The Problem of Metaphor in Religion


  • Language may be used analogically or discursively; these are exclusive categories.
  • Discursive language relies on facticity and perception to convey correspondence truth.
  • Analogical language necessarily relies on coherence and concepts and allows variance in declarations.
  • Every metaphor can roughly translate to three distinct truth claims; while none of the three captures the meaning of the metaphor discursively, the metaphor still intends to convey a knowable truth.
  • To interpret a metaphor involves a number of conceptual and categorical claims, any one of which might be defended on coherence grounds and the combination of which are difficult to defend on correspondence grounds.
  • Since the iron curtain of analogical versus discursive meaning cannot be bridged, figurative language inevitably conveys private connotations even in the simplest, most obvious metaphors.
  • This is not a problem for poetry, which never intends itself to be known discursively, but it is a disqualifying problem for theology, which somehow intends its ubiquitous metaphorical language to communicate correspondence truths about the divine, a necessary ambition because any definition of a divinity must be a universal one.
  • Even if believers or congregants let go of claims to discursive truth about the transcendent God in order to make analogical claims about an immanent God, they cannot do so with metaphorical language because they cannot complete the implied comparison.
  • To control a metaphor, the language maker must understand both sides of the comparison so as to emphasize and also limit the points of comparison between the two objects being related in the metaphor, but in religious language, the metaphor is employed precisely because the speaker cannot know one side and wishes to “climb up” to comprehension by employing the comparison.
  • Consumers of such language understandably blank out on the divine “higher part” of the metaphor and therefore literalize the accessible “human part” and think it discursive, but this asks for too much and violates the iron curtain dividing the two kinds of language.
  • Defenders of such language argue either its intent to be similar to poetry or that the metaphor is a result of revelation, but both defenses render the metaphor’s meaning ambiguous, and given the temptations of immanence, less than knowledge.

One natural fault line concerning our use of language to make truth claims is to see them as being either analogical or discursive, which translates into either figurative or literal. This distinction is not given enough scrutiny, I think. It is inclusive, meaning all elements of each category share the same definitional similarity and no declaration can exist in a third category. All fall into one or the other. The definitions are also exclusive, meaning the distinction between categories cannot be blurred or bridged. No declarative language falls into both categories and none falls into some third, unnamed category.  We can never take literal language figuratively without adding to its meaning more than was intended by the speaker, effectively changing the truth claim. Nor can we take figurative language literally, for to do so would either distort the speaker’s claim or diminish it; we may thus assume that the analogical language adds something to the speaker’s truth claim that ordinary discourse could not capture. These distinctions pose a special problem for those wishing to make religious truth claims.

When we attempt to make our declarations about truth, goodness, or beauty directly and precisely, we use language in the simplest and most accurate way we are capable of. We use discursive language, literal language. We avoid stylistic flourish in favor of plain nouns and verbs. Think chemistry textbooks and Form 1040 instructions. The purpose of using language literally is to communicate with clarity and simplicity some truths about reality. We see this intention made manifest in our respect for facts: simple and unambiguous statements about one small bit of the data of experience that constantly flood our minds (see “Facts are Fluxy Things“). A fact by its very nature must be discursive. It is clear that many other declarations can be true, but knowing them sufficiently to justify them as defensible judgments becomes more and more problematic the further away from simple percepts our declarations wander.

That effort is freighted with two crippling difficulties when we find ourselves tempted to judge beyond facts, which, of course, we must do in every effort to discover the true and pursue whatever we call good. First, language is a blunt and crude hammer with which to drive in the nail of our meaning. Francis Bacon’s warning about the subtlety of nature and the crudity of the language we use to address it comes to mind here. One of the glories of science is its co-option of the only infinite language available to the human mind, numeracy. Compared to the precision of mathematics, ordinary language seems sadly amorphous, which only becomes a problem if we care to state our declarations about reality accurately and carefully. And that introduces the second and more comprehensive difficulty in our use of discursive language, for all that we attempt to say in our literal truth, goodness, and beauty claims can be seen as a kind of comparison, with language standing in for the reality it attempts to capture. To use language literally is to attempt a one-to-one correspondence with the reality we speak of, and even leaving language’s crudity aside, that correspondence is awfully difficult to reach, and even more difficult to verify. This 1:1 correspondence is the holy grail of discursive language. To know we have it is to speak truth. The effort to accomplish this seemingly impossible task is undertaken each time we make a truth claim, something we do a thousand times daily. Most are of this simple, literal, discursive variety. “The cat is on the table.” “I left my umbrella at the market.” That kind of thing.

One difficulty arises when we speak literally about what might be called private reality, as opposed to the public reality available to all perceptions and interpreted by universal reasoning. To say “I am hungry” is quite literal, but because the verification is not available to all, we may call this kind of literal declaration a coherence claim, as opposed to the correspondence claims about cats and umbrellas whose warrant is verifiable to anyone with eyes and a brain (see “What Counts as Justification?“). The only difference is its mode of verification; our coherence claims are justified not by any attempt to establish a 1:1 correspondence to the facts of reality but instead by their logical consistency with an internal schema of truths we have already embraced, a coherence that employs a fundamental principle of reasoning, the principle of non-contradiction. Though such truth claims may be discursive, they are not verifiable by someone outside the perceptual wall of the hungry thinker’s mind.

The distinction between correspondence and coherence becomes much more complex for the other kind of language we use in making our truth claims: figurative or analogical language. But the task here is even more difficult. For every use of such language establishes not one truth claim but three. For example, should I say, “The cat is on the table,” I am claiming a literal truth, one verifiable by direct perception, either mine or yours (I leave aside for the sake of simplicity for the moment modernist questions about the reliability of perception and postmodern issues with the adequacy of language). But if I should say, “The cat on the table looks like a sculpture,” I am now saying three things. (1) that the cat is on the table, presumably in some pose; (2) that sculptures are cat-sized and table-sitting; (3) that a true relationship pertains between the pose of the cat on the table and sculpture, the claim to be verified by undistilled experience: sight (of the cat) and memory (of the sculpture). The sentence about the cat on the table looking like a sculpture is called a simile, which you may remember from your poetry classes being a comparison using “like” or “as.” Not all attempts at truth-telling through figurative language attempt this kind of relationship, but all require the same kind of “unpacking” in order to grasp their truth claims. This is why one definition of poetry that so relies on figurative speech is “intense, compact language.” To grasp its truth claims, we must expand what has been compacted, make literal what has been said figuratively and then from that literality return to the analogical relationship, mentally sorting the ways in which it might hold true before reaching the “aha moment” that finds the intended figurative relationship. And this is only to grasp the intention of the figurative language. Understanding that is only a prelude to employing the necessary act of judgment to determine whether we think the analogy true. When we slow down these complex tasks, we realize how many judgments are involved. That process changes the truth claim depending on whether its referents appeal to correspondence knowledge or coherence within the perceptual wall.

Now bear in mind the exclusive nature of the categories of literal and figurative. This exclusivity demands that the purpose of poetry is precisely to avoid being literal, that literalizing its figurative truth claims is distortional and destructive to the poet’s intent, which is to use language to approach the inexpressible, to speak coherence truths, to use language as a lancet to suggest rather than to define discourse. I am certain that no literary critic would claim that an exegesis of a poem could ever capture its meaning. I am more interested in the opposite problem, which is the use of figurative language by religionists to make the opposite claim: that the analogical language they use somehow communicates a correspondence truth that then somehow becomes literal. This claim violates the “exclusivity clause” separating the two kinds of language. But even if the intent of religionists’ use of analogical language is not to claim a literal truth, they face a more basic problem. If every use of figurative language contains not one but three separate truth claims as illustrated in the cat/sculpture example above, religionists face an insurmountable obstacle in regard to the inevitable second truth claim in their analogies. They cannot know what they claim to know.

Let us first examine their intent, so as not to be guilty of distortion or setting up straw men. When religionists say that God is our Father, or that hell is a place of fire, or that angels fly, is it possible they are using language with a literal intent? That certainly seems unlikely, but if they are, then these correspondence truth claims can be examined and proved false without much effort. It is much more likely that such claims, numerous as they are in all of the world’s religions, are meant figuratively for two reasons: first, that like poetic language, religious language deals with matters of weighty majesty and second, that language is attempting to illuminate facets of reality that are metaphysical and therefore simply not available to literal language.

Such matters are always difficult for correspondence truth claims to address because of the problem of specification: there is simply nothing for the claim to correspond to in ordinary reality. Because these claims must be imperceptible, they cannot be claimed to be facts, at least not correspondently, meaning they are unavailable to public knowledge (to acknowledge coherence truth is also to admit the reality of coherence “facts,” though we should be careful not to confuse such things with the kinds of facts science is capable of revealing; for reasons of terminological clarity, I prefer to call coherentist “facts” by another name: beliefs [see “Knowledge, Trust, and Belief”]). But it is clear that the imperceptible truths religion hopes to establish by analogy, though clearly public, cannot clear the bar of public fact. No cat. No table. No umbrella. No market. Now this is a real problem but not an insurmountable one for correspondence to address in regard to some issues. Take justice, for instance. One cannot point to it or find its literal presence in the world, yet we can claim its correspondence existence if we accept one premise: that there is something common to human reasoning when applied to experience that extracts conceptualizations of justice so that when you and I discuss that concept, our minds are focusing on a similar if not identical conception. Now the nature of that commonality is a matter of some dispute among philosophers, but I hope you will agree that whatever that conception is, we think of its meaning as literal and not figurative. After all, we think of justice and indeed all universal conceptualizations as real and definable ideas we can have access to even if they only exist in the human mind. If we should think of justice as a figurative or coherence conception not subject to common understanding, the 2,200,000 persons in U.S. prisons should immediately file appeals for their release on the grounds that we can’t define what their offenses consisted of. That won’t happen for the simple reason that we have confidence in universal reasoning to stitch together definitions of conceptions whose discursive meaning is decipherable. Their instantiation is available to experience, and though experiences differ, our common reasoning capacity allows us to draw the conceptualization from the aggregate of experiences, the more the better. Indeed, experts exemplify this process for many of their subject fields (see Expertise).But such is not the case with religious concepts. Religionists claim they reference real, if metaphysical, truths. But they also say that because of the problem of specification, they can only be approached analogically. Therefore, they claim, we know their truth through figurative language.

Only we can’t. In any analogy, metaphor, simile, parable, or direct comparison, I must have access to the two things I am relating. Should I say, “The kingdom of heaven is like a city on a hill,” you may infer that I am asserting the same three truths that I assert when I say the cat on the table is like a statue. You must assume I know not only the specific ways in which these two terms are similar, but just as importantly, that I am aware of the ways in which they are different. This is not one of David Hume’s analytic statements, which are tautological. I am saying a is like b rather than that a = b. The figurative part of the language is that I am establishing a relationship. Were I stating a definition or a tautology, I would be speaking literally, and what is more, the literality I establish would be self-evidently true because it is necessary. A bachelor simply cannot be anything other than an unmarried man (at least until we change the definition) and 9+6 simply has to equal 15. These analytic concepts simply cannot be false (though often cited as self-evident “facts,” we should note that these identities are not perceptual and therefore cannot be called facts; that we do is yet another proof that our understanding of fact could use a tune-up). A metaphor is quite another thing, for we are not establishing an identity, not putting an equal sign between the two terms we are relating. A metaphor implies a limited similitude. So while every analogy reveals a similarity, it also mandates some unspoken difference in the two terms that the speaker implies by the terms of the analogy. The two terms, alike in so many respects, are also known to be different. Though the analogy suggests only the points of similarity for reasoning and experience to derive, at least part of that undertaking must involve a clear understanding of the limits of the similarity, meaning the terms’ formal distinctions. The cat may look like a statue because of X, but at every minute of discerning X, the thinker is fully cognizant that one is not the other.

But religionists are quick to acknowledge that this knowledge of likeness and difference is beyond their ability, which is why they resort to figurative language to begin with. They are attempting to reach out to grasp the infinite, to make the mind of God amenable to puny human understanding. That sounds admirable, but the effect is anything but.

For the congregant must make sense of the analogy in what must be an incomplete way. Granted, she grasps part of the comparison. She knows what a father, a son, a lamb, a black stone, a bodhi tree, and a shamrock are. What she doesn’t and cannot know are the accurate relationships they have to the spiritual beings they attempt to illuminate. Yet that is precisely the purpose of the comparison which is launched to give her a clearer understanding of a spiritual reality she cannot see or know discursively.  So in good faith, she seizes on the part of the comparison she does know. The temptation to literality in these cases is so natural as to be almost irresistible, but this temptation must be denied by the exclusivity clause dividing the figurative and the literal.

So what is the damage? Take the simplest of examples. We are told, “God is our Father” or “God is love.” We are attempting to understand this mysterious omniscient and omnipotent being, so we fasten on the “father” and “love” side of the metaphor for the quite simple reason that we can grasp none of the realities of “godness.” Our minds cannot process infinity, eternity, omniscience, timelessness, omnipotence, or any of the other qualities of the being we call “god” regardless of the religious tradition in which the word is used. In truth, to call that unique “greatest good” a being at all is to reduce it to a literal and categorical level that distorts its unique nature. We don’t know a thing about god, but we know what god isn’t, and “being” is one of those categories to which god cannot belong. It cannot fall into any points of literal comparison with other beings. To do so would be a kind of idolatry. So we are thrown out of the literal and into the figurative, thrown back upon the expressed figurative relation. But how can our minds establish a comparative relationship when it can grasp only one of the terms involved? We know how fathers love their children, and we assume the purpose of the metaphor is to attribute the same love to God, that indefinable blank whom we now are informed resembles the father that our own experience has presented to our reasoning. Then, quite naturally, we look around us and confront the infamous problems of evil that has caused so many to lose their faith and the problem of divine hiddenness that has prompted so much agnosticism and so many controversies of religious authority (see “The Fragility of Religious Authority).. But our problem is not that there is evil in the world but that its existence and the ambiguities of God’s presence are incompatible with our understanding of how even a loving human father would act toward his children, much less how the fount of all love would.  Rather than relinquish the metaphor and acknowledge, as C.S. Lewis and Thomas Aquinas did (See C.S. Lewis, Religious Knowledge, and Belief), that God’s love must be in many ways unlike human love, we cling to it and either reject the obvious evidence, embrace yet another divine mystery, or give up faith. This attests to the human need for proper categorization as an essential element of reasoning (see Stereotypes and Categories”). But if the use of figurative language to illuminate God leads to such confusion and error–and ultimately to mystery anyway– isn’t it advisable to simply avoid the metaphor in the first place and say we really have no clarity on what we are talking about?

Now there are two possible counterarguments that I can think of to this critique.

The first is to claim that the use of figurative language in religion is identical in intent to its use in poetry. We are not meant to take these metaphors as anything other than an approach to the unapproachable. As the great hymns lift the heart, so should these poetic devices lift our minds not to a greater understanding but to a deeper contemplation and insight into divine mysteries. In response to such elevated nonsense, I can only doggedly return to the issue of what we are contemplating and what this insight must consist of. Such an argument is a literal statement of nonsense and therefore a temptation to pragmatism or fideism. If the concept of divinity is a black hole of meaning, then we can never penetrate the event horizon of experience, and to claim more is to exaggerate what we can know. This hogwash does a disservice to language, the human mind, and the sincere faith of religionists. It means nothing because either our minds are incapable of rising above the rational (see “The Tyranny of  Rationality) or if irrational, the truths thus revealed are impossible to either frame or communicate, being ineffable (see “Theology’s Cloud of Unknowing”). There is a moral theory called emotivism that claims all of our moral truth claims are merely grunts of approval or disapproval, merely interjections designed to veneer our prejudices with some false universalism. Emotivism sees the real purpose of moral language as comfort, not truth. Are we to take analogical language in religion to be emotivist expressions? If they have more meaning than a grunt of approval, what is it? How are we to translate the incomplete analogy into something the mind can grasp? These metaphors seem to be intended as neither poetry nor song. They seem to claim and believers to possess some correspondence truth claims about the reality we all share. But without the common and universal access to the instantiations of the concept being articulated, and without perceptual access to the ineffable experiences being conceptualized, and without the means of expressing the ineffable, and without access to conceptualizations that must always be rational and discursive in function to allow the listener’s comprehension,  just what can we find to declare other than our own beliefs, so colored by our own desires? The universal use of figurative language in religion is a spur to greater confusion rather than to clarity, and this is hardly a service to the faithful. A cynic might see the entire effort as organized religion’s reach for relevance (see “Religious Knowledge as Mobius Strip).

The other counterargument is worthy of a respectful hearing. It is simply that the speaker does clearly know both sides of the metaphor because of divine revelation communicated through mystic vision. Then the issue becomes less one of knowledge than of reliable transmission, the problem of coherence warrant asked to communicate profound correspondence truth and goodness, of translating belief, with all its freight of desire and private experience, into public knowledge of the true and the good, a problem that can only be resolved by an appeal to trust, first in the person of the beneficiary of the conversion experience and then in sacred texts, traditions, or inheritors of what was ineffable to begin with. The claim of divine inspiration adds its difficulties of revelation and authority. These I have addressed in earlier essays to which we must add one further caveat prompted by this one: even if the speaker fully understands the truth claim she makes because of divine inspiration as a coherence knowledge, her listeners still face the obstacles of translating analogical into literal declarations of truth and goodness (see “Premodern Authority”). That is a bloody road we should not travel again, particularly as a basis for public morality (see Toward a Public Morality).

As a postscript, I should mention that Thomas Aquinas addresses the issue of metaphorical language in Book One (I,9) of The Summa. I have considered his arguments in formulating my own.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s