Graham Harman presents an ontological model that has the merit of being relatively simple and, at first sight quite clear. However this clarity and simplicity are obtained by means of a series of equivocal terms that function as the conceptual frame of his ontology. To attain coherence, Harman will have to decide on many points where he tries to have things both ways. In particular, the Heideggerian term of “withdrawal” has been subjected to an impressive widening of its extension and transformation of its sense, making it a key point of ambiguity and of contradictory determinations. Here, as elsewhere, the simplicity and clarity of Harman’s system are bought at the price of equivocation and incoherence.
Harman tries to maintain both an ontology of withdrawal and an ontology of abundance. He does this by equivocating over the sense of the “withdrawal” of an object, defined as its withdrawal from relation (relational withdrawal) and also as its withdrawal from other objects (objectual withdrawal). These two concepts are quite different, it is not the same thing to claim that objects withdraw from relation and that objects withdraw from contact with other objects.
In his discussions of withdrawal, Harman repeatedly declares that an object is “more” than its relation to another object, when strictly speaking it can only be said to be “other”, in the sense of logical grammar. In relational withdrawal, which seems to be the primary sense of withdrawal intended by Harman, the real object is said to be both more and other than its sensual manifestations conditioned by its vicarious interactions with the sensual realm. This is trivial, as an object is typologically other than a relation, it is not “more“.
In objectual withdrawal, the real object has other qualities than those involved in the interaction with another object. Those other qualities are either not involved , not “in use”, in that interaction (physical withdrawal) or not pertinent to it (logical withdrawal). Once again the object is not “more” than its relation, as this would involve a form of reduction.
The answer to the question as to what hit Newton on the head, thus giving him the idea of universal gravitation, is not that it was some ontological part of an apple, but that it was a particular apple (let’s call it “Mack”). What struck Newton on the head was Mack the apple, not some “slice” of Mack. More generally, an interaction cannot be said to involve only a “slice” of an object that exceeds the interaction, on pain of reducing an object, at least in part, to a series of slices.
The frequently cited example of the fire and the cotton is more complicated than Harman realises. Certainly if we mix phenomenological and physical predicates it can seem plausible to say that in the case of the fire burning the cotton the fire is “more” than its involement in this interaction. Its heat is involved but not its colour. Yet here Harman is conflating the phenomenological and physical domains, mixing elements from two incommensurable regional ontologies.
Certainly, heat construed physically and colour construed phenomenologically (as the colour seen by a human eye) have nothing to do with each other, and in this sense could be said figuratively to withdraw from each other. The only literal sense that withdrawal can have on this construal is merely a restatement of their incommensurability: the heat and the colour of the fire “withdraw” from each other grammatically or logically, just as each withdraws from interactions involving the other. However heat construed physically in terms of energy transfer and colour construed physically in terms of wavelengths of energy emission are one and the same entity.
Harman’s OOO is based on a network of such equivocations. The key terms of his system (object, withdrawal, relation) give an appearance of simplicity and clarity to an underlying logical ambiguity and conceptual incoherence. When Harman is obliged to take into account such ambiguity and incoherence he prefers to ontologise his way out of the impasse, rather than to clarify the differences in logical grammar.