Religious language

Cognitivism and non-cognitivism about religious language#

FoundationAQA 7172 3.3 Metaphysics of God — Religious languageDfE GCE Philosophy subject content

Views of religious language

Fig. 1Cognitivism treats religious statements as truth-apt; non-cognitivism as expressions of attitude or commitment.

The philosophy of religious language asks a prior question, before we debate whether God exists: what does religious language MEAN, and is it even meaningful? The debate is framed by the same cognitivist/non-cognitivist distinction as meta-ethics. Cognitivism about religious language holds that religious statements ('God exists', 'God created the world', 'God loves us') are genuine assertions that express beliefs and are truth-apt — they attempt to describe how things are, and are true or false. Non-cognitivism holds that religious statements are NOT truth-apt assertions of fact but perform some other function — expressing an attitude, a commitment, a way of life, or a moral resolve — so the question of their literal truth or falsity does not arise.

The challenge that dominates the topic comes from LOGICAL POSITIVISM and its criterion of meaning. The positivists (the Vienna Circle) held that a sentence is factually (cognitively) MEANINGFUL only if it is either analytic (true or false by definition, like mathematics and logic) or empirically verifiable (its truth can be tested by sense experience). Any sentence that is neither — that makes no analytic claim and no empirically testable claim — is not false but MEANINGLESS, a pseudo-statement that only appears to say something. The positivists applied this not just to religion but to metaphysics and (in Ayer's hands) to ethics: 'God exists', 'the Absolute is beyond time', and 'stealing is wrong' were all charged with being cognitively meaningless.

This challenge is more radical than atheism. The atheist says 'God exists' is FALSE — which concedes it is a meaningful, truth-apt claim (a cognitivist position). The verificationist says 'God exists' is neither true nor false but MEANINGLESS — it fails even to make a genuine assertion, because no possible experience could verify it. So the verification challenge, if it succeeds, does not merely deny God; it denies that religious (and much metaphysical) discourse says anything at all. This is why the topic is so important: it is a challenge to whether the WHOLE preceding debate about God's existence and nature is so much as intelligible.

The debate divides into two challenges and two kinds of response. The challenges are verification (Ayer: religious statements are unverifiable, so meaningless) and falsification (Flew: religious statements are unfalsifiable, so not genuine assertions). The responses are, broadly, COGNITIVIST (accept that religious language aims at truth, but argue it IS verifiable/falsifiable, e.g. Hick's eschatological verification and Mitchell's partisan, and/or attack the verification principle itself) and NON-COGNITIVIST (accept that religious language is not truth-apt fact-stating, but deny this makes it meaningless — it is meaningful as the expression of a 'blik' or form of life, e.g. Hare). Holding this map steady — two challenges, two kinds of defence — organises the whole topic.

Worked example

Atheism versus meaninglessness

Contrast the atheist and the verificationist responses to 'God exists'.

The atheist
Says 'God exists' is FALSE — which grants it is a meaningful, truth-apt claim that happens to be untrue (cognitivism).
The verificationist
Says 'God exists' is neither true nor false but MEANINGLESS, because no possible experience could verify it.
The difference
The atheist engages the claim; the verificationist dismisses it as a pseudo-statement that fails to assert anything.
The stakes
If the verificationist is right, the whole debate about God's existence is not merely undecided but unintelligible.

Result: The verification challenge is more radical than atheism: it denies that religious statements say anything truth-apt at all, so it must be met before the existence debate can even proceed on cognitivist terms.

Typical mistakes

Treating the verification challenge as a denial that God exists; it is a denial that 'God exists' is MEANINGFUL — a claim about language, not (directly) about God.
Confusing non-cognitivism about religious language (it is not truth-apt) with the claim that it is meaningless; non-cognitivists hold it IS meaningful, just not fact-stating.

Active revision

Explain the difference between cognitivist and non-cognitivist views of religious language. (5-mark AO1 response)

Active recall

Recall the key points — then reveal.

Sources: GCE AS and A level subject content for philosophy (Department for Education)

The verification principle and Ayer#

StandardAQA 7172 3.3 Metaphysics of God — Religious languageDfE GCE Philosophy subject content

The verification principle

Fig. 2A statement is meaningful only if analytic or empirically verifiable; 'God exists' is neither, so is meaningless.

A. J. Ayer, in Language, Truth and Logic (1936), brought logical positivism to an English audience and applied its verification principle to religion. The principle: a statement is (cognitively/factually) meaningful if and only if it is either analytic (a tautology, true or false by definition) or empirically verifiable (there is some possible sense experience relevant to establishing its truth or falsity). All and only such statements say something; everything else is 'literally nonsense' — cognitively meaningless, though it may have emotive or poetic uses. Ayer's target is metaphysics generally: statements about God, the Absolute, the soul, and objective moral values all fail the test, since none is analytic and none is empirically testable.

Ayer draws the distinction between STRONG and WEAK verification to make the principle workable. Strong verification requires that a statement be conclusively established by experience — but this is too demanding, since it would render meaningless not only 'God exists' but also universal scientific laws ('all metals expand when heated', which no finite set of observations can conclusively verify) and statements about the past. Ayer therefore adopts WEAK verification: a statement is meaningful if some possible observations would be RELEVANT to (would count for or against) its truth — if experience could render it more or less probable. Scientific laws pass (observations bear on them); 'God exists' still fails, because Ayer holds no possible observation is relevant to it.

Applied to religion, Ayer's verdict is that 'God exists' and 'God loves us' are not false but meaningless, because no conceivable observation would count for or against them — the theist does not treat any possible experience as confirming or disconfirming God's existence, so the statement makes no genuine empirical claim. Ayer is careful to say this is NOT atheism or agnosticism: the atheist and agnostic both treat 'God exists' as a meaningful proposition (to be denied or suspended on), whereas Ayer holds it is not a genuine proposition at all. He similarly held that talk of a 'transcendent' God is meaningless, and (in his ethics) that moral statements merely express emotion (emotivism), applying the same criterion across metaphysics and ethics.

The verification principle is powerful but faces immediate difficulties that the responses (next section) develop. First, it seems to render meaningless not only religion but also statements we do NOT want to reject — historical statements about the unobservable past, universal scientific laws (on strong verification), statements about other minds, and even ethics. Ayer's weak-verification version tries to save science, but critics argue it is either too weak (letting in 'God exists' as weakly relevant to some experiences) or still too strong. Second, and most damagingly, the principle appears to be SELF-REFUTING: the verification principle itself is neither analytic (it is not true by definition — it is a substantive proposal about meaning) nor empirically verifiable (no observation tests it), so by its own criterion it is cognitively meaningless. This self-refutation objection, pressed hard from the 1950s, is widely thought to be fatal, and it is the reason the verification principle is now generally rejected — though the challenge it poses to religious language remains instructive.

Worked example

Strong versus weak verification

Explain why Ayer moves from strong to weak verification, and what it costs him.

Strong verification
A statement is meaningful only if conclusively established by experience.
The problem
This renders universal scientific laws and statements about the past meaningless, since no finite observation conclusively verifies them — an unacceptable result.
Weak verification
A statement is meaningful if some possible observations would be RELEVANT to (count for or against) its truth; laws now pass.
The cost
Critics argue weak verification is too permissive — some observations (answered prayers, religious experience) seem relevant to 'God exists', so it may let religion back in, or is otherwise hard to formulate to exclude only what Ayer wants excluded.

Result: Ayer needs weak verification to save science, but weakening the principle makes it hard to exclude religious statements without also excluding too much or too little — a tension that, with the self-refutation charge, undermines the principle.

Typical mistakes

Saying Ayer holds 'God exists' is false; he holds it is MEANINGLESS — neither true nor false — which distinguishes him from the atheist.
Ignoring the strong/weak distinction; Ayer needs WEAK verification to save scientific laws, and this concession is important to the evaluation.

Active revision

Outline the verification principle and explain how Ayer applies it to religious language. (12-mark AO1 response)

Active recall

Recall the key points — then reveal.

Sources: GCE AS and A level subject content for philosophy (Department for Education)

Responses: eschatological verification and self-refutation#

AdvancedAQA 7172 3.3 Metaphysics of God — Religious languageDfE GCE Philosophy subject content

The self-refutation of the verification principle

Fig. 3The principle is neither analytic nor empirically verifiable, so by its own criterion it is meaningless.

There are two main cognitivist responses to the verification challenge: attack the principle itself, and argue that religious statements ARE verifiable after all. The first is the SELF-REFUTATION objection, widely regarded as decisive. The verification principle states that a statement is meaningful only if analytic or empirically verifiable. But the principle itself is neither: it is not analytic (it is not true merely by definition — it is a controversial, substantive claim about what meaning is, which many competent speakers reject), and it is not empirically verifiable (no sense experience could test whether the principle is true). So by its OWN criterion, the verification principle is cognitively meaningless — it fails its own test. If the principle is meaningless, it cannot be used to show that religious statements are meaningless; the weapon destroys itself.

Defenders of positivism tried to escape self-refutation in two ways, both problematic. They might say the principle is a DEFINITION or stipulation of 'meaningful' (hence analytic) — but then it is merely a proposal about how to use a word, which the theist can decline to accept, and it loses its force as a discovery about religious language. Or they might say it is a METHODOLOGICAL recommendation rather than a truth-apt statement — but then it is not itself a factual claim, and provides no reason to think religious statements are actually meaningless rather than merely excluded by a chosen convention. Either way, the principle cannot be both meaningful by its own lights and a substantive refutation of religion. The self-refutation objection is the principal reason the verification principle is now generally abandoned.

The second response, from John Hick ('Theology and Verification'), grants that meaningfulness may require verifiability but argues that religious statements ARE verifiable — in principle, if not yet in practice — by ESCHATOLOGICAL VERIFICATION. Hick's parable of the Celestial City: two travellers walk the same road; one believes it leads to the Celestial City, the other believes it leads nowhere. During the journey their experiences are identical, and no observation settles their dispute. But if the road DOES lead to the Celestial City, then upon arrival the believer will be proved right and the sceptic wrong; the issue is verifiable at the journey's end, even though not during it. Analogously, Hick argues, 'there is a God / an afterlife' is verifiable AFTER DEATH: if theism is true, we will have experiences (of God, of survival) that confirm it; so the statement is verifiable in principle, and therefore meaningful, even though it cannot be verified now.

Evaluation weighs both responses. The self-refutation objection is powerful and largely decisive against the strong verification principle as a criterion of meaning; but a defender may reply that SOME connection between meaning and testability survives its failure, so the challenge to religious language (that it makes no testable difference) is not entirely dissolved. Eschatological verification has real force — it shows that 'unverifiable NOW' does not entail 'unverifiable in principle', which is what the positivist needs — but faces objections: it works only if there IS post-mortem experience (which requires the survival of death, itself contested and arguably unverifiable), and it provides only ASYMMETRIC verification (if theism is true it is verified, but if it is false there is no one left to falsify it — no experience of 'nothing after death'), which some argue makes it verifiable but not FALSIFIABLE, exposing it to Flew's separate challenge (next section). A balanced verdict is that the verification principle FAILS as a general criterion of meaning (self-refutation), so it does not show religious language meaningless; and that Hick's eschatological verification plausibly shows religious statements are cognitive and verifiable IN PRINCIPLE, provided one grants the possibility of post-mortem experience — so the verification challenge, though historically important, does not succeed in reducing religious language to nonsense. This shifts the debate to Flew's falsification challenge, which does not depend on the discredited verification principle.

Worked example

Assessing eschatological verification

Explain Hick's eschatological verification and evaluate one objection to it.

The parable
Two travellers dispute whether the road leads to the Celestial City; no observation settles it en route, but arrival (or not) will verify it at the end.
The application
'There is a God/an afterlife' is verifiable AFTER DEATH: if true, we will have confirming experiences; so it is verifiable in principle, hence meaningful.
The strength
It shows 'unverifiable now' does not imply 'meaningless': verifiability in principle suffices, defeating the positivist's demand for present verification.
The objection
It presupposes post-mortem experience (survival of death), itself contested; and it is ASYMMETRIC — verifiable if true, but not falsifiable if false (no one experiences 'nothing'), exposing it to Flew.

Result: Eschatological verification plausibly shows religious statements are cognitive and verifiable in principle, defeating the verification challenge — but only by assuming survival of death, and its asymmetry (verifiable not falsifiable) hands the debate to Flew's falsification challenge.

Typical mistakes

Overlooking the self-refutation objection, which is the decisive reason the verification principle is now rejected; a full answer must state it.
Presenting eschatological verification as verification NOW; Hick's point is verifiability IN PRINCIPLE (after death), which requires the survival of death and is only asymmetric.

Active revision

'The verification principle fails to show that religious language is meaningless.' Assess this claim. (25-mark essay)

Active recall

Recall the key points — then reveal.

Sources: GCE AS and A level subject content for philosophy (Department for Education)

The falsification principle (Flew)#

AdvancedAQA 7172 3.3 Metaphysics of God — Religious languageDfE GCE Philosophy subject content

Flew's falsification challenge

Fig. 4A genuine assertion rules something out; if nothing could count against 'God loves us', it asserts nothing.

Antony Flew ('Theology and Falsification', part of a 1950s 'University discussion') shifts the challenge from verification to FALSIFICATION, drawing on Karl Popper's idea that what distinguishes a genuine empirical assertion is that it RULES SOMETHING OUT — it is incompatible with some possible state of affairs, so that some conceivable observation would count DECISIVELY against it. To assert that something is the case is equally to deny that its contrary is the case; a statement that is compatible with EVERYTHING, that nothing could count against, asserts nothing. Flew's challenge is that religious statements are UNFALSIFIABLE — believers allow nothing to count against them — and so are not genuine assertions at all.

Flew illustrates with John Wisdom's parable of the gardener, which he retells. Two people return to a long-neglected garden and find, among the weeds, some plants flourishing as if tended. One says a gardener must come and tend it; the other denies it. They set a watch, but no gardener is ever seen; they use bloodhounds and electric fences, but detect nothing. Yet the believer keeps qualifying his claim: the gardener is invisible, intangible, undetectable, insensible to shocks — he comes secretly. Eventually the sceptic asks: how does this 'invisible, intangible, eternally elusive gardener' differ from no gardener at all, or from an imaginary one? The believer's assertion has been so qualified that it rules nothing out, and so, Flew argues, it has 'died the death of a thousand qualifications' — it no longer asserts anything.

Flew applies this directly to religious belief. The theist asserts 'God loves us as a father loves his children'. But when confronted with a child dying of inoperable cancer, the theist does not withdraw the claim; instead he qualifies it — God's love is not merely human love, it is inscrutable, it works in mysterious ways, its purposes are beyond us. Flew's challenge: what would have to happen — what conceivable event or amount of suffering — for the theist to admit 'there is no God' or 'God does not love us'? If the answer is 'nothing' — if no possible state of affairs would count against the claim, because every apparent counterexample is absorbed by a qualification — then 'God loves us' is not a genuine assertion; it does not really claim anything, because it excludes nothing. Note that Flew's challenge, unlike Ayer's, does not rely on the discredited verification principle; it is a challenge about whether religious statements have any assertive content, which stands independently.

Evaluation weighs the challenge's force. Its strength is that it identifies a real feature of much religious discourse — the tendency to protect belief from any possible disconfirmation, absorbing every apparent counterexample (especially suffering) by qualification — and it poses a sharp, answerable question (what would count against your belief?) that a genuine assertion should be able to meet. It also connects the problem of religious language to the problem of evil: the theist's refusal to let suffering count against God's love is exactly what Flew targets. Its weaknesses, developed in the responses (Hare and Mitchell, next section), are: first, that falsifiability may be too STRINGENT a criterion of meaning — many meaningful statements (about history, about universal moral principles, arguably about scientific paradigms) are not straightforwardly falsifiable, and Popper's criterion was proposed to demarcate SCIENCE from non-science, not the meaningful from the meaningless, so applying it to rule religion meaningless may overreach; and second, that a theist CAN specify what would count against belief (Mitchell: the theist DOES feel the weight of evidence against faith, and could in principle be defeated by it), or can deny that religious statements are assertions at all while remaining meaningful (Hare's bliks). A balanced verdict is that Flew's challenge is more resilient than Ayer's (it does not rest on the self-refuting verification principle) and exposes a genuine tension in religious discourse, but that it is not decisive: it either presupposes an over-stringent falsifiability criterion of meaning, or is answerable by showing that religious claims DO risk something (Mitchell) or MEAN something without asserting (Hare).

Worked example

Applying the falsification challenge

Show how Flew's challenge works against 'God loves us', and state one weakness of the challenge.

The assertion
The theist asserts 'God loves humankind as a father loves his children'.
The apparent counter
A child dies slowly of cancer — apparent evidence against a loving God.
The qualification
The theist does not withdraw the claim but qualifies it: God's love is inscrutable, mysterious, beyond human comprehension.
Flew's question and a weakness
Flew asks what WOULD count against the claim; if nothing, it asserts nothing. Weakness: falsifiability may be too stringent a test of MEANING (much meaningful discourse is unfalsifiable), and it was designed to demarcate science, not sense.

Result: Flew's challenge exposes a real tendency in religious discourse to resist all disconfirmation, but it is not decisive: falsifiability is arguably too stringent a criterion of meaning, and the theist may either specify what would count against belief (Mitchell) or deny the claim is an assertion (Hare).

Typical mistakes

Conflating falsification with verification; Flew asks what would count AGAINST a claim (falsifiability), and his challenge survives the collapse of the verification principle.
Missing the 'death by a thousand qualifications' point; the objection is that endless qualification empties the claim of assertive content, not merely that it is hard to test.

Active revision

'Religious statements are meaningless because nothing could count against them.' Discuss. (25-mark essay)

Active recall

Recall the key points — then reveal.

Sources: GCE AS and A level subject content for philosophy (Department for Education)

Bliks and assertions: Hare and Mitchell#

AdvancedAQA 7172 3.3 Metaphysics of God — Religious languageDfE GCE Philosophy subject content

Two answers to Flew

Fig. 5Hare reclassifies religious language as a non-cognitive blik; Mitchell keeps it as an assertion held in trust despite counter-evidence.

R. M. Hare responds to Flew from a NON-COGNITIVIST direction: he concedes that religious statements are not falsifiable assertions of fact, but denies that this makes them meaningless. Religious utterances express a 'BLIK' — an unfalsifiable but meaningful way of seeing and interpreting the world, a fundamental attitude or framework that shapes how one experiences everything. Hare illustrates with the parable of the lunatic who is convinced all the dons at his university want to murder him; however many kindly dons he is introduced to, he absorbs each into his conviction ('ah, but he was concealing his true intentions'). His belief is unfalsifiable — nothing counts against it — yet it is clearly MEANINGFUL: it makes a real difference to how he lives and feels, and it differs sharply from the SANE person's blik that the dons are friendly. So there are 'right' and 'wrong', 'sane' and 'insane' bliks, even though none is falsifiable.

Hare's point against Flew is that Flew wrongly assumes religious language is trying to be assertion of fact and failing. Instead, religious belief is a blik — an unfalsifiable interpretive stance — and bliks are meaningful and important (they matter enormously to how we live) even though they are not falsifiable factual claims. It DOES make a difference whether one has the blik that the world is the creation of a loving God or the blik that it is a blind, indifferent mechanism; the two lead to different lives, hopes and responses, so the religious blik is far from meaningless. Hare thus keeps religious language meaningful by reclassifying it as non-cognitive: not fact-stating, but a framework of interpretation.

The objection to Hare is that his account seems to make religious belief SUBJECTIVE and to misdescribe what believers think they are doing. Believers do not think 'God exists' is merely their unfalsifiable way of seeing things, on a par with the lunatic's paranoia; they think it is TRUE, a claim about how reality objectively is. Hare's own example is awkward for him: he calls the lunatic's blik 'insane' and 'wrong', which implies there is a fact of the matter (the dons really are friendly) against which bliks can be assessed — but if bliks can be right or wrong by the facts, they are after all making factual, falsifiable claims, undermining the non-cognitivist reclassification. So Hare either makes religious belief a mere subjective stance (which believers reject) or smuggles back the cognitive, factual dimension he tried to remove.

Basil Mitchell responds to Flew from a COGNITIVIST direction, arguing that religious statements ARE genuine assertions, but ones that the believer holds firm against counter-evidence as a matter of faith and TRUST, not because nothing could count against them. His parable of the partisan and the stranger: in wartime, a member of the resistance meets a stranger who deeply impresses him and claims to be on the resistance's side, urging the partisan to trust him whatever happens. Thereafter the partisan sometimes sees the stranger helping the resistance, sometimes apparently working for the enemy. The committed partisan continues to trust the stranger ('he is on our side') and interprets the ambiguous evidence in that light — but he ACKNOWLEDGES that the stranger's hostile-seeming acts really do COUNT AGAINST his belief; he feels their weight, and does not pretend they are not evidence. His faith is a trusting commitment held DESPITE contrary evidence, not a claim immune to it.

Mitchell's point is that this is exactly the theist's situation with the problem of evil: suffering genuinely COUNTS AGAINST 'God is good', and the believer feels its force, but maintains trust in God on the basis of other grounds (religious experience, the person of Christ, the arguments), interpreting the evil in that light without denying it is evidence. So religious statements are genuine ASSERTIONS (they are falsifiable in the sense that evidence bears against them — the theist admits evil counts against God's goodness), but the believer's faith is a reasonable trust that withstands the counter-evidence rather than pretending it away. This answers Flew: the theist does NOT let 'nothing' count against belief (which would make it vacuous), but neither does he abandon belief at the first difficulty (which would make it not faith). Evaluation weighs Mitchell's cognitivism (which preserves the truth-aptness believers intend, and honestly engages the problem of evil, but must explain when trust-despite-evidence becomes mere irrationality — how much counter-evidence is too much?) against Hare's non-cognitivism (which neatly sidesteps falsification but at the cost of the objectivity believers claim). A balanced A-Level judgement is that the verification challenge fails (self-refutation) and the falsification challenge is answerable — most persuasively by Mitchell's cognitivist account, which keeps religious language truth-apt while honestly admitting that evidence (especially evil) bears against it, avoiding both Flew's 'asserts nothing' and Hare's loss of objectivity. The topic thus connects religious language to the problem of evil (what counts against 'God is good') and to meta-ethics (the cognitivist/non-cognitivist debate), showing that whether religious language is meaningful, and cognitive or not, is continuous with the wider analysis of language and truth across the specification.

Worked example

Mitchell's partisan against Flew

Explain how Mitchell's parable answers Flew, and one objection to Mitchell.

The parable
A partisan trusts a stranger who claims to be on the resistance's side; when the stranger seems to help the enemy, the partisan keeps faith but ACKNOWLEDGES the acts count against his trust.
The application
The theist's 'God is good' is a genuine assertion; evil really does count against it (the believer feels its weight), but faith trusts God despite the acknowledged counter-evidence.
Why it answers Flew
The theist does NOT let 'nothing' count against belief (so it is not vacuous), but neither abandons it at once — faith is trust despite, not immunity to, counter-evidence.
The objection
How much counter-evidence is too much? If the partisan/theist NEVER lets evidence defeat belief, faith collapses back into Flew's unfalsifiable claim; Mitchell must say when trust becomes irrational.

Result: Mitchell answers Flew by keeping religious statements as genuine, evidence-sensitive assertions held in reasonable trust — the strongest cognitivist reply — but owes an account of when steadfast faith shades into irrational refusal to be falsified.

Typical mistakes

Treating Hare and Mitchell as giving the same answer; Hare is NON-cognitivist (a blik, not an assertion), Mitchell is COGNITIVIST (a genuine, evidence-sensitive assertion held in faith).
Missing that Mitchell CONCEDES evil counts against belief; his answer to Flew is that the theist does not let 'nothing' count against faith, but trusts despite acknowledged counter-evidence.

Active revision

'Religious language is best understood as expressing a blik rather than making assertions.' Is this claim true? Defend your answer. (25-mark essay)

Active recall

Recall the key points — then reveal.

Sources: GCE AS and A level subject content for philosophy (Department for Education)

Key points

Atheism versus meaninglessness

Key points

Strong versus weak verification

Key points

Assessing eschatological verification

Key points

Applying the falsification challenge

Key points

Mitchell's partisan against Flew

Sources

Religious language

Key points

Atheism versus meaninglessness

Key points

Strong versus weak verification

Key points

Assessing eschatological verification

Key points

Applying the falsification challenge

Key points

Mitchell's partisan against Flew

Sources