In the discussion of whether empirical evidence is a valid demand in debates about the existence of God, a crucial question arises: Is asking for such evidence a category mistake? The answer depends partly on what is meant by "empirical evidence." A crude demand that God be detected like a planet, molecule, or organism may misunderstand classical theism. But a more serious atheist demand, found in different forms in writers such as Antony Flew, J. L. Mackie, and Graham Oppy, asks whether theism makes any difference to what we should expect in the world, and whether the total evidence is better explained by theism or by naturalism. This article explores the complexities of this issue, aiming to distinguish between a legitimate philosophical point and a potential unfalsifiability dodge.
The Nature of Category Mistakes
Gilbert Ryle popularized the idea of category mistakes in his seminal work, "The Concept of Mind." A category mistake occurs when things are presented as if they belong to one logical type or category, but actually belong to another. An example Ryle famously gives is that of a visitor to a university who, after being shown the various colleges, libraries, and laboratories, asks, "But where is the university?" The visitor's mistake is not that the university is unreal, but that he is looking for it as one more building alongside the others rather than as an institution organized through them.
Applied carefully, the category-mistake point says this: if God is conceived as a non-physical, necessary being rather than as one object inside the universe, then demanding direct sensory observation of God may be like asking for the weight of justice or the color of freedom. However, this does not show that demands for evidence in general are mistaken. Many rational beliefs are supported indirectly: by explanatory power, coherence, metaphysical argument, or by observable consequences. So the category-mistake objection works only against a narrow demand for direct physical detection, not against every demand for reasons.
The Concept of Necessary Existence
Ibn Sina provides an important historical context for this discussion with his concept of necessary existence. For Ibn Sina, a necessary being is one whose non-existence is impossible, while contingent beings exist but could have failed to exist. His argument does not begin by assuming the God of any scripture; it begins from a metaphysical distinction between what exists through itself and what depends on another. Ibn Sina then argues that the Necessary Existent is not contingent, composite, or dependent in the way physical objects are. Whether that argument succeeds is a further philosophical question, but it is not a request to find God as one more item inside the cosmos.
From this perspective, demanding that a necessary being be isolated and measured like a contingent object may misunderstand the claim being made. But the burden does not disappear. The theist still has to argue that contingent reality requires a necessary explanation, that the proposed necessary reality is coherent, and that alternative views - such as brute contingency, a necessary cosmos, or a naturalistic metaphysics - are less adequate. A category mistake in the demand for direct measurement is not itself a proof of theism.
Does This Constitute an Unfalsifiability Dodge?
While the argument that demanding direct empirical observation is a category mistake may hold philosophical merit, it is also important to consider whether this becomes an unfalsifiability dodge. Falsifiability is most at home as a criterion for scientific hypotheses, not as a universal test for all meaningful claims; mathematical, moral, and metaphysical claims are not usually falsified in a laboratory. Still, if a religious or philosophical claim entails that certain things should or should not be observed, then those implications can be assessed. The issue is not whether every claim must be scientific, but whether a claim is being protected from any possible rational pressure.
The strongest atheist worry is not merely that God is called non-empirical. It is that some theistic claims seem to have implications for the observable world - for example, claims about providence, miracles, answered prayer, moral order, or the amount and distribution of suffering - yet retreat into non-empirical language when those implications are challenged. The evidential problem of evil is especially serious here: if theism includes a perfectly good, all-powerful, and all-knowing creator, then apparently pointless suffering is not a trivial objection. A category-mistake reply does not by itself answer that challenge.
Distinguishing the Legitimate from the Dodge
To distinguish between a legitimate category-mistake argument and an unfalsifiability dodge, one should focus less on a speaker's intent and more on the claim's content and evidential commitments. A legitimate argument will define the concept of God independently, explain why direct physical detection is the wrong category of test, and then offer some other rational route: metaphysical argument, explanatory comparison, conceptual analysis, or cumulative-case reasoning. It should also say what would count against it: incoherence in the concept of God, failure of key metaphysical premises, stronger rival explanations, or evidence that sits badly with the attributes being defended.
Conversely, if the claim is used selectively to avoid inconvenient questions while still making empirical claims about the world, it becomes suspect. A thinker cannot reasonably say both that no empirical consideration can ever matter and that the world contains observable signs, interventions, or patterns that support theism. Once a theistic worldview appeals to features of the world as evidence, it must also allow contrary features of the world to have evidential weight. The key is whether the concept of God as a necessary being is used consistently, or whether it is invoked only when criticism becomes difficult.
Conclusion
Demanding empirical evidence for God raises complex philosophical issues that touch on the nature of existence and the limits of empirical inquiry. It may be a category mistake to treat God, as understood in classical theism, as an empirical object alongside other objects. But it is not a category mistake to ask for reasons, arguments, explanatory advantages, or engagement with evidence that appears to count against theism. The strongest position is therefore modest: direct measurement may be the wrong test for a necessary, non-physical reality, but theistic belief still requires rational support and remains open to philosophical and evidential criticism.