The minimal condition for whether or not a philosophy counts as “realist” can be found in what that philosopher thinks can be said of a world in which all humans and rational animals have ceased to exist. Here the dividing line is not between whether or not the philosopher holds that a world independent of humans exists, but rather whether or not certain entities known by humans would exist as they exist even if humans did not exist (I’m still not entirely satisfied with this formulation). Thus, for example, there are many philosophers– if not most –that affirm the existence of an independent world, while claiming that we have no direct access to that world. Kant affirmed the existence of such a world, but claimed that we cannot have any direct access to this world. For Kant we can have a knowledge of objects, but not a knowledge of things. What we have, rather, is a relationship to objects as they appear for us, not as they are in-themselves. When we talk of atoms, then, we are speaking of entities mediated by various forms of intuition, categories of understanding, and activities of reason, and can have no knowledge of whether atoms are really like this independent of us. The realist, by contrast, holds that if we have a genuine knowledge of atoms, what we know of atoms belongs to atoms regardless of whether or not any humans exist to know atoms. In other words, the realist holds that atoms are what they are independent of humans, and that it is not knowledge that makes atoms what they are. By contrast, the anti-realist might accept that there is an independent world, but claims that it is something about our cognition of an object, our knowing of an object, or language about the object, etc., that makes that object what it is. The anti-realist that affirms the existence of an independent world thus distinguishes between objects and things. Things are what is independent of all human existence and beyond any possible knowledge. Objects are things as mediated by various human categories whether in the form of structures of mind, language, history, society, power, and so on.

There can, of course, be hybrid positions. For the realist it seems that there must be a difference between chairs and trees. Once again, presuming that trees are real things, a tree would be what it is regardless of whether there were any humans about to know it. When we speak of chairs, by contrast, it seems that humans must exist for chairs to be chairs. Such too would be the case with money, books, etc., etc. Something of these things would exist were humans to cease to exist, but they would no longer be chairs, money, or books. Here, however, there are good reasons for the radical realist to be suspicious of such claims about hybrids. The argument for attributing a lesser ontological reality to things like chairs, money, language, social instutitions, etc., seems to be based on the insight that these things are dependent on something else to exist. However, does dependency merit the inference that something should be granted a lesser ontological status than something else? If we reflect a bit, the answer seems to be no. Cells are dependent on proteins and the laws governing proteins in order to exist, but cells are no less real for this reason. There can be many circumstances in which proteins exist without cells existing– pre-biotic soups, for example –but this doesn’t diminish the reality of cells when they do exist.

read on!

One of the things that makes these issues so difficult to discuss is the manner in which ontology and epistemology have become so thoroughly imbricated with one another since the seventeenth century. For us today it is a reflexive gesture to assume that questions of the being of things are questions of our knowledge of the being of a thing or of our access to a thing. As Kant put it, “the conditions for the possibility of our experience, are the conditions for the possibility of the objects of experience.” In this way questions of ontology are collapsed into questions of epistemology. The anti-realist acknowledges being independent of humans but claims that we can only ever talk– if we are to avoid dogmatism –of the being of things as they are for-us.

However, it is not clear that this conclusion follows. The problem seems to emerge with Hume. In his discussions of causality, Hume (or his subsequent readers), conflates questions of our knowledge of causality with questions of what causality is. Among many other things, Hume’s key question is that of how we can know relations of necessity among causes and effects. Beginning from the premise that all of our reasoning about matters of fact are reasonings about cause and affect, and that we only ever encounter relations of cause and effect through successions of sensation or one impression following another impression, Hume concludes that we are unable to ever demonstrate a necessary relation between one impression and another impression. I know that when I lit the fire (impression-1) the water boiled (impression-2). However, all I perceive is one impression following the first impression. I do not perceive the relationship of necessity between these two impressions and therefore cannot infer that water must boil for all times in places where the first sensation is given.

I think there are good reasons for questioning Hume’s understanding of what statements about cause in effect are, however, setting aside these criticisms, does it follow that we cannot talk ontologically about causation without falling into dogmatism if Hume is right? In other words, is there not a major difference between a discussion of what causation is whether or not we know any particular causal relation and a discussion of whether we know any particular causal relations? Have we inevitably fallen into dogmatic speculation in discussing what causation is regardless of whether any humans know it? The answer seems to be no. In fact, Hume has to be able to state what causality is regardless of whether or not we know it for his argument to get off the ground. We can concede Hume’s thesis that our knowledge of causality is based on custom wherein relations between sense-impressions arise through constant conjunctions of sense-events (and I don’t), while still saying a good deal about what ontologically causation is whether or not humans have ever come to know a single instance of cause and effect at the level of their knowledge of particulars in the world. For example, we minimally know that causation is a relation of necessity such that if there is a relation of causation between fire and water boiling then water must boil. This is a claim about the being of fire and water– a claim that, not so incidentally, can be mistaken –not a claim about our knowledge of the relationship between fire and water. From here it is but a short leap to speculation about the nature of objects regardless of our knowledge of objects.

In many respects ontological speculation is far more modest than so-called “critical” thought. The ontologist begins from the premise that his claims about the nature of being can be mistaken or are fallable and that he might have to revise his very general model about the structure of the world. The epistemologist, by contrast, demands certainty and, barring our ability to achieve that, skepticism is the only option. In the case of Hume and Kant, this criteria of certainty arises in the shadow of Newton’s great success. Yet there is an irony here, for in their treatment of Newton they approach Newton not as a scientist who actively worked through experimentation, but rather as a body of propositions in a book, the Principia, divorced from the great labor it took to distill these propositions. But absent this highly involved labor where the knower is anything but a passive observer of the world linking impressions, the question of epistemology or how we know becomes horribly distorted, leading to an obscenely underdetermined epistemology based on intuition and concepts, rather than active engagement on the part of material bodies (scientists and their tools) with material bodies.