Category: Theory

Graham Harman, commenting on Jane Bennett’s Vibrant Matter:

It is certainly true that context and relation can affect the reality of an individual thing. It does not follow that each last detail of a context or relation changes the thing that is involved with them. An individual is a kind of filter (or “firewall,” as I often call it) responding to some relation partners but not all. To be affected by something outside us is a special case, even if not a rare one. Countless things happen around us without this entailing that our reality registers each tiny fluctuation in such a way that it changes who we are.

This is the closest I have come to agreeing with Harman about objects as individuals. I still want to argue for promiscuous interrelations among objects, rather than seeing them all as vacuum-sealed; but here, my only qualification would be that I think that every entity makes a “decision,” as Whitehead puts it, as to which “relation partners” (Harman’s phrase, not Whitehead’s) it responds to, and which it ignores. In Whitehead’s parlance, this ignoring another entity could take the form either of what he calls a “negative prehension” (which is a decided refusal) or of the fact that the other entity has only a “negligible” influence on the entity that is making a decision. So, while I think that “to be affected by something outside us” is the general case, rather than a special one, in practice the degree to which an entity is affected is fairly minimal.

Harman further remarks that “Unless a philosophy can account adequately for the fact that not all changes make a difference, then its sense of individuals is too weak.” And again, I mostly agree. But I would argue that this condition is met by Whitehead’s claim that, although in principle an entity is affected by all the other entities in the universe (or at least in its light cone), in many cases  (and probably in the overwhelming majority of cases), this influence is negligible.

I still differ with Harman in thinking, following Whitehead (who in this case is himself following William James), that the existence of an entity is punctual, and that the endurance of an object through time needs to be understood as a succession of entities, with a large measure of inheritance accounting for the continuity. This is why (as I said at the OOO conference last week — but this part of my talk still needs some revision) the question of whether an entity remains “the same” over time is a relative one, a matter of degree.

Discussing object-oriented philosophy, Voyou Desoeuvré suggests:

It’s worth disentangling a number of different ways in which objects could be thought to be “real.” First would be to maintain that objects cannot be reduced to their components, either physical or sensory (that is, there really is a chair over there, not just an aggregate of atoms or sense-perceptions). Second would be maintain that these objects exist independently of human minds, knowledge or perception. Third, this could be expanded to get away from a human/object binary, and so maintain that objects are independent of other objects: that in each interaction of an object with something else, there is something in that object over and above what is involved in that interaction. Fourth, one could universalize this position, saying that, not only is an object never completely involved in any particular relation, but that objects are withdrawn from all relations, that their core being is not involved in any relations at all.

And he goes on to question whether the fourth of these steps is necessary. I find this clarifying. I too go along with Harman on steps 1-3, but I reject step 4. My reason for doing so is not quite the same as Voyou’s, but it is obviously this that I need to work out more fully.

Harman argues, I think, that step 4 is necessary in order not to reduce objects merely to the sum of their qualities. In fact, I am even willing to accept this provision, but still without moving on to step 4. Here is where Harman would probably find my position contradictory or impossible; if I agree to steps 1-3, and if I also agree that an object is not to be identified with the sum of its qualities (which means that it cannot be grasped even as the sum of all the possible encounters other objects could have with it), then am I not forced to accept the object’s independence of relations, as posited in step 4? Otherwise, I am evidently in danger of falling into the infinite regress of Lewis Carroll’s logicist version of Achilles and the Tortoise. Nonetheless, I still find myself resisting at this particular point.

Unsurprisingly, I think that Whitehead successfully pulls off the balancing act that is necessary in order to accept object independence without denying relationality. It has to do with the way that an entity’s prehensions are always partial (or negative as well as positive), and that this bundle of prehensions is only unified (and therefore only generated) by a process of selection among eternal objects.

(God makes the eternal objects available, but he doesn’t have any influence over what selection is made — and this is why I don’t think Harman is entirely right to lump Whitehead with Leibniz as a theological occasionalist. Whitehead’s occasionalism — if that is what it is — may not be as fully secular as Latour’s, but the peculiar marginality of his concept of God is explicitly presented as a critique of Leibniz’s God; it moves precisely in the direction of “secularizing God,” instead of in the much more familiar direction of altogether abolishing him. I think — as I wrote in my book — that there is more to be said for this strategy, even for an atheist such as myself. It may even be a better way to resolve certain problems, such as the infinite regress of mediators, that Harman has with Latour’s own version of secular occasionalism).

More generally, I resist Harman’s move to Voyou’s step 4 because (as I already say in my article on Harman coming out later this year in The Speculative Turn — I think this move omits, or renders inessential, all considerations about an object’s becoming and perishing. That is to say, it doesn’t take the radical temporality of objects or (or of what Simondon calls individuals, or of what Whitehead calls actual entities) seriously enough. Harman resists all talk of potentialities (or of Deleuze’s virtual), and of processes of becoming, as undermining (or of “overmining”) the reality of objects. But I don’t think this need be the case. I would urge the question of potentials and becomings as as a better reason than the one Voyou gives (he interestingly cites Hegel and Berkeley) for staying with steps 1-3 without going on to step 4. What this comes down to, in “speculative realist” terms, is to read Harman together with Iain Hamilton Grant.

Harman is always saying that Whitehead does not at all belong with Bergson, Deleuze, and Grant, because they dissolve objects into flux, whereas Whitehead (despite his relationalism) insists on precisely delimited individual entities. But the whole point is to specify how we flip from one of these situations into the other; and that is something Whitehead, with his dual aspect notion of entities (public/private, mental/physical, etc.), does better (I think) than either Harman (who emphasizes only the closed-off entity) or Deleuze (who emphasizes mostly the flux) is able to do by himself.

What I have here is not (yet) an argument, but only a note to myself about the sort of argument that I want to make. I hope to work this through better by the time of the Object-Oriented Ontology conference coming up at Georgia Tech this spring. I think the differences between Harman, Deleuze, Whitehead, etc, need to be grasped as aesthetic oness (agreeing with Harman — and also, I think, less explicitly, both Deleuze and Whitehead — that aesthetics is indeed “first philosophy.” I am also trying to think more about how neo-vitalism and panpsychism intersect with “speculative realism,” and in particular with those branches of it that do retain (even in what might seem to some to be the vitiated form I am advocating here) object-orientation of some sort.

Nathan of An Un-canny Ontology, responding to the same posts by Levi Bryant that I cited in my Tumblr workblog, asks the question: “What exactly happens during translation? What is translation? And why do some things get translated and others do not?” After mulling over this question for some time, Nathan concludes “that objects predict, expect, or anticipate other objects – they recognize potential.”

Now, I am not sure that this is the right answer — or, at the very least, I would argue that it isn’t all of the answer. Nathan makes this claim because, for instance, “for leafs [sic] to translate photons of light into complex sugars, they must recognize the photons of light as photons of light.” I suppose this is true in a sense: leaves will not — cannot — translate just anything into complex sugars. But I don’t see why “recognition” has to be the precondition. If anything, I’d say that the leaf’s “recognition” of the photon is a consequence of, rather than a precondition for, its “translation” of light into sugar. Re-cognition, and indeed any form of cognition, always comes afterwards; it is the error of cogntivists (which we human beings, unavoidably misunderstanding ourselves, tend to be much of the time) to think that cognition is a ground of action, when actually it is a result of action.

I think that the source of this problem, in Nathan’s account, is the following. He says that ” objects first and foremost recognize each other,” precisely because — here paraphrasing Levi, and also to an extent Graham Harman — “objects translate each other, they change each other without encountering each other directly.” But as I’ve said before, my biggest disagreement with both Levi and Graham is that, for me, objects do encounter each other directly. (Whitehead’s actual entities are a bit like Leibniz’s monads, but actual entities touch each other directly, as monads do not. Cf. also Gabriel Tarde, who posits monads that — unlike Leibniz’s — interact with one another directly).

Levi puts it this way:

One of Harman’s core claims is that objects withdraw from one another or never directly encounter one another. This is the Kantian moment in Harman’s ontology. Where Kant holds that we never have direct access to the thing-in-itself, emphasizing the relationship between mind and thing-in-itself, Harman generalizes this thesis to all relations between things, regardless of whether or not humans are involved. This is precisely why Harman’s ontology, despite being an ontological realism is also an epistemological anti-realism. In my own ontology, I refer to this general feature of things with the concept of “translation”. As Gadamer (and Quine) taught us, every translation is a transformation.  (from this post)

I largely agree with this (as I’ve said before, here and here). I think that it is precisely right to generalize what Kant says about the mind’s encounter with external reality to all interactions between/among objects. However: unlike Levi, I am unwilling to equate Kant’s argument for the cognitive inaccessibility to the thing-in-itself with the thesis that “objects never directly encounter one another.” This is because contact or encounter cannot be reduced to cognitive access. In Kant’s account, we are affected by things-in-themselves, even though we can never know them. This is indeed the source of one of the most-remarked problems with Kant’s thought: he seems to be saying that, in some sense, things-in-themselves cause our perceptions of them, even though he explicitly says that causality is merely phenomenal (i.e. merely produced by the way our minds organize our sensations). There are two ways to resolve this dilemma. One is Hegel’s and Zizek’s way, which absolutizes Mind or Spirit or Subject, by saying that even the inaccessibility of things-in-themselves is in fact posited by the Mind in the first place. Obviously, I find this undesirable. The other alternative — or, more precisely, the move in the opposite direction — consists in distinguishing the way things affect other things from “causality” understood as a Transcendental Category (i.e. roughly, as a form of cognition). Causality, as a cognitive category, isn’t adequate to describe the way that the mind is non-cognitively affected by things-in-themselves. Or — to make the speculative realist generalization — causality, as a cognitive category, isn’t adequate to describe the way that an object affects, or is affected by, another object.This is one way of describing Whitehead’s distinction between “causal efficacy” (what I am calling non-cognitive affectivity) and “presentational immediacy” (which, for Whitehead, means the type of causal connection discussed by Hume and by Kant).

So I agree with Levi and Graham that an object never cognitively grasps any other object in its entirety. (This is what Levi calls epistemological anti-realism). But I disagree with their move of equating this cognitive inaccessibility with the claim that objects never directly encounter one another. My non-vicarious version of ontological realism consists in claiming that objects do directly encounter (or affect) one another — only they do so non-cognitively. This is precisely why our ontology can be realist, even when our epistemology is confessedly anti-realist. The translation that happens in every encounter between objects — i.e. when, in Whitehead’s terms, one object prehends another object — is a direct, but non-cognitive, encounter (in Whitehead’s terms, it is a process of feeling, in which an “actual entity” determines itself by making a “decision” about how it will feel that which moves it to feel. An object functions for another object, Whitehead says, as a “lure for feeling”).

[I know that Levi and Graham won’t agree with my account here, and probably Nathan won’t either. But none of this would have come clear to me — to the extent that it has come clear — if not for my puzzling over what they wrote].

I delivered my paper critiquing Graham Harman at the SLSA conference the other day. But here I want to address one of the ways in which I have been stimulated by Harman’s ideas.

In one of his recent posts, Harman usefully critiques the correlationist claim that you cannot think the unthought, or that “to think things-in-themselves converts them into things-for-us,” because by the very act of referring to something ostensibly outside thought you are therefore bringing it within thought. [This claim is parallel to the equally facile claim that you cannot coherently affirm relativism, because by the very act of affirming it you are thereby making an absolute, i.ee. nonrelative, statement].

But Harman points out that “you don’t just have the options of saying something or not saying it. There is also a way of saying something without saying it: we allude to it.” In this way, we can reference, or refer to, or “point to” something that we cannot access directly, cannot see or say. We are never really stuck with the early Wittgenstein’s dictum that “what we cannot speak about we must pass over in silence” — because in practice we actually are always speaking in various ways towards, around, and about “what we cannot speak about.”

One can relate this to Levi Bryant’s recent suggestion that the object-oriented ontology espoused by him and by Harman consists in an anti-realist epistemology coupled with a realist ontology. Another way to put the common correlationist claim is to say that anti-realism in epistemology (or the simple recognition that things are not altogether as they “naively” appear to us) entails anti-realism in ontology as well. But the possibility of allusion, or of metaphor, or indeed of any non-literal use of language and of other modes of expression (pictures, musical sounds, etc.) allows us to escape the correlationist claim, and to be realists about “things in themselves.”

In his book Guerrilla Metaphysics, Harman writes of allusion and metaphor, and this leads to discussions of humor, tragedy and comedy, and charm, and allure. These aesthetic discussions are among my favorite things written by Harman. Aesthetics, as Harman develops it, is both a way to break out of the charmed circle of correlationist epistemology, and a broader way of discussing how objects interact with other objects on all scales. That is to say, aesthetics is not just a human attitude, but a primordial form of relation and interaction. And this leads Harman to suggest, in a lovely (and justified) hyperbole, that “aesthetics becomes first philosophy” (“Vicarious Causation, in Collapse 2).

Now, I find this sort of approach useful and liberating from my own Whiteheadian point of view. Aesthetics describes what Whitehead calls feelings: i.e. the ways that objects affect, and are affected by, other objects, even (and especially?) when there is no cognition going on. The failure of epistemological cognition does not mean the impossibility of ontological interaction. Aesthetic modes of expression correspond to “vicarious” (in Harman’s sense) as well as to noncognitive (in a Whiteheadian sense) modes of interaction — they are ways of positively expressing “what we cannot speak about.”

So I find Harman extremely valuable on this point of aesthetics — even though I see objects as continually jostling up against one another, “prehending” one another, i.e. primordially relating to one another and defining themselves by means of the multiplicity of their relations — a view which (as I have noted before) is very far from Harman’s vision of objects packed away in vacuums, unable to touch one another except “vicariously.” But Harman’s vicarious relations and Whitehead’s promiscuous ones can both be described aesthetically first of all; the difference between them might even be seen as a difference in aesthetics (a suggestion that I begin to make at the end of my paper, and that I am extending here.) Bruno Latour writes of the different modes of existence; what’s needed, similarly, is an account of the different modes of aesthetic expression, which would also point to different modes of object interactions. In the past I have tended (like most aestheticians) to fall back upon the old opposition between the beautiful and the  sublime (an opposition that Kant codified, but that long pre-existed him); but I think that we need a more nuanced and varied account of aesthetic modes (and presumably, one that would not presume to enumerate all the possibile modes of aesthetics a priori, but would instead work simply by listing and describing, with the understanding that it might always be possible to add new items to the list.

[I should note that the Latour article I linked to above is an account of a long-out-of-print book, Les différents modes d’existence,  by the long-forgotten French philosopher Etienne Souriau; and that, thanks to the efforts of Latour and Isabelle Stengers, the book has just been republished, for the first time in years — I ordered my own copy just the other day — unfortunately, in French only for now].