On Philosophy

October 9, 2006

A Basic Metaphysical Foundation

Filed under: General Philosophy — Peter @ 12:31 am

One of the jobs of metaphysics is to determine what kinds of things exist. This is not to say that metaphysics is meant to be a kind of physics or chemistry or biology, which classifies the things we find, but rather a way of structuring our thought about the world. Once it was thought that things such a properties and colors really existed as independent entities, that somehow a white object had an instance of the universal whiteness, a universal whiteness that extends beyond an individual white object. However, as our knowledge of how the physical world worked advanced this metaphysical approach had to change as it became evident that the physical features of objects were all that was needed to explain our interactions with them, and that moreover the physical world was causally closed.

To escape from lingering trappings of past metaphysical systems I will outline here a modern metaphysics. The first task is to address what exists, what is part of the world. Here I will define what exists as those things that can be observed with our outward directed senses, which I will call observables subsequently. As detailed here we can show that the universe of observables is causally closed. The only reasonable objection to this account of what exists, assuming that we define observable properly [1], is the possibility that consciousness exists but is not observable, as defined here. However, if consciousness isn’t one of the observables then it can’t have a causal effect on the observable world. This would be to assert that the mind is epiphenomenal, a claim we have good reason to think to be false [2]. So if consciousness isn’t epiphenomenal than it must somehow be observable using our outward directed senses, we just don’t know exactly how yet. That objection aside this definition of what exists seems to capture everything relevant. Even if unobservable objects did somehow “exist” I don’t see how it could matter to us, since they couldn’t be a cause of observable events, and thus would be both unknown to us and irrelevant to understanding the world that we live in.

Of course, if the world of observables is causally closed it is important to define what causation is before we proceed further. We say that some thing or things is the cause of some outcome if without those thing or things, holding everything else constant, the outcome would not have occurred [3]. Usually considering causal relations in this way is straightforward, but a dilemma may arise when it is impossible to remove one thing and hold everything else constant. For example if there is a man with two names, X and Y it is impossible to remove X without also removing Y. Of course, as the example reveals, the reason that it is impossible is because X and Y are really identical. In general this is a good test for identity: if it is impossible to remove one of two things without removing the other there must be some identity between those two things, either in part or in whole [4].

Now we can turn to the case of forms / properties existing as independent entities, or entities that extend beyond individual objects, which, as I mentioned earlier, is a possibility that is excluded by the metaphysical theses outlined above. Here I will specifically address color, but the same reasoning can be applied to all “forms”. If color really does exist then it must be an observable, and thus must have causal effects, by the definition of what exists as given above. But we can’t consider what would happen if we remove the color green unless we also remove certain physical properties of the object (specifically surface properties) [5]. Clearly then green or some part of green is identical with the object itself. Let us then consider each part of green individually. First there is the part of the color that is identical with some properties of the object, which might be called an instance of the color by an Aristotelian. If it wasn’t associated with the form or property that extends beyond the object itself we might call it a description. Since it is identical with the object it doesn’t exist beyond the object in any way, and hence can’t be one of the kinds of forms / properties that exist “outside” or independently of objects. If something is to exist that extends beyond the individual object it must be the other part, although it is hard to say what the other part could even be. However if we take away this “other part” there is no reason to expect events to progress any differently. All indications point to the progression of events as governed by physical laws, which don’t take into account these “other parts” since they can’t be identified with any physical properties of the object. Thus the “other part” can’t be considered the cause of anything observable, making it itself unobservable, and hence not something that exists. Thus color, as something that extends beyond individual objects, doesn’t exist.

Some may think that here I am denying the obvious. Don’t we see the same color in many different objects? I would not deny that many object have the same color, but I would deny that our use of the word same in that sentence implies that there is some one color that finds its way into different objects; our use of language is misleading us. The sense in which multiple objects may have the same color is that they have similar properties, specifically they are reflecting similar wavelengths of light [6]. However, just because it is a fact that objects are similar in a certain way doesn’t mean that the similarity as something extra exists. Although it is often convenient to talk about the similarity as something in itself, this is simply a way of talking, not a reflection of the way things really are [7].


1. As those things that can in principle be observed, not things that we can currently observe or are currently observing.

2. For example, from experience. Also see here.

3. More detail here.

4. The common test for identity is if the two things have all their properties in common. Unfortunately using this criterion as a judgment for identity can often be misleading as we are often not aware that the two objects share all of their properties in common. For example, if you weren’t previously aware that X and Y were identical then you might deny that they have all their properties in common, since you believe them to have different names.

5. If you allow that green could be removed without changing any of the physical properties than green has no causal force, since the physical properties are what matter for determining future events. And if it has no causal powers then it can’t be observable, and hence doesn’t exist.

6. Actually the story is more complicated than that, since how we perceive color is also determined by the color of the rest of the environment. Thus is we were to be completely precise the similar properties would actually be combination of reflected light and the states in the mind that it subsequently gives rise to. However since both are observables there is no difficulty, in principle, in saying that the similarity is governed by these properties, it simply takes longer to say.

7. In fact it is hard to even talk about similarity without referring to it as though it was an independent object. For example, if one attempts to describe why two objects are similar we are tempted to say something like “the objects have some of the same physical properties”. In one sense this is a true statement, but on the other it may give the mistaken impression that a physical property is something that exists above and beyond the individual objects.


Create a free website or blog at WordPress.com.

%d bloggers like this: