On Realism

In Reason, Truth and History, Hilary Putnam defines metaphysical realism as the belief that ‘the world consists of some fixed totality of mind-independent objects’ and that there is, in principle, one true complete description of that world. His critique of this position is that (a) there is no way for us to climb out of our systems of reference in order to check that they successfully refer to the world; (b) we cannot rule out the possibility that the world could be conceptually divided up by us in a number of incommensurable ways, all consistent with our experience.

Contemporary speculative realism comes in a variety of forms, including object-oriented ontology and the kind of transcendental realism practised by Meillassoux. Without wishing to collapse these together, I think it is fair to say that they share an absolutising of the real, in the etymological sense that reality/objects are absolved (i.e. severed from) any dependence upon epistemological structures. Our access to the world does not determine that world.

It seems that this contemporary realism is not committed to significant aspects of the position which Putnam critiques. It need not entail a conviction that objects in the world are a ‘fixed totality’. Objects can change or join to form new, irreducibly real objects. The lists of objects which are part of the rhetorical style of OOO encompass radically diverse things, including physical assemblages, social groups and fictional works. Each of these ‘objects’ consists of other irreducible objects and so on. There is not simply one stratum of object.

For Meillassoux, the picture is different. In one respect, the absolute consists of the fact that anything can be different for no reason: there is no founding ontological or transcendental necessity for the order of things. And this is what we can know. So his realism also does not entail that there is one fixed totality, or one complete and true description of things.

Is Putnam’s critique therefore no longer relevant? The problem, it seems to me, is that in a laudable attempt to dethrone an anthropocentric epistemology from philosophy, these contemporary versions of realism are still dependent upon theories of access to and translation of the real. For example, to characterise the absolute in terms of hyper-chaos, as Meillassoux does, implies a judgement whereby chaos and order can be distinguished. It further implies that the possibility of order is a legitimate product of chaos. But if this is the case – if order and our ability make sense of things are themselves possibilities produced by the absolute – then we are in no position to judge that the absolute is ‘ultimately’ chaotic. We do not advance beyond Kantianism, in which the absolute provides the supersensible basis for our knowing of the world, whilst remaining unknowable in itself. Attempts to give an ultimate characterisation of that absolute lead to antinomy. It would be interesting to construct such a Kantian contemporary antinomy, in which the absolute could be rationally proved to be both order and chaos.

In OOO, the concept of translation has been explicitly used by Graham Harman. Objects, he claims, relate to each other indirectly, via translation: taking up the sensual images of other objects, whilst remaining inaccessible to relation in their withdrawn interior. However, if such ‘translation’ is to result in new objects (parts fitted together to make a machine, for instance), we have to ask what is it that constitutes the inaccessible interior of the new object? The answer must be: a system of differences, of translations, of mutual interpretations. So, having dethroned human epistemology from philosophy, OOO has arguably displaced questions of access, translation and interpretation into the absolute ‘itself’.

All of which is to suggest that Putnam’s critique is still worth a look. Access to the world, consisting of translation between objects or the translation of the absolute itself into order and sense, constitutes an irreducible part of that world. The world does not offer itself as one mind-independent order of things.

This is not meant as a dismissal of speculative realisms, but an appeal to move beyond their initial self-definition, as simply radically opposed to anything that smacks of deconstruction or phenomenology, for instance. It is simply not the case that deconstruction is about irony and the prison house of language whereas realism is about seriousness and the great outdoors, any more than it is the case that realism is a stupid scientism. Perhaps issues of access and translation could provide the means for the varieties of continental and post-continental philosophy to begin to talk with, rather than at, one another.

16 thoughts on “On Realism

  1. Edit: a fairly crucial ‘not’ was left out of the final sentence of the penultimate paragraph of this post when first published! It has now been added. Apologies for turning the order of reality upside down.

  2. “But if this is the case – if order and our ability make sense of things are themselves possibilities produced by the absolute – then we are in no position to judge that the absolute is ‘ultimately’ chaotic.”

    I hope this doesn’t come across as nitpicking, but Meillassoux is pretty clear that an everlasting, absolute order without any becoming could appear. The “ultimate chaos” is just the infinite sets of non-contradictory possibilities; that chaos never being actualized or never coming into appearance is one of those possibilities.

  3. “the infinite sets of non-contradictory possibilities” just sounds like a set of possible words or models. As I suggest in my post, this means that hyperchaos would just come down to the claim that non-contradictory statements could be true in some or other possible world. I hope Meillassoux has something more interesting than this in mind.

  4. But surely there is an underlying claim for a hyper-chaos, which does not mean that everything is actually chaotic (whatever that would mean), but that nothing happens for any reason. In which case, no order could be ‘absolute’, since it would be predicated on the permanent possibility of its disappearance for no reason whatsoever.
    One question I have (for which there is probably an answer) is this: Meillassoux is clear (in Spectral Dilemma) that a God could appear. Does this entail that it is possible that a necessary being appear? If so, modal logic will presumably show that a necessary being must actually exist, since if one is possible (is actual in a possible world) it must be actual in all worlds (since necessary being is, by definition, one that exists in all possible worlds). Meillassoux could rule this out, as nothing is necessary, but then we get the problem: is the non-possibility of necessary beings itself necessary? Is the law of non-contradiction itself necessary? Does Meillassoux account for this?

  5. The possible worlds of modal logic are counterfactual. Kripke makes clear that the point is to stipulate them, not to describe them. For Meillasoux it seems it is one world that bears all possibilities. It seems another case of his twisting an epistemological concept into an ontological one (even though I don’t think he metions modal logic).

  6. I don’t see how the fact that absolute stasis is possible means that the absolute is not ultimately chaotic. The latter permits both determinate order and determinate chaos – it does not permit the possibility of swallowing its own tail and becoming ‘hyper-ordered’. In any case, I think the original point is importantly true: if we can distinguish hyper-chaos and order, then each is a determination of being and not only the latter. Meillassoux arguably wants to be seen as sensitive to this sort of equipollence problem, and as going further than the German idealists towards resolving it. I just don’t see *how* he goes any further in this respect. Any determination of the absolute can be opposed by its negation. So if there is something other than the absolute (notice the conditionalisation), how do we account for this distinction without sliding back into dogmatism? This is why there is a tendency towards denying the being of the absolute (amongst those who don’t want to go the other way and deny the being of the non-absolute) – since if the absolute is non-being then it can be distinguished from the being of the non-absolute without being determinately predicated. However, some for whom this remains unsatisfactory propose that we arrest this tendency just before the absolute is rendered as absolute nothingness. What does this get us? Relative nothingness, or: matter. Here I think Nietzsche’s quip is apposite: Schopenhauer only succeeds with dictatorial tone in draping predicates over something completely dark and ungraspable. I wonder the extent to which this verdict applies to all (aspiringly non-dogmatic) materialism.

  7. Regarding modal logic, Meillassoux is not obligated to accept the inference from ‘possibly necessary’ to ‘actually existent’. At least, a case could be made for why anyone coming from the German idealist tradition would want to fiddle with the accessibility relations between worlds – e.g. in light of the non-totalizability of the set of them. Markus Gabriel’s discussion of Schelling is useful here. In any case, I don’t think the possible future God is meant to be a necessary being for Meillassoux. Aren’t necessary beings already ruled out by the necessity of contingency?

  8. @ David Roden and @ Steven Shakespeare:

    Also at the risk of sounding “knit picky” but really in the spirit of good dialogue:
    I think for Meillassoux order “appears” at a meta-level though is not “necessarily” part of surgcontingency. There is no order to the surcontingent save for the temporal structure that it establishes, which is by the “figure” of non-contradiction (basically the principle of process or becoming).
    By that I mean surcontigency is not a non-totalizable set of possibles in such a way that those possibles could form possible worlds. Rather, for example, the “possible” of the divine inexistence subsists as a virtuality – it technically neither “exists” nor “doesn’t exist,” but rather may suddenly appear from an ontological “nowhere,” a true non-existent. The distinction between virtuality and possibility is crucial in this case, though certainly difficult to parse out, as, in part, possibilities are conceptual whereas virtualities are ontological realities that we seem to be able to identify after the fact; that is, conceive of. Although conceiving of a virtuality seems to me to suggest “conception” in a Platonic sense, “to give birth to.”
    This is important, I think, because Meillassoux’s take on non-existent virtualities begs some questions found in the debate between Whitehead and Hartshorne. For Whitehead eternal objects subsist in an ontological nowhere, they are virtual, and may come to be or not. Whereas for Hartshorne there are no individuated particular virtualities, just the virtual capacity for further particularization. This seems to be what Meillassoux wants to say, but doesn’t. The divine inexistence isn’t fully individuated so far as I can tell.
    “Necessity”is rendered in terms of “figures” (at least in The Divine Inexistence). Surcontingency as a ground is necessary itself, but nothing from it arises necessarily, nor is anything other than the surcontigent ground necessary. So of its virtualities – which on my reading it does not seem that those virtualities are strictly individuated in any one way versus another, at least as conceptual possibility – we cannot say that one is necessary or not, or will definitely appear or not.
    The trouble is that even if the divine inexistence appears, would it truly be necessary if its necessity is only taken on after it begins to exist? I somewhat thing true necessity must be non-temporal (incapable of not existing, at any time – past or future). There is an implicit temporalism that rides with Meillassoux’s modal theory.
    Hopefully some of this makes some sense. I welcome comments in response.
    Leon / after nature

  9. I’m wondering if looking at order and chaos dialectically is the way to go. I get the impression (I haven’t read him in a while) that the appearance of order is still a determination of chaos. As we’ve been pointing out, hyper chaos doesn’t mean constant flux.

  10. Meillassoux clearly says in After Finitude that it is necessary that there be a contingent entity (i.e. determinate being). I read this as saying that chaos is intrinsically determined in one way or another. Putting this aside, given that there are determinations *as such*, to say that chaos is indeterminate is to say something determinate about it. Perhaps this is a good gloss on Eric’s suggestion that the order/chaos dichotomy is illusory. But there are other glosses which are less accurate – it is very easy, for example, to slide into the view that, sub specie aeterni, determinacy lacks being.

    Leon’s question (how can something become necessary?) got me thinking about Schelling, whose position in several respects resembles Meillassoux’s. For Schelling God will exist necessarily, but not in a sense of ‘necessary’ that we can presently comprehend. I remember that Christopher Watkin also argues that Meillassoux has no way of ruling out transformations in our understanding of modality that would permit for the emergence of a superior type of necessity that is presently incomprehensible to us. I think the least that can be said about this is that it requires us to think the absolutely determinately and hence dogmatically; Schelling accepts this, but I personally find this option distasteful.


  11. @eric I acknowledge in my post that Meillassoux would be liable to reject the possible worlds interpretation of modality. So we cannot interpret the thesis regarding hyperchaos in terms of the possible words reading of de dicto or de re contingency. Just how we an interpret it, I don’t know, and I’m not sure if it is an obligation we need to discharge since I also reject Meillassoux’s correlationist assumptions. If there is no correlation, contingent or otherwise, then the problem simply does not arise.

Comments are closed.