1. Idols, Icons, and Ways of Seeing

Although I have no particular interest in defending the propriety of icons or icon veneration, it may nonetheless be useful to situate the topic within a broadly polemical context and to begin with one of its most articulate expositors. Near the beginning of his classic Defense of Holy Images (ca. 730), John of Damascus defines an icon as “a likeness of the prototype with a certain difference.”Footnote ³ Given this expansive definition, all manner of visual representations will count as icons, but John is interested primarily in icons that depict particular sorts of subjects and are used in particular sorts of ways. It is precisely this dual particularity, moreover, that renders them potentially controversial. The controversy answers to the fact that icons regularly depict objects which, because of their very nature, might seem insusceptible of visual representation. The paradigmatic case of just this sort of “object” is God, and the problem is easy to see: because the divine nature is invisible, bodiless, and uncircumscribed, any attempt to represent it in the necessarily visible, embodied, and circumscribed medium of the icon risks transgressing the ancient injunction against idolatry.Footnote ⁴ But it is not merely their subject matter that makes icons suspicious; it is also the particular way they are used. Unlike ordinary works of art, icons are not merely observed or admired but venerated: that is, they are accorded honor that one might think the exclusive reserve of God. And if making visual representations of the invisible God is not idolatrous enough, surely venerating such representations is.

John's answer to the first of these objections is concise and compelling. It is true, he says, that the incorporeal God does not admit of visual depiction, but this objection overlooks a crucial point. For now that the invisible God has “clothed himself in flesh” and “deigned to inhabit matter,” and since the incarnate Christ is himself “the eikōn of the invisible God” (Col. 1:15), we ourselves may likewise “venture to draw an image of the invisible God”—not qua invisible, of course, but “as having become visible for our sakes through flesh and blood.”Footnote ⁵

The second objection is more challenging. Suppose the incarnation indeed authorizes visual representations of the divine. What then authorizes the veneration normally afforded those representations? What, in other words, prevents veneration from collapsing into idolatry? To answer this question, the first thing to notice is that icons have two sides: the “image” (the thing we see) and the “prototype” (the spiritual reality to which the image points). The point of looking at an icon, moreover, is not merely to see the image itself, but also to see through the image to the reality at which it aims. Similarly, when one venerates an icon, one does not venerate the image itself but instead the invisible reality to which the image points. It is possible, of course, for this relationship between image and prototype to become perverse and distorted. I might, for instance, grow so enamored of Rublev's icon of the Trinity that I come to venerate it without attending to the invisible reality to which it points. In such cases, when the image supplants the original as the object of veneration, iconicity indeed passes over into its opposite: idolatry. For idolatry, as Aquinas remarks in the Summa, is simply “giving divine worship to whom it should not be given.”Footnote ⁶ Importantly, Aquinas's comment suggests that in fact no object is inherently idolatrous, that nothing is an idol in itself. Rather, things only become idols when they are treated in particular ways—specifically, when they are worshipped as if they were divine. The same goes, mutatis mutandis, for icons. An icon is not an icon in virtue of its physical properties or its material constitution; it rather becomes an icon when it is accorded honor or veneration on the basis of the reality it depicts.

In his classic 1982 account of iconicity, Jean-Luc Marion makes a similar point from a slightly different angle. “The icon and the idol,” writes Marion, “determine two manners of being for beings, not two classes of beings.”Footnote ⁷ To say that idols and icons constitute “manners” rather than “classes” of beings is to say that the difference between them is less like the difference between, say, birds and rocks and more like the difference between Superman and Clark Kent.Footnote ⁸ Birds and rocks, on the one hand, are indeed two classes of beings, substances that inhabit different regions in the taxonomy of existence. Superman and Clark Kent, by contrast, are substantially identical, both with one another and with the Kryptonian Kal El. And yet a distinction must be drawn. What sort of distinction exactly? One way of thinking about it is as a difference between modes of appearance. When Kal El wears the appropriate disguise (jacket, tie, thick-rimmed glasses, etc.), he appears as Clark Kent; when he wears a different disguise (red cape, blue tights, etc.), he appears as Superman. In both cases, the same substance lies behind each mode of appearance. The difference is that when he wears one disguise we see him one way, and when he wears the other disguise we see him in another way.

For Marion, this distinction between modes of appearance goes to the heart of the difference between icons and idols. The thought is that idols are idols and icons are icons not because of some property inherent in them, but rather because they appear to us—because we see them—in a particular way. Specifically, idols are idolatrous because they are merely visible: because when we see them we suppose that the divine is somehow constrained by the visibility of the image, exhausted by what can be seen.Footnote ⁹ The “idol-ness” of the idol is thus constituted not by its visibility, but by the fact that we see it as somehow exhausted by that visibility. And this has a crucially important consequence. Because the divine appears in the idol only in terms of what can be seen, and since what can be seen is a function of the (human) seer, the idol effectively “freezes” God within the confines of the visible and thereby “measures the divine to the scope of the human gaze.”Footnote ¹⁰

Against the idol's limiting or constraining function, the icon works not to restrict the divine to the scope of our gaze, but instead to point beyond itself to the invisible reality of which it is a likeness. So whereas the idol draws our gaze to itself, freezing the divine within the scope of that gaze, the icon, in Marion's words, “summons the gaze to surpass itself,” and so lets “the visible be saturated little by little with the invisible.”Footnote ¹¹ For what we see when we look at an icon, like what we see when look at an idol, is a visible image; but the iconic image, unlike its idolatrous counterpart, simultaneously points beyond itself to an invisible reality that can be neither contained in, nor exhausted by, the scope of the gaze that observes it. Another way of making the same point is to say that icons, unlike idols, require a kind of “double vision”: by inviting us to look simultaneously at them and through them, icons appear both as what they are and as something other than what they are.Footnote ¹² In this sense, and to borrow a term from Marion, the icon is an image “transfigured” by the invisible reality of which it is a visible sign.Footnote ¹³ A transfigured image is of course still a visual image; but, crucially, it is a visual image that has been so saturated by the invisible reality to which it points that it is no longer reducible to its materiality. Like the transfigured Christ, then—and indeed like the incarnate Christ—icons stand at the intersection between the visible and the invisible; and, as such, they invite us to look in two directions at once and hence to see two realities at once: on the one hand, created reality in all its materiality, and, on the other, created reality as mysteriously saturated by the invisible splendor of divine glory.

2. Metaphors, Similes, and Icons

For reasons that should become clear in a moment, the “double vision” required to see an icon as an icon invites an immediate comparison with metaphor. In fact, the relationship between metaphor and vision has a long and important history. In the most basic sense, metaphors are figures of speech that establish an analogy or comparison between two terms by transferring or “carrying over” (metapherein) the meaning of one to the other. Among classical rhetorical theorists, this semantic transference was regularly associated with sight. Quintilian, for instance, remarks that metaphors are designed to “set things before eyes,”Footnote ¹⁴ and Cicero, in the De oratore, claims that metaphors related to sight are “more striking” than those drawn from the other senses because “they place in the eye of the mind such objects as we cannot discern and see by the natural eyes.”Footnote ¹⁵ Both Quintilian and Cicero are deeply indebted to Aristotle, who argues in Poetics 3 that of all poetic devices “the use of metaphor is by far the greatest” because “to metaphorize well is to see the similar.”Footnote ¹⁶ Both poles of Aristotle's definition are crucial. It is not merely that metaphors invite us to “see” but that they invite us to see in a particular way: namely, to see similarities. Perhaps an example will help to clarify the point, so let me turn to a few lines from the third book of the Iliad. The scene is Troy, the house of Priam, and Helen has just arrived:

I begin with these lines for two reasons. First, they are an early and paradigmatic example of what poets do. And what do poets do? Quite simply, they draw comparisons; they detect resemblances; they make similes and metaphors. The second reason has to do with the phrase “seems like” in the final line. The Greek word is eoiken, from the verb eoika, meaning “to be like” or “to seem.” Both share a root with the noun eikōn (icon or image), and both are usually translated into Latin by some form of the phrase similis est, the source of the English word “simile.”

The point is twofold. The first is that one of the things poets do is make metaphors, similes, and comparisons. The second point is that one way of thinking about what poets are doing when they make metaphors, similes, and comparisons is in terms of icons. The connection is not as arbitrary as it might appear at first glance. In classical rhetorical theory, in fact, eikōn is the standard term for simile. In Rhetoric 3.4, Aristotle writes:

This is still the standard way of distinguishing similes from metaphors: the former includes, while the latter suppresses, the explicit term of comparison. There is, of course, significant debate about the precise relationship between the two, not least the question of conceptual priority. (Is a simile an expanded metaphor, or is a metaphor a contracted simile?) This is no doubt an important issue, but since nothing in my argument hangs on answering it, and since even Aristotle admits that any difference between metaphors and similes is “very small,” I shall use the two terms synonymously, focusing on what can be said indiscriminately of both.

And what can be said indiscriminately of both metaphors and similes is that they depend upon the ability to detect resemblances. In a line from Poetics 3 cited earlier, Aristotle remarks that of all poetic devices “metaphor is by far the greatest thing” since “to metaphorize well is to see the similar.”Footnote ¹⁹ It is worth noting that Aristotle characterizes the practice of making metaphors less as a technical skill than as a way of seeing. To “metaphorize” is thus not simply a matter of constructing rhetorical figures, but also of what Aristotle later calls “being metaphorical” (to metaphorikon einai): that is, a way of orienting oneself toward the world. For Aristotle, of course, “being metaphorical” in this sense is a mark of genius, since unlike geometry or mathematics, it can be neither taught nor learned. But we need not accept this (rather bizarre) assumption to appreciate the broader point that one cannot construct the simile “X is like Y” until one has first noticed that X is indeed like Y.

Something similar goes for reading or “seeing” a well-constructed metaphor or simile, though in this case the order of explanation runs in the other direction. Whereas the poet first sees that X is like Y and then constructs a simile or metaphor on the basis of that resemblance, we readers are first presented with the poet's claim that “X is like Y,” which then leads us to see that X is indeed like Y. To draw out the implications of this point, let us return briefly to Homer and consider another question: what is the effect of Trojans’ description of Helen? What is it like to read their language? The first relatively simple point to notice is that the language is about Helen: she is its referent. The poet is therefore trying to make Helen present to us; he is trying, with language, to point to her. But this is only half the story. For if Homer's language indeed points to Helen, it also points beyond Helen. How exactly? By asking us to see her not simply as she is in herself, but also as “like” the immortal goddess. No sooner, in fact, does our gaze settle upon Helen than Homer's language transports us beyond Helen to contemplate the goddess in terms of which she is described.

One way of summarizing this point would be to say that when the Trojans describe Helen as “like the immortal goddess to the eyes,” they are at once “being metaphorical” in Aristotle's sense and simultaneously asking us to “be metaphorical” in Aristotle's sense. For what the Trojans see when they look at Helen—and what they likewise invite us to see when we look at Helen—is both what she is and something other than what she is: both her own particularity and the “immortal goddess” she resembles. But if the Trojans’ vision and ours is indeed metaphorical in Aristotle's sense, it is also “iconic” in the sense I outlined above, albeit in two somewhat different ways. For the Trojans, on the one hand, Helen functions as a relatively straightforward visual icon: a visible image that has been transfigured or saturated by the presence of the goddess. They see Helen, and, through her, they see the “immortal goddess.” For us, on the other hand, both the initial image of Helen (“Helen is…”) and its transfiguration (“…like the immortal goddess”) are provoked by language rather than sight. Even in this latter case, however, the language itself works in such a way as to incite in us precisely the same kind of “double vision” that the Trojans enjoy: an image of Helen which, on the basis of a resemblance (eoiken), is transfigured by the invisible presence of the immortal goddess.Footnote ²⁰ So that when the poet tells us that Helen “seems like an immortal goddess to the eyes,” we, like the Trojans themselves, see two things at once: both Helen and, through her, the immortal goddess she resembles. To the extent that Homer's language invites this sort of vision, it functions precisely like an icon.

3. Poetic Language and the Metaphysics of Creation

It is just this iconic character of poetic language that I would like to press into service for thinking about the Christian understanding of creation. Here, as before, I wish to root my reflections in the concrete practice of a particular poet, but let me begin with the relatively uncontroversial claim that Christian thought always invites us to see creation itself as an icon: that is, as both what it is and as something other than what it is. Citing Saint Paul's remark in Romans 1 that “the invisible things of God have been clearly perceived through the things that have been made” (Rom. 1:20), John of Damascus writes that in created things “we see images [eikonas], which remind us faintly of divine tokens.”Footnote ²¹ In a very basic sense, John's thought is that creation, simply in virtue of its createdness, always points beyond itself. From one angle, the point is almost definitional: part of what it means to see the world as “created” is to see it as standing in some relation or other to its creator. From another angle, however, John's thought is actually much stronger. For it is not merely that creation stands in some relationship to its creator, but that it stands in a very particular sort of relationship: a relationship of similitude or likeness. For John, quite simply, the created order is an eikōn of its creator, and this means that when we look at creation we see in it what we see in every icon: both a material reality and, through it, an invisible reality to which the materiality points.

For Christian thought, the idea that creation resembles God is rooted in the doctrine of the imago Dei, but it is also guaranteed by the causality of creation itself. Aquinas makes the point like this: “Since every agent reproduces itself so far as it is an agent, and since everything acts according to the manner of its form, the effect must in some way resemble the form of the agent.”Footnote ²² Since all effects, on this view, resemble their cause, and since all created things are effects of God, all created things are in some sense “like” God.

But suppose this is true. What follows? Lots of things, of course, but at least these three. The first, as Gilson points out, is that similarity is “co-essential with the very nature of efficient causation,” and hence that all Christian reflection on creation “requires the use of the notions of similitude and participation.”Footnote ²³ To think of the world as created is already to think of it as “like” God. But, second, the similarity between God and creation is only a similarity. In fact, Christian thought has traditionally held that any likeness between God and creation must finally give way to what Erich Przywara called an “ever greater dissimilarity,” a point codified by the Fourth Lateran Council: “inter creatorem et creaturam non potest tanta similitudo notari, quin inter eos maior sit dissimilitudo notanda.”Footnote ²⁴ But this means, third, that since creatures exist by participating in divine being, and since that participation implies “similarity within ever greater dissimilarity,” creatures exist in a state of unresolved tension: they both do and do not resemble that in which they participate.Footnote ²⁵

Of this dynamic rhythm of similarity and difference the icon is but a particular instance. Indeed, to the extent that icons are visible images that point beyond themselves to the invisible reality they resemble, they are simply focalizations or intensifications of what is true of created being in general. But if the visual icon participates in creation's rhythm of similarity and difference, so too does the rhetorical eikōn. It is no mere coincidence that the Lateran Council's language of similitudo trades in the language of rhetoric and poetics, just as it is no coincidence that Aquinas's dictum “omne agens agit sibi simile inquantum est agens” might also be rendered, “Every agent produces a simile of itself insofar as it is an agent.” This is not to say that either Aquinas or the Lateran fathers intended to evoke the connection between metaphysical and rhetorical similitudo. It is simply to say that the vision of creation to which their comments give voice lends itself to a poetic or metaphorical or “iconic” reading. And this, in turn, is simply because what poets are doing when they are “being metaphorical” in Aristotle's sense is what Christian thought always invites us to do. Indeed, if for the Trojans Helen is a simile (eikōn) of the immortal goddess, then for Christian thought, all creation is in some sense a simile of its creator.

Suppose we say, then, that creation in some sense resembles God. We may then be inclined to ask: “In what sense exactly does creation resemble God?” To get some traction on this question, let us leave the theologians for a moment and turn to another concrete poetic example. “The Word,” by the contemporary Nicaraguan poet-priest and liberation theologian Ernesto Cardenal (b. 1925), was first published in 1989 as part of the much larger Cántico cósmico (Cosmic Canticle), a sprawling, visionary text that seeks, in the words of Cardenal's compatriot Iván Carrasco, to recount the story of the cosmos from Genesis to the Apocalypse in a series of “poetic, philosophical, mythical, and religious visions.”Footnote ²⁶ “The Word” appears near the beginning of the collection, and in the most general sense it is a poem about the relationship between creation and language. The central theme is already apparent in the opening lines:

The biblical phrase “In the beginning was the Word,” which alludes to the opening chapters of both Genesis and John, establishes the poem's basic theme: because the Word was “in the beginning,” and because through the Word all things came into existence, all things therefore “exist in word-form.” The last phrase is crucial. It might, of course, be construed as a compressed, poetic way of saying that all things are in some sense “like” words, but Cardenal is sufficiently theologically subtle to permit a more nuanced reading. Recall that for Aquinas creation resembles God because effects in some sense resemble the form of their causes. A form, in turn, is that in virtue of which a substance actually exists as an individual of certain kind.Footnote ²⁸ The form of “fire,” for instance, is what makes fire exist as fire and not (say) as water. Further, just as things exist according to their form, so also act according to their form. The form of fire, to use the same example, is the source of fire's power or capacity to do what fire does (i.e., burn). Causation works on the same principle. Agent x produces effect y in patient z by “communicating” its form to z. So, for example, fire causes paper to burn by communicating its form (fire) to the paper. Causation itself is therefore the communication of form from agent (agens) to patient (patiens). Further, since likeness or similitude consists in having a common form, effects must resemble their causes simply because when a cause produces an effect by communicating its form, cause and effect come thereby to share the same form.

To say, then, as Cardenal does, that “all things exist in word-form” is not merely to say that all things are in some sense “like” words, but also that all things are “like” words because they owe their existence to the Word, because they were “spoken” into being, and because they therefore share a common “form” with the source of their being. Several stanzas later, Cardenal expands and complicates this idea:

Cardenal is referring here to the passage from Romans cited a moment ago, where St. Paul says, literally, that the “invisible things of God have been perceived through the poems” (tois poiemasin – i.e., “through the things that have been made”). I doubt that Paul intended the pun on poiema, but I also doubt that Cardenal is interested primarily in getting his exegesis right. He is instead interested in getting us to see that if all things indeed exist in “word-form,” they exist in the form of a particular sort of word—a poetic word. And what does it mean for Cardenal to say that things exist in the form a poetic word? This, too, he spells out clearly: creation is like a “poem” because “each thing is like a ‘like’, / like a ‘like’ in a Huidobro poem.” A great deal is packed into this phrase, but perhaps we can begin by noting that the poetic character of creation depends in the first instance upon similarity, where similarity should be construed in explicitly rhetorical terms: creation is a poem because each of its parts has the form of a rhetorical device in which one thing is said to be “like” another. Which is just to say that creation has the form of a simile or an icon.

But this simply returns us, somewhat frustratingly, to the original question: how exactly are we to understand the “similitude” or “likeness” Cardenal has in mind? The answer, I want to suggest, lies in the fleeting allusion to “Huidobro.” Vicente Huidobro (1893-1948) was a celebrated Chilean poet and the leading proponent of an avant-garde movement called Creacionismo. As the name suggests, Creationists were keen to sever the link between poetry and mimesis and to foster instead an aesthetic based on what Huidobro himself called “extraordinary situations that can never exist in the objective world.”Footnote ²⁹ Such “situations” are ubiquitous in Huidobro's poetry, but they perhaps nowhere received more sustained expression than in what is widely regarded as his magnum opus: the intensely futurist masterpiece Altazor (1931). Written in Paris between 1919 and 1931, Altazor (literally: Alto, “high”; azor, “hawk”) resists easy thematic summary, but for our purposes, the poem's language is more important than its content. In fact, though it is impossible to be certain, I suspect that Cardenal's allusion to a “‘like’ in a Huidobro poem” is meant to evoke Altazor; and, within Altazor, one could do far worse than the following passage from Canto III:

And so on for thirty-odd lines. So what is going on here? It is not easy to say, but perhaps we can venture a few observations. The first point to notice is that the lines take the form of a series of comparisons: X is like Y is like Z and so on. This is only partly right, of course, since the comparisons themselves form a kind of interlocking chain in which the vehicle of the first comparison is preserved as the tenor of the following comparison: “a kiss (tenor) like a glance (vehicle),” “glances (tenor) like trees (vehicle),” “trees (tenor) like birds (vehicle),” and so forth. This interlaced pattern lends a certain transitivity to the entire sequence: because aspects of each comparison are preserved in each subsequent comparison, it not merely that one member of the series is “like” another member, but also that every member is in some sense like every other member. Similarity is, as it were, passed down the line. One gets the impression, moreover, that the series itself is designed to go on indefinitely: in other words, that one could, if time allowed, add to the series infinitely many comparisons of the form “To x a y like a z.” (The last line, “Etc. etc. etc.,” makes this possibility all the more tempting.) In fact, one might even produce a series that included among its members every aspect of the created order: every proton, every blade of grass, every star in the sky, every act of love. In that case, we would have something like “absolutely similarity”—not because any one thing is absolutely similar to any other thing, but because absolutely everything is in some sense similar to absolutely everything else.

And yet precisely this conclusion raises a problem. Perhaps I can say what I mean by looking closely at just one of Huidobro's comparisons: poets, he tells us, know how to “play a heliotrope like music.” I understand this line in the sense that I understand what each of the words means, at least as they are used in ordinary discourse. But, even so, I do not think I know what it means to “play a heliotrope like music.” And I do not think I know what that means because I cannot see that the terms of the comparison share what I.A. Richards called a “ground,” an inherent similarity that the simile is designed to bring to light.Footnote ³¹ To take a quick Aristotelian example: the simile “Achilles rushes on like a lion” makes sense because the two terms share an identifiable ground (i.e., ferocity of attack), and so it is easy to see in what respect the one resembles the other. I cannot see that the same goes for Huidobro's comparison, simply because I cannot see that there is a common “ground” on the basis of which a heliotrope could be played “like” music. I assume, of course, that the comparison must mean something, that it must make sense somehow, that there must be some “ground” that holds the terms together. And I do so if only because to assume otherwise (i.e., that it is simply meaningless) strikes me as insensitive to the fact that Huidobro's comparison has the form of a simile and that similes are, by definition, figures of speech designed to indicate likeness.

But this analysis, if correct, puts us in a potentially difficult position. On the one hand, the form of Huidobro's utterance (i.e., a simile) seems to entitle us to assume that its terms are indeed “alike” in some way; and yet, on the other hand, the content of the utterance makes it difficult to see what such “likeness” might amount to. Or, put another way, if we are indeed entitled to assume that poets can “play a heliotrope like music,” it is difficult to see how such a thing could be the case. For precisely this reason, reading the line has the effect of inducing a kind of linguistic vertigo: because the words do not make immediate and obvious sense, the mind searches for a kind of stabilizing context. This sense of vertigo is even more pronounced in Spanish. To take just one example, the verb translated “play” (tocar) also means “to touch,” and, since heliotropes cannot be “played” under ordinary circumstances, the most natural way of understanding the first half of the line is simply “to touch a heliotrope.” The rather unexpected appearance of “music” at the end of the line rules out this interpretation, and in an effort to preserve meaningfulness, the mind almost unconsciously replaces “touch” with “play.” But this substitution solves one problem by creating another. For if music can be “played” but not “touched,” then heliotropes can be “touched” but not “played,” and so the cost of preserving the meaningfulness of either half of the simile is to disrupt the meaningfulness of the simile as a whole.

Something similar, I think, can be said of this entire section of Huidobro's poem. To read it, in other words, is to experience language pushed to its limits, to see words employed in ways that are at once familiar and bizarre, recognizable and strange, meaningful and inexplicable. Perhaps most importantly, because the passage takes the form of a series of similes that seem, paradoxically, to defy our ordinary sense of what counts as similarity, the experience of reading it has the effect of loosening our grip on the very concepts of “similarity” and “resemblance.” What I want to propose, moreover, is that if we read Huidobro's poem in just this way, we have a kind of poetic training ground for understanding one of things Christians may mean when they say that creation is an “icon” or a “simile” of God.

Notice, for starters, that to say that we can assume that Huidobro's similes are meaningful even if we cannot say precisely how they are meaningful sounds very much like certain formulations of the doctrine of analogy. Near the beginning of the Summa theologiae, Aquinas famously argues that since we cannot know God in himself but only on the basis of our knowledge of creatures, it follows that we can name God only on the basis of how our words apply to creatures. But since creatures “represent” God only “imperfectly,” it follows that terms predicated of both God and creatures cannot have precisely the same meaning.Footnote ³² And yet if such terms cannot have precisely the same meaning, neither can they have entirely unrelated meanings. The doctrine of creation itself excludes this latter possibility: creatures stand to God as effect to cause, and effects always resemble their causes. Aquinas makes the point by saying that terms predicated of both God and creatures apply neither univocally (in the same sense) nor equivocally (in unrelated senses), but analogously, where analogy should be understand as a middle term between univocity and equivocity. The difficulty, of course, is that the nature of analogical predication makes it impossible to specify where exactly in the “middle” analogy falls—impossible, in other words, to say how terms predicated analogously of creatures and God apply to the latter. For Denys Turner, in fact, this is practically the definition of analogical predication: “A term is predicated analogically of creatures and of God when we know from creatures that it must be true of God too, but also know that how it is true of God must be beyond our comprehension.”Footnote ³³ The reason is straightforward: since we cannot know God in himself but only from creatures, and since creatures represent God only imperfectly, it follows that the precise sense in which terms drawn from creatures apply to God must remain unknowable. In fact, if we could specify the exact sense in which God is (say) “wise,” we would have direct knowledge of the divine essence. And that, of course, is precisely what we do not have.

A similar line of reasoning opens up a helpful way of thinking about the notions of “likeness” or “similitude” that lie at the center of this essay. When Christians say that creation is “like” God, or that the created order is an icon or a simile of its creator, we say it knowing that it is true because the doctrine of creation guarantees that it must be true. At the same time, however, we say it knowing that how it is true remains an incomprehensible mystery. The similarity of creation to creator is thus what I would like to term an “incomprehensible similarity”: a similarity guaranteed by the metaphysics of creation but whose precise nature eludes our comprehension. It is just this sense of incomprehensible similarity that Huidobro's language dramatizes. His similes, though not themselves theological, are nonetheless case studies in how everyday terms become slippery and unstable when removed from their normal frames of reference and employed in unfamiliar contexts. So yes, the reader of Huidobro may say, poets can indeed “play a heliotrope like music” or “empty music like a sack”—but not in the ordinary senses of “play” or “empty,” simply because in the ordinary sense heliotropes cannot be played “like” anything and music cannot be emptied “like” anything. Analogously, Huidobro's similes function as linguistic and theological training in how everyday terms like “good,” “wise,” or “exists” become slippery and unstable when removed from their ordinary creaturely frame of reference and applied to the source of all existence. So yes, the theologian may say, God “exists,” but not in any knowable sense of “existence”; or yes, God is “good,” but not on any “scale of goodness”; or yes, God is “wise,” but not in the same sense that Socrates is wise.Footnote ³⁴

Again, this is not to suggest that Huidobro treats these themes explicitly, or that his poem is in any discernible sense about theology. It is rather to say that his language is, or can be, a kind of spiritual exercise, a poetic education in abiding the tension between “similarity” and “ever greater dissimilarity” that characterizes theological language. To summarize the point a bit more explicitly, if “being metaphorical” requires the sort of double vision I have described throughout this essay, and if the Christian doctrine of creation likewise requires us to see material reality both as what it is and also as an eikōn or similitudo of its creator, then metaphorical language itself functions as a training ground for “seeing” creation in the way Christian thought demands. Within this context, Huidobro's similes function as a kind of internal critique, a way of guarding against the temptation to assume that we have a firm grip on the meaning of “similarity.” And they do so precisely by alerting us to what happens not only to individual terms (“good,” wise,” “music,” “heliotrope,” and so on), but also to entire concepts (“likeness”) when they are called upon to serve purposes, both theological and poetic, that exceed their ordinary scope. So if creation indeed “resembles” its creator, as Christian thought requires, and if poetic language offers one way of thinking about this idea, Huidobro's similes remind us that asking how creation resembles creator leads neither to greater clarity, nor yet to simple confusion, but instead deeper into the imponderable mystery that attends every attempt to speak the ineffably and incomprehensibly transcendent source of existence itself.