One serious objection to cosmological arguments is not simply that we have never observed a "necessary connection" between cause and effect. That would be the familiar Humean point: experience gives us regular succession, not a metaphysical glue. The stronger objection is eliminativist. It says that mature physics has no need for causes at all, at least not as fundamental furniture of the world. If cosmological arguments say that the universe "must have a cause," they are relying on a pre-scientific picture of causes as productive pushings, where one event precedes another and brings it about. Physics, the objection continues, replaced that picture with equations relating states. The word "cause" may remain useful in everyday life, engineering, medicine, and the special sciences, but it does no deep metaphysical work.
The argument in its strongest form
Bertrand Russell gave the classic early version of this complaint in "On the Notion of Cause." Writing before general relativity, quantum mechanics, and modern quantum field theory, Russell argued that the old "law of causality" had largely been displaced in advanced science by functional and differential equations. The point can be extended, with qualifications, to later physics: Newtonian and Hamiltonian mechanics, Maxwell's equations, general relativity, and the unitary Schrodinger equation are not usually written as "A causes B." They relate variables, states, fields, or geometrical structures. Where the dynamics is reversible, a complete later state may in principle allow retrodiction of an earlier state just as an earlier state allows prediction of a later state. The equation itself does not say that the earlier state metaphysically produces the later one.
This point has been developed in different ways by philosophers such as Huw Price and John Norton. Price has argued that our causal asymmetries are closely connected to our temporally asymmetric standpoint as agents embedded in a low-entropy universe. Norton has argued that causation is not a fundamental principle of science but a useful, context-sensitive concept that functions more like folk science than basic ontology. On this view, causal talk is often legitimate in practice, but it should not be inflated into a universal metaphysical law.
The eliminativist argument can be stated as follows:
- Fundamental physics describes the world through mathematical relations between physical states, not through asymmetric productive relations called "causes."
- Many central dynamical equations are time-reversal invariant, or at least mathematically usable in both temporal directions under suitable conditions: given an appropriate complete state, the formalism may allow retrodiction as well as prediction. This must be stated carefully, because quantum measurement, thermodynamic boundary conditions, and known T-violating weak interactions complicate any simple claim that all fundamental physics is time-symmetric.
- The asymmetry of ordinary causation - cause first, effect later - is therefore not built into the deepest physical description.
- At or near any proposed beginning of the universe, ordinary temporal and causal intuitions are especially suspect, since time itself may be part of the system being described and some cosmological models do not contain a classical first instant at all.
- Therefore, cosmological arguments that say the universe must have a cause rest on a folk notion that mature physics has discarded.
This is stronger than saying, "We have not proved that the universe has no cause." It says that the demand for such a cause may itself be confused, like demanding the location of the whole of space or asking what happened before time if "before" already presupposes time.
What the objection gets right
The objection has real force against some popular presentations of cosmological arguments. When someone says, "Everything has a cause, so the universe has a cause," the argument is vulnerable on several fronts. First, the premise is not one that physics states. Second, it is not even what many sophisticated cosmological arguments claim, since they do not say that everything has a cause. Professional critics of theism and cosmological arguments, such as J. L. Mackie and Graham Oppy, do not usually treat this slogan as the strongest target; they engage more careful principles about beginnings, contingency, dependence, or sufficient reason. Third, if "cause" means a prior physical event inside time, then applying it to the origin of time is at least unclear.
The objection is also right to warn against picturing divine causation, or any ultimate cause, as though it were one billiard ball striking another. That model belongs to ordinary experience within the universe. It should not be casually extended to the universe as a whole. Even in ordinary physics, causation is not always a simple matter of one localized object pushing another. Fields, constraints, global boundary conditions, quantum correlations, and spacetime geometry complicate the picture.
Russell's criticism is also a useful correction to a certain apologetic habit: treating common-sense metaphysical intuitions as if they were direct deliverances of science. Physics does not say, in any simple way, "whatever begins to exist must have a cause." Nor does contemporary cosmology require philosophers to speak in the language of production, priority, and necessary connection. A serious cosmological argument must earn those concepts; it cannot merely assume them.
Where the objection overreaches
Still, the eliminativist conclusion is stronger than the evidence supports if it moves from "causation is not fundamental" to "causation is meaningless or intellectually illegitimate." A careful eliminativist may only be denying that causation belongs in the most basic ontology. The reply is that nonfundamental does not automatically mean illusory. Many concepts do not appear in fundamental equations but remain indispensable at other levels of explanation: temperature, computation, genes, organisms, markets, and reasons. Few people infer that because "gene" is not a term in quantum field theory, genetics rests on a primitive illusion.
The eliminativist can reply that causation is different because it pretends to describe a fundamental metaphysical relation. But that is not always how causal language functions. In the philosophy of science, accounts of causation associated with thinkers such as James Woodward, Nancy Cartwright, Wesley Salmon, and Phil Dowe do not simply revive the crude billiard-ball picture. Woodward's interventionist account, for example, analyzes many causal claims in terms of what would change under suitable interventions. Cartwright emphasizes the capacities, powers, and model-specific causal claims used in actual scientific modeling. Salmon's causal-mechanical work emphasized causal processes, interactions, and the transmission of marks; conserved-quantity accounts are associated especially with Dowe and influenced later causal-mechanical debates. These accounts are debated, but they show that serious work on causation is not merely folk superstition.
Nor is physics as cleanly non-causal as the slogan suggests. Relativity theory has light cones and a technical notion of causal structure, distinguishing which events can be connected by timelike or lightlike signals from those that are spacelike separated. Quantum field theory uses microcausality conditions, roughly requiring appropriate observables at spacelike separation to commute, so that the theory does not permit controllable faster-than-light signalling. These are not the old metaphysical "necessary connections" that Hume doubted, and they do not prove that causation is fundamental. They do show, however, that mature physics has not simply expelled every causal or influence-like constraint. Even where dynamical equations are time-reversal invariant, their physical use often depends on boundary conditions, thermodynamic asymmetries, and explanatory practices not erased by formal symmetry alone.
There is also a distinction between temporal causation and explanatory dependence. A cosmological argument need not claim that the cause of the universe is an earlier event. Arguments associated with Avicenna and Leibniz, for example, are usually read as contingency or dependence arguments rather than as claims that the universe was preceded by another event. Some Thomistic arguments are likewise about a here-and-now order of dependence, not merely a first moment in the past. Whether those arguments succeed is another matter. They require controversial premises about contingency, dependence, or sufficient reason. But they are not refuted simply by noting that earlier-later causation may not apply at the origin of time.
This matters because the eliminativist critique sometimes attacks a weaker target than the strongest cosmological arguments. Professional defenders of cosmological arguments often know that temporal priority cannot be assumed if time itself is under discussion. Contemporary defenders such as Alexander Pruss and Joshua Rasmussen, for instance, typically frame central arguments in terms of contingency, explanation, or dependence; William Lane Craig's kalam argument is more explicitly about a beginning, but even it needs a non-question-begging account of what "cause" means when time itself is at issue. These notions are controversial, but they are not identical to the folk picture of one event knocking another into existence.
The strongest eliminativist reply
The best reply from the eliminativist is that this move merely changes the label. If "cause" is replaced with "explanation," "ground," or "dependence," we should ask whether these notions fare any better. Are they part of fundamental physics? Are they objective relations in the world, or projections of our explanatory interests? If the universe is described by a complete physical model, perhaps there is no further fact called its "sufficient reason." Perhaps asking for an external explanation of the total physical order is like asking for a physical location outside all physical space.
This reply is serious, and it succeeds against any cosmological argument that simply assumes a broad explanatory principle without defending it. The principle of sufficient reason is not a theorem of physics. It is a metaphysical commitment, and many philosophers reject unrestricted versions of it because they seem too strong: they appear to rule out brute facts in advance and may create difficult questions about the explanation of the explanatory principle itself, or about why the ultimate explanation is as it is. More modest versions of the principle may be defensible, but then the argument must explain why that version applies to the total physical order rather than only to facts within it.
The eliminativist can also point to quantum cosmology and early-universe models where the ordinary notion of a beginning is not straightforward. Some models describe the universe without a classical first instant; others treat time as emergent or as behaving differently near the Planck scale. It would be irresponsible to use these open research areas as proof that the universe is uncaused. But it is equally irresponsible to speak as if contemporary cosmology straightforwardly confirms ordinary causal intuitions.
What this critique establishes
The eliminativist objection establishes a limited but important conclusion. It shows that naive causal formulations of cosmological arguments are inadequate. If an argument depends on the premise that every beginning must have a prior producing cause, it owes us an account of what that means when applied to the origin of time, and why physics should respect that premise. It also shows that one cannot simply move from everyday causal habits to a First Cause without serious philosophical work.
But the objection does not establish that all cosmological arguments fail. It does not show that reality has no ultimate explanation. It does not show that dependence, contingency, or intelligibility are empty notions. It does not show that theism is false. At most, it removes one crude route to theism: the route that treats the universe as just another object inside the universe needing an earlier shove.
The deepest issue is not whether the word "cause" appears in fundamental equations. It is whether the existence and structure of the physical order call for an explanation beyond the order itself, and what kind of explanation could be coherent. The Russellian critique forces that question to be asked more carefully. It does not answer it by itself.
A fair verdict, then, is this: mature physics does undermine simplistic causal rhetoric, and cosmological arguments must not rely on a folk mechanics of cosmic production. However, the absence of causes from some formulations of fundamental physics is not enough to eliminate all metaphysical explanation. The serious debate begins after the slogan has done its useful work of clearing away the naive picture.