In response to a recent post, Paul Bains raises a number of questions that I believe are worth responding to as they often come up in relation to object-oriented ontology. Paul writes,

Trees = ‘big’ multicellular perennial green plants (which obviously lack semovience/self-movement).

So, at the risk of the deadly repetition here’s the thing:

The class of nonhuman natural objects is not as simple as we might think – e.g. we think it includes ‘trees’ and honycombs rather than ‘periodic oscillations’ (Sonigo) where these apparent entities are subjective categories.

In a previous post LS refers to something like the non-epiphenomenal level of complexity of the organic….
I’m no physicist but I suspect some of them might demur.

There might be many smaller objects but not some many ‘classical’ ones.

Now this is likely to be dismissed as some kind of valorizing of the sub-atomic – but within the domain of ‘natural non-human objects’ the sub-atomic might be the real thing.

Of I would never actually say that – just curious as to how those non-anthropocentric ‘tree-believers’ really, really, believe in trees – The same way they believe in ‘neutrinos’ or ‘black holes’…or God.
Altho there might (for Whitehead?) be a god without any trees.

I really cannot say anything about Sonigo as I know nothing of his work.

image1With respect to Paul’s comment here, there are two particularly relevant claims advanced by object-oriented ontology. The first of these claims is mereological or about part/whole relations. For the object-oriented ontologists, objects contains other objects in much the same way that Russian dolls contain other dolls. The point that a rock contains atoms, electrons, and other particles besides, does not undermine the thesis that the rock itself is an object, nor does it make the rock less real than the particles it contains. While it is indeed true that the rock cannot exist without these particles, the pattern or structure or system that characterizes the rock is nonetheless what characterizes the rock as a distinct object. Here it is worthwhile to think of Zubiri’s characterization of existents or objects as “systems of notes” in his book On Essence.

read on!

This brings me to the second point, which, I believe, is one of the stranger and more provocative claims of object-oriented ontology. All objects are independent of one another. This is where the mereological thesis gets really strange. The particles that the rock contains are themselves independent objects and the rock itself an object independent of the particles that it contains. Thus, while the rock cannot exist without these particles, the rockness of the rock is nonetheless independent of the particles that contain it. Since Paul is familiar with dynamic systems theory and autopoietic theory I suspect he’ll recognize that the logic at work here is similar to that of how the environment/system distinction functions for autopoietic systems. As Maturana and Varela argue, for example, atoms belong to the environment of cells. While cells certainly cannot exist without atoms, atoms do not compose the systematicity of cells.

There are all sorts of good reasons we can evoke for this odd mereological thesis about the object-independence of objects contained in other objects and the object that contains these objects. Thus, for example, the cells that compose a body are constantly dying and new cells are constantly being produced. Nonetheless the body that contains these cells persists. In the case of the rock above, all sorts of atomic events are taking place among its particles where electrons are jumping from state to state and being exchanged with electrons in other atoms, but the pattern of the rock persists. Similarly, the United States, an object, contains all sorts of other objects like persons, trains, trees, and so on, yet these objects are both independent of the United States and the United States persists with the coming-to-be and passing-away of these other objects.

In this connection, I see no reason to demure to the quantum physicist, though I certainly acknowledge the reality of the objects she investigates. The reason for this is simple: There is no reason to grant the smallest as having the most reality or as being true reality simply because it is a condition for other entities. Entities at different levels of scale have emergent properties distinct from the properties of objects at lower level of scale. In the past I have called these emergent properties “logoi” to capture the distinct “logic” or behaviors of these different patterned relations.

In my last post I argued that object-oriented ontologies tend to advocate an anti-realist epistemology while nonetheless advocating a realist ontology. It is worthwhile to rehearse and expand upon this thesis a bit in response to Paul’s comment above. Anti-realist epistemology actually flows directly from the basic claims of object-oriented ontology and one of its central claims regarding the being of objects. In my experience, this basic and central claim is the most ignored and overlooked claim in all discussions of object-oriented ontology.

In Harman’s ontology, this claim is the claim that all objects withdraw from one another such that no two objects ever directly encounter one another. In short, Harman’s move is to take Kant’s thesis about the mind’s relationship to things-in-themselves and generalize it to all relations between objects. In other words, the inaccessibility of objects to mind is not, for Harman, unique to the human-world gap, but is a gap that exists between all objects. Within the framework of my ontology, onticology, I make a similar point using the language of translation or interpretation. As Latour puts it in Irreductions,

What those who use hermeneutics, exegesis, or semiotics say of texts can be said of all [relations among forces]. For a long time it has been agreed that the relationship between one text and another is always a matter for interpretation. Why not accept that this is also true between so-called texts and so-called objects, and even between so-called objects themselves? (modified, The Pasteurization of France, 166)

315px-Blackbox.svgIn my last post I characterized the core idea of anti-realist epistemologies in terms of the concept of black boxes. Where a naive realist epistemology holds that the mind contains a copy of objects as they are in-themselves out there in the world, the anti-realist epistemologist points out that the world presents the mind with inputs that pass through a black box with a particular programming, structure, or software that produces an output that differs from the input. In Kant’s epistemology, the output is called “phenomena” while the input is the “thing-in-itself”. The black box in Kant contains the a priori categories of the understanding and the two forms of intuition, time and space. The key point not to be missed is that knowledge, for Kant, falls not on the side of inputs, but on the side of outputs. What we know is not what the inputs are like, but rather the outputs or phenomena that result from going through the factory of our black box. Arguments among anti-realist epistemologists thus consist in debates about what the black box contains. You get advocates of Kant’s model, you get Derrideans that argue that the black box contains traces and a play of differance, you get Foucaultians arguing that the black box is structured through power, and so on.

Apart from the thesis that the world is composed of objects– a thesis common to Harman, myself, Whitehead, and Latour –this anti-realist thesis about black boxes is at the heart of all genuinely object-oriented ontologies. Where object-oriented ontologies differ from anti-realist epistemologies is that where anti-realist epistemologies sees this input/output structure as unique to the human-world gap, object-oriented ontologies hold that this input/output relation is true of any and all relations between objects. The relation between a leaf and photons of sunlight is not structurally different than the relationship between humans and objects. Just as humans translate the world around them through their various black boxes, the leaf translates photons of sunlight, turning them into complex sugars.

Now one of the criticisms that commonly emerges in response to object-oriented ontology is the critical question of why claims such as these are not dogmatic. What is it that authorizes these claims? Here my response to this challenge is not to adopt the hat of realist epistemology and make the case that we can represent the black boxes of other objects besides humans, but to show that arguments based on human black boxes are themselves speculative. First, the very fact that we have a debate as to what human black boxes contain (categories and forms of intuition, difference, power, etc), shows that we have no direct access to our own black boxes, but rather only arrive at claims about the black boxes presiding over the production of our outputs through indirect inferences. The sadly departed Levi-Strauss will claim that our black boxes contain structures of mind, Lacan will claim they contain the symbolic, Derrida the trace and differance, Foucault structures of power and discourse, Kant a priori categories and forms of intuition, and so on. The key point not to be missed is that our own black boxes are every bit as “withdrawn” as objects themselves. Second, by way of analogy we can make the point that speculation about what our black boxes contain are, as speculations, deeply prone to error. Take the example of computer black boxes. If I examine the output of a computer alone I might be led to make all sorts of erroneous influences. For example, when I notice that a blog contains italic and bold faced fonts I might be led to think there is a category in the programming that produces this output. However, the actual computer code that produces italics shares very little resemblance to a category or the font. The point here is that we can’t hit on accurate inferences about what black boxes contain, but that these black boxes are themselves objects of speculation and indirect inference that are not immanently or immediately accessible.

So what is my argument here? My argument is that all things being equal, if we are speculating about our black boxes, if our claims about our black boxes are not “critical” claims but speculative claims, then there is no reason not to open the door to a generalized speculation that allows us to freely hypothesize about objects independent of humans and how their black boxes function. Notice the strategy of argument here. My move is not to argue, contra the last 200+ years of sophisticated anti-realist epistemology that somehow we have a mysterious immediate access to objects, but rather to show how the anti-realist position contains a speculative core at the heart of its thought. As a result of this super-ninja, surprise judo move that uses the force of my critics own arguments against his onslaught, I thus arrive not at a transcendental idealism but at a transcendental realism. In other words, the question becomes “under what conditions can such-and-such a type of difference be produced?” This conditions are not mind dependent, but instead are attributed to the objects themselves.

Clearly the question arises of what knowledge is within the framework of an object-oriented ontology. Insofar as object-oriented ontology holds that all objects “withdraw” from one another (Harman) or that all objects “interpret” one another (Levi, Latour), or that all objects are black boxes with respect to one another, it follows that knowledge cannot be a representation of objects. Why? Because you cannot represent an object whose inputs disappear behind their outputs. My hypothesis at this point is that knowledge consists in know-how with respect to producing differences. To know is not to represent an object but rather to have know-how as to evoking differences within various objects under particular conditions. To know an object is to know the differences it is capable of producing under specific conditions.