Wolfendale and Object Oriented Philosophy: Some Remarks and Questions

Wolfendale and Object Oriented Philosophy:
- Some Remarks and Questions -
__________________________________________________________

Following up on some of Pete’s excellent commentary on OOP over at deontologistics, some further issues may be raised about the status of withdrawal in both Graham and Levi’s respective accounts. Pete raises doubts about whether OOP has provided justification for the thesis that ordinary knowledge is finite and inadequate, i.e. in order to advance the thesis that real objects are always in excess of our knowledge of them (which Pete calls the excess principle of withdrawal; the other one being independence: that things exist independent of their relations). The worry arises for Harman’s account in particular, since it is supposed to be the phenomenological ground for the purchase of intentionality extended to all objects. If phenomenologically grounded withdrawal cannot be established, then it becomes less clear how Harman can make good of the epistemological claim that access to things themselves is impossible.

Pete argues that at this juncture Harman brings in unjustified ontological assumptions, which can ultimately vitiate his own position. Interestingly enough, Pete centers his argument in a single assumption, an assumption which he claims may also be used to undermine Meillassoux’s own 'hijacking' of the correlationist argument. For this, Pete claims that what is common to both speculative realism and Harman's OOP is a certain way of understanding what it is for something to be in-itself, and in order to map this common understanding he distinguishes between two possible senses of what in-itselfness could be taken to mean:

Mind Independence: Something is in-itself is if can exist independently of the existence of ‘minds’.

Attitude Independence: Something is in-itself if the way it is is independent of the way we take it to be.

Pete writes:

“I’d suggest that Meillassoux and Graham adopt the former, more traditional way of understanding in-itselfness, although in Graham’s case it modifies to the claim that something is in-itself if it can exist independently of its relations to other entities (insofar as all entities effectively play the role of ‘minds’). The important point is that the former is an ontological way of conceiving in-itselfness, whereas the latter is purely epistemological (insofar as it says nothing about existence).”

The central claim here is that adopting Mind Independence leads both into an ontological grounding of epistemology, and to correlationism finally. But he has certainly not established that both theses are necessarily incompatible with each other, nor shown where Meillassoux or Harman would clearly stand with respect to both theses: he merely asserts they both endorse mind independence.

Now, Meillassoux certainly adopts Mind Independence, insofar as not only the in-itself can exist in the radical contingency of all beings, but since a contingent entity must exist. Pete then claims that in establishing the general contingency of all there is Meillassoux must deny that any knowledge of particular entities is possible. He writes:

"In each case, the fact that we cannot know anything about particular entities functions as an important premise in an argument that establishes something about entities in general. What thus unites speculative realism and OOP is that they hold that all knowledge about particular entities is relative/uncertain/impossible, but that we can thus know something absolutely about the structure of entities as such, namely, the ontological fact which underlies this relativity/uncertainty/impossibility."

But this seems a bit baffling, although I could be wrong in my reading here: Meillassoux does not think it is the absence of any knowledge about particulars which entails knowledge about them in general, it is the knowledge that every particular could not exist that grounds the absoluteness of facticity and solicits its extension to all beings. This he does by saying that any attempt to relativize the absoluteness of contingency invariably results in its re-absolutization at a higher level. Certainly, this entails no entity or quality can be said to be necessary, and therefore such knowledge is disavowed. But it by no means entails that nothing can be known of particular entities; on the contrary, it simply annihilates the gap between the thinkability of an absolute and thought so as to facilitate the purchase of being by thought. Once absolute contingency is deemed thinkable and necessary, nothing prevents from deriving knowledge which would shed light about particular entities or their qualities, so long as they don’t imply their necessity, i.e. the violation of absolute contingency.

In other words, Meillassoux would most certainly reject the Attitude Independence thesis, insofar as there is no necessary epistemological gap between how we take entities to be and the way they are; i.e. we can take the in-itself to be as radically contingent, but immanently thinkable. The problem for Meillassoux becomes rather to elucidate the structure of this coincidence of world and word, being and thought, since it is left provisionally as an unexplained legibility based on the congruence between the formal inferential rules of logic and mathematics, and the structure of thought itself. This congruence between thought and thing, objective structure and conditions of thought, follows from what appears to be underwritten by the Badiouan endorsement of the Parmenidean thesis that being and thought are the same. The problem for Meillassoux thus pertains to the essential obscurity of the relation between inferential logico-mathematical discourse and thought, a clearly ontological relation which is apparently anchored in Badiou’s own mathematical Platonism, which is itself articulated around the dissolution of the form/content distinction (See The Concept of Model) and the thesis that mathematics is ontology.

Although Meillassoux admittedly restricts the ontological scope of mathematical discursivity, it should be noted nevertheless that, like Pete surmises, this remains unexplained in his preliminary account of correlationism in After Finitude. (Pg. 103). Whatever else Meillassoux specifically anchors on set-theoretical strictures (and in particular the Zermelo-Fraenkel axiomatic Badiou draws upon), it is clear these underlying ontological commitments also underlie his epistemological positions: for example, in appealing to Cantor’s theorem to advance the non-thinkability (and impossibility) of totalizable being. Since the latter is critical for Meillassoux’s arguments against the possibility of applying probabilistic reasoning to the universe as a whole, in his way to assert the ontological contingence of the laws of nature, Pete does well to take notice on these shortcomings.

On the other hand, Graham would apparently endorse both positions. Like Pete states, it is clear he defends the Mind Independent existence of the in-itself from thought, insofar as the object withdraws from all relations, and so the real object/qualities remain forever withdrawn from their sensual doubles. But he would also, I take it, defend the Attitude Independence thesis, since the way any object appears to me, or to any other object, is in sensual (intentional) form, and these phenomena are absolutely separated from real objects and qualities.

On Levi and Withdrawal

Here I’d like to fast forward to Pete’s series of posts answering Levi and presenting his many qualms with the latter's version of OOP. One of the main targets of Pete’s critique pertains to the status of how ‘translation’ operates in Levi’s account, in order to advance the excess thesis; which as we stated above, claims that objects exceed knowledge of them. In Levi’s particular account, this comes down to the claim that communication between distinct systems is necessarily destined to fail, since every actual manifestation or 'perturbation' which occasions a change in another system involves a process of translation which necessarily depends on system-specific features (its ‘virtual structure’ in Levi’s own Deleuzean words).

Although Levi’s precise adoption of the Deleuzean frame has not been made excessively clear (indeed, I am hoping Democracy of Objects will deliver on this account), we can stipulate that this entails that a) every object has a unique virtual structure or ‘proper being’, and b) translation invariably ‘involves’ the whole of the object’s virtual structure, The reason why both (a) and (b) are necessary is that if two identical systems exist then direct communication appears possible, insofar as the information transmitted between two systems could result in identical information, and nothing prevents us from saying transmission has taken place without loss or excess. But if direct communication is possible on occasion, withdrawal is threatened wholesale.

On the other hand, Levi would also need to say two systems cannot achieve direct communication on the basis of some common features which can be the only ones involved in a particular relation or transmission. Levi must either require additional metaphysical criteria for why failure would obtain in such cases (since dissimilar features of the system would presumably not take part of the relation), or else stipulate that the entire being of the system is implied in every relation. This last point is needed to make local manifestation invariably system-specific insofar as the total set of features involved in a translation process will be unique to the system performing the translation, even if common features in their virtual structure are shared, and so the actual manifestation will also be always unique. But of course this seems to entail that no two systems can ever share a common quality, nor that common qualities could ever be shared among two different systems. The problem is that since we have not yet been given an account of the general architecture of an object’s being or virtual structure, it seems entirely obscure whether we can gauge similarity is possible at all between systems and their qualities. It also becomes incredibly hard to measure where exactly translation in their causal interaction entails information loss, i.e. that the excess thesis holds. In order to gain greater clarity on this juncture let’s review what Pete writes:

“Translation is the manner in which substances transform the perturbations of other substances according to the substance’s own endo-structure or internal organization, producing new qualities as a consequence.

In essence, Levi holds that causal interactions between objects involves some form of communication between the proper being of each object, but that this communication is indirect, or subject to translation, insofar as the information involved is dependent upon the specific virtual structure (i.e., proper being) of the affected (or receiving) object. The withdrawal of each object’s proper being in relation to the other thus consists in the system specificity of information.”

Here we can raise innocently our questions again. First, we should ask straightforwardly: can there be two identical systems? If so, why would the information imply withdrawal at all, given that the virtual structure of the two systems would by hypothesis be identical and therefore the translation of an actual manifestation into another system would result in the same manifestation, or output? Perhaps an account of temporal separation between endogenously identical objects could do the trick, but it’s far from clear Levi’s account is equipped to take such a move. If Levi wants to deny that translation between identical systems results in identical outputs, and therefore that there is still something like an ‘excess’ involved, he obviously needs an account of how the processing of information based on a system’s actual state may produce different states or actualizations once translated into another identical system. Nothing of that sort seems available at this point, so this thesis seems resolutely implausible.

Alternatively, if Levi wants to deny that two systems are ever identical, and that this accounts for the inevitable withdrawal of all entities, he needs an account of how non-identity in virtual structure entails excess upon translation, i.e. why it is not possible for two non-identical systems to at least sometimes translate information identically or, put differently, for the translation process to result in identical manifest system states in two non-identical (virtually specified) systems.

Finally, if Levi wants to deny that a manifestation or actual system-state is identical to the translated information of a system, he needs an account of the difference not only between virtual structure and actual manifestation, essence and accident, but between the manifest accidents and the manifest information manifested upon contacting perturbations in the environment. Since this position runs against the idea that information produces new system states, Levi would presumably simply say information is identical to a manifest system-state. But then he needs the account to show how information exactly entails loss in its translation from system to system, which of course, as Pete notes, presupposes a standard for direct communication which fails every time. Since it is out of the question that two systems could be completely identical, the question rises whether at least they could share some features. If so, we might stipulate translation involves the actualization of the virtual being of the system, and that every system has a distinct virtual structure, while claiming that direct access would be possible in the way of a communication which only involves shared features between systems in their being, which would entail identical local manifestations. In order to deny this, not only must the virtual structure of every system be unique, but every manifestation must be native to the system.

The problem is that if all systems are distinct, and all manifestations are not only system specific but unique to the system, then for Levi there would be no shared qualities or manifestations between objects. Each and every object entails its own set of manifestations, and it becomes increasingly hard to see what a description of an object or its qualities would be like given any descriptive recourse would undoubtly have to be utterly specific to the system. Whether objects are absolutely dissimilar in their being or admit of similarities at the virtual level, Levi must at least claim every act of translation between two entities will involve dissimilar features in order to save the withdrawal thesis.

An alternative would be to say that there are common qualities insofar as they can be separated in accordance with general features, while it nevertheless remains the case that every instantiation is unique and specific to the system. Thus two systems might exhibit similar abstract features, while nevertheless preserving an inexhaustible void of difference in its depth differing every manifestation from all others. This indeed would be a more predictable Deleuzean line.

But in order to avoid the virulent proliferation of qualities and their non-transmutability, Levi must then produce an account of how not only objects have virtual structure and accidental manifestations, but how manifestations themselves have essential and accidental features as such. This means that for Levi the distinction between the accidental features of an object and its differential strata cannot be said to be a mere function of representation, since presumably the pure being has particularity and is endowed with definite qualities, i.e. is not just a virtual differentiating field without objective specificity. Presumably, then, it would be the difference in these accidental elements which always differ form system to system, which would be left to ground the possibility of generic qualities to be expressed between different systems, thus saving the withdrawal thesis, i.e. translation always involves loss insofar as no two systems and system states are ever identical; even if they might be sufficiently similar so that the essential aspects of a quality are actualized from system to system. So the gap is not so much between the object and its manifestations, as much as how these manifestations produce information in other objects.

Following with his analysis, Pete writes:

“Levi initially described withdrawal as being a split between proper being and local manifestation, much as Graham describes it as a split between real objects and sensuous objects. However, the way it is described here suggests that it is a split between the local manifestation (or perturbation) and the information it produces. This is reinforced by Levi’s suggestion that information plays the same role in his system that sensuous objects play in Graham’s.“

Levi must say that information is just a manifestation understood as caused by a perturbation; but that there is no categorical difference between the two. A good question arises as to whether all manifestations are then caused by perturbations or whether there are system-intrinsic induced manifestations possible. This of course pertains to the core of questions of causality in Levi’s version of OOP: are all relations and causal chains between systems, or are there also relations and chains within systems, i.e. between their virtual proper being and actual manifestations as system-states? How would these latter occur, seeing they would seem to be relations that would not be indirect insofar as they wouldn’t require translation between different systems, and therefore no information in Levi’s sense? Seeing Levi would probably choose to say all system-states or manifestations correspond to complex exchanges with the environment (and thus causal relations upon perturbations) there seems to be a chicken-egg problem insofar as all manifestations presuppose their actualization as information, and therefore that all system states which serve as perturbations for other systems to translate themselves presuppose a perturbation which actualized them, etc. To quote Pete again, at length:

“The difference is that substance must have something like properties that are distinct from the qualities it manifests, because there must be particular features of objects’ virtual structure that distinguish them from one another, which are nonetheless distinct from their actualised qualities. Whatever is analogous to sensuous qualities for Levi must be part of the information communicated in causal interaction… the key issue is whether Levi can account for this idea of directness and indirectness in informational terms. Levi seems to have done this by appealing to the idea that information is system specific. However, Levi has actually used the metaphor of translation, and the corresponding talk of directness in order to cash out the idea of system specificity”

While I agree mostly with Pete here, I find the claim in bold to be a bit baffling. System specificity is cashed out in terms of an objects endo-structure or virtual features; while a standard for directedness has simply not been described, unless one draws the intuitive idea that a direct communication would be the manifestation of identical qualities between two systems, which we have seen entails a host of problems. Pete does well to ask then what would non-system specific information be? Here we might surmise a few things.

Presumably, essential properties of manifestations which can be identified across objects/systems, even if every manifestation has accidental manifestations which are necessarily system-specific. This would mean information is always system specific, but still exhibits general features. If not, as we considered before, every manifestation or quality would be system-specific and there would be no ground for common properties between objects. As we have seen Levi has no grounds to say these general features are any only representational on pains of avoiding de-specifying the object wholesale, which would make of systems nothing but the Deleuzian virtual diced up into a plethora of objects, and where what an actual object or quality would be must be necessarily meaningless. But this seems to deliver Levi’s account to the same kind of problems I described about Harman previously on individuation, insofar it seems to make of reality nothing but an anonymous multiplication of possible entities on the basis of infinitely conceivable actualizations brought about by conception. In any case, it seems Levy must elucidate his relation to Deleuze more expressly to make sense of all these claims. In any case, Pete seems to arrive at somewhat similar conclusions on this point:

“One could claim that although the qualities produced are similar to those that brought them about, they are not for that matter identical However, in order to do this, one either has to claim that all qualities are haecceities (i.e., purely unique), which would be metaphysically disastrous, or one has to claim that it is the fact that they are qualities of a different object (and thus part of a different set of qualities) which makes the cause and effect non-identical. This amounts to saying that for their to be direct contact between entities, one entity would have to become identical to the other.”

Lets fast forward again and dive into Pete’s discussion of Levi’s account of meaning and its structure. The latter argues that knowledge should be understood primarily as production and not mere reporting; knowledge involves grasping information so as to produce a new actual state rather than the transmission of an already existing one. Pete disagrees violently with this reading, but I think his objection doesn’t quite cut it here:

“This [claim that knowledge involves primarily production and capacity of using knowledge to produce new states of affairs] could be interpreted as the claim that although knowledge can be communicated in linguistic form, you only grasp this knowledge if you can actually put it to use. However, this would need to be reconciled with the claim that much newly produced knowledge has no immediate use at all, at least by ordinary standards. Theorems discovered at the forefront of mathematics are understood prior to the discovery of any application they may have, and this means that understanding them cannot be a matter of understanding their application. Similarly, at the forefront of physics their are often different theories that have no difference in real world practical applications (to the extent that we can’t even experimentally choose between them), but we can nonetheless understand how they differ in content.

The important point here is that if one is going to try and reduce knowing-that to knowing-how, you need to actually specify precisely what kind of knowing-how the former consists in. If you don’t do this, then you’ve simply abandoned the distinction rather than explaining it.”

This objection doesn't work, insofar as it rests on an excessively narrow sense of production and use. Since Levi’s distinction between reporting and producing is primarily anchored in the distinction between repetition and innovation, the procedures which produce local manifestations apply to newly developed theories in experimental sciences just as revolutionary political movements. Doing theory is a form of production, and as long as it can be distinguished from merely repetitive activity, it necessarily concerns actualizing local manifestations. The key question is, then, whether Levi is actually equipped to advance the distinction between repetitive and non-productive reporting and innovative productive inventing. Since even ordinary knowledge transmission, in non-experimental science for instance, must involve translation, Levi is thereby forced to distinguish between translations which produce new manifestations from merely repeated ones. But given the withdrawal thesis, it seems his account of information-system is ill equipped to make this distinction. As we have seen, for Levi systems are unique and specific, and every communication is miscommunication insofar as they produce unique information. To this extent, every local manifestation is decidedly a new system-state, produced by a perturbation.

Here, however, we must look into Levi’s account of particular systems which involve meaning: since in these cases Levi seems to allow for common meaning between different systems. It would then be a matter of distinguishing which meanings are simply reported in the sense of translating them in such a way they are preserved as the same meanings after the translation process, and those translation processes which produce new meanings. This seems coherent enough, although I’m not aware if Levi actually endorses this possibility. In any case, this does seem to retain the anthropocentric dimension of human-knowledge which allows meanings to be communicated between systems, and non-human communication in which qualities are never transmitted, but information inevitably indirectly produces new-states. The problem is that this seems to awfully hypostasize human knowledge into the sole case in which something like a direct relation is possible, insofar as meanings can be transmitted between systems as such.

Now, Levi might want to say the dimension of withdrawal nevertheless remains in much the same way as quality transmission remains in non-human relations between objects. For example, we can admit that different systems have similar qualities, while preserving their uniqueness to their native systems, by distinguishing essential from accidental features of qualities themselves: systems might share features in their virtual structure or actual manifestations with other systems, while each might nevertheless has distinctive features which affect the process of translation and thus the account for the distinctiveness of local manifestations. In the same way, meanings might be translated from system to system in such a way they are qualitatively preserved (report), while we may easily imagine a process of production which involves previously unspecified local manifestations and meanings. This scenario seems to collapse the presumably unique character of meaning which is human into relations between all objects.

But since in Levi's account only humans produce meanings insofar as only humans engage in reflexive consideration of alternate possibilities, to claim knowledge is produced in the same way new system states occur seems to either radically collapse the distinction between the human domain of meaning and the more broadly non-human domain of translation. Indeed, the 'subjective form' of meaning production which constitutes knowledge seems no longer separable from natural processes, or general relations between different entities. This seems a thesis Levi would be happy to endorse, insofar as it places objects and humans in the same metaphysical footing and preserves both the possibility of novel production, repetitive transmission, and withdrawal. However, Levi has no real account of how the 'productions' put to use in novel-meaning events are distinguished from reports in the more basic strata of proper being of the object (virtual), local manifestation (actual) and in the translation of information between the two. How are reports categorically distinct from 'productions' (or novel actualizations)? Even if Levi claims the crucial distinction lies in how meaning-producing systems are endogenously specified to discern information in the projection of alternative possibilities, how this projection allows the proper distinction between report and production remains obscure. We have stipulated how Levi could reply, but this is all sheer speculation at this point, since the account of novelty is missing.

It is no surprise then that Levi’s account faulters precisely at the point of specifying what specific know-how reduces knowing-that. This is where the distinction between human and non-human systems becomes unsustainable, insofar as at this juncture either can only be described as know-how in the sense of information processing systems, which as we have seen, share the same structural features in communicating meaning as non-meaningful events.