The Five Tensions of Graham Harman’s Ontology

Although it’s a short book, The Quadruple Object by Graham Harman probably has more terminology than anything else he’s written. In it, he includes condensed summaries of all his major arguments about Heidegger, occasionalism, causation, correlationism, and levels of scale. Then he introduces new key details about his title theme, the “quadruple object” of real and sensual objects and qualities. As he argues, the main interest of this model is not to be found in the four individual poles but in their possible pairings. That’s ten named permutations.

Four of them are special: the engine of change in Harman’s cosmos. They are the object-quality pairs. He ends up distinguishing between their everyday states and special events that produce tension. That’s eight possibilities!

Or is it? I will argue there can only be five according to the picture of the quadruple object that Harman himself lays out. This will then lead me to make one major-ish claim about his concept of “time” that differs from other critiques of his time/space distinction. (A teaser: I think what he calls “simulation” in some places and “confrontation” in others is exactly what his model means by time.) Keep in mind that I am just trying to come to terms with someone else’s ideas here, and I think of my arguments in Part 2 here as the conceptual version of line editing or fan theory rather than substantive philosophical disagreement.


I can’t give all the background anyone will need, but here’s a bit. For Harman, the world at the most general level is nothing but objects. The word “object” does not mean whatever you or I might normally think it means. It isn’t a correlate of a subject, for example, or an immortal abstraction “subsisting always in a self-same condition of being.” That quote comes from Rebekah Sheldon in an essay called “Form / Matter / Chora,” and Peter Gratton has a similar take. Gratton supports his belief that Harman is Plato 2.0 with reasons that would take another paragraph or two to address, but Sheldon can be answered by Harman’s direct prior contradiction of her claim. “Every relation, every object, is already born into a state of disintegration—for it immediately sets up an internal space in which the union of its parts is no longer entirely unified, in contrast with its sleek unity from the outside” (Guerrilla Metaphysics, 202).

What is an object, then? Harman’s object is any perishable thing that resists the context of its interior (historical conditions, pieces) and the context of its relations (effects, functions, perceptions). In other words, objects are singular things that can be viewed from different perspectives and can withstand some changes to their parts while remaining the same being.

Irreducibility to context does have consequences. For example, it means that objects must be endlessly divisible into other objects. A more pressing consequence is that the interconnection between objects is not a foregone conclusion. Causation has to be earned in OOO rather than asserted by fiat. For how can individuals that are independent of their own parts and relations be said to influence each other? Harman ultimately needs to articulate some indirect form of influence between objects. He needs a medium where all the action happens.

His model of the quadruple object addresses this issue. It is based on the intersection of two dualisms. One is between the real and sensual. The “real” is that which exists independently of any relation. A real object can only enact its own being. No other beings have a direct line of access to it. But they do touch something. An object translates whatever it encounters in terms of its own form. Harman’s term for such translation is the “sensual.” Indeed, the sensual object is the sole point of exterior contact for real objects.

The other polarity is between objects and qualities. An object is unified, but it is also specific, with a plurality of features that differentiate it from other things, and some of those features can change over time without destroying the object. Harman holds that this object-quality polarity persists in both sensual experience and the real, though the two situations are not precisely analogous (which I will get to shortly). Combining the two dualisms yields the quadruple object.

Now, the basic problem for OOO is how real objects can reach the real in experience since they have no direct contact with one another. The answer comes down to the dynamics between objects and qualities. As Harman writes: “In order for something to change in the status quo, the bond between object and quality must be dissolved and a new one produced.” There are four possible object-quality pairings: SO-SQ (“time”), RO-SQ (“space”), RO-RQ (“essence”), and SO-RQ (“eidos”). Note that they all take place in the context of experience, i.e., an attachment between a real object and a sensual object. These pairings are not just facets of an isolated object as viewed from the outside, but the interior workings of causation. Causation at the scale of an object results from special forms of translated reception, or aesthetics. Objects are only “active” or “free” via their parts, that is, through the composite attachments that form them.

2. THE TENSIONS: Four Plus Four?

I’ll try to characterize these tensions briefly.

(1) Harman calls SO-SQ “time” because it describes a unity that endures through a multiplicity of profiles or adumbrations. (That’s Husserl’s term, which already connotes a temporal dimension with the idea of “shading off” as opposed to an instantaneous configuration.)

(2) RO-SQ is “space” because it is the palpable distancing of the object, an accessible signal alluding to a receding reality (and this goes back to the critique of Heidegger’s notion of temporality in Tool-Being).

(3) RO-RQ is “essence” because it draws the thing together with what it essentially is. But keep in mind both terms are not directly accessible: we cannot know an essence.

(4) Finally, SO-RQ is “eidos” because, like Husserl’s notion of eidetic variation, it alludes to the features that truly make an experienced object what it is. Again, though, these features are never directly accessible in Harman’s view, not even to an intellect as Husserl claims.

One of Harman’s most interesting choices in The Quadruple Object is to divide the four tensions into two states: a normal everyday “banal” state, and a disruptive tension-producing state. Perhaps the most useful outcome of this intuitive distinction is that it pushes Harman to explain how tensions involving a sensual object (SO-SQ and SO-RQ) work differently from those involving a real object (RO-SQ and RO-RQ). As I’ve implied, I think the symmetrical unfolding of the tensions into eight total possibilities should be abandoned. But let’s see how it works first.

Normally, a real object is cut off from its parts and relations. The RO-tension arises in special circumstances. With RO-SQ, for instance, “real object and sensual qualities will meet only when fused.” Fusion either happens or doesn’t. There is no in-between.

Turning now to the SO-tensions, everything gets inverted. The sensual object is normally bound together with its qualities. In this banal state, it is said to persist in a clean continuum with its shifting profiles. It is taken for what it seems to be. But in special circumstances the object-quality bond splits. In other words, the object-quality relation becomes “at issue,” difficult, dissonant. Fission (or friction, if you prefer) is like fusion insofar as it either happens or doesn’t. In such moments, qualities call attention to themselves and stand tensely next to the object.

Right away, we should see that there is no difference between the two banal SO-tensions. In Quadruple Object, Harman explains that the sensual object is normally fused with its own eidos (real qualities) just as it is with its accidents (sensual qualities). Thus in its banal state the true nature of the sensual object is supposedly/falsely known, or taken for granted. Harman’s term for the explicit form of SO-RQ is “theory.” But since SO-RQ in its banal state is the same as the banal form of SO-SQ, we seem to have too many terms. Time (SO-SQ) matches one tension, while eidos and theory match another. In other words, eidos and theory are the same thing. The only difference between them is the connotation that one is active and the other passive. Activity is exactly what Harman is trying to explain, though, so this is a terminological rather than an ontological distinction—or else it is a separate and unexplained ontological distinction altogether.

Next, we come to the RO-tensions. Harman himself writes that “there is actually no banal state of tension when a real rather than sensual object is involved” (p. 108), and I am inclined to believe him. That’s because the real object is normally separated from other things. The RO-tensions must be produced, therefore we have no way of distinguishing between their normal and explicit states. Insofar as there is any tension at all, “space” matches “allure” (RO-SQ), and “essence” matches “causation” (RO-RQ). Once again, the only difference is the connotation of passivity on one side and activity on the other, but this is just terminological. Space is every bit as active as allure!

Overall, then, we have exactly one banal tension, the SO-SQ link. Harman calls it “time,” in which the shimmer of accidents is mediated by an object. As Harman explains on p. 102, this means the object “must always be accompanied by a swirling patina of sensual qualities and a not yet articulated core of real ones.” Harman here takes up the strict line he holds in Guerrilla Metaphysics, where objects cannot be a bare unity and qualities cannot subsist independently of object mediation. This is Object-Oriented 101 and will only get the stamp of approval from readers such as myself. But I also think that at moments like this Harman forgets the context of causation that guides his banal vs. explicit distinction in Quadruple Object. Harman in the above quote and elsewhere sometimes seems to describe the object-quality tension per se, rather than the banal (versus explicit) species of the tension in the context of an attachment.

What am I making such a big fuss about here? Well, the five vs. eight issue is largely a matter of rhetorical confusion that I think Harman has quietly (and expertly) smoothed over starting with Weird Realism. But it has allowed a slightly deeper confusion to persist. I think Harman is wrong to call the banal version of SO-SQ “time,” and not for any standard reason such as my wanting to protect another definition of time or my wanting to re-inject Harman’s version with anticipation, history, or whatever. Indeed, I accept his implicit point that concepts such as protention and retention ought to fall under the tension of “space” and that all the rest amounts to a tussle over semantics. (This becomes explicit in The Rise of Realism, a recently published conversation with Manuel DeLanda, in Part V.)

My reasoning that SO-SQ in its banal form cannot be called “time” is internal to OOO. Object mediation only ever becomes at issue when the object stands in some kind of tension with its accidents. Otherwise, objects are “elements,” a wonderful term from Guerrilla Metaphysics. That is, objects and qualities are compressed together. A banal cup of coffee in my experience does not endure through a patina of swirling accidents. The cup is part of a patina; it participates in the endurance of some other object of experience, such as the composition of a paragraph at my computer or my drinking from the cup. And insofar as it endures in my experience, the cup is an object with qualities, an explicit SO-SQ and thus a product of fission. In such moments where my cup has time, it is also “simulated” or “confronted.” When Heidegger complains about presence and others dismiss objects as being too abstract or atemporal, what they really have in mind in terms of object-oriented thought is the banal form of SO-SQ, which indeed refers to an “objectified” atemporal lump of presence.


To summarize: Harman lists four normal or banal tensions and four explicit versions of them. Against this, I believe there is just one banal tension where an object compresses with its qualities. Whatever we want to call it (heck, “banality” works for me, or Heidegger’s “presence” if I understand it correctly), the term “time” should be reserved for the explicit version of SO-SQ. I’ve recently been convinced that a better term might be “duration,” from Bergson. In any case, I think the model described above remains true to Harman’s object-oriented philosophy, and simply aims to clarify the tensions. Fusion always comes with fission in OOO, so time and space (or duration and allure) should now be presented as active partners. Even the dullest vanilla endurance of an object in experience points to changes taking place in the configuration of the world. Once more: “Every relation, every object, is already born into a state of disintegration—for it immediately sets up an internal space in which the union of its parts is no longer entirely unified.”

Am I totally content with this revision? I don’t know. One pressing issue is the binary yes/no nature of both time and space. If time is now to be understood as a yes-or-no affair, how could it be called continuous? Harman of course does not have to worry about this question unless he were to accept the above critique, because his presentation of OOO implies that banality and continuity are synonymous and “time” only becomes significant by way of its spatial fissure.

Note, however, that something similar to the above question could already be posed about Harman’s presentation of space. If space must be “produced,” how is it maintained? I think all I have done is leveled the field so that an answer to one also addresses the other.

Here are some initial thoughts on the matter. (Of course, keep in mind with all of this that space and time are not meant in a literal sense, whether explicitly as measured-space and clock-time or implicitly as practical-space and felt-time.) In the context of the Quadruple Object, to say time is binary is NOT to say that it should no longer be called continuous. Rather, time is binary in the sense that an enduring attachment to another object either happens or doesn’t. The gradual shading of a sensual object describes the interior of an active SO-SQ tension, not the banal mode of that tension. But I am not certain what it means to say duration occurs simultaneously with allure.


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 )

Google photo

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

Twitter picture

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

Facebook photo

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

Connecting to %s