Infinite regress

An infinite regress is a chain of reasoning in which each element is explained, caused, or justified by appeal to a prior element, and this process continues without terminus — there is no first or foundational member that anchors the chain. Infinite regress arguments have been among the most powerful tools in the philosopher’s repertoire since antiquity, employed in metaphysics to argue for the existence of a first cause or necessary being, in epistemology to challenge the possibility of justified belief, and in philosophy of religion to underwrite several of the classical arguments for God’s existence. The fundamental question posed by any infinite regress is whether the endless chain is genuinely explanatory — whether an infinite series of dependent explanations can constitute a complete explanation — or whether the regress demonstrates that some independent, self-explanatory foundation is required.^{1, 12}

The philosophical significance of infinite regress lies in its recurrence across multiple domains. The same structural problem appears in discussions of causation (what caused the universe?), epistemic justification (what justifies this belief?), moral grounding (what makes this principle binding?), and ontological dependence (what grounds the existence of contingent things?). In each domain, the regress threatens to undermine the possibility of a satisfactory explanation unless some principled stopping point can be identified.¹⁴

Types of regress

Philosophers distinguish several varieties of infinite regress, each involving a different kind of dependence relation. A causal regress is a chain in which each event is caused by a prior event: event A is caused by event B, which is caused by event C, and so on without end. A justificatory or epistemic regress is a chain in which each belief is justified by appeal to a prior belief: belief P is justified by belief Q, which is justified by belief R, with no terminal belief that is self-justifying. An explanatory regress is a chain in which each fact is explained by a prior fact, with no self-explanatory fact at the base. A constitutive regress concerns the composition or grounding of entities: if every entity is constituted by prior entities, and those by still prior entities, the chain of constitution has no foundation.^{12, 14}

Not all infinite regresses are considered problematic. Some philosophers distinguish between vicious and benign regresses. A vicious regress is one in which the completion of the chain is a necessary condition for the item under consideration to have the property in question (existence, justification, explanation), so that the impossibility of completing an infinite chain renders the property impossible. A benign regress, by contrast, is one that can be acknowledged without threatening the property in question — for example, the regress of natural numbers (every number has a successor) is infinite but does not prevent any particular number from being well-defined. The distinction between vicious and benign regresses is itself a matter of philosophical dispute, with disagreement over what makes a regress vicious rather than merely infinite.^{10, 12}

Aristotle’s argument against actual infinities

Aristotle (384–322 BCE) is the earliest philosopher to develop a systematic argument against infinite causal regresses. In the Physics and the Metaphysics, Aristotle argues that every chain of movers (causes of motion) must terminate in an unmoved mover — a being that causes motion without itself being moved. His argument rests on the claim that an infinite regress of movers would leave the entire series without an adequate explanation for why motion exists at all: if every mover is itself moved by a prior mover, then no member of the series is the source of motion, and the series as a whole has no source.^{1, 2}

Aristotle distinguished between potential and actual infinity. A potential infinity is a quantity that can always be extended but is never actually infinite at any given moment — for example, the natural numbers are potentially infinite because one can always count higher, but no one has counted to infinity. An actual infinity, by contrast, would be a completed collection of infinitely many elements existing simultaneously. Aristotle accepted potential infinities but rejected actual infinities as metaphysically impossible. This distinction became foundational for later arguments against infinite causal regresses: if the causal history of the universe constitutes an actual infinity of past events, and actual infinities are impossible, then the universe must have had a beginning, and therefore a first cause.^{1, 4}

Infinite regress in the cosmological arguments

The infinite regress occupies a pivotal role in the classical cosmological arguments for God’s existence. Thomas Aquinas (1225–1274), in the first three of his Five Ways, explicitly argues that an infinite regress of efficient causes, movers, or contingent beings is impossible, and concludes that there must be a first cause, a first mover, and a necessary being — all of which he identifies with God. In the Second Way, Aquinas writes: “In efficient causes it is not possible to go on to infinity, because in all efficient causes following in order, the first is the cause of the intermediate cause, and the intermediate is the cause of the ultimate cause… Now to take away the cause is to take away the effect. Therefore, if there be no first cause among efficient causes, there will be no ultimate, nor any intermediate cause.”³

P1. Every effect has an efficient cause.

P2. An infinite regress of efficient causes is impossible.

P3. Therefore, there must be a first efficient cause.

C. This first cause is what everyone calls God.

The Kalām cosmological argument, revived by William Lane Craig in 1979, places the impossibility of infinite regress at the centre of its case for the universe’s beginning. Craig argues that an actually infinite number of past events is metaphysically impossible because actual infinities lead to absurdities (such as those illustrated by Hilbert’s Hotel), and that a beginningless series of past events would constitute an actual infinity. Therefore, the series of past events must be finite, the universe must have begun to exist, and whatever begins to exist has a cause — a cause Craig identifies as God.⁴

The “turtles all the way down” metaphor

The most famous popular illustration of infinite regress is the “turtles all the way down” metaphor, which has circulated in various forms since at least the nineteenth century and is often attributed (apocryphally) to a dialogue involving either Bertrand Russell or William James. In the standard version, a philosopher or scientist explains that the Earth is a sphere orbiting the Sun, to which an elderly listener objects that the world actually rests on the back of a giant turtle. When asked what the turtle stands on, the listener replies: “It’s turtles all the way down.” The metaphor captures the intuitive dissatisfaction with infinite regress: each level of explanation merely pushes the question back one step without ever providing a satisfying foundation, yet the interlocutor is content with the infinite chain as a complete answer.^{5, 6}

The metaphor is philosophically illuminating because it highlights the tension between two competing intuitions. On one hand, many people find it deeply unsatisfying that the explanatory chain never reaches a foundation — it seems that something must ultimately hold everything up. On the other hand, if each turtle is fully supported by the turtle below it, one might argue that every member of the series is adequately explained, and the demand for something beyond the series to explain the whole is unjustified. This tension lies at the heart of the debate over whether infinite causal regresses are genuinely explanatory.

Agrippa’s trilemma

In epistemology, the infinite regress of justification generates a problem known as Agrippa’s trilemma (also called Münchhausen’s trilemma or the Agrippan trilemma), named after the ancient Pyrrhonian sceptic Agrippa, whose arguments are preserved in Sextus Empiricus’s Outlines of Pyrrhonism. The trilemma observes that when asked to justify any belief, one faces exactly three options, each of which is apparently unsatisfactory: (1) the justification regresses infinitely, with each justifying belief itself requiring justification from a prior belief, and no end is ever reached; (2) the justification is circular, with the chain eventually looping back to the original belief, such that A justifies B, B justifies C, and C justifies A; or (3) the justification terminates at some point in a belief that is accepted without justification — an arbitrary stopping point.⁸

Each horn of the trilemma has generated a corresponding theory of epistemic justification. Foundationalism, the historically dominant response, accepts option (3) but denies that the stopping point is arbitrary: foundationalists hold that certain beliefs are “basic” or self-justifying — perceptual beliefs, self-evident logical truths, or incorrigible beliefs about one’s own mental states — and that all other beliefs are justified by chains of inference that ultimately rest on these foundational beliefs. Coherentism, associated with philosophers such as BonJour, accepts a version of option (2) but denies that the circularity is vicious: coherentists hold that beliefs are justified not by linear chains of inference but by their membership in a coherent system of mutually supporting beliefs, where coherence itself provides the justificatory ground. Infinitism, most notably defended by Peter Klein, accepts option (1) and argues that an infinite chain of reasons can provide genuine justification, provided that the chain is non-repeating and that each reason is available to the epistemic agent upon reflection.^{9, 10, 14}

Responses to the infinite regress in metaphysics

The major responses to the causal infinite regress in metaphysics can be grouped into three broad categories. The first, championed by Aristotle, Aquinas, and Craig, holds that an infinite regress of causes is impossible, and therefore a first cause or necessary being must exist to terminate the chain. The arguments for impossibility vary: Aristotle appeals to the metaphysical impossibility of actual infinities; Aquinas argues that without a first cause, there would be no intermediate or ultimate causes; and Craig argues that an actually infinite number of past events leads to paradoxes (such as Hilbert’s Hotel, in which a hotel with infinitely many occupied rooms can always accommodate new guests by shifting every existing guest to the next room).^{3, 4, 11}

The second response, associated primarily with David Hume, holds that an infinite regress of causes is unproblematic because each member of the series is adequately explained by its predecessor, and the demand for an explanation of the series as a whole is unjustified. In Part IX of the Dialogues Concerning Natural Religion, Hume’s character Cleanthes argues that if each member of the causal series has an explanation (namely, the preceding member), then every member is explained, and there is nothing left over to explain. The demand for a cause of the entire series, over and above the causes of its individual members, commits what Hume’s character suggests is a version of the fallacy of composition — the error of assuming that what is true of each part must be true of the whole.⁵

Mackie elaborates on Hume’s point: “If we explain each particular event by reference to a prior event, then the series of events is fully explained. To ask for a further explanation of why there is a series at all is to demand an explanation of a different kind — an explanation that goes beyond the causal relations within the series to ask why there is something rather than nothing. But it is not obvious that this question has an answer, or that the absence of an answer is an intellectual deficiency.” On this view, the infinite causal regress is not a problem to be solved but simply a description of how things are, and the demand for a first cause reflects a philosophical preference rather than a logical necessity.⁶

The third response, which may be called the brute fact approach, holds that the existence of the causal series (or the universe, or the fundamental laws of nature) is a brute fact that simply has no explanation. On this view, the principle of sufficient reason — the claim that everything has an explanation for its existence — is not a necessary truth but an assumption that may be false. The universe may simply exist without a reason, and the demand for a complete explanation of why there is something rather than nothing may reflect an unjustified assumption that reality is fully intelligible. Pruss has argued extensively against this view, contending that accepting brute facts undermines the possibility of explanation more broadly, but critics maintain that the costs of accepting a brute fact are less severe than the costs of accepting an unexplained necessary being.^{12, 13}

Mathematical versus metaphysical infinity

The debate over infinite regress intersects with the broader question of whether actual infinities are possible. Modern mathematics, since Georg Cantor’s work in the late nineteenth century, treats actual infinities as well-defined mathematical objects: the set of natural numbers is actually infinite, the real number line contains uncountably many points, and the theory of transfinite cardinal numbers provides a rigorous arithmetic of infinities. In this mathematical framework, there is nothing paradoxical about actual infinities — they are the foundation of modern set theory, analysis, and topology.¹¹

Craig and other defenders of the Kalām argument respond by distinguishing between mathematical and metaphysical infinity. They accept that actual infinities are mathematically consistent but deny that they can be instantiated in the physical world. The paradoxes of Hilbert’s Hotel — a hotel with infinitely many rooms that can accommodate infinitely many new guests without evicting anyone — are mathematically valid deductions from the axioms of set theory, but Craig argues that they describe situations that are metaphysically absurd. A real hotel that could accommodate new guests in the way Hilbert’s Hotel does would violate our deepest intuitions about the relationship between parts and wholes, and this absurdity is evidence that actual infinities cannot exist in the real world.^{4, 11}

Morriston and Oppy have challenged this distinction. Morriston argues that the supposedly absurd consequences of actual infinities are not genuinely absurd but merely counterintuitive, and that our intuitions about finite collections should not be expected to apply to infinite ones. One-to-one correspondence between a set and a proper subset (such as the natural numbers and the even numbers) is a defining feature of infinite sets, not a paradox to be avoided. If actual infinities are logically consistent — as the mathematical framework demonstrates — then the claim that they are metaphysically impossible requires a justification beyond appeals to intuition, and such justification has not been provided. Oppy similarly argues that the “absurdity” of Hilbert’s Hotel reflects the unfamiliarity of infinite arithmetic rather than any genuine impossibility, and that a consistent story can be told about a world containing actual infinities without contradiction.^{7, 11, 15}

Contemporary significance

The infinite regress remains one of the most active areas of inquiry in contemporary philosophy. In metaphysics, the debate over whether the universe requires a first cause or a necessary being to terminate the causal regress continues to generate substantial literature, with the Kalām argument, the Leibnizian cosmological argument, and their various criticisms occupying a central place in the philosophy of religion. In epistemology, the trilemma remains a foundational challenge, and the debate between foundationalism, coherentism, and infinitism shows no sign of resolution. In the philosophy of science, the question of whether fundamental physical laws require explanation or constitute brute facts is a version of the same structural problem.^{10, 12, 14}

The enduring power of the infinite regress as a philosophical tool lies in its capacity to reveal hidden assumptions about explanation, dependence, and completeness. To ask whether an infinite chain of explanations is satisfying is ultimately to ask what counts as a genuine explanation — whether explanation is a local relation between adjacent items in a chain or a global property that requires the chain as a whole to be grounded in something outside it. This question, first raised by Aristotle and still debated twenty-four centuries later, shows no sign of being resolved, suggesting that it touches on something fundamental about the structure of human reasoning about the world.^{1, 5, 12}