Category: Books

Today I read The Prince and the Wolf, the short book from Zer0 that transcribes a discussion between Graham Harman and Bruno Latour, held at the London School of Economics in 2008, and organized and introduced by Peter Erdelyi. I found the book very helpful in further pursuing the questions about Harman’s object-oriented ontology that I have been mulling over for several years. This is largely because of the context we have Latour responding to Harman’s reading of him, which suggests different directions for debate than any I have thought of myself, or come upon elsewhere. I haven’t the time to think through all of the stuff I read — so this posting will just mention briefly a few of the key points that emerge from the book, before I forget them.

Basically, Latour objects to Harman’s characterization of him as a relationist, by saying that he doesn’t understand (or doesn’t accept) Harman’s entire opposition between objects/substances and relations. Where the question of whether objects can be defined by their relations, or on the contrary have hidden nonrelational cores, is crucial for Harman, Latour suggests rather that this is a both/and, not an either/or. It is precisely because things are singular, that they need mediators, relations via translation and transportation, in order to have an effect, or assert their presence in the world. So it’s not a question of whether objects are defined by intrinsic substantial natures or by merely relational qualities, but rather that it is precisely to the extent that objects are singular and irreducible to external common measures that they need to establish modes of relationality.

Latour accepts Harman’s definition of him as an occasionalist, and as the first secular occasionalist. This is because, for Latour, all alliances among things are contingent, and can always be broken or articulated differently. However, it still doesn’t seem to me that causation, or contact among entities, is as problematic for Latour as it is for Harman. Harman affirms occasionalism because, given his notion of sel-subsistent objects, sealed off from one another, the fact that objects do affect one another cannot be taken for granted, but needs a special explanation. I don’t see that this is a problem for Latour — he sees objects making alliances and networks, entering into confederations or fights and oppositions, as being the usual course of things; it isn’t in need of special explanation.

This also is an issue in Harman’s reading of Whitehead, which comes up briefly in the book because of Latour’s overt Whiteheadianism. Harman says that Whitehead is also an occasionalist, and not a secular one, because Whitehead requires eternal objects mediated by God in order for things to affect one another. This seems to me to be wrong. In his doctrine of causal efficacy, Whitehead presents entities as affecting one another directly, without mediation, all the time.

This is the whole point of Whitehead’s critique of Hume. Whitehead says that, if Hume were correct in claiming that no connections among events or entities can be detected in the world, then it would be impossible for such connections to be detected in the mind either — there could be no habit or stability of mental associations. Hume in fact assumes, in the case of the mind, the very causal links that he denies to the world outside the mind. But this is unacceptable, once we reject the Cartesian dualistic notion that the mind is somehow separate from the world. Whitehead says in effect that it is impossible to actually disavow causal efficacy. I accept Harman’s brilliant observation that Hume’s scepticism is really just the flip side of Malebranche’s occasionalism — but my conclusion from this is that, if we accept Whitehead’s argument against Humean scepticism, then this is an argument against occasionalism as well. For Whitehead, an entity cannot ever exist apart from its connections, even though the entity itself is not reducible to these connections.

As for eternal objects and God in Whitehead’s cosmology, it seems to me that they are not deployed in order to answer the question of how things can influence other things. Rather, they are there in order to answer a quite different question: that of how novelty is possible, of how creativity takes place, of how things can be something other than just repetitions of previous things. Harman observes that, “for Aristotle… causation itself isn’t really a problem; there are no gaps between things.” I would claim, contra Harman, that the same is true of Whitehead. The problem for Whitehead is not the occasionalist one of how to bring unconnected things together, but rather the one of how to produce gaps, discontinuities, and changes in a world in which everything (every actual entity) has a reason, which reason is always another actual entity (or a number of them).

In other words: Harman rejects Aristotle’s belief that “there are no gaps between things,” while he seeks to revive an Aristotelian notion of substance. Whitehead, as is well known, utterly rejects Aristotelian substance, but like Aristotle he doesn’t have a problem with things touching and affecting one another. Actually, it is a bit more complicated: for Whitehead – contra Bergson – “there is a becoming of continuity, but no continuity of becoming.” Both the continuity and the gaps in continuity have to be produced, and have to be accounted for. Reality, for Whitehead, is atomistic — but this does not mean nonrelational. I think that Whitehead would probably reject Harman’s basic duality between objects and relations in much the same way that Latour does.

To get back to Latour — he says in The Prince and the Wolf that he is not as much of an actualist as Harman makes him out to be, precisely because he does not conceive things in “punctual” terms. Where Harman seeks to revive a notion of substance in order to get away from the contemporary overvaluation of relations, Latour poses the issue quite differently. Several times in the book he says that, precisely because we can no longer accept the notion of substance, the question that exercises him the most is one of subsistence. “Once substance has been excluded, subsistence comes to the fore.” For Harman, things are substances, in their basic being, regardless of whether they subsist or not. For Latour, things cannot be substances at all, and this is why the question of their subsistence is such an important one. Indeed, Latour hints that his still-unpublished exploration of different modes of being (under the influence of Souriau) is really about different ways of subsisting. There are multiple modes of being, because there are multiple ways in which entities, without being substances, nonetheless subsist over time (and also, I would suspect, through space).

Latour adds that what he now sees as the defect of his early treatise “Irreductions” (part of the Pasteur book) is that it is in fact too “punctual” — it presents as points what are really vectors. Now, “vectors” is very much a Whiteheadian term as well — Whitehead insists on the vector quality of existence — and for Latour, vectors are important because they involve both movements of translation and transportation, and processes of subsistence. Harman objects that vectors are only spatial, not temporal, a movement outward but not a movement forward in time — Whitehead’s and Latour’s vector picture has little to do with Bergsonian duration. Harman is right regarding Bergson specifically, but I don’t accept Harman’s further inference that therefore there is no real temporality in Latour: I think it is just that Latour is following Whitehead’s physics-inflected sense of spacetime, rather than Bergson’s radical duality between time and space. The movement of the vector is as irreducible to the kind of temporality of present instants that Harman describes as it is to Bergsonian continuity of becoming. For Latour (as for Whitehead, and in contrast to Harman) everything has “descendants and ascendants” [I suspect that what Latour meant by the latter word was “antecedents”].

And this, coming near the end of the volume (page 108), is perhaps the crux: Latour claims that “every single entity is expectant of a next step.” Harman responds: “Not expectant, but it becomes a possible mediator of other two entities.” Latour responds that he does intend the stronger meaning that Harman rejects: “No, but for itself, we are talking about the thing itself. It is expectant, is it not?” Harman says no, where Latour says yes. As for me, this is precisely where I side with Latour (and Whitehead) against Harman. Things are indeed “expectant,” because they feel what they prehend, and in turn set down conditions for what will prehend them, i.e. ways in which they will (expect to) be felt. Such is the vector character of experience for both Whitehead and Latour; it is also the “physical intentionality” at the heart of George Molnar’s conception of “powers.”

I just finished reading George Molnar’s extraordinary book Powers. Reading an analytic philosophy book like this one reminds me, once again, that I am not a philosopher, even though I frequently write about philosophical texts. Good analytic philosophy tries to provide basic logical grounds or arguments for all of its assertions — something that I am incapable of doing. And it almost totally ignores what is interesting about classical philosophical texts: which is the implications of the metaphysical assertions. The point is that I am sure that any good analytic philosopher could point to the logical errors or ungrounded assertions in great speculative metaphysicians such as Plato, Aristotle, Descartes, Leibniz, Kant, and William James. But such errors do not negate what is genuinely challenging and thought-provoking about these thinkers. (I am crudely echoing, here, some of the remarks that Graham Harman has made many times on his blog. But then, Harman really is a philosopher, albeit of the continental rather than analytic kind — which means that he is doing the kind of thing the aforementioned great speculative philosophers do, rather than what the analytics do. I, in contrast, am doing something that is closer to speculative fiction than to speculative metaphysics. I respind to philosophy texts in the same way that I do to science fiction novels). 

Nonetheless, although his book is mostly concerned with the usually analytic statement of the particular arguments needed to establish his assertions, Molnar’s metaphysical assertions are themselves fascinating and suggestive, and contribute a lot to current debates in “speculative realism.” (Indeed, I came upon Molnar in the first place because he was mentioned in the context of SR by Ben Woodard). (Molnar is also footnoted in the introduction of The Speculative Turn, in connection with Iain Hamilton Grant’s attempt to produce a “powers” metaphysics; but Grant himself doesn’t seem to mention Molnar, either in his essays in that volume or in his own book Philosophies of Nature After Schelling).

Molnar’s basic argument is that things (or OOO’s objects) possess causal powers that are ontologically real, and not just confined to the instances in which they are manifested. Salt contains the power of being soluble (dissolveable) in water; this power is a veritable property of the salt, even if it never encounters water and never actually gets dissolved. In insisting that powers are actual independently of their manifestation (even if they can only be described in terms of their manifestation), Molnar rejects the skeptical (empiricist, and especially Humean) hypothesis that talk of powers has no meaning apart from the conditional statement that, e.g., if the salt is put into water, then it will dissolve. The classic Early Modern reproach to medieval philosophy was to ridicule the latter for allegedly saying, for instance, that opium puts people to sleep because it has a dormative power — and to claim that this sort of explanation is utterly meaningless. Molnar is arguing, in effect, that opium really does have something like a “dormative power.” This is not to deny that such a power can be analyzed, e.g., in terms of particular neurochemical events that take place in the brain of somebody who has smoked opium. But such an analysis of the “dormative power” does not get away from the attribution of powers, since it simply replaces the power of opium per se with a more detailed account of the powers possessed by particular molecules in the composition of opium. 

In this way, Molnar asserts a realist ontology, one that is directed against the skeptical empiricism of the whole tradition derived from Hume (and one still adhered to by a large number of analytic philosophers today). The parallels with speculative realism go further; Molnar insists, as much as Graham Harman does, that a thing, or an object, is not just a bundle of properties or characteristics, but exists in its own right apart from and in addition to these. (Although Molnar, unlike Harman, endorses the basic scientistic move of reducing objects to their ultimate subatomic constituents, he doesn’t make the claim that this somehow renders objects of the sort that we can see and touch illusory). 

In this way, Molnar offers something like the actualism, and the “flat ontology,” insisted upon by Delanda, by Latour, and by OOO (in contrast to the eliminativist impulses, both of many analytical philosophers, and of Ray Brassier or other more scientistically-inclined speculative realists). But there’s a difference. Molnar writes: “While ontologically there is nothing over and above individuals and their properties (actions), causally there is.” (George Molnar). The insistence on actual causality, and on actual relations (causality being one form of relation), makes for a significant difference between Molnar and Harman. Contra Harman, Molnar rejects any sort of “occasionalism”; he insists that causality is direct — and not merely “vicarious.” Like Harman and against Deleuze, Molnar claims that powers, even when they are not being exercised, are entirely actual qualities of things — they cannot be regarded as “virtual” or “potential.” They fully exist even when they are not manifested in particular events, as a result of particular relational encounters. But against Harman, Molnar insists that relations are as primary an ontological category as things or objects are. 

To put this another way: Harman, in his critique of Latour, opposes the Deleuzian notion of the virtual (together with related notions of the potential) to what he sees as Latour’s “Megarian” actualism. Although he applauds this actualism, he rejects what he claims is Latour’s relationalism, or denial that his discrete entities have any nonrelational substance. But Molnar adds another option to this picture. For Molnar, things do have a substantial reality that is outside of, and anterior to, relations — but this substantial actuality is largely composed of “powers,” or of causal abilities to do things (and thereby to interact relationally with other substances). There is nothing besides individuals and their properties; but since many of these properties of individual things are powers, they make direct causality possible, i.e. when they do contingently encounter other things or substances, they produce real effects.

Molnar asserts that “laws of nature” are supervenient upon the powers of actually-existing things. Against post-Humean skepticism, “laws of nature” are objective features of the world, not mental impositions. This thesis is therefore, once again, realist and anti-correlationist; it affirms that reality is mind-independent and human-independent.  But, in opposing Hume, Molnar also implicitly opposes Quentin Meillassoux’s return to, and alleged solution of, “Hume’s problem.” Something like Leibniz’s law of sufficient reason, or Whitehead’s ontological principle, is preserved against Meillassoux’s all-too-Humean insistence that anything can happen with no reason whatsoever. This is because, for Molnar, it is not that things obey pre-existing laws of nature (which is the thesis that Meillassoux rejects), but rather that “laws of nature” are themselves the consequence of the actual powers actually possessed by individual entities. We might say therefore, that Molnar’s powers are like Spinozian/Deleuzian abilities to affect, and to be affected by, other things. (The Spinozian part of Deleuze, unlike the Bergsonian part, does not involve virtuality). 

In addition to all this, Molnar claims that powers need not be grounded, and indeed that the ultimate powers of things are ungrounded. He argues this on an empirical, rather than a priori basis: the subatomic particles of which, according to contemporary physics, the universe is composed, do not seem to possess any grounding. An electron or a photon is nothing over and above its powers. If the powers of “composite” or everyday objects are themselves grounded (e.g. in physical, non-dispositional properties of these objects), the grounding does not continue downwards infinitely, but ultimately meets the ungroundedness of the powers of elementary particles.

Now, this might well be the place where OOO thinkers would argue that Molnar reveals himself to be a scientistic reductionist after all, but I think that such a criticism would not be entirely fair. This can best be understood, perhaps, by looking at the role that ungrounded powers play in Iain Hamilton Grant’s metaphysics (see Grant’s response to Harman in The Speculative Turn; this is also the place where Ben Woodard, as cited above, associates Molnar with Schelling and Grant). The crucial point we can take from Molnar is that powers need not be grounded in order to be real; and this makes for a crucial step in Grant’s argument, against Harman, that one can trace the anteriority of forces that generate objects, without thereby “undermining” objects and reducing everything to some sort of undifferentiated blob. From another direction, Molnar’s sense of ungrounded powers might also be used to defend Latour’s ontology against Harman’s criticisms. When objects are understood as possessing intrinsic powers, they can be separate and actual without being “withdrawn” in Harman’s sense. Objects possess real forces, which they exert against other, equally real forces being deployed by other objects. Without going so far as to make the difficult claim that Latour and Grant can be reconciled with one another, I think that they both can be defended against Harman’s various criticisms of them on the basis of an appeal to something like Molnar’s insistence upon the actuality, and not-needing-to-be-groundedness, of causal powers.

There’s also another, weirder direction in which one could take all this. For Molnar, subatomic entities like electrons and photons have intrinsic powers, but they don’t have any intrinsic qualities other than their powers. Indeed, this is precisely what he means when he asserts that their powers are ungrounded. If the powers of salt and opium and human beings and (to use Harman’s examples) tar and hailstones are grounded, this is because such entities have intrinsic qualities that are not powers, in addition to their intrinsic powers. I think, however, we can reduce the difference between subatomic entities and the sorts of entites that we can apprehend directly by adopting some form of panpsychism (as I have argued before — of course, Molnar would have hated this). That is to say, I want to argue for a thesis that Molnar explicitly rejects, but which is not incompatible with his main points. The thesis is what Molnar calls “dual-sided theory”: “all properties [of objects] have something about them that is irreducibly and ineliminably dispositional [i.e. is a power], and something (else) about them that is irreducibly and ineliminably non-dispositional or ‘qualitative’… A power is only a face/facet/side of a property that also has a qualitative face/facet/side.” Molnar rejects this thesis primarily because he doesn’t think that subatomic particles (or “field-densities”) have a qualitative side: they are only dispositional (they only have powers without any “grounding” or innerness). But a major argument of 20th century panpsychists, from Russell on to Strawson, is precisely that all entities must have an inner as well as an outer side, even if physics only gives us the latter. For panpsychism, there is a qualitative or experiential dimension to everything, including electrons and photons; just as there is a “dispositional” dimension, or the intrinsic possession of powers, to everything. Such a dual-aspect theory would grant interiority to subatomic particles, while also suggesting that the interiority of mesocosmic and macrocosmic entities need not be thought of as the “ground” of these entities’ powers, but as coextensive with them. Such an account both rescues Molnar’s overall argument from the vestiges of “smallism,” while at the same time preserving the intrinsicality and independence of objects without asserting that they are “withdrawn,” and without asserting that their causal relations are merely “occasional” or “vicarious.” For me, this is a way of taking Harman’s questions seriously, while at the same time giving more credence to the assertions of Latour (on the one hand) and Grant (on the other hand) than he is willing to; and of taking Meillassoux’s critique of correlationism seriously, without accepting his claims that mathematics = the absolute, and that things can and do happen for no reason. The occasionalism of both Harman and Meillassoux is rejected in favor of a Whiteheadian duality of determination and decision.