The argument under discussion is not the familiar slogan "everything has a cause." Serious contemporary contingency arguments usually do not say that, and many theists explicitly deny it. The stronger version is framed in terms of metaphysical dependence or grounding: some things exist in virtue of, or because of, other things. A table depends on its parts; a set depends on its members; a mental state may depend on a physical state; a composite fact may depend on more basic facts. The claim is not temporal causation but ontological priority.
On this approach, defended in different forms by philosophers such as Alexander Pruss, Joshua Rasmussen, Robert Koons, and others, the world contains dependent realities. Dependent realities, it is argued, cannot be the whole story. If every dependent item gets its being or explanation from something else dependent, then the entire structure seems ontologically "needy." So the chain of dependence must terminate in something not dependent in the relevant way: a necessary, self-sufficient, ultimate ground. A further argument is then needed to identify that ground with anything recognizably theistic rather than with an impersonal necessary reality; the bare existence of a terminus would not by itself establish God.
The infinitist challenge
The objection is that the move from dependence to an ultimate ground relies on a controversial metaphysical principle: well-foundedness. In this context, well-foundedness says that grounding or dependence is not infinitely descending: any dependent item or fact is ultimately grounded in something that is not itself dependent in the same way. This need not mean that there is one simple bottom layer shared by everything; it may mean only that dependence terminates somewhere rather than being endlessly deferred. Without that principle, the argument has no guarantee of a stopping point.
Metaphysical infinitists about grounding deny that this is compulsory. Einar Duenger Bohn and Matteo Morganti are among the philosophers who have explicitly explored or defended the coherence of infinite metaphysical dependence; Ricki Bliss and others have also pressed worries about assuming that all regresses of ground are vicious. The broader literature on "gunk" and "junk" is relevant here. A world is often called gunky if every part has proper parts: there are no simples at the bottom. A world is called junky if every thing is a proper part of something larger: there is no all-inclusive whole at the top. These ideas are not usually presented as established descriptions of our world, but as models that challenge the assumption that metaphysical structure must terminate neatly.
The infinitist's point is modest but powerful: why think dependence must bottom out? Infinite sequences are not automatically contradictory. The natural numbers do not have a largest member. Theists themselves often accept an endless future. Many philosophers accept actual infinities in mathematics. So, the objector says, merely pointing to an infinite regress is not enough. One must show that this particular regress is vicious.
Stated at its strongest, the objection is this:
Grounding-based contingency arguments require the premise that chains of metaphysical dependence are well-founded. But well-foundedness is not a neutral axiom accepted by all sides. It is a substantive and disputed thesis in metaphysics. If infinite dependence is coherent, then dependent things may be grounded in further dependent things without any ultimate ground. The argument for a necessary foundation therefore fails unless the theist supplies an independent reason to reject infinite grounding.
This is a serious objection. It is not an internet-level dismissal of cosmological arguments. It targets one of the deepest structural assumptions in many refined versions of them.
What the objection gets right
The objection is right that well-foundedness cannot simply be treated as obvious. In contemporary metaphysics, grounding is widely discussed, but its formal principles are contested. Some philosophers think grounding is irreflexive, asymmetric, and well-founded. Others question one or more of these constraints. There is no universally accepted theory of grounding that a contingency argument can simply borrow without argument.
The objection is also right that not all regresses are vicious. Some infinite chains seem harmless. If every event has a later event, that does not obviously generate contradiction. If every integer has a predecessor or successor, mathematics does not collapse. So the critic is justified in demanding more than the bare phrase "infinite regress."
It is also important that the theist cannot infer a foundation merely from the fact that our explanations come to an end in practice. Human inquiry often stops for pragmatic reasons: limited evidence, limited interest, limited cognitive capacity. But a practical stopping point is not a metaphysical stopping point. To argue for an ultimate ground, one needs more than the psychology of explanation.
Where the infinitist case is less decisive
Still, the infinitist objection is not itself a refutation of grounding-based contingency arguments. It shows that one premise is controversial. It does not show that the premise is false.
The comparison with an endless future, for example, has limited force. An endless future is an infinite sequence ordered toward later and later times. A grounding chain is supposed to be an order of ontological priority: one thing obtains in virtue of another. These structures need not obey the same rules. A future event can occur after the present without making the present depend on it. But a grounded item is, by definition, metaphysically posterior to what grounds it. The worry about an infinite grounding regress is not merely that it is infinite; it is that nothing in the series ever possesses the relevant self-sufficiency that the members are said to lack.
A familiar analogy helps, though it should not be overused. Imagine an infinite series of borrowers, each with no money of his own, each borrowing from another. Does this explain how anyone has money? The infinitist will rightly say that analogies with money are causal and social, not metaphysical. But the analogy gestures at a real question: can derivative being be fully accounted for by endlessly more derivative being? The answer is not settled by noting that infinite series are coherent in mathematics.
Nor do gunk and junk, by themselves, settle the matter. If a gunky world is metaphysically possible, it shows at most that composition need not bottom out in indivisible particles. But grounding-based contingency arguments are not always arguments for simples. A theist can grant that physical or mereological structure might be infinitely divisible while still denying that ontological dependence as such can be infinitely deferred. Likewise, junk challenges the idea of an all-encompassing whole, but the contingency argument need not depend on a highest composite object. It may instead concern dependent facts, or the fact that dependent beings exist at all. The infinitist can then dispute whether there is a further "totality" fact needing explanation, but gunk and junk alone do not decide that dispute.
This means the infinitist needs to be careful. Showing that some infinite structures are coherent does not show that an infinite structure of total ontological dependence is coherent. The relevant question is not "Can there be an infinite chain?" but "Can there be an infinite chain where every member is dependent in the very respect that calls for grounding?"
The strongest foundationalist reply
The strongest reply, whether offered by a theist or by any defender of metaphysical foundationalism, is not to shout "regress!" but to defend a principle about explanation or dependence. It may be put like this:
If dependent reality forms an order in which every item is explained only by something equally dependent, then each local link may be accounted for while the existence of a derivative order as such remains unexplained. Infinite internal explanation would then not amount to ultimate explanation.
This reply appeals to a version of the Principle of Sufficient Reason, or to a more restricted principle: contingent or dependent reality requires an adequate ground. The point is not that every explanation must be psychologically satisfying. It is that, if dependence is genuinely derivative, adding infinitely many derivative links may still leave unanswered why there is any derivative order rather than none. That "if" matters: the restricted explanatory principle must be defended, not treated as self-evident.
Some defenders of fundamentality also argue that grounding is conceptually tied to priority, and that priority is harder to understand if there is no base. If grounding tells us what is more basic than what, then an infinite descent of ever more basic items may seem to prevent anything from being basic in the relevant sense. Infinitists can reply that priority can be comparative without requiring an absolute minimum, just as numbers can be ordered without a largest member. So the foundationalist should not claim that non-well-founded grounding is plainly incoherent; the better claim is that it makes the explanatory role of grounding more difficult to understand.
That is a real pressure point. The infinitist wants to keep the language of dependence while denying that dependence needs an independent terminus. The foundationalist asks whether this leaves us with dependence that never finally depends on anything non-dependent. If the whole chain is described as "in virtue of" further dependent items, but there is no ultimate item or fact in virtue of which the order obtains, the foundationalist will argue that the explanatory demand remains unmet. The infinitist's task is to explain why that demand is optional, excessive, or already satisfied by the infinite structure.
The best infinitist rejoinder
The infinitist can respond that this assumes precisely what is at issue. Why must explanation be ultimate in order to be adequate? In many domains, explanation is local and relational. One fact explains another, which is explained by another, and so on. There may be no final, system-external explanation, but it does not follow that the local explanations are defective. The infinitist can also insist that "adequate explanation" should not be defined in a way that simply builds in foundationalism.
The infinitist may also resist talk of "the collection" of all dependent things. Perhaps there is no totality that stands in need of one further explanation. Or perhaps the infinite structure itself is a legitimate explanatory pattern, not an unexplained heap. If every dependent item has a ground, the demand that the entire chain have an additional ground may be accused of adding a requirement the original principle did not justify.
This rejoinder has force. Many contingency arguments depend on a bridge from "each dependent thing has a ground" to "the whole dependent structure has a ground." That bridge cannot be assumed without defense. Philosophers have long warned against illicitly transferring features of members to a collection. Even if every member of a series is explained, it does not automatically follow that the series as a whole needs the same kind of explanation.
Still, the rejoinder also has a cost. If the infinitist says that local explanation is enough, then the dispute shifts to the standards of metaphysical explanation. The theist will say that the original question was never merely whether each item has a predecessor, but why there is any dependent order at all. The infinitist may reject that question as illegitimate or unanswerable, but that is a substantive metaphysical stance, not a neutral default.
What this establishes
The infinitist objection establishes something important: grounding-based contingency arguments cannot simply help themselves to well-foundedness. If the argument requires the claim that dependence must bottom out, that claim needs defense. It is not a deliverance of logic alone, nor is it a consensus result of contemporary metaphysics.
But the objection establishes less than some popular presentations suggest. It does not show that infinite metaphysical dependence is possible in the relevant sense. It does not show that the demand for an ultimate ground is confused. It does not show that naturalism is true, or that theism is false. It shows that one route from dependent reality to an ultimate ground is blocked unless further argument is supplied.
The most reasonable verdict is therefore conditional. If well-foundedness is false, or if infinite grounding chains can provide adequate metaphysical explanation, then many grounding-based contingency arguments fail at a central step. If, however, dependence is the kind of relation that requires a non-dependent terminus, then the infinitist alternative is not enough to block a move to some ultimate ground. Even then, theism has not yet been established. A further argument would be needed to show that the terminus is one, necessary, explanatorily ultimate, and personal or otherwise God-like, rather than an impersonal fundamental fact, law, field, structure, or plurality of such grounds.
The debate is not about whether one can imagine "turtles all the way down." One can. The question is whether an endless hierarchy of dependent turtles explains why there are turtles at all, or whether it merely postpones the issue indefinitely. On that question, neither side is entitled to victory by stipulation.