Argument

Argument from Science as Faith-Based (Guillen)

Intro

The standard contemporary atheist rhetorical posture treats science as the "evidence-only" epistemology and religion as the "faith-based" epistemology. The asymmetry is supposed to settle who has the epistemic high ground: science / atheism wins by default. Michael Guillen's first argument in Believing Is Seeing (Tyndale Refresh, 2021) is a meta-epistemological defeater of this asymmetry. Science itself rests on unprovable starting assumptions that no scientific procedure can establish. Atheism, as a worldview, similarly rests on unprovable starting assumptions. The science-vs-faith framing is not a difference in kind; it is a difference in advertised self-description.

This page lays out the argument in debate-prep shape.

In full

The argument: "Science depends on five unprovable axioms (the existence of an external reality, the rationality of that reality, the uniformity of nature, the reliability of the senses, and the trustworthiness of human reason). These cannot be proved scientifically; any scientific attempt to prove them must assume them. They are accepted on faith, in the strict sense of 'trust in an unprovable proposition.' Therefore science is faith-based at its foundations. Atheism, as a worldview, rests on additional unprovable starting assumptions (e.g., materialism, the rejection of supernatural causation, the meaning-equivalence of subjective experience to neural activity). The atheist's claim that science is evidence-only while religion is faith-only is a category mistake. Both rest on faith; the question is which faith fits the evidence we then accumulate."

The pattern is the same Polanyi-Kuhn-Plantinga critique of naive empiricism applied to the contemporary atheist debate.

Argument structure

#	Premise
P1	Every belief system, including science, begins from axioms that cannot be proved from within the system. (Gödel's incompleteness theorems formalize this for mathematics; the regress argument formalizes it for epistemology generally.)
P2	Science depends on at least five unprovable axioms: (a) an external reality exists, (b) that reality is rationally structured, (c) the laws of nature are uniform across space and time, (d) the human senses reliably contact reality, (e) human reason reliably draws conclusions from sensory data.
P3	These five axioms cannot be established by scientific method without circularity (any test presupposes them). They are accepted on faith, in the sense of trust in an unprovable proposition.
P4	Atheism, as a worldview, adds further unprovable axioms beyond the scientific ones: materialism (only physical things exist), naturalism (no supernatural causation), and an account of consciousness that reduces subjective experience to neural activity.
P5	Therefore the atheist's framing of "science is evidence-only, religion is faith-only" is a category mistake. Both rest on faith at the foundation; the question is which set of foundational faiths best fits the cumulative evidence.
C	The science-vs-faith asymmetry collapses. Once the asymmetry collapses, the meta-debate shifts from "which side has the evidence" to "which set of foundational commitments best accommodates the cumulative discoveries of physics, cosmology, neuroscience, and mathematics."

Per-premise affirmative case

P1, every belief system starts from unprovable axioms

The regress argument for epistemological foundationalism: every justification for a belief either bottoms out in unjustified beliefs, regresses infinitely, or runs in a circle. Foundationalism takes the first option (some beliefs are properly basic); coherentism takes the third (justified beliefs form a self-supporting web). Both positions concede that pure provability-from-scratch is impossible. Gödel's incompleteness theorems (1931) formalize the same point for mathematics: any sufficiently rich formal system contains true propositions it cannot prove from within. Science inherits the limitation.

P2 and P3, the five scientific axioms

The five axioms of scientific practice that no experiment can establish:

An external reality exists. Solipsism (the position that only one's own mind exists) cannot be empirically refuted; science assumes its falsity as a starting condition.
The external reality is rationally structured. The natural world can be described by mathematics, logic, and conceptual analysis. There is no a priori reason this should be so; on a purely random universe it is not predicted. Yet science presupposes it.
The laws of nature are uniform across space and time. The principle of induction (that future observations will resemble past ones) cannot be justified non-circularly (Hume's problem of induction). Science assumes uniformity to function.
The human senses reliably contact reality. That the experience of seeing a tree corresponds to a tree-in-itself is a metaphysical commitment, not an empirical finding. Berkeleyan idealism cannot be empirically refuted.
Human reason reliably draws conclusions from sensory data. That the cognitive faculties that produce our beliefs are truth-conducive requires a further metaphysical posit. Plantinga's evolutionary argument against naturalism (1993, 2002, 2011) develops this: if our cognitive faculties evolved purely by survival-and-reproduction selection, there is no reason they should be truth-conducive at the level of abstract scientific reasoning.

P4, atheism adds further unprovable axioms

Atheism's additional axioms beyond science's five:

Materialism / physicalism: only physical / material entities exist.
Naturalism: no supernatural causation operates in the world.
Mind-brain identity or reduction: subjective conscious experience is identical with, or reducible to, neural activity.
Methodological naturalism as exhaustive: the scientific method, restricted to natural-causation explanations, exhausts the legitimate epistemological options.

None of these can be established scientifically. They are philosophical commitments the atheist makes prior to engaging the evidence.

P5, the asymmetry is rhetorical, not substantive

The popular New Atheist framing of science as evidence-only and religion as faith-only is a rhetorical move, not an accurate description of the underlying epistemology. Both science and theism start from foundational commitments; both interpret evidence in the light of those commitments. The legitimate question is not "which side has the evidence?" (both sides do, with their respective interpretive frames), but "which set of foundational commitments best accommodates the cumulative discoveries of modern science?"

Anticipated objections and rebuttals

Objection 1: "The five 'axioms' are pragmatically self-validating because they work; that's the only kind of justification science needs."

Rebuttal. Pragmatic self-validation concedes the point. The objection grants that the axioms cannot be proved but works on them anyway. That is what faith means in the relevant philosophical sense: trust in unprovable propositions that one then acts on. The atheist position has been reframed as a faith-position; the objection is no longer an objection to the argument.

Objection 2: "Even granting that science is faith-based, religion adds more unprovable axioms still. Science's faith is leaner."

Rebuttal. The objection grants the central conclusion (science is faith-based) and shifts to a comparative-parsimony claim. The comparative question is then which set of foundational commitments fits the evidence, which is what Guillen's other four arguments (physics, cosmology, neuroscience, mathematics) take up. The retreat to comparative parsimony already concedes the meta-epistemological asymmetry.

Objection 3: "This is just presuppositionalism with a science gloss."

Rebuttal. It draws on the same epistemological insight Cornelius Van Til and the presuppositional tradition emphasize (Romans 1:18-23, the suppression of truth via foundational commitments), but it is not a presuppositionalist argument in the technical Van Tilian sense. It is the broader Polanyi-Kuhn-Plantinga critique of naive empiricism, applied to the contemporary New-Atheist context. The presuppositional resonance is a strength, not a weakness; multiple philosophical traditions converge on the same meta-epistemological observation.

Objection 4: "This proves only that atheism is faith-based, not that Christianity is true."

Rebuttal. Correct, and the argument does not claim more than this. It is a meta-epistemological defeater of the science-vs-faith asymmetry, not a positive case for Christianity. The positive case is the substantive evidence in the other four arguments. P5 is explicit: the meta-debate shifts; it does not get decided here.

Live-cite kit

Philosophical: Michael Polanyi, Personal Knowledge: Towards a Post-Critical Philosophy (Chicago, 1958); Thomas Kuhn, The Structure of Scientific Revolutions (Chicago, 1962); Alvin Plantinga, Where the Conflict Really Lies: Science, Religion, and Naturalism (Oxford, 2011); Plantinga, Warrant and Proper Function (Oxford, 1993)
Scientific-philosophical: Stephen Meyer, Return of the God Hypothesis (HarperOne, 2021); Michael Guillen, Believing Is Seeing (Tyndale Refresh, 2021), chs. 6-8
Scripture: Hebrews 11.1, faith as "the assurance of things hoped for, the conviction of things not seen"; Romans 1.18-21, the Pauline suppression-of-truth doctrine
Aphorism: "There is no view from nowhere. Every view is from somewhere. Honest science admits its somewhere; honest theism does the same."