Here is where I find that some very rudimentary questions can be raised, in spite of Levi’s recent proclamations about how OOO has been circumspect in providing support for their claims. The first obvious observation concerns the status of real objects. Since every time I think about, type in, or generally relate to my computer, either in practice or theory, the real object in relation to me withdraws, how do I know that it is, in fact, one real PC that is withdrawing and not a multitude of PC-Parts, or of qualitatively different real objects altogether? More specifically, since every time I think/act towards my PC this will be towards a sensual distortion of the object, how can I ever know anything about the structure of real objects as such?
There’s much more there (Daniel is prolific), so read the rest of the post here. In my view, it is crucial to distinguish between what I call “epistemological realism” and “metaphysical realism”. Epistemological realism is a thesis about knowledge. It is the claim that the way we represent objects is the way objects are. Here representations and objects mirror one another. Metaphysical realism, by contrast, is the thesis that objects are mind-independent and exist in their own right.
All epistemological realisms are also, necessarily, metaphysical realisms, but not all metaphysical realisms are epistemological realisms. This is the case with object-oriented ontology and follows, as Daniel notes, from the very logic of its metaphysical claims. Object-oriented ontology is a metaphysical realism without being an epistemological realism. If this is so, then it is because objects withdraw from any and all relation. Insofar as knowledge is a form of relation, it follows that objects withdraw from knowledge alone. This is to say that knowledge cannot be a mirror or identical representation of objects.
Here, then, we seem to encounter a contradiction. If knowledge cannot be a mirror of the world by virtue of the fact that objects withdraw, what warrants the claim that the world is composed of objects at all? Isn’t OOO overstepping the limits of its own knowledge, claiming to know that which it itself claims cannot be known? And if this is so, doesn’t this entail that OOO isn’t even entitled to claim that the world is composed of objects, nor that we can even know whether objects are withdrawn?
The best arguments I’ve come across for the thesis that the world is composed of objects are found in Roy Bhaskar’s early work, A Realist Theory of Science. Bhaskar’s strategy is transcendental (in his early work he refers to his position as “transcendental realism”). A transcendental argument is an argument that argues that certain conditions x must be the case in order for something else to be possible or intelligible. Saussure’s Cours is an example of a transcendental argument (though he does not use this terminology himself). Saussure asks “what are the conditions under which communication is possible?” Here communication is the practice to be rendered intelligible. What is sought is what must be the case in order for communication to be possible or take place.
Saussure’s answer is that in order for communication to be possible there must be a shared system of language within which those who communicate are situated. Now note, language, for Saussure, is a very strange thing. Language is not what is spoken. No one has ever heard language. Language cannot be directly observed. Nor, as strange as it may sound, can language be spoken. Rather, language is that shared code, that shared system, that allows communication to take place. In this regard, language, Saussure argues, is the condition under which speech is possible. What the linguistic seeks to investigate, in Saussure’s model, is this system or code, and this system is inferred through the investigation of speech and other texts.
Bhaskar is engaged in a similar project. Bhaskar’s transcendental question is “what are the conditions under which scientific practice is intelligible?” This question needs to be refined in two ways. First, Bhaskar asks “what must the world be like for scientific practice to be intelligible?” Bhaskar’s thesis is thus that the conditions under which scientific practice are intelligible are ontological. In order for scientific practice to be intelligible the world, he argues, must be a particular way. Second, in asking after the conditions under which scientific practice is intelligible, Bhaskar is, in fact, asking for the conditions under which experiment is intelligible. For Bhaskar, the key question is not about the reliability of truth-claims in science, but about scientific production. He is not asking how scientific propositions map or hook on to the world, but about what transcendental conditions lie behind our experimental activity.
I will not go into all the details of Bhaskar’s arguments because his arguments are complicated and lengthy. Bhaskar argues that there are five conditions for the intelligibility of our experimental practice; that is to say, there are five ways the world must be for experimental practice to be intelligible: 1) it must be possible for generative mechanisms [objects] to be out of phase with their qualities, 2) the world must be differentiated, 3) the world must be structured, 4) the world must be hierarchically stratified, and 5) the generative mechanisms that populate the world must be intransitive. I’ll discuss each of these conditions briefly in their turn.
Being Out of Phase
What are we doing when we conduct a scientific experiment and why are we doing it? When we conduct an experiment we place entities in a controlled or isolated setting. But why do we do this? We wish to observe, of course, what they do or how they behave. Yet why is it necessary to place them in a controlled or isolated setting? Bhaskar argues that scientific experimentation is necessary because it is possible for entities to be out of phase with their properties or the events of which they are capable. In other words, outside of the experimental setting entities can fail to display the nature of the acts of which they are capable. They can be dormant, but also their powers can be disguised or suppressed by the intervention of other entities. Experiment thus places entities in closed settings so that it might discover the acts of which a particular sort of entity is capable. Were entities not capable of being out of phase with their properties and the events of which they were capable there would be no need for this detour through experiment.
This first condition is the foundation of my distinction between local manifestation and virtual proper being. Within my framework, all local manifestations are acts or activities of objects or generative mechanisms. Here what we ordinarily refer to as properties or qualities are doings on the part of objects. The red of a billiard ball, for example, isn’t a quality that that billiard ball has or is. Rather, the red is something the billiard ball does. It is an activity on the part of the billiard ball. To manifest is to do something. This doing, in its turn, is dependent on all sorts of interactions. For example, lighting conditions. The key point about local manifestation is that it can fail to occur. We can, for example, turn out the lights. At this point, the activity of “redding” ceases to take place. The ball ceases to locally manifest itself as red. However, the powers of the object, its capabilities, its virtual proper being remains. We infer powers through local manifestations, yet these powers are nothing like the local manifestations. The powers of an object always far exceed any of its local manifestations. Here, already, we encounter the withdrawal of objects in a variety of forms.
The World is Differentiatied
As we saw in the previous section, experiment involves situating entities in controlled and isolated settings. If this is to be possible, it follows that the world must come in chunks. The entities of the world must be differentiated or independent. Were this not the case, then it would not be possible to isolate entities so as to conduct controlled experiments on them. Thus, while we always encounter entities in relations of one sort or another, we must conclude that these relations are of an external, rather than internal nature. In principle it must be the case that entities can be severed from whatever relations they happen to currently entertain. For example, the condition under which fifty yards of fabric (one of Marx’s favorite examples) can indifferently exist and interact in either a barter system or a capitalistic system is that its being is independent of either of these relational networks. Here then we have grounds for an object-oriented ontology.
Entities are Structured
In order to act and locally manifest themselves in the world entities must have an internal structure that maintains itself throughout the duration of its existence. We should take care, however, not to conclude that these structures are fixed or static. Each structure or substantial form is capable of endless variations and mutations. In this respect, we should think of structure not geometrically but topologically. The clip below gives a nice sense of the nature of structure:
Here we encounter instances where, in each instance, one and the same structure undergoes a series of variations. In each instance (there are three different topologies in the video), the structure topological remains the same while the local manifestations perpetually shift and change. These structures, of course, can undergo variations that reach bifurcation points where new structures or substantial forms come into being. In other words, they can reach points where, as a function of reigning conditions, new objects come into being.
Entities are Hierarchically Structured
When Bhaskar refers to the hierarchical structuration of the world he is referring to emergence. Scientific inquiry proceeds by showing how higher level laws are based on lower level laws. For example, one phenomenon might be explained by electricity, yet electricity, in its turn, is explained through quantum mechanics. Alternatively, the motion of the planets or the movement of our red billiard ball might be explained through gravity, yet we then need to explain gravity through the elusive Higgs-Boson particle (if it’s found to exist). Entities are nested within entities. The important point here, however, is that the hierarchical stratification of the world is not eliminative. Higher scale entities have an autonomy, independence, or existence of their own characterized by their own generative powers. They cannot exist without the lower scale entities, nor do these higher order powers contradict the powers of the lower scale entities, but they are autonomous entities in their own right, irreducible to these lower scale entities.
Entities are Intransitive
The claim that entities are “intransitive” is Bhaskar’s fancy way of saying that entities are independent of mind, society, language, or experience. They exist in their own right. The term is to be contrasted with what Bhaskar calls the “transitive”. The transitive dimension of knowledge, by contrast, refers to the social dimension of knowledge: the practices by which knowledge is produced, the social setting (investigated by figures such as Pickering or Latour) in which knowledge is produced, inherited knowledge in the form of theories, texts, prior findings, institutions in which knowledge is produced, etc., etc., etc.
The claim that the generative mechanisms or entities of knowledge are intransitive is the claim that they are independent of the minds, society, language, or experience through which they are known. A transitive verb is a verb that requires both a subject and an object as, for example, in the sentence “Tom threw the ball.” Were the objects of knowledge purely transitive they would have no existence independent of the investigators and society through which this knowledge is produced. Bhaskar rejects this hypothesis again on the grounds of experimental practice. Were the objects of knowledge purely transitive there would be no need for experimentation as entities would not be out of phase with the properties or acts of which they are capable. This possibility of being out of phase requires that entities be independent of the knowledge that discovers them.
The point here is not that the transitive dimension does not exist, but that the entities aimed at by this transitive dimension are independent of that transitive dimension. Take the example of my acquaintance with Graham. I get to know Graham transitively in a variety of ways. I gradually come to know him through my experience of him, talking to him, listening to him, reading what he writes, observing how he behaves, and so on. Moreover, my knowledge of Graham is informed by a broader historical, social, and philosophical background that influences how I interpret him. However, Graham himself is intransitive to how I come to know him and to this background horizon. Graham exists as an entity in his own right independent of any of this that I bring to the fore in getting to know him.
A first point to draw from the foregoing is that these claims are entirely agnostic with respect to the question of what entities exist. The foregoing makes no claims as to whether it is atoms, subatomic particles, organisms, stars, baseballs, etc., that exist. These are questions for actual inquiry and cannot be answered a priori. All that is said here is that if scientific practice takes place then the world must be such that entities can be out of phase with their qualities or the acts of which they are capable, that it must be differentiated, that entities must be structured, that it must be stratified, and that the objects of knowledge must be intransitive to that knowledge. Here, incidentally, it’s worth pointing out that goo theorists and relationists are obligated to respond to this argument. They can, of course, argue that scientific practice doesn’t take place, that it is a pseudo-practice, and therefore dismiss the entire transcendental argument altogether. Alternatively, if they concede that scientific practice does take place, they are obligated to explain how it is possible in that instance where the world is composed of pre-individual goo.
Second, fallibility is built into this transcendental account of knowledge. Daniel asks “how do we know that the computer is an object and not just an aggregate of computer parts?” Insofar as OOO is a metaphysical realism, not an epistemological realism, the only answer is that we’re not entirely certain. It could be that the computer is merely a collection of objects and not an object in its own right. I don’t claim certainty. Many philosophers, influenced– probably unconsciously –by a Cartesian tradition seem to treat certainty as a model of knowledge. Absent certainty, then the conclusion seems to be that we are entitled to dismiss any ontological claims. This strikes me as a very odd premise. Certainties are few and far between, but that doesn’t entail that there aren’t degrees of probability and likelihoods.
Third, however, we are obligated to provide arguments as to why we believe it is likely that this or that is an object. When I make the claim that, for example, the Coca-Cola Corporation is an object, I am making both a very strange claim and one that ought to be defended with reasons. Here my argument is that the Coca-Cola Corporation both possesses powers and is interacted with by other entities in ways that are irreducible to any of its parts. The Coca-Cola Corporation is capable of doing things that its parts are not capable of doing. Likewise in the case of the computer. I believe the computer is itself an object, that it can’t be reduced to a mere aggregate, because it has powers and is capable of doing things that its parts are not capable of doing.
One might here evoke the “causal redundancy” argument, claiming that when such claims are made we fall into a causal redundancy that parity does not require. That is, when we claim that the computer has causal powers we seem to be claiming both that the parts of the computer cause such and such and that the computer causes such and such. Yet aren’t we multiplying entities needlessly when we do this? Can’t we say, instead that it is merely the parts that do whatever it is that we say the computer is doing? And if this is the case, can’t we merely abandon the idea that the computer exists altogether? The problem is that the structure or organization of the computer is not nothing. It is by virtue of this structure or organization– what I call “endo-structure” or “endo-consistency” –that the computer has the powers that the computer has. The computer cannot exist without its parts– though it is not unimportant that those parts can be replaced –but neither can it be reduced to the parts. The computer has the powers it has by virtue of that structure, organization, or substantial form. The parts, of course, are a necessary condition for the computer being a computer, but they are not a sufficient condition nor can the computer be eliminated in favor of its parts. At any rate, back to editing.