OVER THE PAST TEN YEARS, people in all manner of disciplines have turned to things: to matter, stuff, obdurate objects. Often loosely grouped under the rubric “new materialisms,” these strains of thought have captured the imagination of artists and critics alike. The art world just can’t quit them, apparently—a perverse situation, since art and art history have, of course, already devoted hundreds of years to thinking precisely about objects as objects. But are things really as they seem? In the following pages, scholar ANDREW COLE takes the measure of the two new-materialist philosophies that have come to dominate the art-world conversation, arguing that object-oriented ontology and speculative realism are beset by contradictions, misguided assumptions, and outright fallacies.

A BRICK HOUSE CRUMBLES in the village of Veselovka, Russia, just a few miles from Kaliningrad. It’s said that Immanuel Kant had something to do with this house back when the region was part of Prussia (and when Kaliningrad was known as Königsberg), but what, exactly, is not clear. Ambiguities such as whether the philosopher really lived here didn’t stop someone from regarding the house as his and tagging it with the declaration КАНТ ЛОХ. These words, spray-painted in green and garnished with a groovy heart and a cute flower beneath, were translated in English-speaking media as “Kant is a moron.”

You rarely hear the words irony and Kant used in the same sentence, but what’s ironic about this vandalism is the fact that the house isn’t Kant’s—the existing structure dates from the nineteenth century. Only the foundations are contemporary with the philosopher, who lived in the area in the late 1740s. What we have here, I think, is a vivid illustration of how the critique of Kant—whether inscribed in graffiti or couched in academic prose—usually misses its mark. You will often hear contemporary critics say that Kant is a moron owing to this or that failing of his, but this assessment almost always involves a misreading—a misidentification, as it were—of his philosophy. In such cases, the foundations of Kant’s system remain untouched and solid as ever. You see, even in death Kant is the reigning All-Destroyer—Der Allzermalmende, as his friends called him, ribbing him for his annoying habit of exiting debates completely unscathed and triumphant.

Yes, Kantian moral philosophy leaves something to be desired, as when the philosopher exemplifies the categorical imperative by asking readers to imagine having sex near the gallows—easy to say for a person who never got laid. But Kant’s epistemology, in particular his insight into how we experience the world, remains foundational. He tells us that ours is a world of phenomena, the infinite array of objects and events we experience, and he says also that the world is composed of noumena we cannot experience, the equally infinite number of things that exist, and processes that transpire, apart from our minds thinking them. These two domains are radically different but nonetheless linked, inasmuch as noumena are the basis for the phenomena. Of course, there’s far more to Kant’s “critical philosophy” than that, as we’ll soon see. For example, we can’t ignore such famously unfraught topics as “thinking the unthinkable.” But this is the gist, and enough to get us going.

Our interest here is in showing that Kant doesn’t crumble like his ersatz house (though props to the house for lasting this long). In fact, Kant’s ideas remain a crucial component of recent philosophies that try hard to vitiate his philosophy. Object-oriented ontology is one such philosophy, as is its cousin, speculative realism. What is object-oriented ontology, however? You might surmise that it’s a return to the object qua object—a renewed focus on the composition, vitality, materiality, autonomy, wonder, and durability of objects large and small, near and far. In this sense, you could say that any discipline or practice is “object oriented,” including not only art history and criticism but also architecture, graphic design, museum studies, archaeology, science and the philosophy of science, book history, literary criticism and rhetoric, and the culinary arts—indeed, any field of study whose subject is objects. This crude understanding of object-oriented ontology also applies to speculative realism,which may explain why both have become irresistibly appealing to the art world.

But object-oriented ontology, as it happens, isn’t all that. Instead, it is, well, an ontology—and, as such, involves a set of theses about All That Is. Let’s dive in, surveying its three major tenets. First, everything is an object, including you and each of your thoughts. Second, and accordingly, no object relates to any other object, because the universe itself is devoid of all relation. Why is there no relation in the universe? It’s because objects sever relations with every other object and withdraw into themselves to become self-subsisting, autonomous beings. It’s also because relation is typically a human mode of apprehending, describing, and interacting with the world. Given that not every object is a human, though every human is an object, you can’t have an object-oriented ontology if humans are at the center of it. Such an anthropocentric object-oriented ontology would be a contradiction in terms, because objects are not a means to our ends: They are meaningful whether or not we perceive them. Third—and finally—all objects are equal and, ontologically speaking, on the same plane. You, a speck of flea shit, an electric chair, and a solar flare are all equal objects.

NOT EVERY SCHOLAR, critic, curator, or practitioner adheres to these major points, much less retails them, when declaring an “object-oriented” approach to this or that field of study or aesthetic endeavor. You see this quite a lot: People follow a trend, but only in spirit. For that reason, it might be helpful to think about this new philosophy as a philosophy—to look at the letter of its laws and see how it fares against the likes of Kant, the All-Destroyer. Why this focus on Kant? To be sure, object-oriented ontology builds on the work of several thinkers (Heidegger, Husserl, Whitehead, Latour, even Deleuze), but it devotes much of its energy trying to get out from under Kant. Does it succeed?

That sounds like a rhetorical question. But it’s earnest. For no sooner does Graham Harman, the founder of object-oriented ontology, start to divide objects into different kinds with various sorts of qualities than we begin to wonder whether we aren’t in Kant’s domicile. Harman establishes the “basic elements of an object-oriented metaphysics” in his short, lucid book The Quadruple Object (2011), arguing that, “while there may be an infinity of objects in the cosmos, they come in only two kinds: the real object that withdraws from all experience, and the sensual object that exists only in experience. And along with these we also have two kinds of qualities: the sensual qualities found in experience, and the real ones Husserl says are accessible intellectually rather than through sensuous intuition.”1 So we have a distinction between the “real object” and the “sensual object.” The former is withdrawn, autonomous, and free of all relation, and the latter is available to our perception. Hasn’t this been thought already by Kant, whose insight in the Transcendental Analytic put forward in his Critique of Pure Reason (1781/1787) concerns precisely the difference between the thing-in-itself and the objects of sense perception, the noumena and the phenomena?

The answer is a resounding yes. If we join Kant at the table of categories, we can see precisely how. Kant presents this table quite early in the Critique, gathering concepts from Aristotle, as well as a few of his own, to portray the forms of experience, "all the elementary concepts of the understanding in their completeness.”2 You can regard these forms, or categories, as filtering mechanisms within our minds that instantly translate the random data of the world into coherent and spontaneous experience. Thanks to the categories working in the background, we don’t need to pause and decode our experience every time we look out the window and see a distant tree blowing in the wind, or close a door and hear the latch click. Kant calls these forms “possible experience,” because without them experience would not be possible; we would simply be bombarded by meaningless data. Instead, our minds are always ready to render the world into experience, so that we already know that the swaying tree is actually large but far away, not small but close at hand.

One form of possible experience, Kant shows us, is “relation.” This means, importantly, that relation, as we experience it, is never a quality of things-in-themselves, or noumena. Nor are any of the other categories inherent to noumena. In fact, Kant is very specific about this, saying that these categories “have been incautiously converted from being criteria of thought to be properties of things in themselves.”3 It is technically incorrect to say (as Harman and his followers do) that Kant imagines things-in-themselves as already correlated—i.e., already related—to the subject, or that objects wait around for the subject to hoover up their qualities, as it were, and exhaust them. Nothing of the sort. In fact, Kant’s primary suggestion here is that objects are only partially correlated to our minds, and are so only when they make themselves available to our experience.

However, in object-oriented ontology, and in alleged rebuke to Kant, we are to understand that “reality is free of all relation” and pay obeisance to the “founding principle of object-oriented philosophy,” which “is the insight that . . . objects make no direct link whatsoever” to us or other objects.4 But this is about as Kantian as you can get—and here’s the kicker that exposes a contradiction in this philosophy: We are to turn around and adopt relation as the supreme philosophical category anyway. According to Harman, we can confidently claim that there are “a number of different kinds of relations . . . in the cosmos: ten of them, to be exact.”5 Objects—be they sensual objects, real objects, sensual qualities, or real qualities—enter into relations, after all. These relations are assigned a variety of intriguing names for novelty’s sake, such as fission, fusion, sincerity, allure, theory, and confrontation. And these latter three—allure, theory, and confrontation—are named “tensions” and are illustrated on the page by broken, squiggly lines that crisscross one another as they connect to the four poles of the sensual and the real like so many entangled Slinkys.6

Amid all the excitement about object-oriented philosophy, no one has paused to work out how talk about these new terms for relation is supposed to improve radically on the concept of “relation” in the history of philosophy. The problem is that the original sins of “relation” are not rendered entirely clear in Harman’s and his followers’ writing, apart from glib remarks about poststructuralist relationality, systems theory, and human observation. There’s really no need to overturn the concept of relation in the cursory manner of the object-oriented ontologists, because there’s already plenty in the history of philosophy since Aristotle to instruct us that relation is not always human or correlational, reciprocal, or even fixed or permanent, or anything more than a “moment” of relating that’s always vanishing by dint of becoming and decay. That’s why philosophers in the late Middle Ages commonly distinguished between relationes reales, relations among all entities apart from human perception, and relationes rationis, those relations we’ve reasoned out in our inspection of the world. Kant, for his part, knew that relation is not only aesthetic (what Aristotle derided as the “said-of” of relation; i.e., that relation is what we make of it). Rather, he understood that the problem of relation is exactly the same as the problem of the thing-in-itself: There are relations in the noumenal world, but we cannot think them directly because we have access only to phenomenal relations, the imperfect representations of noumenal relations. The human version of relation, in other words, isn’t the same as noumenal relation, and isn’t the only kind of relation. This idea is all over Kant’s lectures in metaphysics, which none of the object-oriented ontologists seem to know.

The epistemological gambit of object-oriented ontology is to say that object relations are thinkable because they are real, even if withdrawn and unknowable. Realism is obviously what you could call this philosophy—or, as Harman has it, “weird realism.” But realism (weird or otherwise) is a point of view about the world—a human point of view that requires the world to be accessible to us and structured in such a way that we can think it. It’s here that Harman’s ten modes of relation reveal themselves to be equivalent to Kant’s forms of “possible experience.” These ten modes guarantee in advance that, say, an object somewhere will be “sincere” to another object at some point in time, or that an object somewhere will “confront” another object three days from now. Even if we aren’t on the scene, somewhere in Ohio, observing an object indifferently “theorizing” another object, we can know that objects are doing things with other objects and will continue to do so behind our backs. Now, one might say that Harman has simply extended the Kantian forms of possible experience to objects, which thus experience other objects in multifarious ways. That would be partly right, for—according to this philosophy—objects themselves have experiences, as you will see below. But there’s more: The fact that we can also think these object relations means that the relations are already thinkable—already correlated to our minds and thus already something we know about the world. The much maligned “correlationism” that object-oriented ontology hopes to expunge from its thinking is in fact its preeminent feature.

THIS QUESTION of thinking versus knowing is an important one, because it points to further Kantian problems lingering at the center of a purported anti-Kantianism. For example, Harman urges us to reject the idea that “we cannot think something without thinking it”—that is, to reject the notion that we can only think what is available to us as phenomena we experience.7 He asks us instead to perform a thought experiment by thinking what you cannot think, such as the “tree outside of thought.”8 Here, Harman attempts to get the human mind out of the picture entirely by resorting to a realism that assures us that we can think objects as those objects are, outside our minds.

These are intriguing claims because they are Kantian at heart. Kant was obsessed with precisely these questions of what we can think and what we can experience, what is intelligible to us and what is knowable. You could say that one feature of his intellectual biography, ever since the writing of his inaugural dissertation, shows Kant offering a variety of opinions about how we can think the unthinkable noumena. He tells us early in the Critique of Pure Reason that “to think an object and to know an object are by no means the same thing.”9 This distinction between thinking and knowing is crucial for Kant, for it bespeaks the difference between thinking what you cannot experience firsthand, such as the cinnabar outside of thought, and knowing, or “cognizing,” the cinnabar in thought, the cinnabar as you experience it.

Kant goes on to expand the possibilities of thinking what we cannot experience or know. In fact, his entire Critique of Judgment (1790) is nothing if not an exercise in extending the possibilities of thinking noumena of various kinds: positive, causal, worldly, natural, human, and divine. He tells us that to think these noumenal realities, no matter how mundane or sublime, you have to make up your own concepts and, in short, use whatever imaginative means you have at your disposal. He suggests we can think the unconditioned from our vantage point in the sensibly conditioned, using various media as the bridge from here to there: language, poetry, art, analogy, math, and allusion. This is precisely what Harman and his followers claim to have invented: We can think the unthinkable if we adopt “allusion” or other “oblique approaches” to the object world we cannot directly experience.10 Again, that’s about as Kantian as you can get. Similarly, the speculative realist Quentin Meillassoux says that objects have “mathematizable properties” that exist equally in thought as outside it—real properties we can think as they really are. Having studied a few semesters of advanced calculus, I admit that Alain Badiou is right in saying that Kant was terrible at math. But he nevertheless tried his hand at it, attempting to construe extension (after Descartes) as just one way to think the noumenal or, as he also terms it, the “supersensible” (Übersinnliche) world. Meillassoux, for his part, broadens the Cartesian insight about extension into a whole thesis about our mathematical perception of objects, but writes as if Kant had never attempted this himself.11

The only answer I have for why two voracious readers of philosophy, from whom many others receive lessons about the history of Western thought, would exhibit such a partial view of Kant is that either their reading of Kant is incomplete or they know this about Kant already but aren’t telling. I’ll abstain from answering definitively, because—believing in dialectics—I think it’s a little of both. Harman’s Kant is only the Kant of the Critique of Pure Reason; Meillassoux’s Kant is the Kant of the critical philosophy. Meillassoux in fact embraces a “pre-critical” philosophy in order to “revise decisions often considered as infrangible since Kant.”12 That we are supposed to look to precritical philosophy as an end run around the critical Kant is a compelling idea for theorists who can sense the persistent médiévalité or “dogmatism” of modern thought. But—and this is a pronounced problem in theory more broadly—it is wrong to uphold the distinction between the pre- and postcritical, insofar as Kant himself violates the distinction left and right in his work up to the very end. Contra Meillassoux, we can embrace the critical Kant and still think the unthinkable, if we so wish.

THERE'S MORE TO the curriculum of object-oriented ontology. Each object, no matter what it is, is abstracted in the same way. Each, that is, conforms to a template: All objects have insides and outsides, interiors and exteriors, depths and surfaces, and—especially—essences and accidents. It’s here, in the ontologist’s very idea of the object, that another contradiction in the philosophy appears. As Heidegger, Merleau-Ponty, and Deleuze have shown, each in his own particular style, the construct of inside/outside in any ontology, whether it concerns objects or subjects or both, is a function of the old subject/object dualism, which is a dualism precisely because there is a sovereign subject around to proclaim what makes the cut, what qualifies as an object or not—and whose very perspective on that object determines what is “inside” and what “out.” The point isn’t that object-oriented ontology unwittingly centers an autonomous subject at the heart of objects simply in the way it tells us to look at objects. Rather, it’s that Heidegger had already created a philosophy whose very purpose was to destroy these old ontological constructions of essences and accidents, interiors and exteriors. Such terms, he thought, obstruct the genuine thinking of Dasein, or being there in the moment. This is important because object-oriented ontology claims to be a Heideggerian philosophy based on select passages in Being and Time (1927). Likewise, the reader (i.e., Harman) who could discover in Heidegger’s opaque essay “The Thing” (1950) a schematic for “quadruple” objects must have missed this philosopher’s poetic discussion of the jug in that very same paper, along with his caution about realist perspectives on objects: “Science represents something real, by which it is objectively controlled. But—is this reality the jug? No.”13

Regarding anthropomorphism and anthropocentrism, object-oriented ontology can’t seem to avoid these perspectives after all: The aforementioned modes of relation—fission, fusion, allure, sincerity, theory, confrontation, and so forth—are themselves human-centric. Some of these terms read like authorial observations on the fickle characters in a novel—Dickensian descriptions of who is sincere, who is confrontational, who is alluring, and so forth—while others, such as theory, denote contemplation in its most humanly reclusive form. Yet other terms, such as fusion, recall problems my friend Charlie is trying to solve in the plasma lab where astrophysicists are creating a miniature star. Fusion evokes the mysterious supersensible processes in our sun as much as it conjures up the human effort to duplicate and harness these processes before we destroy the planet with our capitalism and carbon emissions. Maybe it’s just me, but fission prompts a similar thought—this time about whether an atomic bomb is really “equal” to a doorknob, or the blasts at Hiroshima and Nagasaki “equal” to a unicorn. This term requires some serious epoché, or mental bracketing, for us to picture the pure object relation it’s meant to describe.

Can we bracket our thoughts in this fashion, when the father of this practice, Husserl, could never do it himself, describing (for example) a piece of paper on his desk but never thinking the paper as the paper really is? To be sure, these new terms for relation in object-oriented ontology are meant to help us decenter ourselves as we reflect on objects relating to other objects, forcing us to realize that a rock cares not a whit about the difference between a nuclear weapon and a doorknob when it “confronts” them. But we access this object relation by thinking about the name of the relation, and all of these names are decidedly human, cultural, social, and literary—that is, the names are in fact predicates, chock-full of meanings you can’t unthink or bracket. Contemplate fusion or fission, for example, and you’ll soon encounter the problems of production and technology right where they shouldn’t manifest—in the object relation. Or think about confrontation and you’ll eventually face the politics of what it means to “be” autonomous, right where issues of freedom or necessity shouldn’t appear—in the object relation. This problem extends to the ethics of why we should even think a thoroughly reified world, as called for by object-oriented ontology, or why we should even fantasize about objects as scholastic assemblages, Erector Sets of the imagination.

AT THIS STAGE, you could throw up your hands and just admit that Kant was right: There are object relations, yes, but we can’t really know or describe them in detail, only allude to them in our inevitably human way. Or you could press on, chalking up these considerable difficulties in naming to the problem of language and solving them by taking a page from Heidegger, who uses neologisms to refresh the addled language of philosophy—though who really wants to hear more jargon? You could also consult your local analytic philosopher, who will tell you that metaphysical mistakes are mistakes in natural language: Artificial languages, anyone? In any event, object-oriented ontology may, as a philosophy, want to decenter the human, but as a language—and perforce as a way of thinking—it expands the human into all relations, raising serious political and ethical questions along the way, but never answering them.

Names, characters, objects, and, of course, quirky lists of things, like aardvarks, baseball, and galaxies; or grilled cheeses, commandos, and Lake Michigan—these (“Latourian litanies,” as they are called) salt the prose of every object-oriented ontologist. They humanize the philosopher. This is a people-person’s philosophy, after all, in the sense that objects are people too. Or, to be technical as well as to the point: Objects are subjects. In describing the life of objects, Harman offers up what he calls a “speculative psychology,” which holds that “primitive perception is found even in the nethermost regions of apparently mindless entities.”14 “It is not true that the psychic pertains only to the animal,”15 he argues, so what’s needed is a category of experience “applicable to the primitive psyches of rocks and electrons as well as to humans.”16 The effort here is to extend consciousness to the object world and to regard experience itself as the result of objects grinding up against one another: “Experience is nothing other than [the] confrontation of an experiencing real object with a sensual one.”17 These quotes speak for themselves and confirm what the great modernist poet Marianne Moore once told us: “It is human nature to stand in the middle of a thing.”

Marx knew a thing or two about human nature as well—especially our tendency to personify objects. In his great work Capital, he speaks about commodities and, in a memorable passage, talks about that table you just have to have, especially now that it’s on sale and would look so good in the front room. (Who cares if we never have guests over for dinner. I need this table!) The table has a certain allure, Marx tells us, thanks to its “metaphysical subtleties” and complex of “properties.” We don’t know how those properties satisfy our needs and wants—what is it about the wood? or the shape? Likewise, we can’t fathom what those properties tell us about how the table was made—by whom and under what labor conditions. But the table has meaning for us nonetheless and becomes ensouled under our gaze. We so admire the table as a commodity that it magically “changes into a thing” like no other. “It not only stands with its feet on the ground, but . . . stands on its head, and evolves out of its wooden brain grotesque ideas, far more wonderful than if it were to begin dancing of its own free will.”18 This is what Marx calls the “mystical character of the commodity”—mystical because we can only think the object’s inner properties by personifying it in a focus so narrow that we ignore the larger drama, the greater historical process that makes a commodity a commodity, an object an object, and capitalism capitalism. This is, in short, the metaphysics of capitalism, then as now.

Ours is a time when schools of interpretation ask us to personify and caricature objects as autonomous and alive—whether they are the objects who “speak” in the new so-called vibrant materialism, or objects who fuss and act up in actor-network theory, or objects with “primitive psyches” in object-oriented ontology. Is this really the way to think at this moment? For Marx, at least, this way of thinking about objects is what keeps capitalism ticking. To adopt such a philosophy, no questions asked, is fantasy—commodity fetishism in academic form. To identify such philosophy as the metaphysics of capitalism is theory, ever attentive to history’s impress on our imaginations, whatever we may dream.

Andrew Cole, a professor of English at Princeton University, is the author of The Birth of Theory (University of Chicago Press, 2014).


1. Graham Harman, The Quadruple Object (Alresford, UK: Zero Books, 2011), 49.

2. Immanuel Kant, Critique of Pure Reason, trans. Norman Kemp Smith (New York: St. Martin’s Press, 1965), 116.

3. Ibid., 118.

4. Harman, The Quadruple Object, 47, 128.

5. Ibid., 108.

6. Ibid., 107, 114–15.

7. Ibid., 62.

8. Ibid., 65–66.

9. Kant, Critique of Pure Reason, 161.

10. Harman, The Quadruple Object, 98.

11. Quentin Meillassoux, After Finitude: An Essay on the Necessity of Contingency, trans. Ray Brassier (London: Bloomsbury, 2013), 76, 31, 35.

12. Ibid., 26.

13. Martin Heidegger, “The Thing,” in Poetry, Language, Thought, trans. Albert Hofstadter (New York: Harper & Row, 1971), 170.

14. Harman, The Quadruple Object,103.

15. Ibid., 110.

16. Ibid., 103.

17. Ibid., 133.

18. Karl Marx, Capital: A Critique of Political Economy, vol. 1, trans. Ben Fowkes (Harmondsworth, UK: Penguin, 1976), 163–64.