it does not settle the rival-ultimate question. A second strand of the wager (Pascal’s counsel that one may cultivate belief by acting as if, until assent follows) meets the account of the will at §8.9, which shows the will cannot install a conviction its field withholds; it is taken up where that result bears on inculpable nonbelief, in the divine-hiddenness entry (E.13). The strong claim that belief is uniquely rational does not survive without further decision-theoretic and theological premises.

5.2 Doctrines of the Divine Nature

This cluster gathers the doctrines of the divine nature: the constructions built to keep the divine ground ultimate without regress, simple without composition, and independent of the structure it grounds. Divine simplicity (§5.2.1) holds the ground free of all composition; the via negativa (§5.2.2) secures it by denial, saying only what it is not.

The question is the one the chapter presses throughout: can the doctrine hold its ground apart from structure without relying on structure to do so? The finding is again relocation rather than refutation. Each doctrine, in securing the ground’s independence, reintroduces the very structural conditions it was meant to set the ground above.

5.2.1 Divine Simplicity

Divine simplicity is the doctrine that God has no parts, not only no physical parts, but no metaphysical composition of any kind: no distinction between God and God’s attributes, none between God’s essence and God’s existence. Where the other classical counter-moves contest how structure applies to God, simplicity denies that it applies at all.

It is not austerity for its own sake. It is meant to secure divine ultimacy by ruling out composition, dependence, and metaphysical parts in God: were God assembled from features more basic than itself, aseity would already be lost, for the more basic features would ground God rather than vice versa. The doctrine has two tightly linked expressions, simplicity proper and actus purus (pure act without potency), which classical theism treats as mutually entailing. Strict Thomistic simplicity takes both to their limit: Deus est suum esse, God’s essence just is God’s existence, without distinction.

The doctrine has a long history and a contested one. Augustine (De Trinitate VI) articulates an early form: God’s attributes are identical to God’s being. Maimonides (Guide of the Perplexed I.50–60) presses the most austere version: the divine attributes are equivocal with creaturely attributes, and positive predication of God is strictly impossible. Avicenna (Metaphysics VIII) develops the essence-existence identity later inherited by Aquinas. Aquinas (ST I.3–I.4) gives both doctrines full rigor and argues their mutual entailment. The doctrine was contested even within its home tradition: Duns Scotus¹⁶⁹ introduces the formal distinction as a middle ground, arguing that God’s attributes are formally distinct though really identical, an internal concession that the strict reading generates problems.

Contemporary defenders: Dolezal’s God Without Parts¹⁷⁰ defends strict simplicity; Brower¹⁷¹ and Stump’s Aquinas¹⁷² refine it through truthmaker and phased-identity variants; Feser’s Aquinas¹⁷³ defends the act/potency distinction as the metaphysical basis of simplicity; Kerr’s Aquinas’s Way to God¹⁷⁴ reconstructs esse as the central Thomistic category; Pruss¹⁷⁵ supplies contemporary PSR groundwork; Leftow’s God and Necessity¹⁷⁶ develops an alternative modal-theistic framework preserving key simplicity commitments; Rogers’s Perfect Being Theology¹⁷⁷ defends a contemporary Anselmian simplicity.

Against the doctrine: Plantinga’s Does God Have a Nature?¹⁷⁸ presses the charge that strict simplicity either identifies God with a property (ontologically incoherent) or abandons the doctrine; Hughes’s On a Complex Theory of a Simple God¹⁷⁹ develops the analytic critique at length; Mann’s “Divine Simplicity”¹⁸⁰ is sympathetic but catalogs the unresolved problems; Mullins’s The End of the Timeless God¹⁸¹ presses the combination of simplicity and timelessness as internally inconsistent; Wolterstorff’s “Divine Simplicity”¹⁸² rejects the doctrine in favor of a richer divine psychology.

The structural account sits closest to the Plantinga–Mullins wing in finding a deep tension between strict simplicity and active divine predicates. Its narrower claim is not that every simplicity doctrine is immediately contradictory, but that any account on which a simple God grounds, donates, sustains, knows, or wills must explain how those relations acquire determinate content without the structural roles the account identifies.

Against the structural account, simplicity is the counter-move with the longest pedigree. The canonical line is run end-to-end at §2.1, which defends Necessary Possibility and S5 invariance; Cartesian Certainty for change’s bilateral witnesses is supplied at §2.2; the necessary modal constraint, tying the possibility of change to T, S, and Φ, is established at §2.4; that constraint is lifted to the unconditional Triconditional Necessity of the three Conditions at §2.5; and those Conditions are identified with Nature through the Triconditional machinery at §2.6. Simplicity, in its strict form, denies the application of that whole structure to God.

The classical theist answers that a maximally ultimate being is precisely one that does not instantiate the structural roles: a being with no composition cannot be assembled from parts the way structured entities are. One coherent being lacking relational non-identity, causal persistence, and temporal succession is all the position requires.

Pressure turns on whether that one being is coherently specifiable. Nine questions take up the objection: one at the level of the commitment, eight across its forms. They test three connected matters: whether the divine subject can be distinctly specified, whether it can stand in a real relation to creation, and whether its predicates retain differentiated content. The replies test whether any form avoids dissolving into ∅ or reinstating the roles under a different name.

First. The commitment question asks what simplicity delivers beyond its negative specification. Simplicity names the absence of composition without naming the activity the doctrine performs with it. Negatively specified, the doctrine rules out parts, composition, potency, and any distinction between essence and existence. A position that did nothing beyond negation would be inert. Simplicity earns its place because the theist uses it to affirm: that God grounds creation, donates existence, sustains the created order, and is fully actualized without potency. Each operative verb is a positive predicate. Ground requires an asymmetric dependence relation and a distinction between grounder and grounded; donate requires directed source–recipient content, though not necessarily a temporal transition in the donor; sustain requires an enduring or counterfactually stable dependence relation; fully actualized must retain content beyond bare self-identity. These are thin role-level demands, not yet an identification with loaded physical Time, Space, and Substance. The doctrine is specified by negation and cashed out by positive structural operations. The remaining eight questions test whether any form of the doctrine can accommodate the affirmations the doctrine is used to make.

Second. The relational-extension form tests whether a being with no spatial extension can be a distinct existent. Strict Thomistic simplicity holds that God has no parts, no composition, no distinction between essence and existence. God is not assembled from features; God is God’s features, wholly and undividedly. Structural categories (distinction, succession, persistence) apply to creatures, not to the ground of creatures.

The absence of spatial extension does not by itself prevent numerical distinctness: abstract objects, if admitted, supply an obvious countermodel. The sharper question is whether the simple being has enough determinate identity and relational content to be distinguished from creation and to ground it without composition. Pressed to this limit, the worry is whether the doctrine leaves in place the structural conditions under which a distinct existent can be picked out at all.

Even granting existence as a hypothesis, a simple being collapses the architecture required for causal power. For any being to be a causal agent (to ground, sustain, or relate to creation), it must be a distinct relatum capable of entering into relationships. Interaction requires a between. Simplicity removes the between. The result is a Point-State — a would-be existent contracted to a dimensionless point, with no extension across which anything could be held apart from anything else: no relational manifold to host distinct coordinates, no gap between divine will and divine act, no distinction between divine subject and object of knowledge. Without Space in its distinction-grounding role, no distinct relata; without distinct relata, Substance’s interaction has nothing to connect. The advocate who strips all distinction performs one in the act of stripping: individuation requires exactly what absolute simplicity removes.

The classical theist has one reply left here: God’s distinctness from creation is transcendent, not positional; numerical, not at a distance; God distinct as Being Itself rather than as one relatum standing off from another. Grant the reply for distinctness taken statically; it still cannot be purchased for the verbs the doctrine refuses to drop. To ground, donate, or sustain is more than merely to differ from the recipient: it requires determinate asymmetric dependence. Metaphysical grounding need not involve motion or transit across a spatial interval. The burden is therefore to explain the direction and content of the relation without introducing composition, not to supply a literal crossing. The move that rescues the static distinctness forfeits the active verbs, and those verbs were the doctrine’s entire reason for asserting the distinctness to begin with.

Third. Brower’s truthmaker simplicity and Stump’s phased-identity simplicity refine non-compositional distinctness. Simplicity does not require God’s properties be strictly identical in the strong Thomistic sense; it requires only that no composition grounds God. The properties may be distinct in some sense but not compositional; what is ruled out is metaphysical dependence, not distinctness as such.

The refinement concedes the structural point while relabeling the relation. If God’s properties are distinct in some sense but non-composite in another, then some relational structure holds among them: a structure that distinguishes the properties while unifying them. That is exactly what relational non-identity within a unified manifold consists in. Space at the role-definition level does not require a unified whole be dissolved into its parts; it requires parts be distinct within the whole. Brower’s truthmaker simplicity and Stump’s phased-identity simplicity both concede distinct relata within a unified divine being, a description of a structurally unified substance with distinct aspects. The structural conditions the argument names are retained in the refinement; what changes is the vocabulary under which they are described.

Fourth. Dolezal’s apophatic form pairs strict simplicity with the claim that creaturely concepts cannot fully articulate what it amounts to, and holds that structural categories do not apply to the uncreated ground. The categories of our thought (including the structural roles the argument invokes) apply to creatures, not to the uncreated ground. Simplicity is a truth affirmed of God even though the categories we use to describe God fall short; the structural objection applies a creaturely measure to a non-creaturely reality.

Apophatic simplicity makes the position immune to objection only at the cost of making it immune to assertion. If the structural categories do not apply to God, neither do the categories of existence, power, will, or ground, each of which the theist affirms. The apophatic move is selectively applied: negation is invoked against the argument’s categories but suspended for the categories the tradition wants to preserve. Either the negation is global, in which case the doctrine says nothing about God at all, or it is selective, in which case its selectivity requires a justification apophaticism was meant to avoid. Selective apophaticism does work the doctrine needs done while exempting itself from the obligation to justify where the negation stops. The structural analysis does not apply a creaturely measure. It applies the measure any affirmation of existence requires.

Fifth. The actus purus form presses simplicity through its metaphysical architecture. God is pure act: the complete actualization of all perfections with no unrealized potential. Whatever has potency is partly not actual; whatever is pure act has no potency. God is therefore pure act and the unconditioned ground of everything partly actual.

Pure actuality generates its difficulty more specifically than the crude charge that “no potency” means “no power.” The Thomist is right to reject that formulation: potency in the act/potency scheme is not active power, and pure act is meant to be fully actual causal power rather than inertness. The problem is not active power as such. The problem is the transition from not-exercising to exercising, from no creaturely effect to creaturely effect, from source alone to source-related-to-recipient. If the effect is contingent, the exercise of power must be related to that effect in some determinate way; if the exercise is eternal and unchanged, the effect seems eternal with it; if the effect is not eternal, some relational asymmetry has entered the account.

Feser’s refinement distinguishes pure act from static immobility: pure act is not absence of causal capacity but fully realized causal capacity, a being whose power is entirely actualized. The refinement sharpens the structural question rather than removing it. An act that genuinely produces, sustains, or donates being is an exercise of causal capacity. It requires determinate source–effect difference, asymmetric dependence, and a bearer of the causal role. Those thin requirements do not by themselves establish spatial extension, temporal succession, or physical substance; the stronger identification needs the independent arguments supplied elsewhere. Sustains is itself a maintenance relation: to sustain is to hold something in being across the relevant interval or dependence-structure. An atemporal pure act can be asserted; what remains missing is how it produces, sustains, or donates without the structured relation those predicates require.

What the Way–Possibility Convertibility (§3.2) makes inescapable here is the way-side of the predicate. Pure act forecloses potency, and the Thomist can refuse the potency framing without contradiction: the act/potency scheme is internal Thomistic vocabulary, and actus purus declines the predicate on the Thomist’s own terms.

The way-predicate is not declinable. The Thomist’s own characterization of God as ipsum esse subsistens (subsistent being itself, a determinate manner-of-being rather than dependent or composed) is itself a way of being. To say God’s mode of being is subsistent rather than dependent, simple rather than composed, necessary rather than contingent is to specify a way under each predicate (direction, difference, link). By the Convertibility, whatever bears a way bears the structural conditions a way decomposes into. The exit actus purus offers against potency is not available against way: the Thomist commits to God’s way of being as part of the doctrine’s positive content, and cannot then deny the structural conditions that way-talk presupposes.

Two Thomist replies deserve direct hearing. Dolezal’s God Without Parts¹⁸³ apophatic move holds that way is creaturely vocabulary, predicated of God only analogically and not univocally, so the way-predicate does not carry its structural conditions across the analogy. The Fourth question above answers this in the general case: selective apophaticism cannot disqualify way while retaining subsistent, simple, actus purus, and ipsum esse as positive content. Either all the positive predicates the doctrine asserts of God fall under the apophatic disqualification (in which case the doctrine asserts nothing), or none of them do, in which case way stands with the rest. Brower’s¹⁸⁴ and Stump’s Aquinas¹⁸⁵ truthmaker-identity reply holds that one simple truthmaker can ground distinct predicates (subsistent rather than dependent, simple rather than composed) without those predicates introducing real distinctions in the truthmaker. The general form of that reply is met at the Seventh question below; the point that bears here is way-specific: as soon as the truthmaker is the condition under which way of being is correctly predicated, the truthmaker bears a way under the Convertibility, regardless of whether the multiplicity of senses is located on the creaturely-concepts side.

The same wedge applies to creation specifically. Classical theism reads creation as either a way from God (efficient donation outward, source-to-recipient) or a way through God (sustaining medium across which created being is held). Both readings require God to occupy a way-role: the former requires source-direction (T) and source/recipient distinction (S); the latter requires medium-as-carrier (Φ). The Convertibility forecloses both, because neither way-from nor way-through can be predicated of a being held outside the structural conditions ways decompose into. The doctrine cannot drop both readings, because dropping both abandons creation as a relation between God and creatures, which the doctrine is constructed to maintain.

The underlying critique (that classical simplicity is in tension with creation as a real relation) is well-trodden ground: Mullins’s The End of the Timeless God¹⁸⁶ presses it as the modal-collapse argument; Schmid’s “From Modal Collapse to Providential Collapse”¹⁸⁷ sharpens the tension for atemporalist simplicity by showing that avoiding modal collapse forces providential collapse; Wolterstorff¹⁸⁸ rejects simplicity partly on the ground that it cannot accommodate God’s responsive relation to creatures. The standard Thomist reply invokes the doctrine of mixed relations (real on the creature side, conceptual on the divine side, ST I.13.7) together with simultaneous causation (treated at §5.1.3); the atemporalist’s further resource is the Stump–Kretzmann eternal-present account,¹⁸⁹ whose ET-simultaneity is offered to dissolve the modal-collapse and timelessness tensions by denying the temporal sequence they presuppose; its eternal-present mechanism is handled at §4.2.1, where the Boethian eternal present and ET-simultaneity are met directly.

The way-foreclosure framing isolates what these replies must still concede: even a relation merely conceptual quoad nos requires some way-role on the divine side to terminate the creaturely relation, and a pure terminus with no way-role at all is ∅ articulated in classical vocabulary rather than an alternative to it.

Feser’s Aquinas¹⁹⁰ and subsequent work extends the actus purus defense through the Thomistic doctrine of the analogy of being (analogia entis): predicates like “ground,” “donate,” and “sustain” apply to God not univocally (importing creaturely conditions) nor equivocally (as mere homonyms) but analogically: in a mode proportioned to the divine being, tracking a real divine attribute without importing the full creaturely profile. This is a refined attempt to hold the positive predicates and the simplicity negations simultaneously: the analogy licenses the affirmations while the simplicity doctrine strips the structural conditions those affirmations normally presuppose.

The structural reply addresses the analogical mode directly. Analogical predication requires a real proportion grounded in something real about the divine subject: the predicate must track a real divine correlate that stands in the appropriate proportion to its creaturely counterpart. A feature that distinguishes God-sustaining from God-not-sustaining is a relational position, Space in its distinction-grounding role. To be the source rather than the recipient of sustaining is to occupy a directed role, Time in its ordering sense. To be the medium through which sustaining propagates is Substance’s persistence role. The analogical mode does not eliminate these structural conditions; it reasserts them under a mode of predication that is itself a structural fact about the proportion between divine and creaturely attributes. The doctrine of analogy, taken seriously, presupposes the structural conditions it is designed to moderate. The same cluster recurs at §5.1.3 (Feser’s act/potency machinery under direct scrutiny) and §5.2.2 (the analogical-apophatic complementarity); the structural question at each is whether an analogical-relational claim can be grounded without the structural conditions any relational claim presupposes.

The eminent-containment retreat. One form of the analogical reply is not yet exhausted. Pressed past predication to its ground, the classical theist may hold that the perfections are present in God not as ways God is (not as direction, difference, or link) but eminently: as the positive plenitude in which every perfection pre-exists in a higher mode than its creaturely realization (Aquinas, ST I.4.2). On this retreat God is neither a being among beings exhibiting the structural roles nor bare being, but positive fullness containing what creatures have in structured form, without the structure. The retreat concedes that way-talk carries the conditions the Convertibility names, and for that reason declines to predicate way of God at all: what is affirmed is content, not a way of being.

The Convertibility reaches the retreat through what positive content is. Content that is genuinely positive is determinate: it is thus and not otherwise, the fullness of being rather than its privation or limit. Determinacy is contrastive: to be the fullness rather than the privation is to exclude the privation, and a content that excluded nothing (compatible with every state, including its own absence) would not be content but an idle word (omnis determinatio est negatio, in its minimal logical form). To exclude is to stand in relational non-identity to what is excluded, and relational non-identity is Space in its distinction-grounding role, the same feature the analogical reply already conceded must distinguish God-sustaining from God-not-sustaining.

A defender may resist the step from determinacy to Space by pointing to abstracta: the number seven is fully determinate (seven and not eight) yet stands in no spatial relation, so determinacy alone seems not to force relational non-identity. The example is granted (the status of abstracta is treated in its own right at §4.2.4 and conceded at §5.1.7), and it does not rescue the retreat. The contrastive determinacy of an abstract object describes what the structural conditions afford without grounding, donating, or sustaining anything; it is determinate the way a value is determinate, not the way an agent is. The eminent content the retreat needs is not inert in that way: it is the positive plenitude from which creation is grounded and sustained, and a determinacy that does grounding work distinguishes the source from what it grounds in the very act of grounding it. That distinction is relational non-identity in its concrete, distinction-grounding role, Space. The dilemma therefore re-presents one level up: either the eminent content is abstract-like, in which case it grounds nothing and the doctrine forfeits the creation, donation, and sustenance the retreat was invoked to preserve, or it does the grounding work, in which case its determinacy is the concrete contrastive determinacy that reinstates Space.

Eminent content is therefore not a tertium quid beside way-of-being and bare being. Either it is determinate, in which case it is contrastive and reinstates the structural condition under the name eminence; or it excludes nothing, in which case it is content-free and is ∅ after all. The tertium quid the retreat needs would have to be positive content that is neither contrastive nor empty, and there is no such option to occupy.

Before that premise is named, its reach is worth marking, because it is far narrower than the result it secures: it is load-bearing only here, against the eminent-containment retreat. Every form of the doctrine that keeps the active verbs (ground, donate, sustain) was already closed by the way-foreclosure, which needs no truthmaker premise at all; only the retreat that drops the verbs for pure content reaches this far, and only it forces the question of what then distinguishes that content from ∅. The closure rests on one premise worth making explicit: that what bears no feature discriminating it from ∅ is ∅, a discernibility principle for being. The classical theist’s only coherent refusal is to hold that divine positive content is really distinct from the Null State while bearing no discernible differentiator from it, a primitive positivity distinct from nothing by no feature at all. That refusal is not free here, because §2.7 precludes the Null State on exactly this ground: a putative something with no feature to distinguish it from nothing is nothing described as if it were something. The eminent-containment retreat and the Null State stand or fall together; the doctrine cannot reject the principle for ∅ and keep it for God.

The principle is not a stipulation the argument helps itself to; it is the truthmaker demand the doctrine already honors. God is not nothing is either true in virtue of something on the divine side (a fundamentum in re, some feature answering to the distinction), or it is a distinction drawn in our concepts alone, with nothing in God answering to it. The seventh question below presses the first horn for the attribute-predicates: good and omnipotent are true of God only if they track something real about him. The existence-predicate is not exempt. If a fundamentum distinguishes God-existing from the Null State, that feature is a discriminating position (relational non-identity, Space in its distinction-grounding role), and the structural condition stands reinstated under the name fundamentum. If nothing answers to it, then nothing on the divine side makes God is not nothing true rather than false, and a distinction with no foundation in re is exactly the empty, merely-verbal distinction the tradition refuses when it insists the divine names are not synonyms.

So the primitive-positivity refusal is an exemption, and a local one: it asks that a single predication (God is, God is not nothing) be released from the truthmaker demand the doctrine accepts for everything else it says of God, and the release falls exactly where the doctrine needs it.

The exemption is neither free nor Thomist. A reader who rejects truthmaking outright can decline the demand; the classical theist, having already grounded the divine names in what is real about God, cannot withhold it for the one predication on which the simplicity of God and the reality of the Null State are made to part.

What stays genuinely open is one premise, and it is worth scoping exactly, because it is narrower than full truthmaker commitment. The closure needs no truthmakers for all truths, and none for negative or general truths, the contested precincts of truthmaker theory. It needs only that a single positive existential (God is, God is not nothing) be true in virtue of something on the divine side: that there be a difference in re between God-existing and the Null State.

Nor can the Thomist slip the demand by recasting this as a negative truth: God is not nothing read as the bare denial of a competitor, exempt the way negative and general truths are exempt. That recasting is foreclosed by the doctrine’s own center: Deus est suum esse, God’s essence is God’s existence. On that thesis God is is made true by the divine esse itself, the most positive fundamentum the tradition recognizes, the act in virtue of which anything whatever exists; it is the one existential the doctrine can least afford to make costless. A theist who downgrades God is to an exempt negation has not reached for the negative-truth exemption but denied that God’s essence is esse, which is to abandon simplicity at its root. That is the least controversial corner of the doctrine, and it is one the classical theist already occupies. Aseity is the claim that God’s existence is grounded in God rather than in another; a theist who denies that God is has an ontological ground has not escaped the argument but abandoned aseity, the classical project’s load-bearing commitment. So the closure holds as firmly as aseity, and no more firmly: which is as firmly as the doctrine needs, since aseity is not negotiable for the position under examination.

Sixth. The ipsum esse subsistens form holds that being itself precedes structural characterization. Ipsum esse subsistens (the sheer act of existing, prior to any structural characterization) is the version of the doctrine that goes deepest. God is not a being among beings but Being itself, the unconditioned ground in virtue of which beings are. T, S, and Φ are features of creatures, of beings that have existence. The source of that existence is not itself a being with structural features; it is the pure act of existing, which precedes every structural distinction. The stratification proposal makes this precise: created being exists within T, S, and Φ; uncreated being stands at a prior level; the two levels are related by the act of esse; God donates existence to creatures, and the donation is neither a structural event (which would place God in the structure) nor a non-event (which would leave the donation inexplicable). It is a transcendent act that stratifies being itself.

The stratification does not escape the structural conditions. It relocates them to the interface. For God to donate existence to creatures, there must be a relation between donor and recipient, and a relation is a structural fact. (The simultaneous-causation mechanism classical theism offers to explain this relation is §5.1.3.) Donation is a causal relation connecting source to recipient across an ontological transition: distinction between donor and recipient (S), direction from source to product (T), structural persistence across which the donation runs (Φ). The Thomist who claims God’s being is prior to structure must invoke structure to explain how that being relates to anything, which is to invoke structure to explain how what is supposedly prior to it grounds anything at all. Read charitably, the Thomist’s deepest insight (that being is not a property added to an essence but the fundamental act making anything real) is precisely what the structural picture affirms. The disagreement is about whether that act is identical to the structural conditions or prior to them. The above establishes that the “prior to” reading is either vacuous or covertly structural.

Seventh. The multiple-descriptions form tests whether a strictly simple being can ground distinct true predicates. Strict simplicity holds that God has no real distinctions: God’s goodness, power, and knowledge are not separate properties but identical with God’s simple essence. The multiplicity is in our creaturely concepts, not in God.

This generates a dilemma the doctrine cannot discharge. If “God is good” and “God is omnipotent” describe genuinely distinct divine attributes, then the attributes are distinct, and distinction is composition. If they are not descriptions of genuinely distinct attributes (if both sentences describe the same simple fact), then they are informationally equivalent: swapping them out changes nothing, and the divine names the tradition treats as distinct turn out to be synonyms. The classical response (the multiplicity is in our concepts, not in God) does not close the gap. Even if the multiplicity is on the concept side, those concepts must be tracking something real about God to be true of him. “God is good” is true in virtue of something. If that something is numerically identical to what makes “God is omnipotent” true, the truthmaker-identity defender¹⁹¹ replies that one simple truthmaker can ground distinct predicates the way one Venus grounds both “morning star” and “evening star”: same referent, distinct senses. The reply preserves distinct content only as long as the senses do no metaphysical work. As soon as “good” and “omnipotent” require even minimally different aspects of the divine essence to be the conditions under which the predicates apply correctly rather than incorrectly (as they must, on pain of becoming synonyms), those aspects are real features the predicates discriminate among. Discrimination among real features within one being is what relational non-identity within a unified manifold consists in.

The dilemma re-presents at the level of aspects: either the aspects collapse and the predicates are synonyms whose multiplicity is unprincipled, or the aspects are real and strict simplicity has conceded the structural condition under the name “aspect.” The doctrine of analogy (§5.2.2) addresses the linguistic question of how predicates apply; this question precedes it: what the predicates must track if they are to have multiple applications at all. A being of which only one thing can be truly said (because all descriptions collapse into one simple identical fact) is not a being the theological tradition has described. It is a being that exhausts itself in a single predicate the tradition has not been able to name.

Eighth. The softenings (Leftow, Scotus) preserve assertability by conceding enough structure to make the doctrine statable. A kindred softening is Leftow’s¹⁹² divine-modality framework, which preserves key simplicity commitments while allowing some distinctions within God: specifically, between God’s nature and God’s will. The softer version does the work kindred deflations always do: it concedes enough structure within the divine being to make the being assertable, and the conceded structure is what the argument identifies as T, S, and Φ at the appropriate level. The line is independent of the apophatic reading’s fate: grounding modality in divine nature and will places necessary-possibility-of-change inside God, which Modal Identity (§2.6.6) identifies with T, S, and Φ. Scotus’s formal distinction, anticipating the contemporary softenings by seven centuries, makes the concession inside the Catholic tradition itself. Open-theist softenings (Hasker, Sanders) concede still more (temporal duration, genuine response, real relation to creatures) and concede the structural commitments most explicitly. The softenings rescue assertability at the cost of simplicity: the doctrine that remains is a structural doctrine with the word simplicity still on the label.

Ninth. The essence/energies form is the Eastern counterpart to the softenings, and it concedes the structural point more openly than any other while most insisting it has not. Gregory Palamas (Triads III.2), systematizing the hesychast tradition, distinguishes within God the imparticipable essence (ousia, utterly unknowable, beyond participation) from the uncreated energies (energeiai), the divine operations in which creatures genuinely participate: the deifying grace and the light of Tabor through which theosis occurs. The distinction is held to be real, in re and not merely in our concepts, yet to introduce no composition: the energies are God wholly in act, not parts of God, so simplicity is claimed to stand. Lossky’s The Mystical Theology of the Eastern Church¹⁹³ and Bradshaw’s Aristotle East and West¹⁹⁴ develop the position as the one account securing both divine transcendence (the essence withdraws absolutely) and real communion (the energies are genuinely God and genuinely shared), the pair strict Western simplicity is pressed on at the Fourth and Fifth questions. Taken at its strongest, it is the most direct rival to the structural account in the neighborhood, because alone among the forms it does not deny distinction in God: it affirms one and asks to keep simplicity anyway.

That affirmation is where the form supplies the structural conditions rather than evading them. The distinction between essence and energies is either real or not. If it is merely nominal (a distinction in our mode of access, essence-as-hidden against energies-as-revealed), it collapses into the strict simplicity of the Third and Fourth questions and inherits their verdict, at a cost Palamas cannot pay: theosis becomes participation in the bare essence, which the doctrine exists to deny. If the distinction is real, then God internally carries the inventory the argument names: a relational non-identity between essence and energies (S), a directed procession from essence as source to the energies as its uncreated activity (T in its ordering sense, directed origin without clock-time), and a participation across which the energies are conveyed to creatures (Φ, the carrier-role of an interval something crosses). A real distinction held in God, with a directed origin and a participable activity, is structure under the name energies.

The defender answers that this is a distinction sui generis: neither real as between two things nor merely logical, but the Palamite middle. That middle is the formal distinction of the Eighth question pressed one tradition over: Scotus’s formal distinction and the essence/energies distinction are the same concession made in two tongues, and the structural conditions never required separability; a real but inseparable distinction, with a directed order and a carried activity, is the paradigm of an articulated structure rather than an exemption from one.

The plurality of the energies presses the point a second time, more bluntly. The energies are many (wisdom, power, foreknowledge, providence, deifying grace) and each is held to be uncreated. Either they depend asymmetrically on the essence, which is a directed dependence-order within the divine (T) and an internal articulation simplicity was meant to exclude, or they are co-primitive, in which case the uncreated is an irreducible manifold (S) and the unity that secured aseity is already surrendered. The doctrine multiplies the uncreated; the argument has only to register the multiplication. Read at its most careful, the essence/energies distinction does not bypass the structural conditions classical simplicity strips from divine being: it is the form that supplies them itself, uncreated, and names the supply energies.

A being denied every determinate distinction, dependence direction, and stable bearing role could not support simplicity’s positive predicates. Mere lack of spatial extension or temporal succession, however, does not establish that stronger absence.

What converges. The nine questions survey the principal defenses considered here, and each returns to a common burden: the doctrine is specified by negation and cashed out by positive structural operations, so that strict identity secures at most a static distinctness that leaves the doctrine’s active verbs without relational content; non-compositional distinctness retains the structural conditions under a different label; selective apophaticism does work it exempts itself from justifying; pure act still owes an account of the transition from source alone to source-related-to-recipient; ipsum esse subsistens relocates structure to the donation interface; multiple true descriptions of a strictly simple being collapse into informational equivalence or reinstate real distinction; the softenings concede the structural commitments explicitly; and the essence/energies distinction either collapses into the strict simplicity it meant to refine or, taken as real, supplies the structural conditions itself under the name energies.

All the ways necessary for possibility are held in place by the structural conditions classical theism strips from divine being. The doctrine asks the reader to grant every condition possibility requires while denying every one of them to the being possibility was supposed to terminate in.

What the doctrine calls simplicity may preserve a minimal structural floor whenever it retains determinate active predicates or real relations to creation. A defender may instead posit an essence-prior or apophatic ground; that move is not answered by merely repeating the floor, but owes an account of how the proposed ground and its relation to creatures are determinately specified. The chapter’s conclusion is therefore a burden: strict simplicity does not by itself deliver a structure-free ground, and its strongest versions must discharge the relation and predication problems developed here.

Sibling theistic counter-constructions: Rasmussen’s Limits-plus-Isomorphism derivation (§5.1.1), Swinburne’s probabilistic simplicity (§5.1.4), and the PSR-demand variant (§4.1.3). The downstream severance this defeat produces for classical predication (how the attributes come apart once simplicity no longer holds them together as one act) is §6.2.

5.2.2 Via Negativa

The via negativa (the “negative way” of approaching God by denying predicates) is among the oldest theistic responses to the structural pressure §3.5 makes general. Pseudo-Dionysian unknowing (Mystical Theology 1; Divine Names 1), Maimonidean predicate-denial (Guide of the Perplexed I.50–60),¹⁹⁵ and the later apophatic tradition (Eckhart, Sermons)¹⁹⁶ all approach God by stripping predicates away until what remains is supposed to exceed every structural category. Contemporary reconstructions¹⁹⁷ argue that apophatic discourse does not merely deny predicates but performs a linguistic self-subversion that transcends the predicate/denial framework altogether.

The general point has already been pressed at §3.5: the act of negating the structural conditions instantiates them in the very performance of the negation. The apophatic application brings the abstract pressure to ground. An apophatic move that subtracts time from divine being uses time to make the move, distinguishes what is subtracted from what remains, and persists as a negating act. The successive negation of T, S, Φ (Pseudo-Dionysius’s ladder of unknowing, Maimonides’s series of attribute-denials, Eckhart’s abyss beyond God) does not reach a standpoint outside structure; it traces structure’s contour from the inside, marking each condition by the act that tries to remove it.

The contemporary self-subversion reading marks a necessary distinction. Sells’s Mystical Languages of Unsaying¹⁹⁸ and Turner’s Darkness of God¹⁹⁹ argue that the apophatic move is not a simple negative proposition but a recursive undoing of its own enunciation. Granted in full: the practice of self-subversion is an activity unfolding through succession, distinction, and a continuing practitioner. That establishes an act-side structural fact. It does not by itself establish that the intended referent of apophatic discourse is structured, because the apophatic theorist may deny that the discourse represents a determinate object at all.

A parallel front: the immutable-thinker dilemma. The apophatic discipline pressures classical theism from one direction: every act of negation instantiates what it negates. A parallel pressure runs along the divine attribute of immutability itself. The classical-theistic God is, on the standard view (Aquinas, ST I.9; Anselm, Proslogion 18–22), immutable, atemporal, and omniscient.

The three attributes pull against one another when the question is whether God thinks. Either divine knowledge involves successive cognitive episodes, or it is one timeless act. The first option conflicts with strict immutability. The second is the classical answer and cannot be dismissed merely by assuming that every instance of thought must be episodic.

The classical responses to this horn, Boethian eternity (Consolation of Philosophy V.6); the Stump and Kretzmann ET-simultaneity reconstruction (a relation of co-presence between eternity and time without temporal succession on the divine side);²⁰⁰ Leftow’s Time and Eternity;²⁰¹ Helm’s Eternal God,²⁰² all attempt to dissolve the succession by relocating divine cognition outside time. The relocations preserve immutability by separating the mode of divine knowing from creaturely succession. They face a burden of explaining how one timeless act bears differentiated, tensed, and contingent content, but the proposal is not closed simply by redescribing that mode as non-successive.

If God has no cognitive state or relation at all, omniscience is lost. But dispositional and non-occurrent models of knowledge show that knowing need not consist in a stream of episodes. The live issue is what positive cognitive relation remains and whether it can distinguish its contents.

The Thomistic actus purus names fully actual actuality without potency rather than activity-as-transition; on that strongest reading (§4.2.1 develops the eternal-present and actuality-potentiality discipline; §5.2.1 the classical divine simplicity case), the conditions §3.5 establishes are instantiated not in the mode of divine cognition (which is unitary act, not succession) but in the content divine cognition cognizes. ‘Eternal knowledge’ (Aquinas, ST I.14) is content-bearing: omniscience predicates determinate content over articulated objects (S-distinguishability); divine knowing of what-is-the-case inherits the temporal articulation of what-is-the-case (T at the level of what is known); the unity of the eternal act sustains its content (Φ as the structural carrying of the act, not the duration of a process).

Nor does recasting divine omniscience as second-personal knowledge (acquaintance with persons rather than grasp of propositions) lift the result clear of the conditions: the persons known are themselves distinct from one another and from their knower (S) and articulated across the times of their lives (T), so personal knowing bears determinate content no less than propositional knowing, and the content it bears carries the structure the relocation of mode was meant to escape. The Aristotelian and Thomistic frameworks (Metaphysics Λ 1072b14–30; ST I.3) shift where T, S, and Φ are instantiated; they do not extract divine cognition from the structural conditions any cognition’s content requires. The classical-theistic account therefore faces a mode–content tension rather than a deductively closed dilemma: succession in the divine mode threatens immutability, while a timeless unitary mode owes an account of differentiated and temporally articulated content.

Feser’s Aquinas²⁰³ and Five Proofs²⁰⁴ develop the most sophisticated contemporary integration of the apophatic and analogical modes: apophasis disciplines the excess of positive predication, and the analogy of being (analogia entis) supplies the positive content that apophasis alone cannot provide. On Feser’s reading, “God is not composite” (apophasis) pairs with “God is subsistent being itself” (analogy), the two modes functioning as complementary faces of a single theological discipline.

The structural reply tracks both poles. The apophatic pole has been addressed above: every act of negation instantiates T, S, and Φ in the performance. The analogical pole is addressed in the engagement with the analogia entis at §5.2.1: analogical predication requires a real proportion grounded in something real about the divine subject: the predicates must track a real divine correlate that stands in the appropriate proportion to its creaturely counterpart, and such a correlate must at least support determinate difference, direction, and stable bearing. Whether those roles warrant the stronger labels Space, Time, and Substance is a further step rather than a consequence of analogy alone.

The complementarity places structural conditions on the act side and a determinate-content burden on the object side. It does not by itself prove that the apophatic referent instantiates the same structure as the discourse used to approach it.

What converges. The chapter’s stance on the via negativa is disciplined rather than eliminative. The apophatic tradition tracks the inadequacy of univocal projection from creaturely concepts to divine being. Its enacted negations are structurally conditioned, but that act-side fact does not settle whether apophatic discourse succeeds in referring beyond determinate predication. The remaining issue is therefore not refuted by retorsion; it is a burden of apophatic semantics and theology. The general structural result sets the terms of that burden, while the religious and existential significance of the apophatic remainder is carried forward to §6.8.3.

5.3 Non-Classical Ultimates

This cluster gathers the accounts of the ultimate that are not classically theistic but make the same bid against structure: Spinozist pantheism (§5.3.1), Neoplatonism (§5.3.2), process theology (§5.3.3), and Madhyamaka emptiness (§5.3.4). Each names or reconceives ultimacy — one infinite substance, the One beyond being, creative process, or emptiness. Some place the ultimate prior to familiar structure; others identify ultimacy with an immanent or dependent order rather than trying to transcend it.

The single question carries over: can the account reach its ultimate without leaning on the structure it was meant to stand free of? The finding is relocation rather than refutation. The candidates divide: articulated priority claims incur determinate-role burdens; immanent and process accounts may be compatible with the structural floor while revising ultimacy; strict apophatic and anti-thesis positions leave an open semantic or dialectical seam.

5.3.1 Spinozist Pantheism

Spinoza’s Ethics (Part I) identifies God with Substance: the single, infinite, necessarily existing reality of which all finite things are modes. Spinoza’s God is not a person standing outside Nature but the totality of what is, expressed under the attributes of thought and extension (and infinitely many other attributes we do not access). On this view, God is the one necessary substance expressed through attributes and modes. That makes Spinozism a close comparison for the book’s necessary structural ground, but not an automatic identification with it.

The Spinozist position is the non-classical counter-construction closest to the argument’s own commitments. Unlike the classical theist, the Spinozist does not place God beyond Nature. The Spinozist can therefore regard T, S, and Φ as necessary features of the one substance’s modal order. This is compatibility, not identity: Spinoza’s attribute theory, necessitarian derivation of modes, and parallelism go beyond the book’s structural result.

The alliance should be scoped, though. Spinoza’s attributes are thought and extension expressed co-equally (the order and connection of ideas the same as the order and connection of things) whereas T, S, and Φ make Physicality fundamental and derive mind downstream, as recursive structural depth yields phenomenal interiority (Chapter 8). That downstream derivation of mind is a commitment this account owes rather than a result it can claim here: the defensive case that experience requires no non-structural ingredient is made at §4.3.1, and the positive account of how structural depth yields interiority is earned, not assumed, in Chapters 7–8.

On the immanence-and-necessity axis the convergence is genuine; on the attribute taxonomy the present account is a physicalist necessitarian, not a parallel-attribute substance monist, and the parallelism that is Spinoza’s third distinctive demand is precisely what it does not inherit.

The half-necessitarian pressure. A sharper Spinozist challenge presses not the identification but the prior consistency of the necessity verdict. If the engine behind the argument (the derivation in §§2.6–2.7 that closes the necessity of T, S, and Φ) carries any modal force, Spinoza’s own reasoning suggests it should close modal space at the level of the modes as well.

Spinoza holds that finite modes follow from infinite Substance with the same logical necessity as theorems follow from axioms; contingency, on his view, is an epistemic effect of partial perspective, not a structural feature of reality. A half-necessitarian reading (necessary structure, contingent particular configurations) appears to split what Spinoza holds together.

This is not a worry read off the primary text alone: Melamed (Spinoza’s Metaphysics),²⁰⁵ the most careful contemporary reconstructor of Spinoza’s substance monism, defends the strict-necessitarian Spinoza against the moderate readings that would soften the modal claim, holding that on Spinoza’s system there is no genuine contingency anywhere, the finite modes included, and that the appearance of contingency is wholly an artifact of inadequate ideas. The challenge is therefore pressed in its strongest current form: if the argument’s modal engine carries the force claimed for it, the most defensible reading of Spinoza says it should leave no contingency standing at the level of the particular configurations either.

The reply is that the argument’s modal engine is not a rationalist PSR that cascades from necessary ground to necessary consequence. It is a more confined demonstration: that T, S, and Φ are conditions on any concrete existent, derived from the incoherence of ∅ (the maximally undifferentiated configuration in which none of the conditions obtains), not from a principle requiring that whatever follows from the ground must follow necessarily. The modal PSR is confined at §4.1.3 to the existence of contingents as a class, and the necessity claim is located at the structural level alone by §2.8.

What is necessary is that there be a reality satisfying T, S, and Φ; what particular reality satisfies them is not determined by the necessity of the conditions.

The half-necessitarian position is therefore stable: the necessity runs to the conditions, not to the configurations meeting them. Spinoza’s deeper necessitarianism requires a further principle (that the specific attributes and modes are uniquely determined by Substance) that the present argument does not need and does not accept.

The argument grants the Spinozist’s structural commitment. What it contests is the identification. Deus sive Natura (God-or-Nature: Spinoza’s formula identifying the two) is offered as a term-substitution. If “God” is just another name for the necessary ground, the substitution is harmless and the disagreement is verbal.

But the historical theological work the term God is asked to do (providing an object of worship, a source of moral obligation, a being to whom prayer is addressed, a guarantor of cosmic justice) does not transfer to the structural ground. The structural ground is not a being to whom one prays, a moral agent, or a loving person.

What is established. The Spinozist who accepts this pares the divine vocabulary down to something closer to reverence for what is. That position is coherent and may be theologically valuable; it is not what the analytic theology literature means by classical or personal theism. The argument’s verdict is that Spinozism avoids the specific burden of locating God beyond structure because it identifies divinity with immanent Nature. The cost is the abandonment of most of what the term God was historically asked to deliver.

5.3.2 Neoplatonism

Plotinus (Enneads)²⁰⁶ and the Neoplatonic tradition place the ultimate ground beyond being itself. The One (τὸ Ἕν) is not a being among beings, not even the supreme being; it is the source from which being emanates, lying beyond all predication, including the predicates of existence and unity in the strict sense. Plotinus is willing to push apophasis (the via negativa: approaching the divine by denying predicates rather than affirming them) past where classical theism stops. Where Aquinas’s God is being-itself (ipsum esse), Plotinus’s One is prior to being and to esse alike.

The position relocates the challenge rather than escaping it. If the One is the source from which Reality emanates, the account posits an ordering or dependence relation between source and derivative. Plotinian emanation is not ordinarily a temporal process, so succession cannot simply be assumed. The sharper burden concerns determinate sourcehood, derivative difference, and asymmetric dependence: how these obtain without turning the One into one relatum among others.

Nor does the henological motive for positing the One (that a manifold of conditions must derive from a prior unity to be one structure rather than a heap) find purchase here: T, S, and Φ (the three structural conditions, borne by Time, Space, and Substance) are not an aggregate awaiting an external unifier but one Nature, identical under the Triconditional Identity (§2.6.10), so the unity the One is invoked to supply is already internal to the structural ground.

The same analysis covers the inward arm of the Neoplatonic cycle. The reply to this point runs on procession (the outward emanation from the One) but the reversion by which all things turn back toward their source fares no better: reversion likewise expresses an ordered dependence of the derived upon its source, not necessarily a clock-time journey. It requires intelligible difference and direction, while the Neoplatonist owes an account of how those roles apply analogically without compromising the One’s priority.

Plotinus is alert to the difficulty, and the difficulty has able contemporary defenders, so the apophatic reply must be taken at full strength. Gerson (Plotinus)²⁰⁷ reconstructs the language of emanation as systematically analogical rather than literally predicative, and Perl (Thinking Being)²⁰⁸ reads the One not as a further being standing behind beings but as the sheer intelligible presence in which beings consist, so that talk of a source names a relation of dependence internal to appearing rather than a process undergone by some super-entity.

The Enneads insist that the language of emanation, derivation, and procession is improper: that nothing can really be said about the One, including that it is the source of anything. The strategy of strict apophasis preserves the One’s transcendence by withdrawing every predicate, including the predicate of being a ground.

The consequence should be stated with care. If the One is not asserted to stand in a determinate grounding relation, then the present argument has no object-level ground claim to refute. It can register that the philosophical practice of unsaying is structurally enacted, but that act-side point does not establish the One’s structural character. If a Neoplatonist instead retains determinate claims of emanation, sourcehood, or dependence, those claims incur the grounding-role burden developed in this chapter. The engagement therefore yields a fork: apophatic silence leaves a theological or contemplative stance outside this chapter’s positive grounding claims; articulated emanation enters the dialectic and must explain its relation.

The contemporary readings do not escape this. Analogical predication is still predication: an analogy holds between two terms in a respect, and to sustain it is to deploy distinction and relation, the very conditions at issue. And Perl’s immanentist reconstruction concedes the point rather than evading it, for a presence in which beings consist, a dependence internal to appearing, just is a structural-dependence relation described from the inside, T, S, and Φ under another name, not a way past them.

What closes. The chapter closes only the articulated priority claim: a One described as emanating, grounding, or standing in determinate dependence relations incurs the structural burden of those relations. Strict apophatic silence is not thereby contradicted; it declines to supply the kind of ground claim this chapter adjudicates. The remaining disagreement is whether such silence can support the explanatory work Neoplatonism often asks of the One.

5.3.3 Process Theology

Process theology, developed by Charles Hartshorne (The Divine Relativity;²⁰⁹ A Natural Theology for Our Time)²¹⁰ from Whitehead’s metaphysics in Process and Reality,²¹¹ abandons the classical attributes in favor of a God who is temporal, related, and changing. Hartshorne’s central technical move is dipolar theism: God has two natures, a primordial nature that grasps the eternal possibilities and a consequent nature that grows with the world’s becoming. Divine perfection is reconceived. Omniscience is knowledge of all that has become, not of what will be (which is genuinely indeterminate). Omnipotence is the maximal persuasive power compatible with creaturely freedom, not coercive control. Eternity is everlastingness within time, not timelessness outside it.

John Cobb and David Griffin’s Process Theology: An Introductory Exposition²¹² gives the tradition its most influential systematic statement, and the position has been extended by Marjorie Suchocki (God–Christ–Church;²¹³ The End of Evil),²¹⁴ Joseph Bracken (The Triune Symbol;²¹⁵ Society and Spirit)²¹⁶ on social-trinitarian process theism, Roland Faber (God as Poet of the World)²¹⁷ on theopoetics, and Thomas Jay Oord (The Uncontrolling Love of God)²¹⁸ on essential kenosis.

The position’s significance for the present argument is that it is the theistic family closest to the present account: both reject the classical attributes of timelessness, immutability, simplicity, and impassibility; both insist on the reality of time, change, and relation as features any adequate metaphysics must accommodate.

Where they part is over personhood and grounding. The process theist holds that the supreme exemplification of process is itself a person (the divine consequent nature is conscious, loving, and responsive) and that this person is the necessary ground of the world’s becoming. The present argument is silent on the first claim and supplies a limited structural challenge to the second. Two distinct questions need separating, and each requires a distinct reply.

The first question is whether, within Nature, there exists a maximally inclusive person who experiences every event and to whom every event matters. On this the present argument is silent. It establishes Nature’s necessary features; it does not settle which configurations within Nature actually obtain.

The choice between a process-theistic Nature (with a maximally inclusive divine person) and a non-process-theistic Nature (without) is downstream of what the argument settles, and turns on substantive empirical, phenomenological, and theological considerations rather than structural ones.

The phenomenology of religious experience, the problem of evil and theodicy, the coherence of divine consequent-nature consciousness, the question whether anything in nature exhibits the universal-prehension structure Whitehead’s metaphysics requires (prehension being Whitehead’s technical term for the non-cognitive grasping by which each occasion of experience takes up its predecessors): these are the considerations that bear on the first question, and the present account does not adjudicate them.

Hartshorne’s argument from divine relativity, Cobb and Griffin’s case for the religious adequacy of process theism, Suchocki’s development of the eschatological consequent nature can stand or fall on their own terms. The present account neither requires nor forbids them.

The second question is whether such a being could be what classical theism wanted God to be: the necessary ground of T, S, and Φ (the three structural conditions, borne by Time, Space, and Substance) rather than a configuration within them. On this the present argument is not silent, but its conclusion is limited: a process-theistic God clearly realizes temporal, relational, and active roles; whether those roles preclude every sense of divine ultimacy depends on the further theory of grounding and aseity the process theist offers.

Hartshorne’s dipolar theism is the test case. The primordial nature, on Hartshorne’s account, grasps the eternal possibilities timelessly; the consequent nature grows with the world’s becoming. The question is which nature is supposed to ground the world’s structural conditions.

If the primordial nature grounds them, the process theologian must explain how an abstract pole that orders possibilities grounds concrete becoming; this does not by itself collapse into classical theism. If the consequent nature grounds them, its growth with the world raises a circularity or mutual-dependence concern, though a process theorist may explicitly reject one-way classical grounding.

Cobb and Griffin’s reply²¹⁹ is that the dipolar God is the ground of the order of becoming rather than of the structural conditions of becoming, but the distinction does not save the position. Order presupposes the structural conditions, and a God who grounds order while presupposing the conditions is not the necessary ground the classical project required.

The dilemma tightens into a single result, one the attribute analyses of Chapter 6 rest on rather than re-derive. What makes the two poles poles of one God, on Hartshorne’s own account, is a relation between them: the consequent nature prehends, includes, and grows from the primordial nature, and the primordial nature is the abstract ground the consequent nature concretely actualizes.

That relation is not static co-instantiation but a directed dependence: grounding and prehension run from the primordial ordering of possibilities to the consequent nature’s concrete inclusion of the world.

The pole relation requires determinate difference, asymmetric order, and a bearer that unifies the poles. These are the book’s thin structural roles. Identifying them with loaded spatial interval, temporal succession, and physical substance requires further argument, but dipolar theism plainly does not describe a wholly structureless ultimate.

The same role-floor applies to the dipolar reconception of perfection as responsiveness. Perfect inclusion of and affectedness by the world require determinate receptivity and relation. This locates the divine consequent nature within an articulated process, but it does not by itself settle whether that process can be metaphysically fundamental.

Hartshorne’s God avoids the classical tensions of timeless simplicity by instantiating the structural roles at both poles and in their relation. The remaining question is whether such instantiation makes God merely an inhabitant or instead a necessary, co-constitutive feature of the process order. This result is deployed attribute by attribute across Chapter 6: each section names the local form the pole-relation takes for its predicate and refers the structural verdict back here.

The second question is taken up at length in Chapter 6. The framework-invariant point is established at §6.1: a temporally embedded, causally extended, relationally constituted deity instantiates the structural conditions and therefore owes an account of whether its relation to them is dependence, identity, or co-constitution. It is closed as a binding cost at §6.10 (Horn 3): a God who exists in time and acts within the structure is a God who depends on the structure for that position.

Process theology avoids the charge that its divine candidate lacks determinate temporal, relational, or active roles. It carries a serious aseity burden: a temporal, relational God appears dependent on the conditions of the process it inhabits. Process theists may respond by revising aseity rather than denying dependency; the present account records that cost without treating the revision as a contradiction.

Suchocki’s eschatological development of the consequent nature does not change this: an eschatological consummation is still a temporally extended event, and an eschatologically consummated God is still a configuration within Nature. Bracken’s social-trinitarian extension multiplies the divine persons but does not relocate them outside Nature. Oord’s essential kenosis specifies the modality of divine action without altering the metaphysical embedding. Each contemporary development of process theology preserves the central commitment (God is temporal, related, changing) and each, on the present analysis, preserves the consequence: such a God instantiates T, S, and Φ and owes the same account of whether its relation to them is dependence, identity, or co-constitution.

A kindred position worth marking is open theism (Pinnock et al., The Openness of God;²²⁰ Hasker, God, Time, and Knowledge).²²¹ Open theism shares with process theology the rejection of timeless omniscience and the affirmation of genuine creaturely freedom, but typically retains creation ex nihilo and divine independence in ways process theology does not. The present reply raises the same question: an open-theist God who exists in time and knows the past while leaving the future open is structurally embedded, and the theorist must explain how that is compatible with the proposed grounding role. The aseity question is the same question for both positions.

Process theology is the theistic position best positioned to accept the argument’s structural commitments without contradiction, but the cost is a substantial revision of classical aseity (the divine attribute of existing from oneself alone, independent of anything else for one’s being), and aseity is the load-bearing commitment of the classical project. Process theology provides a coherent way of being a theist after the present account; it does not provide grounds for accepting theism, and it does not preserve what the classical project was distinctively trying to secure.

Whether the further commitments of process theology are independently warranted (whether the consequent nature exists, whether divine persuasion is the right account of God’s action, whether eschatological consummation is the right account of history) are questions for the philosophy of religion rather than for the present account. The present account leaves the first question open; on the second, it establishes the structural instantiation and records the aseity cost.

What is established. The process theologian who accepts the structural argument has a coherent position: a God who realizes temporal, relational, and active roles at maximum scope. This is not the classical aseity claim, and it requires either treating God as a configuration within Reality or articulating a revised account of ultimacy compatible with divine process. Hartshorne’s move is therefore a genuine theological revision, not a mere preservation of the classical project. The present account records the cost: process theology does not straightforwardly preserve the grounding role classical theism assigned to God. Whether its revised religious and metaphysical picture is independently preferable remains a live philosophy-of-religion question.

5.3.4 Madhyamaka Emptiness

Nāgārjuna’s Mūlamadhyamakakārikā (c. 150 CE)²²² develops the doctrine of śūnyatā (emptiness) as a sustained critique of any view that attributes intrinsic nature (svabhāva) to anything whatever. Nothing exists from its own side; everything arises in dependence on conditions. The tetralemma (A; not-A; both A and not-A; neither A nor not-A) exhausts the options and refutes each. The result is not nihilism but the deconstruction of any positive metaphysical thesis, including the theses of substance, time, and space as intrinsically existing.

The challenge is radical in a way no other challenge the argument faces is. The Madhyamaka position is engineered to dissolve precisely the kind of structural claim the present book defends, and contemporary Madhyamaka scholarship has sharpened the challenge rather than softened it. Garfield,²²³ Siderits,²²⁴ and Westerhoff²²⁵ defend a Madhyamaka (particularly in its Prāsaṅgika form) that is not the conciliatory Madhyamaka of Tsongkhapa’s reading but a thoroughgoing rejection of all metaphysical theses.

On this reading, the prasaṅga (consequence-only) method is dialectical: the Mādhyamika (the adherent of the Madhyamaka school) does not advance any positive position, including the position that things arise dependently, but only draws absurd consequences from the positions of others.

The tetralemma is not a fifth alternative but the closure of all alternatives. The Mādhyamika has no thesis,²²⁶ and the present argument, having theses, is precisely the kind of view prasaṅga targets.

Two replies must be distinguished, because they engage Madhyamaka at different scopes.

The narrower reply concerns svabhāva. Svabhāva in the technical sense is intrinsic, independent, self-existing nature. Nāgārjuna’s arguments are devastating against the claim that any phenomenon has nature in that sense. They are not directed against (and do not refute) the claim that phenomena exhibit structural conditions that obtain dependently. The present argument here has never claimed svabhāva for T, S, or Φ (the three structural conditions, borne by Time, Space, and Substance). The structure is mutually constituting: each role requires the other two, and none has independent existence apart from the others. In the technical Mahāyāna vocabulary, T, S, and Φ are empty of svabhāva and dependently arisen with respect to one another.

The narrower reply observes the convergence: the pratītyasamutpāda (dependent arising) Nāgārjuna names is structurally isomorphic to the present argument’s mutual requirement.

The wider reply must engage the Prāsaṅgika who refuses the convergence, who insists that even “structural conditions that obtain dependently” is svabhāva in disguise, because the assertion takes a stable form, picks out determinate roles, and survives translation across linguistic and conceptual schemes. On the strict Prāsaṅgika reading defended by Garfield and Westerhoff, the present argument is doing what prasaṅga targets: advancing a positive thesis (the Triunity holds across all possible worlds) under cover of denying svabhāva in name only.

The present argument resists the wider critique only at a limited point. It does not depend on assimilating Madhyamaka to a conciliatory reading; it distinguishes the practice of prasaṅga from the object-level thesis the Mādhyamika declines to assert.

Prasaṅga is a dialectical method: the Mādhyamika draws absurd consequences from an opponent’s premises. The drawing-of-consequences is an activity. It occurs through succession, differentiates premises from conclusions, and persists through a reasoner or practice. Prasaṅga therefore exhibits T, S, and Φ at role-resolution in its operation.

That act-side fact does not force the Mādhyamika to accept an object-level thesis that the structural conditions exist inherently or independently. It establishes instead that the conditions of the dialectical practice are not dissolved by the practice. The structural account may claim those conditions are necessary; the Prāsaṅgika may deny that this is a final metaphysical description. Their dispute remains over the status of the positive structural thesis, not over whether the practice of reasoning takes place in a structured manner.

The four corners, run explicitly. The catuṣkoṭi is built to exhaust the options for any thesis and refute each in turn; run against □(T∧S∧Φ), it reaches no exit.

To affirm the first corner (that the conditions hold) is to perform an assertion that unfolds over time, distinguishes what is affirmed from what is not, and persists through the one affirming: the affirmation instantiates T, S, and Φ.

To affirm the second corner (that the conditions do not hold) is to perform a denial carrying the same three features, so the denial instantiates what it denies.

The third corner, both, must differentiate the respect in which the conditions hold from the respect in which they do not (Space), hold the two together in one bearer (Substance), and articulate them in sequence (Time); it presupposes the conditions in order to state its two conjuncts.

The fourth corner, neither (the refusal to occupy any corner, and the Prāsaṅgika’s own dwelling) is itself a configuration: the withholding unfolds over time, sets itself apart from assertion, and persists through the one who withholds.

The tetralemma closes the space of positions one might take toward the modal claim; it does not close the space of conditions on taking any position at all. Each corner, the empty fourth included, is occupied by an activity that exhibits the very structure the corner was meant to dissolve: the catuṣkoṭi cannot be turned against □(T∧S∧Φ) except from within it.

What is established. The Tsongkhapa-Gelug reading distinguishes the negation of svabhāva from the negation of conventional existence: the structural conditions are conventionally real and ultimately empty of intrinsic nature, and the argument is compatible with that reading. The stricter Prāsaṅgika reading remains a live disagreement. The present account can establish that Madhyamaka practice is structurally enacted; it cannot settle the wider anti-thesis position by treating that enactment as a contradiction. The narrower reading yields accommodation; the wider reading leaves a genuine dispute over whether structural necessity is a positive metaphysical thesis or only a conventionally indispensable description.

5.4 What the Rival Ultimates Carried

This is the chapter’s reckoning. The rival ultimates do not all end with one verdict. Some descriptions subtract the conditions under which a determinate ground or relation can be specified, and on the book’s analysis dissolve into ∅. Some preserve directed dependence, relational non-identity, and real grounding content under analogical, process, or apophatic vocabulary. Some remain live only by positing an essence-prior or transcendent ground whose relation to the structural domain requires independent metaphysical defense.

The point is not that theistic philosophers reasoned carelessly — they reasoned with great care, and the tradition’s repairs are among the most carefully built constructions in the literature. The question is which verdict each repair earns: incoherence, relocation, compatibility, explanatory cost, or a remaining open seam.

The preceding sections examined the tradition’s main repairs: the foundation and contingency arguments (Rasmussen, the Kalām), the simplicity constructions (classical divine simplicity, Swinburne), the modal and ontological arguments (Plantinga, the modal-grounding argument), the cosmological and design inferences (simultaneous causation, fine-tuning, intelligent design), and the apophatic, absolute-monist, and process traditions (the via negativa, Neoplatonism, Spinozist pantheism, Madhyamaka emptiness, process theology). Each arrived with its own apparatus and its own point of pressure.

A recurrent pattern from Chapter 4 emerged here: some constructions presuppose structural roles they were built to transcend, while others revise ultimacy or leave an explanatory seam rather than collapse.

The Spinozist identified God with immanent Nature and accepted structural articulation rather than classical transcendence. The Neoplatonist placed the One beyond being and described emanation. The process theologian made God temporal and embedded: the revised theist’s option, named differently. Plantinga’s maximal greatness raises a compossibility and possibility-premise burden; the strongest negative classical attributes are additional commitments rather than part of that definition. The Madhyamaka (Nāgārjuna’s middle-way Buddhist school) emptied svabhāva (Sanskrit: “own-being” or self-existence) and left the dependent-arising structure intact. The resulting verdicts differ: compatibility for some immanent accounts, a grounding burden for articulated transcendence, and an open dialectical seam for strict apophasis or anti-thesis positions.

Three interlocutors deserve to be taken slowly, because each presses back in a different way: the revised theist, who accepts the structure and keeps the name; the Thomist identifier, who accepts the result and claims the tradition reached it first; and the Kalām objector, who grants the structure as actual and asks for a cause outside it.

Consider the revised theist. The philosopher who abandons classical theism and redefines God as temporal, extended, and causally embedded (a necessary person who instantiates the structural roles rather than lacking them) has escaped ∅. That much the argument grants. The revision places God within the structural order. That creates a burden for classical aseity, but embeddedness alone does not prove non-fundamentality: the revised theist may propose a necessary, co-constitutive, or priority-bearing configuration and must defend that grounding account.

Consider the Thomist identifier. Distinct from the revised theist, this interlocutor does not modify the classical attributes. The identifier accepts the argument’s structural result and identifies it with God under the classical description intact: Thomas’s ipsum esse subsistens, Feser’s Pure Act, Dolezal’s being-itself: these, the identifier says, just are □(T∧S∧Φ) under another name. If the argument proves a necessary structural ground, Thomism was right. The ground exists and deserves the name the tradition gave it. The book has not refuted classical theism; it has renamed it.

The identification requires what it denies. Ipsum esse subsistens was specified by Thomas as being without temporal succession, without spatial extension, without material composition: the via negativa’s long labor was precisely to mark what the necessary ground lacks. The structural account specifies the ground by what it instantiates: succession, relational non-identity, causal persistence. These are opposing specifications of the ground’s character, not two descriptions of one fact. The identifier cannot hold both.

Drop the classical attributes to preserve the identification and the move collapses into the revised theist’s option above. Keep the classical attributes and the argument’s result (that a ground lacking T, S, and Φ is ∅) reapplies without remainder. No third position.

Analogical predication does not supply a non-T/S/Φ articulation. The Thomist reply at this junction is that the “requires” in the reduction above smuggles in univocal predication: it treats the source–recipient relation as if it had to be articulated in the same kind of temporal, spatial, and physical terms that creaturely causal relations require, whereas the Thomist insists on analogical predication for God-talk.

Grant the point: God-talk is analogical, and the divine act is not one more event inside the world.

The structural claim is not that the divine act must be creaturely T/S/Φ, but that any determinately contentful “source–recipient” relation must still carry (i) an ordering (before/after, priority/posteriority, dependence direction), (ii) a non-identity between relata (this rather than that), and (iii) a persistence or invariance that makes it the same source across the relation. Those are exactly the role-definitions of T, S, and Φ.

If the Thomist denies all three at once (no ordering, no non-identity, no persistence/invariance) then the “relation” has no determinate content and collapses into the Null State-style limit: not a higher articulation, but no articulation at all. If the Thomist grants that the relation has determinate content (which requires all three), the via negativa no longer specifies a ground lacking T/S/Φ; it specifies a ground instantiating them under analogical description, which is just the revised-theist relocation in different vocabulary.

So the analogical/univocal distinction does not open a third option: it either leaves the reduction intact (no determinate relation, hence ∅) or reintroduces the structural roles under another name (hence in-Nature rather than beyond it).

Thomist-Identifier Variant · R14 · Result of the Drop/Keep Dilemma

The Thomist identification of □(T∧S∧Φ) with the classical ipsum esse subsistens cannot be sustained. Either the classical attributes are dropped, in which case the identifier collapses into the revised theist (configured within Nature, not above it); or the classical attributes are kept, in which case the ground specified by what it lacks reduces to ∅ under the Null State definition (§2.7.1). No third option. The local Chapter 5 summation rests on the Chapter 2–4 machinery already established; the Result form anticipates Supernatural Preclusion (§6.1) as a variant routing through Horn 3 of the Trilemma Closure, where the same result is restated under Supernatural Preclusion once that result is in place downstream. (§2.7.1, §6.1)

Proof (Thomist-Identifier Variant). Notation: Drop = drop the classical attributes; Keep = keep the classical attributes (the two options developed above).

1. Drop ∨ Keep: exhaustive over the identifier’s options on the classical attributes.

2. Drop → collapses into revised theism: the structural ground is configured within Nature, not above it.

3. Keep → a ground lacking T, S, Φ is ∅; the argument’s result reapplies without remainder.

∴ the Thomist identification cannot be sustained. ∎

The identifier who presses anyway applies the tradition’s name to a ground specified by the features the tradition defined its own ground against. The structural ground is specified by what it instantiates: succession, relational non-identity, causal persistence; ipsum esse subsistens is specified by what it lacks. The identification holds only if one of the two specifications is silently dropped.

Consider the Kalām objector. The Kalām is the contemporary research program dedicated to denying necessity-of-T on independent grounds: the conceptual impossibility of an actual infinite past (Hilbert and traversal arguments), the empirical claim that the universe has a past boundary (BGV), and the cosmological inference from beginning to a transcendent personal cause.

It fits the chapter’s pattern differently from Chapter 4’s deflationary, change-skeptical, and modal-variants clusters: the Kalām does not arrive at the structural conditions by presupposing them in its denial; it accepts T, S, and Φ as actual and contingent and asks for an originating cause outside the order.

The chapter pressures the Kalām along a different seam. §5.1.2 distinguishes objections to an actual infinite past, evidence for a finite cosmic boundary, and the further inference to a timeless personal cause. A first temporal boundary need not be a moment prior to T, and §5.1.6 no longer treats an ordinary historical Fall as requiring such a categorial origin. The composition view of Temporality is marked at §7.2.2, separating the quantitative question of how far the past extends from the metaphysical question of what grounds the order.

The destination the Kalām pursues (a necessary ground for the contingent series) is adjacent to where the structural argument lands; the line taken to that destination, and the personal-cause specification of what is found there, are what the chapter contests.

The same structural ground constrains accounts of miracles. It does not by itself establish their logical impossibility, because classical theists can construe a timeless cause as producing temporal effects without undergoing a temporal episode. (This local summation draws only on the Chapter 2–4 machinery already in place: the Principle of Modal Invariance and the structural-condition requirement on any act of intervention; the parallel to the Supernatural Preclusion of §6.1 is anticipatory, restated under that result once it is established downstream.)

A miracle is ordinarily a change in a state of affairs attributed to divine intervention. The effect has temporal, spatial, and bearing conditions. The disputed question is whether the cause must instantiate those conditions in the same mode or can ground the effect timelessly.

The Principle of Modal Invariance, if established, forecloses an effect that literally lacks all structural conditions. It does not without further premises foreclose an unusual event within those conditions or settle the mode of its cause.

A being defined as atemporal, aspatial, and immaterial faces an interface problem: the account must explain how it stands in a determinate causal relation to physical events without simply assuming that cause and effect share one mode of being. “Intervention” is a content-laden notion. The event described as intervention occurs to something, at a time and place, through T, S, and Φ. Whether its cause must share that mode is precisely the interface question, not a premise the analysis may assume. The structural analysis therefore does not foreclose miracles merely from the classical attributes. It imposes a burden on the causal interface and on the evidential identification of an event as divine intervention.

The difficulty pressed here is chiefly structural, though several arguments also carry evidential burdens: classical claims must retain determinate content and explain their relation to the structured domain.

The book inherits from these repairs the question it turns: not whether a necessary ground exists — the argument grants that it does — but whether what stands there is a being beyond Nature or the structure of Nature itself.

The theistic rivals carried the question further than anyone. Their active grounding descriptions commonly carry ordered dependence, relational non-identity, and real grounding content. A ground defined by the absence of every such condition dissolves on the book’s analysis; a ground that grants them has conceded a structured ground. What remains open is not merely whether God is its name, but whether an essence-prior or transcendent proposal can make its grounding relation intelligible without collapsing into either a structured relation or an unexplained primitive. That is the residual burden Chapters 5 and 6 continue to test.

---

1. Joshua Rasmussen, How Reason Can Lead to God: A Philosopher’s Bridge to Faith (IVP Academic, 2019).

2. Avicenna, The Metaphysics of the Healing (Translated by Michael E. Marmura. Provo, UT: Brigham Young University Press, (c. 1020) 2005), https://search.worldcat.org/title/56012718.

3. G. W. Leibniz, “Monadology,” in Philosophical Essays, Translated by r. Ariew and d. Garber. Indianapolis: Hackett ((1714) 1989), sec. 32, https://philpapers.org/rec/ARIGWL.

4. Alexander R. Pruss, The Principle of Sufficient Reason: A Reassessment (Cambridge University Press, 2006).

5. Robert C. Koons, “A New Look at the Cosmological Argument,” American Philosophical Quarterly 34 (2): 171–192, 1997, https://www.jstor.org/stable/20009892.

6. Koons, “A New Look at the Cosmological Argument.”

7. Alexander R. Pruss and Joshua L. Rasmussen, Necessary Existence (Oxford University Press, 2018), https://doi.org/10.1093/oso/9780198746898.001.0001.

8. David Pizzi, “The Impossibility of Nothingness: A Foundational Argument for Necessary Existence,” 2025, https://philarchive.org/rec/PIZTIO-2.

9. Pizzi, “The Impossibility of Nothingness.”

10. Graham Oppy, Arguing about Gods (Cambridge University Press, 2006).

11. al-Kindī, On First Philosophy (Fī Al-Falsafa Al-Ūlā) (Trans. Alfred L. Ivry. Al-Kindī’s Metaphysics. SUNY Press, 1974., c. 850).

12. al-Ghazālī, The Incoherence of the Philosophers (Tahāfut Al-Falāsifa) (Trans. Michael E. Marmura. 2nd ed. Brigham Young University Press, 2000., c. 1095).

13. William Lane Craig, The Kalām Cosmological Argument (Macmillan (London) / Library of Philosophy; Religion, 1979), https://doi.org/10.1007/978-1-349-04154-1, ch. 1; Herbert A. Davidson, Proofs for Eternity, Creation and the Existence of God in Medieval Islamic and Jewish Philosophy (Oxford University Press (New York), 1987), https://search.worldcat.org/title/242351292 traces the Islamic transmission.

14. Craig, The Kalām Cosmological Argument; Koons, “A New Look at the Cosmological Argument”; Pruss, The Principle of Sufficient Reason; Pruss and Rasmussen, Necessary Existence; Rasmussen, How Reason Can Lead to God.

15. Oppy, Arguing about Gods.

16. William Lane Craig and James D. Sinclair, “The Kalam Cosmological Argument,” in In w. L. Craig & j. P. Moreland (Eds.), the Blackwell Companion to Natural Theology (Wiley-Blackwell), Pp. 101–201 (2009), sec. 3, https://doi.org/10.1002/9781444308334.ch3; William Lane Craig and J. P. Moreland, Naturalism: A Critical Analysis (Routledge (London & New York), Routledge Studies in Twentieth Century Philosophy, 2000), https://www.routledge.com/Naturalism-A-Critical-Analysis/LaneCraig-Moreland/p/book/9780415758352.

17. Arvind Borde et al., “Inflationary Spacetimes Are Incomplete in Past Directions,” Physical Review Letters 90 (15): 151301, 2003, https://doi.org/10.1103/PhysRevLett.90.151301.

18. Anthony Aguirre and Steven Gratton, “Steady-State Eternal Inflation,” Physical Review D 65: 083507, 2002, https://doi.org/10.1103/PhysRevD.65.083507.

19. Sean Carroll, From Eternity to Here: The Quest for the Ultimate Theory of Time (Dutton (New York), 2010), https://search.worldcat.org/title/318411645; Sean Carroll, The Big Picture: On the Origins of Life, Meaning, and the Universe Itself (Dutton (New York), 2016), https://search.worldcat.org/title/999399320.

20. Craig and Sinclair, “The Kalam Cosmological Argument,” sec. 4.

21. Robert C. Koons, “A New Kalam Argument: Revenge of the Grim Reaper,” Noûs 48 (2): 256–267, 2014, https://doi.org/10.1111/j.1468-0068.2012.00858.x.

22. Alexander R. Pruss, Infinity, Causation, and Paradox (Oxford University Press, 2018), https://doi.org/10.1093/oso/9780198810339.001.0001.

23. Nicholas Shackel, The Form of the Benardete Dichotomy, The British Journal for the Philosophy of Science 56(2): 397–417, 2005, https://doi.org/10.1093/bjps/axi121.

24. Joseph C. Schmid and Alex Malpass, Benardete Paradoxes, Causal Finitism, and the Unsatisfiable Pair Diagnosis, Mind 134(534): 397–421, 2025, https://doi.org/10.1093/mind/fzae057.

25. Andrew Ter Ern Loke, God and Ultimate Origins: A Novel Cosmological Argument (Palgrave Macmillan / Springer (Cham), 2017).

26. Wes Morriston, “Must the Beginning of the Universe Have a Personal Cause?: A Critical Examination of the Kalam Cosmological Argument,” Faith and Philosophy 17, No. 2: 149–169, 2000, https://doi.org/10.5840/faithphil200017215; Wes Morriston, “Creation Ex Nihilo and the Big Bang,” Philo 5(1): 23–33, ahead of print, 2002, https://doi.org/10.5840/philo2002512.

27. Morriston, “Must the Beginning of the Universe Have a Personal Cause?”; Wes Morriston, “Doubts about the Kalām Cosmological Argument,” in In Debating Christian Theism, Ed. J. P. Moreland, Chad Meister, and Khaldoun a. Sweis, 20–32. Oxford: Oxford University Press (2013), https://spot.colorado.edu/~morristo/NewKalamCritique.pdf.

28. Thomas Aquinas, Summa Theologiae (Benziger Brothers, (1265–1274) 1947), I.19.3, I.19.7.

29. Koons, “A New Look at the Cosmological Argument.”

30. David Hume, Dialogues Concerning Natural Religion (London (publisher not named), 1779 (posthumous), 1779).

31. Paul Edwards, “The Cosmological Argument,” The Rationalist Annual for the Year 1959 (London: Pemberton), Pp. 63–77, 1959.

32. Koons, “A New Look at the Cosmological Argument,” sec. 3.

33. Pruss, The Principle of Sufficient Reason.

34. Hume, Dialogues Concerning Natural Religion.

35. Pruss, The Principle of Sufficient Reason, chs. 2–7.

36. William L. Rowe, The Cosmological Argument (Princeton University Press, 1975), https://philpapers.org/rec/ROWTCA-9.

37. J. L. Mackie, The Miracle of Theism: Arguments for and Against the Existence of God (Clarendon Press / Oxford University Press (Oxford), 1982).

38. Bede Rundle, Why There Is Something Rather Than Nothing (Clarendon Press (Oxford University Press), 2004).

39. Adolf Grünbaum, “Theological Misinterpretations of Current Physical Cosmology,” Philo 1 (1): 15–34, 1998, https://doi.org/10.5840/philo1998113.

40. I. Kant, Critique of Pure Reason (Translated by P. Guyer; A. W. Wood. Cambridge: Cambridge University Press, (1781) 1998), A603/B631.

41. Kant, Critique of Pure Reason, A426/B454.

42. Arthur Schopenhauer, On the Fourfold Root of the Principle of Sufficient Reason (Doctoral dissertation (rev. 1847); trans. E. F. J. Payne (La Salle, IL: Open Court, 1974), 1813), https://www.gutenberg.org/ebooks/50966.

43. Oppy, Arguing about Gods.

44. Jon McGinnis, Avicenna (Oxford University Press (Oxford & New York), 2010); Peter Adamson, “From the Necessary Existent to God,” in In Interpreting Avicenna: Critical Essays, Ed. Peter Adamson, 170–189. Cambridge University Press (2013), https://doi.org/10.1017/CBO9781139047890.010.

45. Eleonore Stump and Norman Kretzmann, “Eternity,” The Journal of Philosophy 78(8): 429–458, ahead of print, 1981, https://doi.org/10.2307/2026047.

46. William Lane Craig, God, Time, and Eternity: The Coherence of Theism II: Eternity (Kluwer Academic Publishers, 2001), https://doi.org/10.1007/978-94-017-1715-1.

47. Nicholas Wolterstorff, “God Everlasting,” in In Clifton Orlebeke and Lewis Smedes (Eds.), God and the Good: Essays in Honor of Henry Stob (Grand Rapids, MI: Eerdmans, 1975), Pp. 181–203 (1975), https://philpapers.org/rec/WOLGE.

48. Alan G. Padgett, God, Eternity and the Nature of Time (St. Martin’s Press, 1992).

49. R. T. Mullins, The End of the Timeless God (Oxford University Press, 2016).

50. Morriston, “Creation Ex Nihilo and the Big Bang”; Wes Morriston, “Beginningless Past, Endless Future, and the Actual Infinite,” Faith and Philosophy 27(4): 439–450, ahead of print, 2010, https://doi.org/10.5840/faithphil201027444.

51. Morriston, “Creation Ex Nihilo and the Big Bang.”

52. Edward Feser, Aquinas: A Beginner’s Guide (Oneworld, 2009), https://search.worldcat.org/title/421814088.

53. Edward Feser, Five Proofs of the Existence of God (Ignatius Press, 2017), https://ignatius.com/five-proofs-of-the-existence-of-god-fpegp/.

54. Mullins, The End of the Timeless God.

55. Padgett, God, Eternity and the Nature of Time.

56. Alfred J. Freddoso, “God’s General Concurrence with Secondary Causes: Why Conservation Is Not Enough,” Philosophical Perspectives 5: 553–585, 1991, https://doi.org/10.2307/2214109.

57. Richard Swinburne, The Coherence of Theism (Clarendon Press, 1977).

58. Richard Swinburne, The Existence of God (Clarendon Press, 1979); 2nd ed. Richard Swinburne, The Existence of God (2nd Ed.) (Clarendon Press, 2004).

59. Richard Swinburne, Simplicity as Evidence of Truth (Marquette University Press, 1997).

60. Swinburne, The Existence of God (2nd Ed.).

61. Elliott Sober, Ockham’s Razors: A User’s Manual (Cambridge University Press, 2015).

62. Michael Huemer, “When Is Parsimony a Virtue?” The Philosophical Quarterly 59 (235): 216–236, 2009, https://doi.org/10.1111/j.1467-9213.2008.569.x.

63. Mackie, The Miracle of Theism.

64. Herman Philipse, God in the Age of Science? A Critique of Religious Reason (Oxford University Press, 2012), https://doi.org/10.1093/acprof:oso/9780199697533.001.0001.

65. Oppy, Arguing about Gods.

66. Richard Dawkins, The God Delusion (Bantam Press (London), 2006), https://search.worldcat.org/title/68965666.

67. The Existence of God, Swinburne, The Existence of God (2nd Ed.), ch. 2; Mind, Brain and Free Will, Richard Swinburne, Mind, Brain and Free Will (Oxford University Press, 2013), ch. 1.

68. A. N. Whitehead, Process and Reality (Macmillan, 1929); Charles Hartshorne, The Divine Relativity: A Social Conception of God (Yale University Press (New Haven), 1948), https://yalebooks.yale.edu/book/9780300028805/the-divine-relativity/.

69. Philip Clayton, God and Contemporary Science (Edinburgh University Press / Eerdmans (Grand Rapids), 1997), https://search.worldcat.org/title/39178216; Arthur Peacocke, All That Is: A Naturalistic Faith for the Twenty-First Century (Fortress Press, 2007).

70. William Hasker, God, Time, and Knowledge (Cornell University Press, 1989), https://search.worldcat.org/title/God-time-and-knowledge/oclc/18835334; Clark H. Pinnock et al., The Openness of God: A Biblical Challenge to the Traditional Understanding of God (InterVarsity Press, 1994).

71. Mackie, The Miracle of Theism.

72. Huemer, “When Is Parsimony a Virtue?”

73. Dawkins, The God Delusion.

74. Arguing about Gods, Oppy, Arguing about Gods; The Best Argument against God, Graham Oppy, The Best Argument Against God (Palgrave Macmillan, 2013); Atheism and Agnosticism, Graham Oppy, Atheism and Agnosticism (Cambridge University Press, 2018).

75. Hume, Dialogues Concerning Natural Religion.

76. Timothy McGrew et al., “Probabilities and the Fine-Tuning Argument: A Sceptical View,” Mind 110 (440): 1027–1037, 2001, https://doi.org/10.1093/mind/110.440.1027.

77. Brandon Carter, “Large Number Coincidences and the Anthropic Principle in Cosmology,” in In m. S. Longair (Ed.), Confrontation of Cosmological Theories with Observational Data (IAU Symposium No. 63), d. Reidel (Dordrecht/Boston), Pp. 291–298 (1974), https://doi.org/10.1007/978-94-010-2220-0_25.

78. Nick Bostrom, Anthropic Bias: Observation Selection Effects in Science and Philosophy (Routledge (New York), 2002), https://doi.org/10.4324/9780203953464.

79. Swinburne, The Existence of God (2nd Ed.).

80. Robin Collins, “The Teleological Argument: An Exploration of the Fine-Tuning of the Universe,” in In w. L. Craig & j. P. Moreland (Eds.), the Blackwell Companion to Natural Theology (Wiley-Blackwell), Pp. 202–281 (2009), https://doi.org/10.1002/9781444308334.ch4.

81. Luke A. Barnes, “The Fine-Tuning of the Universe for Intelligent Life,” Publications of the Astronomical Society of Australia 29 (4): 529–564, 2012, https://doi.org/10.1071/AS12015.

82. Swinburne, The Existence of God (2nd Ed.).

83. A. Plantinga, Where the Conflict Really Lies: Science, Religion, and Naturalism (Oxford University Press, 2011).

84. Carter, “Large Number Coincidences and the Anthropic Principle in Cosmology.”

85. Bostrom, Anthropic Bias.

86. McGrew et al., “Probabilities and the Fine-Tuning Argument.”

87. McGrew et al., “Probabilities and the Fine-Tuning Argument.”

88. Saul A. Kripke, Naming and Necessity (Harvard University Press, 1980).

89. David J. Chalmers, “Does Conceivability Entail Possibility?” in In Conceivability and Possibility (Eds. Gendler & Hawthorne), Oxford University Press (2002), https://academic.oup.com/book/49672/chapter/422537279.

90. McGrew et al., “Probabilities and the Fine-Tuning Argument.”

91. Victor J. Stenger, The Fallacy of Fine-Tuning: Why the Universe Is Not Designed for Us (Prometheus Books, 2011).

92. Carroll, The Big Picture.

93. Collins, “The Teleological Argument.”

94. Carter, “Large Number Coincidences and the Anthropic Principle in Cosmology.”

95. Bostrom, Anthropic Bias.

96. Collins, “The Teleological Argument,” sec. 6.

97. Barnes, “The Fine-Tuning of the Universe for Intelligent Life.”

98. Swinburne, The Existence of God (2nd Ed.).

99. Bradley Monton, “God, Fine-Tuning, and the Problem of Old Evidence,” The British Journal for the Philosophy of Science 57(2): 405–424, ahead of print, 2006, https://doi.org/10.1093/bjps/axl008.

100. Michael Ikeda and Bill Jefferys, “The Anthropic Principle Does Not Support Supernaturalism,” in In m. Martin & r. Monnier (Eds.), the Improbability of God (Prometheus Books, Amherst, NY), Pp. 150–166 (2006), https://philpapers.org/rec/IKETAP.

101. Ikeda and Jefferys, “The Anthropic Principle Does Not Support Supernaturalism.”

102. Paul Draper, “Pain and Pleasure: An Evidential Problem for Theists,” Noûs 23(3), 1989, https://doi.org/10.2307/2215486.

103. Stenger, The Fallacy of Fine-Tuning.

104. Michael J. Behe, Darwin’s Black Box: The Biochemical Challenge to Evolution (Free Press (New York), 1996), https://search.worldcat.org/title/832583127; Michael J. Behe, The Edge of Evolution: The Search for the Limits of Darwinism (Free Press (New York), 2007), https://search.worldcat.org/title/136958644.

105. William A. Dembski, The Design Inference: Eliminating Chance Through Small Probabilities (Cambridge University Press (Cambridge Studies in Probability, Induction; Decision Theory), 1998), https://doi.org/10.1017/CBO9780511570643; William A. Dembski, No Free Lunch: Why Specified Complexity Cannot Be Purchased Without Intelligence (Rowman & Littlefield (Lanham, MD), 2002), https://lccn.loc.gov/2001031791.

106. Stephen C. Meyer, Signature in the Cell: DNA and the Evidence for Intelligent Design (HarperOne (New York), 2009); Stephen C. Meyer, Darwin’s Doubt: The Explosive Origin of Animal Life and the Case for Intelligent Design (HarperOne (HarperCollins), 2013).

107. Plantinga, Where the Conflict Really Lies, ch. 9.

108. William Paley, Natural Theology; or, Evidences of the Existence and Attributes of the Deity (R. Faulder, 1802).

109. Elliott Sober, Evidence and Evolution: The Logic Behind the Science (Cambridge University Press, 2008), https://doi.org/10.1017/CBO9780511806285.

110. Sober, Evidence and Evolution.

111. Dennis R. Venema and Scot McKnight, Adam and the Genome: Reading Scripture After Genetic Science (Brazos Press, 2017).

112. Alvin Plantinga, “Supralapsarianism, or ’o Felix Culpa’,” in In Peter van Inwagen (Ed.), Christian Faith and the Problem of Evil (Grand Rapids, MI: Eerdmans), 1–25 (2004), https://philpapers.org/rec/PLASOO.

113. A. Plantinga, The Nature of Necessity (Oxford University Press, 1974).

114. Charles Hartshorne, The Logic of Perfection and Other Essays in Neoclassical Metaphysics (Open Court (La Salle, IL), 1962), https://search.worldcat.org/title/710740.

115. Robert Merrihew Adams, “The Logical Structure of Anselm’s Arguments,” The Philosophical Review 80 (1): 28–54, 1971, https://doi.org/10.2307/2184310.

116. Graham Oppy, Ontological Arguments and Belief in God (Cambridge University Press, 1995).

117. Jordan Howard Sobel, Logic and Theism: Arguments for and Against Beliefs in God (Cambridge University Press, 2004).

118. Carlos van Hamme, “Why Modal Logic Cannot Ground a Necessary Personal God: Structural Limits of Modal Necessity and the Failure of Necessary Existence,” 2025, https://philarchive.org/rec/VANWML.

119. Alvin Plantinga, Warrant and Proper Function (Oxford University Press, 1993).

120. Brian Leftow, God and Necessity (Oxford University Press, 2012).

121. Alvin Plantinga, “Two Dozen (or so) Theistic Arguments,” in Appendix in Deane-Peter Baker (Ed.), Alvin Plantinga (Cambridge: Cambridge University Press), 203–227 (2007), https://www.cambridge.org/core/books/alvin-plantinga/appendix-two-dozen-or-so-theistic-arguments/484B8F7BD70B64055472A59FBEE5D6A1.

122. Thomas V. Morris and Christopher Menzel, “Absolute Creation,” American Philosophical Quarterly 23(4): 353–362, 1986, https://www.jstor.org/stable/20014160.

123. Cornelius Van Til, The Defense of the Faith (Presbyterian; Reformed Publishing Co., 1955).

124. Greg L. Bahnsen, Van Til’s Apologetic: Readings and Analysis (Presbyterian; Reformed (P&R Publishing), Phillipsburg, NJ, 1998), https://www.prpbooks.com/book/van-til-s-apologetic.

125. Morris and Menzel, “Absolute Creation.”

126. A. Plantinga, Does God Have a Nature? (Marquette University Press, 1980).

127. Leftow, God and Necessity.

128. Greg Welty, “Theistic Conceptual Realism,” in In Paul m. Gould (Ed.), Beyond the Control of God? Six Views on the Problem of God and Abstract Objects (New York: Bloomsbury Academic, 2014), Pp. 81–96 (2014).

129. Til, The Defense of the Faith.

130. Bahnsen, Van Til’s Apologetic.

131. Morris and Menzel, “Absolute Creation.”

132. Plantinga, Does God Have a Nature?

133. Leftow, God and Necessity.

134. Welty, “Theistic Conceptual Realism.”

135. William Lane Craig, God over All: Divine Aseity and the Challenge of Platonism (Oxford University Press, 2016), https://global.oup.com/academic/product/god-over-all-9780198786887.

136. René Descartes, Letters to Mersenne (April–May 1630), Letters to Mersenne, 15 April / 6 May / 27 May 1630 (AT I; trans. CSMK III), 1630.

137. Harry G. Frankfurt, “Descartes on the Creation of the Eternal Truths,” The Philosophical Review 86 (1): 36–57, 1977, https://doi.org/10.2307/2184161.

138. Edwin M. Curley, “Descartes on the Creation of the Eternal Truths,” The Philosophical Review 93 (4): 569–597, 1984, https://doi.org/10.2307/2184828.

139. Elliott Sober, Philosophy of Biology (Westview Press, 1993).

140. Gila Sher, Epistemic Friction: An Essay on Knowledge, Truth, and Logic (Oxford University Press, 2016).

141. Morris and Menzel, “Absolute Creation.”

142. Plantinga, Does God Have a Nature?

143. Leftow, God and Necessity.

144. Welty, “Theistic Conceptual Realism.”

145. Leftow, God and Necessity, chs. 10–12.

146. Plantinga, Does God Have a Nature?

147. Descartes, Letters to Mersenne (April–May 1630).

148. Frankfurt, “Descartes on the Creation of the Eternal Truths.”

149. Curley, “Descartes on the Creation of the Eternal Truths.”

150. Leftow, God and Necessity, ch. 4.

151. Sober, Philosophy of Biology.

152. Sher, Epistemic Friction.

153. Til, The Defense of the Faith.

154. Bahnsen, Van Til’s Apologetic.

155. Robert Merrihew Adams, Finite and Infinite Goods: A Framework for Ethics (Oxford University Press, 1999), https://doi.org/10.1093/0195153715.001.0001.

156. C. Stephen Evans, God and Moral Obligation (Oxford University Press, 2013), https://academic.oup.com/book/6961.

157. John E. Hare, The Moral Gap: Kantian Ethics, Human Limits, and God’s Assistance (Clarendon Press / Oxford University Press (Oxford), Oxford Studies in Theological Ethics, 1996), https://global.oup.com/academic/product/the-moral-gap-9780198269571.

158. Craig and William Lane, Reasonable Faith: Christian Truth and Apologetics (Crossway, 2008).

159. David Baggett and Jerry L. Walls, Good God: The Theistic Foundations of Morality (Oxford University Press, 2011); David Baggett and Jerry L. Walls, God and Cosmos: Moral Truth and Human Meaning (Oxford University Press, 2016).

160. George I. Mavrodes, “Religion and the Queerness of Morality,” in In r. Audi & w. J. Wainwright (Eds.), Rationality, Religious Belief, and Moral Commitment, Cornell University Press: 213–226 (1986), https://philpapers.org/rec/MAVRAT-2.

161. Mackie, The Miracle of Theism.

162. Erik J. Wielenberg, Robust Ethics: The Metaphysics and Epistemology of Godless Normative Realism (Oxford University Press, 2014).

163. Blaise Pascal, Pensées (Guillaume Desprez, 1670).

164. Jeff Jordan, Pascal’s Wager: Pragmatic Arguments and Belief in God (Clarendon Press / Oxford University Press (Oxford), 2006).

165. William James, “The Will to Believe,” The New World 5: 327–347, 1896, https://en.wikisource.org/wiki/The_Will_to_Believe.

166. Alan Hájek, “Waging War on Pascal’s Wager,” The Philosophical Review 112 (1): 27–56, 2003, https://doi.org/10.1215/00318108-112-1-27.

167. William G. Lycan and George N. Schlesinger, “You Bet Your Life: Pascal’s Wager Defended,” in In j. Feinberg (Ed.), Reason and Responsibility (Wadsworth) (1989), https://philpapers.org/rec/LYCYBY.

168. Plantinga, The Nature of Necessity.

169. John Duns Scotus, Philosophical Writings (Edited; translated by Allan Wolter. Indianapolis: Hackett, (c. 1300) 1987), https://search.worldcat.org/title/829102022.

170. James E. Dolezal, God Without Parts: Divine Simplicity and the Metaphysics of God’s Absoluteness (Pickwick Publications, 2011), https://search.worldcat.org/title/955548946.

171. Jeffrey E. Brower, “Simplicity and Aseity,” in In the Oxford Handbook of Philosophical Theology, Edited by Thomas p. Flint and Michael c. Rea, 105–128. Oxford: Oxford University Press (2009), https://doi.org/10.1093/oxfordhb/9780199596539.013.0006.

172. Eleonore Stump, Aquinas (Routledge, 2003).

173. Feser, Aquinas.

174. Gaven Kerr, Aquinas’s Way to God: The Proof in de Ente Et Essentia (Oxford University Press, 2015).

175. Pruss, The Principle of Sufficient Reason.

176. Leftow, God and Necessity.

177. Katherin A. Rogers, Perfect Being Theology (Edinburgh University Press, 2000).

178. Plantinga, Does God Have a Nature?

179. Christopher Hughes, On a Complex Theory of a Simple God: An Investigation in Aquinas’ Philosophical Theology (Cornell University Press, 1989).

180. William E. Mann, “Divine Simplicity,” Religious Studies 18 (4): 451–471, 1982, https://doi.org/10.1017/S0034412500014360.

181. Mullins, The End of the Timeless God.

182. Nicholas Wolterstorff, “Divine Simplicity,” Philosophical Perspectives 5: 531–552, 1991, https://doi.org/10.2307/2214108.

183. Dolezal, God Without Parts.

184. Brower, “Simplicity and Aseity.”

185. Stump, Aquinas.

186. Mullins, The End of the Timeless God.

187. Joseph C. Schmid, “From Modal Collapse to Providential Collapse,” Philosophia 50 (3): 1413–1435, 2022, https://doi.org/10.1007/s11406-021-00438-z.

188. Wolterstorff, “Divine Simplicity.”

189. Stump and Kretzmann, “Eternity.”

190. Feser, Aquinas.

191. Brower, “Simplicity and Aseity”; Stump, Aquinas.

192. Leftow, God and Necessity.

193. Vladimir Lossky, The Mystical Theology of the Eastern Church (James Clarke & Co., 1957).

194. David Bradshaw, Aristotle East and West: Metaphysics and the Division of Christendom (Cambridge University Press, 2004), https://www.cambridge.org/core/books/aristotle-east-and-west/51AB6B683806BF9A3969F990A49DC9C3.

195. Kenneth Seeskin, Searching for a Distant God: The Legacy of Maimonides (Oxford University Press, 2000); Josef Stern, The Matter and Form of Maimonides’ Guide (Harvard University Press, 2013).

196. Jean-Luc Marion, God Without Being: Hors-Texte (University of Chicago Press (trans. T. A. Carlson), 1991); Denys Turner, The Darkness of God: Negativity in Christian Mysticism (Cambridge University Press, 1995); Oliver Davies and Denys Turner, Silence and the Word: Negative Theology and Incarnation (Cambridge University Press, 2008), https://www.cambridge.org/core/books/silence-and-the-word/B99D0F676A0B5FF6C85FC88787F95F45.

197. Michael A. Sells, Mystical Languages of Unsaying (University of Chicago Press, 1994); Turner, The Darkness of God.

198. Sells, Mystical Languages of Unsaying.

199. Turner, The Darkness of God.

200. Stump and Kretzmann, “Eternity.”

201. Brian Leftow, Time and Eternity (Cornell University Press (Ithaca, NY & London), 1991).

202. Paul Helm, Eternal God: A Study of God Without Time (Clarendon Press / Oxford University Press (Oxford & New York), 1988), https://search.worldcat.org/title/18350008.

203. Feser, Aquinas.

204. Feser, Five Proofs of the Existence of God.

205. Yitzhak Y. Melamed, Spinoza’s Metaphysics: Substance and Thought (Oxford University Press, 2013).

206. Plotinus, The Enneads (Harvard University Press (Loeb Classical Library), c. 270 CE).

207. Lloyd P. Gerson, Plotinus (Routledge, 1994).

208. Eric D. Perl, Thinking Being: Introduction to Metaphysics in the Classical Tradition (Brill, 2014).

209. Hartshorne, The Divine Relativity.

210. Charles Hartshorne, A Natural Theology for Our Time (Open Court (La Salle, IL), 1967).

211. Whitehead, Process and Reality.

212. John B. Cobb Jr. and David Ray Griffin, Process Theology: An Introductory Exposition (Westminster Press (Philadelphia), 1976).

213. Marjorie Hewitt Suchocki, God-Christ-Church: A Practical Guide to Process Theology (Crossroad, 1982).

214. Marjorie Hewitt Suchocki, The End of Evil: Process Eschatology in Historical Context (State University of New York Press, 1988).

215. Joseph A. Bracken, The Triune Symbol: Persons, Process and Community (University Press of America (Lanham, MD), 1985).

216. Joseph A. Bracken, Society and Spirit: A Trinitarian Cosmology (Susquehanna University Press / Associated University Presses (Cranbury, NJ), 1991), https://search.worldcat.org/title/231401902.

217. Roland Faber, God as Poet of the World: Exploring Process Theologies (Westminster John Knox Press (Louisville), 2008), https://search.worldcat.org/title/761456566.

218. Thomas Jay Oord, The Uncontrolling Love of God: An Open and Relational Account of Providence (IVP Academic (InterVarsity Press), 2015).

219. Jr. and Griffin, Process Theology, ch. 2.

220. Pinnock et al., The Openness of God.

221. Hasker, God, Time, and Knowledge.

222. Nāgārjuna, Mūlamadhyamakakārikā: The Fundamental Wisdom of the Middle Way (Oxford University Press, (c. 2nd c. CE) 1995).

223. Jay L. Garfield, The Fundamental Wisdom of the Middle Way: Nāgārjuna’s Mūlamadhyamakakārikā (Oxford University Press (New York), 1995), https://global.oup.com/academic/product/the-fundamental-wisdom-of-the-middle-way-9780195093360; Jay L. Garfield, Empty Words: Buddhist Philosophy and Cross-Cultural Interpretation (Oxford University Press (New York), 2002), https://search.worldcat.org/title/45603979.

224. Mark Siderits, Personal Identity and Buddhist Philosophy: Empty Persons (Ashgate, 2003); Mark Siderits, Buddhism as Philosophy: An Introduction (Hackett / Aldershot: Ashgate, 2007).

225. Jan Westerhoff, Nāgārjuna’s Madhyamaka: A Philosophical Introduction (Oxford University Press, 2009).

226. Westerhoff, Nāgārjuna’s Madhyamaka, ch. 8.