Cognition and Decision in Nonhuman Biological Organisms

My edited volume, Cognition and Decision in Nonhuman Biological Organisms, has just been published as part of the new Living Books About Life series from Open Humanities Press.

I’m excited about the entire Living Books About Life series. It represents a new form of collaboration between scientists and scholars in the humanities. And it is entirely open access as well. Each volume contains a number of crucial science articles, collected (or curated) and introduced by a humanities scholar.

My own volume covers topics such as “free will” in fruit flies, moods and emotional tones in bees, and more generally processes of affect, cognition, and decision found not just in animals, but in other sorts of organisms (trees, slime molds, bacteria) as well.

When the biologist and science fiction writer Joan Slonczewski, in her recent novel The Highest Frontier , envisions plants that display a sense of humor, and that can learn to resolve “Prisoners Dilemma” situations with mutual cooperation, she isn’t extrapolating all that much from what we actually already know about “mental” operations even in entities that have few or no neurons.

Harman on Metzinger

I have just a few quick notes on Graham Harman’s article on Thomas Metzinger, recently published in the latest issue (volume 7, # 1) of the open-access journal Cosmos and History (available here).

I should mention first of all that I have not read Being No One, which is Metzinger’s magnum opus, and the book on which Harman is commenting. Instead, I have read Metzinger’s volume that is in effect a popularization of his ideas — The Ego Tunnel — together with Metzinger’s precis of Being No One, available here (pdf). So I am aware that my own capacity to enter into this debate is somewhat limited. In particular, I cannot make commentary in the sort of way that David Roden does here.

But I do have a few brief reflections, from within my limited grasp of the topic. In the first place, I am largely in sympathy with Harman’s overall critique of scientism and eliminativism. One can take the results of neurobiology seriously, and use them to rethink questions about mind and consciousness, without having to attack so-called “armchair” philosophizing and “folk” psychology, epistemology, etc. And one should be especially suspicious of what Harman calls “the entire ‘ominous’ dimension of Metzinger’s book, which has made it so especially appealing to nihilistic younger males who enjoy breaking things into pieces.” (I’m not sure it is fair to charge Metzinger himself with this, but it does apply to certain aspects of Metzinger’s reception). This goes along with my general sense that Metzinger is not discrediting the very notion of a “self,”, so much as he is describing what a “self” actually is. The notion of a deep, substantial self that Metzinger attacks is something of a straw man; the “phenomenological self-model” that Metzinger opposes to this straw man itself is our consciousness or selfhood.

Also, the discovery of the ways in which our “self,” or sense of self, is hallucinatory, sel-contradictory, mistaken about itself, and so on and so forth, need not entail the earth-shattering conclusion that all values have been nullified, that life is suddenly devoid of meaning, that it is impossible for us to go on as before, etc. Once we have divested ourselves of an excessive anthropocentrism, the discoveries of recent brain science, like the prior discoveries of Copernicus and Darwin, no longerhave to reduce us to nihilistic desperation.

Despite the rhetorical rejection, in Metzinger and others, of anything “armchair” or not backed up by scientific research, it remains the case (as I think Harman effectively shows) that science alone (even given the recent quite powerful and interesting advances in cognitive psychology and neurobiology) underdetermines the answers to philosophical questions, which means that “armchair” reasoning cannot be dispensed with — indeed, Metzinger himself (as Harman points out) frequently and unavoidably participates in this.

One can see the underdetermination of philsophy of mind by empirical research if one reads Metzinger alongside the very different philosopher of mind Alva Noë. Noë’s ambitious philosophy of mind book, Action in Perception, was published just a year after Metzinger’s Being No One; and Noë’s popularizing book, Out of Our Heads, was published the same year as Metzinger’s Ego Tunnel. The works of these two thinkers overlap and interfere with one another in quite intriguing ways: they often refer to precisely the same empirical research, from which they draw diametrically opposed philosophical conclusions. I think that this juxtaposition is itself significant; both Metzinger and Noë have quite interesting things to say, and I wouldn’t want to categorically maintain that one is right and the other is wrong. (I am temperamentally inclined more to Noë’s position than to Metzinger’s, but for this very reason I think it is crucial to read them with and against one another, in order to get a grasp on what is being argued by both of them).

Getting back to Harman’s review of Metzinger: I am also inclined to agree with Harman about the high value and interest of many of the ways that Metzinger does employ empirical brain research for philosophical ends. This is especially the case with what Harman calls the science-fictional aspect of Metzinger’s look into the multiple sorts of mental activities that have been too simplistically been grouped together under the rubric of “consciousness.” As Harman writes. 

[Metzinger] thinks that decomposing the self into numerous complicated dimensions makes the self less real, when in fact it makes the self so much more real than before. By showing how much complexity is underway in our supposedly simple selves, Metzinger leads us to conclude not ‘well then, the self is just a sham in the end’, but ‘think of how many different and bizarre selves we might create, or which might already exist among animals or on other planets!’… [In Metzinger’s account] the human is just another bizarre species whose experience is generated by specific constraints, just as reptiles, insects, and extraterrestrials might have different lives from ours at this very moment.

We can find an attention to what might be called Metzingerian possibilities in such SF novels as Scott Bakker’s Neuropath and Disciple of the Dog, and in Peter Watts’ Blindsight. There are also a number of science fiction and fantasy novels that delve into what I am inclined to call Jamesian explorations of entirely nonhuman consciousnesses (I’d mention, just off the top of my head, such books as Justina Robson’s Living Next Door to the God of Love, and N, K, Jemsin’s The Hundred Thousand Kingdoms). But all in all, the different kinds of conscious experience described by Metzinger constitute a rich vein of speculation, that more SF writers ought to explore. 

There are other aspects of Harman’s critique of Metzinger that I am less happy with. Unsurprisingly, Harman chastizes Metzinger for “underming” and “overmining” objects, instead of accepting their full reality. Harman especially objects to the ways that Metzinger claims that all is “process,” and that therefore fixed objects (or Aristotelian substantial forms) are illusory. Metzinger says, in Harman’s paraphrase, that what we perceive as objects are really just the illusory results of reifying our own perceptual process, or freezing it in time. At the risk of opening up an old (and at this point boring) debate, I will repeat my own Whiteheadian sense that, indeed, all “things” are “really” processes. But for me, this doesn’t mean that things (or Harman’s objects) are thereby “undermined” by something else that is more essential than they are. For the fact that objects are “reifications” of processes doesn’t mean that they are illusory, or even that they aren’t basic. For the endurance of things, or their establishment of an “identity,” as a result of “reification” (which I think would better be called, in Whiteheadian parlance, social transmission and inheritance) is something that is perfectly real in and of itself. Endurance is an accomplishment, a singular and specific achievement in every case.

Moreover, this endurance is not something that happens (as Metzinger seems to claim, at least according to Harman) in our perceptual process, but actually in reality itself, in the very things which we are in process of perceiving. I am entirely in accord with Harman in rejecting the “simple a priori dogma that if something has causal antecedents, then only those antecedents can have independent reality” — to the extent that Metzinger buys into such a dogma, he is wrong, and Harman’s criticisms are justified. But it remains more of an open question for me than it does for Harman to what extent Metzinger is actually guilty of this; and my accord with Harman in rejecting this “simple a priori dogma” is part of the reason why I find ascriptions of process to be much more acceptable than he does. 

The biggest question for me that is raised by Harman’s review has to do with the relation between epistemology and ontology. Harman agrees with Metzinger in positing “autoepistemic closure” for all entities: this is the claim that we are never in direct contact with reality, since what we “perceive” is really just our own construction or simulation. But Harman goes on to criticize Metzinger for trying to somehow sidestep this clsoure in order to assert the objectivity and truth of scientific knowledge (as opposed to “folk” beliefs). I agree with Harman that, when the eliminativists belittle human cognition in general, but praise the positive and objective knowledge embodied in science, they are in effect contradicting themselves. Indeed, as Harman suggests, they are still being anthropocentric, since they see scientific knowledge of the world (whether in the form of mathematicization, as with Badiou and Meillassoux, or more generally with the results of experimentation, as with Metzinger and other scientistically inclined analytic philosophers) as a uniquely privileged instrument of making contact with other entities in the world, sharply different from the way that (to use one of Harman’s old examples) hailstones make contact with tar.

But for me, this is not just a problem of epistemology. I would say, against Harman, that of course we are always in direct contact with reality — since we are a part of this reality, rather than being separate from it (i.e. rather than being “withdrawn”). We are not caught in some Cartesian or Humean mental prison, familiar only with our own sense impressions (orfamilar only with our own languages, in the 20th century version of this line of thought). The point, however, is that this contact cannot be reduced to, or captured as, “knowledge.” The contact is not epistemological; when it comes just to epistemology, Harman and Metzinger are correct. But our contact with other entities is not restricted just  to relations of knowledge. Harman is right to say that my concept of a tree, however full and nuanced, will never be equal to the tree itself. But this does not negate the fact that the tree has “touched” me, and I have “touched” it, non-cognitiviely and unconceptually. 

Metzinger claims that “we are never in any direct epistemic contact with the world surrounding us even while phenomenally experiencing an immediate contact”: Harman, quoting this claim of Metzinger’s, wants to convict him of a contradiction,or of an “attempt to have it both ways.” For we cannot both be cut off from contact and be in immediate contact. However, here I think that Metzinger is more in the right, because he is talking about two different sorts of contact. “Phenomenal” contact need not, and cannot, be reduced to “epistemic” contact. Contact among entities is ontological, not epistemological — and this other dimension, which Metzinger at least senses as a problem, is omitted entirely from Harman’s account, when he says that, because we do not actually know other entities, or even ourselves, therefore all all entities must “withdrawn” from one another — and even from themselves.

I think that this all hinges upon questions of intentionality — and this is where I reach my own limits. Harman criticizes Metzinger’s attempt to resolve the contact paradox by calling upon intentionality — he criticizes Metzinger for misunderstanding both Brentano’s and Husserl’s theorizations of intentionality. “Contra Metzinger’s misreading,” Harman says,” the intentional for Brentano does not mean leaping outside the mental sphere and making direct contact with the real. Intentionality intends intentional inexistence, not something lying behind that inexistence.” Harman therefore does not think that Metzinger succeeds in “actually find[ing] a way to jump outside the phenomenal capsule and make some sort of contact with the real.” He particularly rejects the idea that experimental scientific knowledge represents such an actual contact.

But I am not convinced that Metzinger is wrong when he argues that mental states “intentionally contain an object within themselves.” I am more inclined to think that this is indeed what happens — as Whitehead puts it, “an actual entity is present in other actual entities” (PR 50). This presence is not cognized, and cannot be equated with Heidegger’s “presence-at-hand.” And Metzinger does have some sense of this, even though he is wrong (here I agree with Harman) to turn this into the unique guarantor for scientific knowledge, and for nothing else. Intentionality — including Molnar’s “physical intentionality” — has an important role to play here; even if this is not Brentano’s version of intentionality, nor Husserl’s. (I am trying, instead, to yoke intentionality to Whitehead’s sense of “prehension”). At this point, I no longer see very clearly — this is where I am stuck right now, and what I am trying to work my way through. And both Harman and Metzinger give me hints for this, even if I am ultimately not willing to follow either of their paths. 

 

Hyperbolic Futures

My essay “Hyperbolic Futures” attempts to think about the ways that speculative fiction (i.e. science fiction) works in relation to speculative finance (of the sort that has screwed us over in the last several years). I take a look back at my 2003 book Connected, Or What It Means To Live in the Network Society, and think about what has changed in the world, and in SF’s relation to the world, since then. And I discuss two recent, great SF novels in particular: Richard K. Morgan’s Market Forces, and Lauren Beukes’ Moxyland.

The article was published in the excellent SF journal Cascadia Subduction Zone, published four times a year by Aqueduct Press. Each issue is published both in hardcopy and in pdf, and the pdf version is released free on the Internet six months after intial publication. The issue that includes my article (volume 1, # 2) is now available for free download, here.

Panpsychism And/Or Eliminativism

Here is the text of the talk I gave twice last month: at the OOO III conference in New York, and at the SLSA conference in Kitchener, Ontario.

PANPSYCHISM AND/OR ELIMINATIVISM

Today I would like to think, in a cosmopolitical frame, about the recent philosophical movement known as Speculative Realism. It would be better, actually, to speak of speculative realisms, in the plural; for the four thinkers who spoke at the initial Speculative Realism conference, at Goldsmiths in London in 2007, in fact have vastly different positions and programs. And still more varieties of speculative realism have been enunciated since. What justifies uniting these diverse new modes of thought is that they have a common starting point. The four original speculative realists — Quentin Meillassoux, Ray Brassier, Graham Harman, and Iain Hamilton Grant — all reject what Meillassoux calls correlationism, or what Harman characterizes as “the bland default metaphysics that reduces objects to our human access to them.” In what follows, I will consider what positive positions this initial rejection commits us to.

Correlationism is defined by Meillassoux as the doctrine according to which “we never grasp an object ‘in itself’, in isolation from its relation to the subject.” Kant’s transcendental idealism is correlationist, and so is Husserl’s noetic-noematic structure. For correlationism, a mind-independent reality cannot exist, because the very fact that we are thinking of such a reality means that it is not mind-independent after all. From the correlationist point of view, Meillassoux says, “thought cannot get outside itself in order to compare the world as it is ‘in itself’ to the world as it is ‘for us’, and thereby distinguish what is a function of our relation to the world from what belongs to the world alone.” In correlationism, as Brassier puts it, “since it is impossible to separate the subjective from the objective, or the human from the non-human, it makes no sense to ask what anything is in itself, independently of our relating to it.” Or in the words of Harman, under correlationism “everything is reduced to a question of human access to the world, and non-human relations are abandoned to the natural sciences.” In other words, “the correlationist holds that we cannot think of humans without world, nor world without humans, but only of a primal rapport or correlation between the two. For the correlationist, it is impossible to speak of a world that pre-existed humans in itself, but only of a world pre-existing humans for humans.” As Harman sarcastically summarizes the position, correlationism assumes that “what is thought is thereby converted entirely into thought, and that what lies outside thought must always remain unthinkable.”

Correlationism is not the same thing as the “bifurcation of nature” denounced by Whitehead; the critique of the correlation and the critique of the bifurcation arise from different needs and concerns. Nonetheless, the two are not unrelated. It is only when our experience has been sundered in two that we could ever think of the need for a correlational structure in order to put it back together again. Modern Western thought, from Descartes through Locke and on to Hume, partitioned the world between primary and secondary qualities, or between objectively extended objects, on the one hand, and merely subjective “psychic additions” (CN 29), on the other. This culminated in the crisis of Humean skepticism, which Kant resolved by arguing that the unknown realities “out there” must be organized in accordance with the conditions imposed by our minds. We have viewed the world through a correlationist lens ever since.

Correlationism might seem to be at odds with everyday common sense; most people, if you asked them, would unhesitatingly affirm that things outside us are real. Remember Dr. Jonson, who kicked a rock, and claimed thereby to have refuted Berkeley. Nonetheless, the idea that the world is necessarily beholden to our ways of shaping and processing it has indeed been the “default metaphysics” of the West for more than two centuries, ever since Kant. To reject the correlationist consensus is to risk being accused of “naive realism.” Now, in fact, no version of speculative realism actually maintains the “naive” thesis that we can somehow have direct, unmediated access to a reality that is simply “out there” and apart from us. However, I also agree with Harman that we should be suspicious of any argument that disparages something by characterizing it as “naive.” For there is something disingenuous about such an accusation. Usually, the critics of “naive realism” are not urging us to adopt a more robust or sophisticated sort of realism instead. Rather, they are making the underhanded rhetorical suggestion that all realism is unavoidably naive. This critical sleight of hand really works to reinforce the solipsistic primacy of thought thinking only about itself. It’s a way of refusing and denying any movement towards what Meillassoux calls “the great outdoors, the eternal in-itself, whose being is indifferent to whether or not it is thought.”

In any case, the basic speculative realist thesis is the diametrical opposite of the “naive” assertion that things in themselves are directly accessible to us. For the key point, rather, is that the world in itself — the world as it exists apart from us — cannot in any way be contained or constrained by the question of our access to it. “Man” is not the measure of all things. We habitually grasp the world in terms of our own pre-imposed concepts. We need to break this habit in order to get at the strangeness of things in the world: that is, at the ways that they exist without being “posited” by us, and without being “given” or “manifested” to us. Even the things that we have ourselves made possess their own bizarre and independent existence. If philosophy begins in wonder — and ends in wonder, too, as Whitehead insists — then its aim should not be to deduce and impose cognitive norms, or Concepts of Understanding, but rather to make us more fully aware of how reality escapes and upsets these norms.

This is why any true realism must be speculative — despite the fact that “speculation” has been held in ill repute for most of the past century. For, confronted with the real, we are compelled to speculate: that is, to do precisely what Kant told us that we cannot and must not do. Pace Kant, we must think outside of our own thought; and we must positively conceive the existence of things outside our own conceptions of them. In Eugene Thacker’s terms, it is not enough to just consider the (objective) world-in-itself, in its difference from the (subjective) world-for-us. We must also actively explore what Thacker calls the world-without-us: the world insofar as it is subtracted from, and not amenable to, our own concerns. We learn about the world-for-us through introspection, and the world-in-itself through scientific experimentation. But we can only encounter the world-without-us obliquely, through the paradoxical movement of speculation.

Speculative realism is therefore as far removed from post-Kantian “critical” thought as it is from “naive” or unreflective thought. It rejects, not only the “default metaphysics” of continental anti-realism, but also (and perhaps more importantly) what Jon Cogburn calls “neo-Kantian ‘realism of the remainder’ type realisms… the view that the real is some inarticulate and inarticulable mush.” Slavoj Žižek, for instance, proposes that human subjectivity marks a unique rupture in the fabric of being. In the light of this continuing human exceptionalism, the Real can only be regarded negatively. It is nothing more or less than the traumatic remainder of a primordial split. The Real is what’s left over from our separation from it. Since this Real resists all of our symbolizations, Žižek says, it cannot be characterized at all. For Žižek as much as for Kant, then, articulation and determination can only be found on the side of human access. Kant, after all, never denied that there was such a thing as a nonhuman real. He maintained that things-in-themselves must really exist; he only insisted that we could not know anything positive about them, or say anything meaningful with regard to them.

Let me rephrase all this as a formula. Philosophers have only described the correlationist circle, in various ways; the point, however, is to step outside it. The aim of speculative realism, as Meillassoux puts it, is to break the circle, and once more reach “the great outdoors, the absolute outside of pre-critical thinkers.” Early Modern philosophers like Spinoza and Leibniz exhibit a freedom, boldness, and daring that are scarcely imaginable today. More precisely, the question posed for speculation is how to attain this “pre-critical” freedom without reverting — as Meillassoux says we must not do — to any sort of pre-critical or pre-Kantian metaphysical “dogmatism.” How, Meillassoux asks, can we “achieve what modern philosophy has been telling us for the past two centuries is impossibility itself: to get out of ourselves, to grasp the in-itself, to know what is whether we are or not”?

In order to get beyond Kant’s assertion of unknowability, or contemporary philosophy’s disappointing “realism of the remainder,” it is necessary to propose some sort of positive, speculative thesis, alongside the negative (anti-correlationist) one. More precisely, every variant of speculative realism must maintain both a positive ontological thesis, and a positive epistemological one. The ontological thesis is that the real not only exists without us and apart from our conceptualizations of it, but that it is actually organized or articulated in some manner: on its own, without any help from us. The epistemological thesis is that it is in some way possible for us to point to, and speak about, this organized world-without-us, without thereby reducing it yet again to our own conceptual schemes.

What distinguishes the various speculative realisms from one another is that they all propose different ways of stepping outside the correlationist circle. However, I think that all these approaches do have something in common. They all return to the very starting point of correlationism — Kant’s so-called “Copernican revolution” in philosophy — in order thereby to redistribute Kant’s original terms differently. This redistribution of terms opens up a place for renewed speculation.

Meillassoux himself follows such a strategy. He disrupts correlationism from within, by establishing that the Kantian correlation of thought and being is itself contingent (or “factial”) rather than necessary. Where Kant in his Paralogisms of Pure Reason demonstrates that certain fundamental metaphysical propositions are undecidable, Meillassoux traces this undecidability back to a more fundamental contingency — which turns out to be necessary in its own right. Kant argues that the sort of logic which works in particular, limited empirical circumstances is no longer valid when applied to the world conceived as a totality. Meillassoux follows a nearly identical line of argument when he shows that probabilistic reasoning, valid when applied to “objects that are internal to our universe,” cannot be applied “to the universe as such.” The difference, of course, is that Meillassoux draws upon Cantor’s theory of transfinites (which was unknown to Kant) in order to show that any sort of totalization is a priori impossible. This radicalization of Kant’s own argument opens the way to a new kind of absolute knowledge, one that is free from Kant’s strictures against it.

Iain Hamilton Grant similarly returns to the Kantian moment of decision, and orients it otherwise, when he reconstructs and revitalizes Schelling’s critique of Kant. The Kantian transcendental argument becomes a principle of genesis and productivity, rather than one of a priori necessity. In consequence, thought does not, and cannot, posit or legislate the nature of appearance. Rather, thought is itself generated through a process that is antecedent to it, and that forever exceeds its grasp. It is a “necessary truth,” Grant says, that “antecedence is non-recoverable.” Somewhat like Meillassoux’s ancestrality, Grant’s antecedence cannot be recuperated in any sort of correlation. And yet, the “unthought” of an infinitely productive Nature is not sheer negativity (as it remains for Hegel and for Žižek). Rather, it is an active composition of powers or forces.

For his part, Harman proposes what I would like to call (echoing Derrida on Bataille) a “Kantianism without reserve.” This consists in extending the gap between phenomena and noumena to the experiences of all entities. We can no longer specially privilege human beings (or rational beings in general), because every object encounters all other objects phenomenally only, as “sensual objects,” without being able to reach those entities as they in themselves, noumenally, as “real objects.” No object can ever entirely know (grasp or comprehend) any other object; indeed, an object cannot even really “know” itself. But Harman points out that we can, and do, allude to objects — indeed, we are doing this almost all the time. We refer to objects that we do not know by designating them metaphorically, or indirectly. In this way, we can be aesthetically moved by objects, even when we do not (and cannot) actually know them. Indeed, such “vicarious” affection is a crucial mode of contact among entities. (It roughly corresponds to what Whitehead calls “conceptual prehension”). In this way, for Harman as for Whitehead, “aesthetics becomes first philosophy.”

Brassier’s physicalist revision of the Kantian distinction between phenomena and noumena can be contrasted with Harman’s aestheticist one. Brassier converts Kant’s “transcendental idealism” into a “transcendental realism,” by asserting “the transcendental presupposition of an extra-conceptual difference between concept and object.” That is to say, the real as such is non-conceptual; and the difference between the real and our concepts of it cannot itself be conceptualized. Our concepts are always inadequate to the objects that they refer to, and that they futilely endeavor to circumscribe. Physical science is a way of exploring this gap between concept and reference — even if it can never bridge the distance altogether. Rather than thought imposing its categories on the real, Brassier says, “the reality of the object determines the meaning of its conception, and allows the discrepancy between that reality and the way in which it is conceptually circumscribed to be measured.” Kant’s own defense of scientific objectivity is thus transformed into a more robustly realist form than Kant himself was able to provide. Physical science is grounded in the inevitable failure of any correlation between thought and the world, rather than in the necessity of such a correlation. The nonconceptual remainder is no longer mute, as it was for Kant and Žižek; rather, scientific experimentation allows it (or forces it) to speak.

I have been insisting upon the Kantian background of all these speculative realist projects, even though speculative realist thinkers themselves often describe what they are doing in very different ways. I have done this because Kant’s Copernican revolution in philosophy — or rather, his “Ptolemaic counter-revolution,” as Meillassoux insists — itself establishes correlationism and anthropocentrism on the basis of its own critical self-reflexivity. We should stop to think for a minute about how strange this is. According to Kant, thought does not discover its accordance with the world by reaching out towards the world. Rather, it is precisely when thought reflects back upon itself, when it engages in the critique of its own powers and limits, that it is suddenly brought into correlation with being. It is only by focusing back upon itself, to the exclusion of all else, that thought comes into correspondence with something that lies outside it, and beyond it. And it is this strange knot of thought and being — mirrored within thought itself by the pre-established harmony of inner-directed self-reflection with outer-directed intentionality — that speculative realism strives to undo.

In order to untie this knot of thought and being, it is necessary to dislodge the self-reflexivity of thought in one way or another. Thought needs to be radically problematized, from ouside — instead of grounding and validating itself by means of its own purifying autocritique. The anthropocentrism of our “default metaphysics,” which Harman rightly finds objectionable, rests almost entirely on the dubious presupposition that human beings are uniquely rational, uniquely possessed of subjectivity and interiority, uniquely capable of thought and/or language. Such a position was radically undermined by Darwin. Whitehead entirely removed the need for it, by elaborating an analysis of prehension that applies equally to all actual entities. And human exceptionalism is even less tenable today, now that we know that not only chimpanzees and parrots, but also slime molds and bacteria, communicate, calculate, and make arbitrary decisions.

But in fact correlationism is not reducible to humanism, nor to notions of subjectivity. As Meillassoux writes: “we must emphasize that the correlation of thought and being is not reducible to the correlation between subject and object.” Even the freeing of thought from subjectivity and from representation — Meillassoux gives the example of Heidegger — does not suffice to undo correlationism. And further, even the deconstruction and dissolution of the humanist subject does not really get us away from anthropocentrism: at best, it merely replaces this with an impersonal noocentrism or logocentrism.

In order to step outside the correlationist circle, Meillassoux insists that we must displace thought (and language) altogether. We need to adopt a stance, he says, “which takes seriously the possibility that there is nothing living or willing in the inorganic realm.” If we are to reject the phenomenological notion of “the givenness of the world,” then we must recognize the existence of “a world capable of subsisting without being given” to us or to any other perceiver: a world that is “capable of existing whether we exist or not.” Reality for Meillassoux is “totally a-subjective.” We must “think a world that can dispense with thought, a world that is essentially unaffected by whether or not anyone thinks it” (emphasis added).

I think that we need to take this radical purgation of thought from being seriously. Anti-correlationism can plausibly lead to radical eliminativism, as Meillassoux’s formulations at least suggest, and as Brassier argues much more forcefully and straightforwardly. For such an account, matter must be entirely impassive — devoid of life, initiative, or active force — in order that it not be affected by thought. And sensation and perception need to be downgraded — or even abolished — because (like anything carnal) they imply an interaction between an observer and something being observed.

In his quest to guarantee the independence of being from thought, Meillassoux goes so far as to reintroduce into philosophy the explicit distinction between primary and secondary qualities. Meillassoux privileges mathematical formalism at the expense of perception and sensation: this is the only way to “remove the observer,” leaving behind just those properties that an object has in and of itself. “All those aspects of the object that can be formulated in mathematical terms,” Meillassoux writes — and only those aspects, we might add — “can be meaningfully conceived as properties of the object in itself.” Radicalizing Badiou’s dictum that mathematics is ontology, Meillassoux argues that it is exclusively through “the mathematization of nature” that physical science indubitably allows us “to know what may be while we are not.” In effect, Meillassoux resolves the bifurcation of nature by brutally amputating the subjective side of the duality.

Brassier’s arguments are similar to Meillassoux’s, but even more far-reaching. Once we accept that the difference of objects from the concepts we have of them is itself non-conceptual, and not to be subsumed by thought, then we are forced to come to terms with “a world that is not designed to be intelligible and is not originarily infused with meaning.” This leads us inexorably to the “truth of extinction,” the inevitable extermination of all thought in the future course of the universe. For Brassier, even more than for Meillassoux, the recognition of a (past or future) time without thought must radically devalue thought in the present — including even the thought of this recognition itself. Unless we were to embrace some bizarre form of extreme idealism (thought without being?), we would seem to be condemned by the rejection of correlationism to a regime of being without thought. Undoing the Kantian nexus of thought and being leads us, in this case, to the conclusion that thought is epiphenomenal, illusory, and entirely without efficacy. Where Western science has traditionally seen mere matter as passive and inert, Brassier — following Thomas Metzinger and Patricia and Paul Churchland — argues that, once we are rid of an unjustified anthropocentrism and narcissism, we must view human beings in this manner as well.

Brassier pushes this grim logic all the way to the end, proclaiming an “extinction of meaning that clears the way for the intelligibility of extinction. Senselessness and purposelessness are not merely privative; they represent a gain in intelligibility.” There is something impressively bracing about such militant nihilism, even if I am unwilling to give it the last word. But, once we accept the anti-correlationist argument, what other alternatives can there be? Must the radical annihilation of meaning and purpose be the price we pay for understanding the real as it is, apart from us?

In contrast to Brassier, Meillassoux evades the radical consequences of eliminativism by arguing for the absurd, radical emergence ex nihilo, at some point in the history of the universe, first of life and then of thought. As Harman makes evident in his recent exposition and partial translation of Meillassoux’s otherwise unpublished manuscript The Divine Inexistence, Meillassoux insists — against all of modern biology — both that life is radically discontinuous with mere matter, and that thought is radically discontinuous with mere life. Meillassoux thus maintains the Cartesian picture of matter or extension as passive and inert, while providing an escape clause in the form of the absolutely contingent and unforeseeable coming-into-existence first of life and then of thought, both of which are irreducible to matter. This restores human exceptionalism with a vengeance. The violent audacity of Meillassoux’s reversal reminds me, once again, of Kant. Just as Kant lets God back in through the back door, as it were, in the Second Critique, after having eliminated him in the First Critique by destroying the ontological argument for his existence, so Meillassoux rehabilitates life and thought in The Divine Inexistence, after having expelled them, together with the principle of sufficient reason, in After Finitude.

I am not willing, myself, to travel this route with Meillassoux. Despite his demonstrations of the contingency of the correlation, and of the impossibility of transfinite totalization, I cannot see any justification for abandoning the principle of sufficient reason. Harman observes that Meillassoux has two objections to the principle. The first objection is that it implies an infinite regress of causes, unless we bring it to an end by arbitrarily positing a First Cause or Unmoved Mover. The second objection is that it implies that effects are reducible to their causes; and if this were the case, then novelty would be impossible. But Harman replies that there is nothing wrong with conceiving an infinite regress; and that an effect can well exceed its causes, without thereby being entirely independent of those causes.

Both of Harman’s points are in accordance with Whitehead’s revision and restatement of the principle of sufficient reason in the form of what he calls the ontological principle. According to this principle, everything that exists — every actual entity — has a reason (or more than one) for being what it is; and these reasons are themselves actual entities in their own turn. “Actual entities are the only reasons; so that to search for a reason is to search for one or more actual entities” (PR 24). There is no First Cause independent of this process; even God is a particular actual entity, the reasons for whose existence reside in other actual entities. Whitehead insists that nothing is ever entirely determined by its causes. An actual entity must decide how it receives and responds to the causes that feed into it. And every such decision introduces at least a modicum of novelty into the universe. But the ontological principle also states that no entity can ever be entirely free from its antecedent reasons; “there is nothing which floats into the world from nowhere” (PR 244).

Beyond this, the real problem with Meillassoux’s and Brassier’s accounts is that they both assume that matter in itself — as it exists outside of the correlation — must simply be passive and inert, without meaning or value. But isn’t this assumption itself a consequence of the bifurcation of nature? It is only an anthropocentric prejudice to assume that things cannot be lively and active and mindful on their own, without us. Why should we suppose that these are qualities that only we possess, and that we merely project them upon the “universe of things” outside us? Eliminativist arguments thus start out by presupposing human exceptionalism, even when their explicit aim is to humble and humiliate this exceptionalism. If you take it for granted that values and meanings are nothing but subjective human impositions, then it isn’t hard to conclude that they are ultimately illusory, for human beings as well as for other entities.

What’s needed is an alternative way of unbinding the Kantian knot of thought and being. And this is what Whitehead offers us, following William James. Rather than brutally purging the physical universe of anything like thought — an enterprise as absurd as it is ultimately impossible — James and Whitehead urge us to recognize the commonness and ordinariness of thought. They do not contest thought per se, as the eliminativists do, but only its self-reflexive self-privileging, its claim to specialness and preeminence. Isabelle Stengers observes that James engaged in “a deliberate project of the ‘depsychologization’ of experience in the usual sense of conscious, intentional experience, authorizing a clear distinction between the subject and its object.” In this way, James “denied the privilege of occupying center stage to reflective consciousness and its pretensions to invariance.” Or, as James himself puts it, the reified entity known as consciousness “is fictitious, while thoughts in the concrete are fully real. But thoughts in the concrete are made of the same stuff as things are.”

James’ thesis is both monist (since everything is “made of the same stuff”) and pluralist (since there are many thoughts and many things, which cannot be gathered together as one). But it is anti-dualist, and opposed to the bifurcation of nature. Indeed, James positions his thesis in explicit opposition to what he calls the “neo-Kantian” doctrine that “not subject, not object, but object-plus-subject is the minimum that can actually be.” In this way, James is an anti-correlationist avant la lettre.

James’ characterization of experience provides the “prototype,” as Stengers says, for Whitehead’s “actual occasions.” These are always “bipolar,” with conjoined “physical” and “mental” poles (PR 108). This means that thought is an immanent attribute — or a power — of being itself, and of each individual entity that exists. Nothing could be further from the post-Kantian (or correlationist) sense of thought as something that would approach being from without, and that would strive (successfully or not) to be adequate to it. For Whitehead, every entity immanently experiences something; or better, every entity is an experience. This does not mean, however, that every entity is conscious. Whitehead insists that “consciousness presupposes experience, and not experience consciousness” (PR 53). Timothy Morton makes a more concrete, but somewhat similar, point when he suggests “that there is something that my mind does that isn’t that different from what a pencil does when it rests on a table… It’s not that pencils have minds, it’s that minds are pencil like.” In this way, thought is common and humble, rather than rare and preeminent.

Nonconscious experience is not an oxymoron; it’s simply that more things are felt than are known. Whitehead writes that “the primitive form of experience is emotional — blind emotion” (PR 162). It is only in a few rare cases that this emotion is subsequently elaborated into self-conscious cognition. Emotional feeling, Whitehead says, is always “felt in its relevance to a world beyond”; but “the feeling is blind and the relevance is vague” (PR 163). Primordial “vector feeling,” the physical movement or “transmission” from one thing to another, is undoubtedly the raw material out of which the whole drama of correlationism was constructed. But in its non-cognitive, or pre-cognitive, blindness and vagueness, thought as Whitehead describes it happens, or passes, without any epistemological warrant. It makes to sense for thought to be correlated to a world outside itself. For thought is already a constituent — think of it as a sort of flavoring — of the very world that it is supposed to be “about,” and whose objects it is supposed to “intend.”

We might think here of George Molnar’s claim for the existence of what he calls physical intentionality. The commonly held doctrine, deriving from Brentano, is that intentionality is an exclusive mark of the mental or psychological; indeed, intentionality is generally held to provide the definitive principle of a “demarcation between the psychic and the physical.” Against this, Molnar argues that “something very much like intentionality is a pervasive and ineliminable feature of the physical world.” Rejecting commonly-held Humean or nominalist assumptions, Molnar is a thoroughgoing realist about the physical powers or dispositions of things. He insists that “physical powers, such as solubility or electrical charge, also have that direction toward something outside themselves that is typical of psychological attributes.” Of course, physical intentionality, so described, cannot be conscious; it does not have any semantic or representational content. But Molnar argues that mental intentional states are not necessarily semantic or representational either. Pain, for instance, “is directed towards its intentional object” — the location where it is felt — “without representing (symbolizing) its object.” Although Molnar does not himself put it this way, the result of his argument is to detranscendentalize intentionality. That is to say, intentionality becomes an implicit relation, or a potential for becoming, within the world — rather than being an underlying principle or structure of correlation.

Molnar admits that this point of view might lead to what he calls the “threat of panpsychism.” And he pushes away this “threat” by replacing intentionality “with another criterion of demarcation” between mind and matter. The only other available criterion, however, is precisely “the capacity for consciousness” — which Molnar embraces while acknowledging “that this position has its own distinctive difficulties.” If we accept that thought (or feeling, or experience) need not be conscious, then we might well be led to abandon the demarcation between mind and matter altogether.

Although Molnar is unwilling to embrace panpsychism, I propose that it gives us a good way to avoid the problematic baggage both of consciousness and of phenomenological intentionality. In this way, panpsychism might be a promise, rather than a “threat.” The non-eliminativist way of escaping the correlationist circle is to recognize the sheer ubiquity of thought in the cosmos. We don’t need a criterion of demarcation, because there is nothing to demarcate or separate. Once we understand “thought” in Whitehead’s deflationary sense, rather than in Kant’s grandiose one, we discover that it is everywhere, rather than nowhere.

We can take an inverted clue here from Meillassoux. If we reject his thesis of the radical emergence of thought out of nothingness, then we must rather conclude that thought is always there already, in the very place where he claims that “there is nothing living or willing.” This is basically Galen Strawson’s position. Strawson argues that radical emergence is impossible; “experiential phenomena cannot be emergent from wholly non-experiential phenomena.” Strawson regards eliminativism as absurd, “because experience is itself the fundamental given natural fact… there is nothing more certain than the existence of experience.” But since experience cannot float into the world from nowhere, our only alternative is to accept that reality is experiential, all the way down.

Panpsychism, no less than eliminativism, undoes the Kantian knot. Precisely because panpsychism claims that thought is always already present everywhere, it does not grant to thought any special foundational or reflexive privileges. If mind is intrinsic to being, then it exists in and for itself, apart from any question of what it might be correlated with. For panpsychism, everything is mindful, or has a mind; but this does not necessarily entail that everything is “given” or “manifested” to a mind.

To conclude, I need to bring this discussion back to the initial speculative realist thinkers. Neither Harman nor Grant is a full-fledged panpsychist, but they are both inclined strongly in the panpsychist direction. This is evident from their essays in David Skrbina’s anthology of contemporary panpsychist thought, Mind That Abides. Grant indeed argues for “panpsychism all the way down, that is, without exception”; but in doing so, he complicates the question of emergence. Everything is in some sense minded or mindful, he says, but this mindedness is not there at the beginning. Rather, it necessarily but belatedly arises out of the antecedence of nature’s productive powers. For his part, Harman sees mentality, or experience, as an inevitable component of any relationship, or interaction among objects. But since he claims that objects are “withdrawn,” existing apart from all relations, he doesn’t attribute mentality or experience to these objects in and of themselves. There are undoubtably objects, he says, that remain “dormant,” never entering into relation with anything else. Hence, for Harman, “even if all entities contain experience, not all entities have experience.”

Despite these qualifications, I think that we are left with a clear alternative. If we are to reject correlationism, and undo the Kantian knot of thought and being, no middle way is possible. We must say either (along with Harman and Grant) that all entities are in their own right at least to some degree active, intentional, vital, and possessed of powers; or else (along with Meillassoux and Brassier) that being is radically disjunct from thought, in which case things or objects must be entirely divested of their allegedly anthropomorphic qualities. When we step outside of the correlationist circle, we are faced with a choice between panpsychism on the one hand, or eliminativism on the other.

As a coda, I would like at least to mention — without further comment — some of the newer versions of speculative realism, beyond the four presented at the 2007 Goldsmiths conference. Ben Woodard’s “dark vitalism,” Reza Negarestani’s “dark materialism,” and Eugene Thacker’s “horror of philosophy” all seem actually to combine the most extreme tendencies of both panpsychism and eliminativism — however oxymoronic such a conjunction might appear. For these thinkers, the world-without-us is alien and actively hostile to human life and thought. If nothing else, such projects are further signs that we are beginning to think speculatively and cosmologically again — after a century in which, with the lonely exception of Whitehead, such efforts were viewed with suspicion and derision.