In discussions of speculative realism, a lot of ink has been spilled over debates about correlationism or whether or not it is possible to think being apart from thought, but, within an object-oriented framework, it seems to me that a far more profound ontological issue is at stake of which the issue of correlationism is only a subset. Here I hasten to add that I owe this conclusion to Graham Harman. The real issue is not whether or not whether or not it is possible to move beyond correlationism or whether being can only ever be thought in relation to thought and whether thought can only ever be thought in relation to being, but rather whether or not relations are internal or external. The thesis that relations are internal is admirably summed up in Karen Barad’s statement that “relata do not precede relations” (Meeting the Universe Halfway, 334). The thesis that relations are external is the claim that entities can break with whatever relations they happen to entertain to other entities at a particular moment and enter into new relations with other entities. Based on these two conceptions of relations, we can thus see why the issue of correlationism is a subset of the question of whether relations are internal or external. Correlationism is the thesis that the relation between mind and world or culture and world is always an internal relation, such that thought is necessarily inseparable from world and world is necessarily inseparable from thought. It is a specific variant of a broader thesis that all entities are internally related or inseparable from one another. We can see that if relationism or internalism is true, it is impossible for it to avoid correlationism, for in a world where everything is internally and inseparably related it would necessarily follow that all beings must be internally related to minds or culture. This, however, is not the issue I wish to focus on.
Before proceeding to discuss these issues of relation in more detail, it is first important to address what this debate is not about, as I’ve sensed there has been some confusion regarding these matters.
First, the debate over whether relations are internal or external is not a debate about whether or not relations exist. All sides are agreed that there are relations. The question is whether these relations are characterized by inseparability or separability. Can something break with its relations or is its being necessarily determined by its relations such that it has no minimal ontological independence from these relations?
Second, the question is not whether or not relations are important or whether they make a difference. Whether or not my cat is related to oxygen or food or certain atmospheric pressures makes a big difference for my cat. All sides, I think, can agree that the sorts of relations that obtain between entities make a tremendous difference to those entities.
Thus, third, the question is not whether we should focus on the analysis of objects or relations. In other words, the debate isn’t whether we should focus on objects, decontextualized, independent of any milieu, or whether we should focus on milieu/relations. Rather, the discussion is whether or not relations are internal to beings such that they are inseparable from their relations, or whether they are external in the sense that entities can break with relations. As I have said with respect to my own work on many occasions in print and publicly, it is not so much objects or substances that interest me, but what happens when a substance, machine, or object (they’re all synonyms to me) either enters into new relations or breaks with its existing relations. In my view, internalism just does not do a good job of responding to this question because it presupposes that beings are already internally related and thereby forecloses the question of what takes place when relations are either severed or forged.
Finally, fourth, it’s important to be clear as to what sorts of relations we’re talking about here. Clearly there are some relations that are of a purely internal. A parent cannot be a parent without a child. There is no left without right. There is no North without South, etc. A triangle can only be a triangle through relations between three points. We can call these “diacritical” relations. Diacritical relations are relations that only exist internally. Indeed, I would argue that every substance has its “diacritical domain” defined by what I call its “endo-relations”. This is its internal relations– like the three points of a triangle –without which it would not be what it is. The debate here is not between internal relations of this sort, but about relations between entities, substances, objects, or machines (again synonyms); or what I call “exo-relations”. Is the relation that my cat entertains to all other entities in the universe like that of the three, inseparable points of a triangle such that it cannot break with any of these relations, or does my cat possess some minimal ontological autonomy that allows it to break from whatever relations it currently entertains, such that it is able to enter into new relations? That’s the question.
With this question in mind, let’s return to Barad’s exemplary articulation of relationism, that “relata do not precede relations”, and examine why we might be wary of such a thesis. What is Barad saying? On the one hand, she appears to be saying that entities, “relata”, are generated out of relations. We have the relation first, and then the entities second. On the other, she appears to be claiming that these entities, “relata”, can have no subsistence or being apart from that relation. They are what they are only in and through this relation such that they can have no being independent apart from this relation.
In my view, this thesis is a catastrophe for ontological thought, empirical investigation, and concrete practice. To see this, let’s start first with empirical investigation. What is it we’re doing when we empirically investigate the world about us? Are we simply looking and recording what we see? Certainly this sort of observation that we find in fields like botany and zoology is of great importance, but it only makes up a small subset of what is meant by “empirical observation”. Empirical observation is not a simple “looking”– a view all too prevalent in a history of philosophy dominated by theoria and the denigration of the servant in Plato’s Meno as its paradigm –but rather is far more defined by acting. In the sciences, we do not simply look at entities, but rather act on them by perturbing them in a variety of ways by seeing how they behave under these conditions of pressure, temperature, light, when encountering these other chemicals or substances, when grown in these particular soils, etc. In other words, our practice consists in varying the relations between entities to see what they do when placed in these new milieus or contexts.
Nor is this variation of relations restricted to experimental practice in the sciences. There’s an increasing body of research ranging from developmental psychology to pedagogical theory (you’ll see some of it in the forthcoming issue of O-Zone) that suggests that this is how children learn as well. Children learn not by simply looking and listening, but by acting on entities in their environment and by varying their relations to other objects, to see what happens when these relations are varied. Both the experimental setting in the sciences and the phenomenon of learning suggest a very different conception of what an object, substance, entity, or machine is. A machine or entity is not a “bundle of qualities or impressions”, but rather is a collection of powers capable of producing certain effects under a variety of different conditions or relations. An object is a factory or set of potentialities, not a set of qualities. It is through interacting with objects or machines or through observing their interactions with other objects or machines that we begin to build up a diagram or cartography of what an object is. That cartography consists in building up a profile of what an object can do.
Yet if it were the case that “relata do not precede relations”, none of this would make sense. In the experimental setting it would not make sense to act on substances to see what they do or vary relations between substances to see what happens, because entities would be completely exhausted by their relations at this particular moment. Rather than saying garlic does this when it sits in a basket for three weeks or garlic does this when baked with olive oil in tin foil for twenty minutes at 350 degrees, we would instead be forced to say that the garlic in the basket and the garlic in the oven are two entirely distinct entities. These variations in relations would teach us nothing about garlic and the powers of garlic or what I call its “virtual proper being”. We can see just how strange such a thesis is in the domain of ecological thought. If it were true that “relata do not precede relations”, it’s hard to see why we would be concerned with what happens to bird egg shells when exposed to DDT, or aboriginal critters in the Australian eco-system when cane toads are introduced, or what happens to streams and water supplies when certain chemicals from fracking are introduced because, if this thesis is followed through, these are entirely new entities. Indeed, to take Barad’s favorite example, it is very difficult to understand what quantum mechanics could possibly be doing when acting on particles to locally manifest them in terms of position or vector, if there weren’t some excess of the particle in relation to how it’s acted on in the experimental setting. It is only where entities that possess some minimal ontological autonomy pass through a variety of different relations that these investigations make sense.
It is here that we encounter the disturbing political consequences of relationism as well. If relationism/internalism is true, then it is incredibly difficult to understand what projects of emancipation could possibly be about because insofar as “relata do not precede relations” there would be no beings to emancipate because those beings calling for emancipation would possess no independence apart from the social field of relations in which they’re enmeshed. Likewise, if it’s true that “relata do not precede relations”, it’s difficult to see how deprivation from things such as food, services, resources, goods, etc., could be a political issue insofar as talk of deprivation presupposes some ontological independence from currently reigning relations.
We understand what motivates internalist/relationist claims: All too often people believe they can act on entities in their environment without affecting other entities in their environment. We see this in many technological and scientific practices that seem to think that doing this thing here will not affect that thing there. We see this in many economic discussions where people seem to suppose that people are entirely self-made and are the result of their own “grit”. The corrective then becomes attending to relations and avoiding what Hegel called “thinking abstractly“. In our frustration with those who think abstractly, we then envision a metaphysics in which all things are necessarily related as an antidote. By all means, we should avoid abstract thinking in the Hegelian sense. Yet we must avoid the converse abstraction of seeing all relations as internal, lest we fall into an equally debilitating position. The idea that relations are internal undermines our sense of the fragility of relations, that they can be all too easily broken, and that the destruction of these relations often has incalculable destructive consequences. Yet the idea that all relations are internal also undermines our hope in the possibility of producing alternative worlds through the forging of new relations and the severance of old ones. Internalism makes a stab in the right direction, but does not yet fully reflect its own internal presuppositions and commitments.