Does Wittgensteinian fideism (D. Z. Phillips) preserve the meaning of religious language at the cost of emptying it of cognitive content?

This question touches on a central debate in analytic philosophy of religion since the 1960s. D. Z. Phillips—Wittgenstein's indirect student through Rush Rhees—developed a reading of religious language that claims to preserve its meaning by removing it from the realm of factual claims. Did he succeed in this endeavor, or was this "preservation" achieved at the cost of evacuation? The debate is complex and deserves careful analysis.

Inadequate Responses to Avoid

From some defenders of Phillips:

"Phillips preserves authentic religious meaning away from metaphysical distortion." Oversimplification. Phillips doesn't claim to historically recover "original meaning," but rather offers a normative philosophical reading of what religious language should be. Confusing description with prescription weakens the position.

"Religious language in Phillips retains its full power." A claim requiring scrutiny. "Power" here means expressive and practical force, not cognitive force. Phillips is explicit that religious language doesn't describe facts, and this is an evacuation of a certain kind of power.

"Metaphysical criticism of Phillips presupposes realism." Circular fallacy. Even from within a non-realist perspective, Phillips can be criticized: Is his reading internally coherent? Does it preserve what believers themselves claim is important in their faith? Criticism isn't necessarily from a realist position.

From some critics:

"Phillips denies God's existence." Misunderstanding. Phillips doesn't "deny" God's existence in an atheistic sense, but rather redefines what "God's existence" means so that it's not a factual claim. The difference matters: the former is a metaphysical position, the latter a position in philosophy of language.

"Wittgensteinian fideism is merely disguised atheism." Reductionist. The position is more complex: Phillips attempts to preserve religious practice and its life-meaning while rejecting its interpretation as claims about reality. This is neither traditional atheism nor traditional faith, but a third position deserving evaluation on its own terms.

Why These Responses Are Inadequate

They share a methodological error: treating Phillips's position either as traditional faith in new language, or as atheism with a religious mask. Both characterizations miss the radical nature of his project: fundamentally redefining what religious language means, so that it's not claims about how the world is, but expressions of how we live in the world.

Structure of Phillips's Position

Wittgensteinian Foundation. Late Wittgenstein developed the idea of "language games" (Sprachspiele): each domain of life has its own linguistic rules. The mathematical language game differs from the ethical language game differs from the religious language game. The philosophical error is attempting to apply rules of one game to another.

Phillips's Application. Religious language is an independent language game with internal rules. The statement "God exists" isn't like "the chair exists"—the former doesn't describe a fact in the world, but expresses a life attitude, an existential commitment, a way of seeing the world as a whole. Attempting to convert it into a factual claim misunderstands the nature of the religious language game.

Phillips's Applied Examples:

- Prayer: Not "conversation with a being in heaven" but a practice that transforms the self, an expression of surrender and submission, a way of seeing life sub specie aeternitatis.

- Immortality: Not "continued existence after death" but a way of living that transcends limited temporal perspective, seeing life from the perspective of eternity now.

- Miracles: Not "violations of natural laws" but seeing ordinary events with eyes of faith, perceiving the sacred dimension in daily life.

- Judgment Day: Not a "future event" but realizing that every moment has eternal weight, that life is lived before the Absolute.

The Alleged Gain. Phillips claims this reading preserves religious language from:
1. Conflict with science (because it makes no empirical factual claims)
2. Problems of philosophical proof (because it needs no proof)
3. The problem of evil (because "God" isn't an agent in the world in a causal sense)

The Basic Criticism: Evacuation of Cognitive Content

Kai Nielsen's Objection. If "God exists" doesn't mean there's a real being independent of our consciousness, what's the difference between it and "God doesn't exist"? Phillips responds that the difference lies in life attitude, but Nielsen replies: life attitude presupposes belief in some reality, otherwise it becomes empty performance.

Richard Swinburne's Objection. Ordinary believers actually believe God is a real being who created the universe and responds to prayer. Phillips's reading doesn't "preserve" their faith but replaces it with something entirely different. This isn't interpretation but radical revisionism.

John Hick's Objection. Religious language in Phillips loses its capacity to distinguish between truth and falsehood. If "God loves me" is merely an expression of life attitude, how do we distinguish between someone truthful in this expression and someone lying? Loss of realistic criterion leads to loss of criterion altogether.

The Deeper Response: The Problem of Commitment

Peter van Inwagen posed a decisive objection: genuine religious commitment requires believing that something is true about reality, not just about one's attitude toward it. The believer who sacrifices life for faith does so because they believe God actually exists, not because they're expressing a "life attitude." Phillips empties faith of its driving force by converting it into mere subjective stance.

Phillips's Defenses and Their Limits

Phillips responds that critics assume a false dichotomy: either factual claim or mere subjective expression. He says there's a third level: practical meaning that's neither purely subjective (because it's tied to communal practices) nor realistic in a metaphysical sense.

But this response faces a problem: even communal practices usually presuppose beliefs about reality. Communal prayer presupposes someone being prayed to, pilgrimage presupposes actual sacredness of place. Attempting to separate practice from realistic belief seems artificial.

Contemporary Developments

The "Expressive Realism" current attempts a middle position: religious language is primarily expressive but refers to a transcendent reality that cannot be described literally. This preserves the realistic dimension without falling into naive literalism.

"Critical Practicalism" accepts Phillips's insights about practice but insists that religious practices contain implicit realistic commitments that cannot be denied without distorting the practice itself.

Final Assessment

Phillips provided a philosophical service by alerting us to the complexity of religious language and its non-reduction to simple factual claims. But his radical project of completely removing it from the realm of factual claims appears to:

1. Not preserve what believers want preserved—most believers want to say God actually exists, not just that "God exists" expresses an attitude.

2. Empty religious language of its driving force—willingness to sacrifice for faith presupposes belief in factual truth, not mere commitment to life attitude.

3. Create tension with religious practice itself—many religious practices (prayer, seeking miracles, hope for afterlife) presuppose the reality of what they discuss.

From the Perspective of Rational Preference (rajḥān ʿaqlī)

Phillips's project represents an important challenge to naive religious realism, but in attempting to avoid the philosophical problems of realism, it falls into bigger problems: emptying religious language of content that makes it important to believers. Rational preference inclines toward a middle position: recognizing the complexity of religious language and its expressive and practical dimensions, while insisting it also contains factual claims about reality.

Where We Stand in This Debate Today

The discussion about Wittgensteinian fideism didn't stop with Phillips's death (2006), but was renewed in the last decade in new forms. Between 2020 and 2026, three main currents emerged:

First, the "Modified Wittgensteinianism" current that accepts Phillips's insights about the independence of religious language games but reintroduces a minimal realistic dimension, represented in the works of Stephen Mulhall and some Swansea College students. Second, the rising interest of analytic philosophy of religion in practice and rituals (practice turn) influenced by Mark Rieken and Jonathan Kvanvig, where religious language is read as practice with both cognitive and expressive dimensions, without the sharp dichotomy Phillips fell into. Third, the return of religious realism in more sophisticated forms with Andrew Moore and Roger Pouivet, absorbing Phillips's criticism of naive literalism without abandoning cognitive content.

The philosophically sound position today: Phillips's contribution in revealing the complexity of religious language remains valid and influential, but his radical project of completely removing it from the cognitive realm hasn't withstood objections from actual religious practice nor developments in post-Wittgensteinian philosophy of language. The prevailing direction moves toward a synthesis combining practical sensitivity with realistic commitment—and this precisely aligns with the method of cumulative rational preference (rajḥān ʿaqlī tarkumī) adopted by this site.