http://cscs.umich.edu/~crshalizi/reviews/brainchildren/

Of course, Dennett’s idea of compressibility is framed in Turing complexity terms which assumes classical computation which itself is a framework which can already be superseded by quantum computation, etc., but the basic idea is appealing to me: simply that, in some way, one can say it is reasonable to assume that reality as we encounter it has some patterns which can be approximated by models or processes which require less specification than the raw percepts.

I agree with this completely. However, as you point out, decomposing the world into objects (no matter how cleverly designed to avoid the error of naive correspondence theory) doesn’t seem to me to be needed to get that effect. At the very least, as an intuitive device it seems to me to impose more structure than is needed, in a rather awkward and unnatural fashion.

>I am not necessarily hinting towards the
>‘flux vision’ but more towards what I could
>call a ’speculative nihilism’.

True, even the notion of a “flux” adds a bit too much to the picture, as that sort of image implies a kind of substance which flows, which has a kind of continuity, extent, and so on. That’s why I like Dennett’s even more minimalist condition: that reality is patterned in a way which is computable, i.e., you can attempt to describe the patterns in fewer bits than you have in your observational data sets. That alone implies some independence from the purely subjective without having to posit very much additional structure, even a “flux”.

http://speculativeheresy.wordpress.com/2009/04/20/ontic-structural-realism/

He raises a lot of the concerns I have with object-oriented ontology. In particular, he says:

‘What Ladyman and Ross add to this is twofold. On the one hand, they add the expertise to pronounce that there is a “convergence” in philosophy of physics towards the idea that there are no such things as individual entities in the study of fundamental physics. Undertaking an in-depth and extensive look at the various work being done in contemporary physics, they argue persuasively that individual things don’t exist. Rather, what exists are ‘real patterns’ – temporal and spatial patterns which are mapped by the mathematical structure of scientific theories. (ETMG, 120; also, cf. Daniel Dennett’s ‘Real Patterns‘ essay) Now patterns, in Dennett’s formulation, must be capable of being captured in a smaller amount of bits than the original data set from which they came….

What this all means for OOP is that the notion of real patterns articulated by Dennett and extended by Ladyman and Ross is a useful conceptual tool for understanding the irreduction principle. It provides, in a computational manner, explicit criteria for formulating the reality of objects in the special sciences. But as Ladyman and Ross remind us, “one makes a metaphysical mistake if one reifies these essences and imagines that they are the real constituents from which the world is fashioned.” (241) In other words, it seems to me that while OOP can be a productive theory of ontic relations and potentiality, it may not be sustainable as an ontological theory.’

Dennett’s formulation, i.e., representing discernible patterns in terms of computability — this seems very general and powerful, and a clean and elegant way to express the idea that one may be able to “refer” in some sense to reality — that is, in the sense that one can discern patterns which are computable, describable in fewer bits than the input (i.e., one could say that you can compress sense data into fewer bits in some sense). This seems like a perfectly reasonable thing to talk about, and is obviously quite consistent with quantum mechanics and other physical theories and interpretations.

Note that of course, I agree that even in these relatively fundamental models, you have “particles” and so forth —- but these particles can be seen to be radically interconnected with other particles so that it may be impossible to tease them out from either each other or the observer in any coherent fashion. Object-oriented models, to be consistent with physics, have to admit that the statistics of an object’s potential to be observed in a particular manner may change as the result of changes to objects scattered across the universe, even backwards in time (i.e., quantum entanglement), or they may have to admit that there is a fundamental entanglement of the observation with the observer (which actually is friendlier to the notion of locality, ironically). But if you have to have objects with the property that they have no beingness that is entirely separable from everything else in the universe, then why bother making objects a fundamental part of your ontology at all? It seems quite strained, and you end up with objects that aren’t very object-like (i.e., they have no clear boundary and no clear beingness separate from other objects), or you can treat them as seems far more reasonable to me as convenient fictions imposed upon reality, but which may have some “reality” in the sense that reality admits itself to mathematical description in the form of computability.

As for how to get objects from a flux; again, I reference the wonderful work of Brian Cantwell Smith, who deserves far more fame than I think he has gotten for his work. He goes into great detail into this question and I believe comes up with a very plausible story, without making very many metaphysical assumptions whatsoever (though his arguments tend to assume a classical world, they are nevertheless quite illuminating).

There’s a distinction, I think, between “thinking primarily in terms of the subjective” and recognizing that one’s ontological theory is a conceptual model of the world. Being a conceptual model means translating between what’s “out there” and what goes on in our heads that allows us to understand things. That’s one of the reasons why I think cognitive science is so important to ontology.

I agree with you that there are cultural factors at work in the forms that our theories take, but I think that the “object” distinction is pretty firm in all humans, as evidenced by their languages. Quantum mechanics talks about quarks and their charm, electrons and their spin. Objects (particles) are not only present, but they’re also given properties using very conceptually rich and common words that don’t apply literally.

In Prince of Networks, Harman says that the problem of “object” philosophies is explaining how objects connect, and the problem of “flux” philosophies is explaining “why the world is not a single molten whole, devoid of regions”. Kind of puts one in mind of the old two-slit experiment.