Phenomenalism

The "phenomena" in the philosophical term "phenomenalism" are the phenomena of consciousness, and specifically, sense data (or, perhaps, acts of awareness of sense data). I won't try to be precise because I'm describing a group of views with incompatible fully precise versions. Phenomenalism is something like the view that all there are are phenomena, and in particular, the view that in some sense material objects are made or composed of or reducible to some sort of arrangements of sense data.

We are finally in the Chapter in which Ayer presents his theory of how we "constitute" material things given sense data, the purpose for which we've been working on clarifying what sense data are all semester. The view is something like a species of phenomenalism, and so it is worth defending it from standard criticisms of phenomenalism.

It is usually assumed that a phenomenalist takes material objects to be made up of sense data (or some related sorts of things) in much the same way the physicist takes material objects to be made up of subatomic particles or the idealist takes them to be made up of ideas. And that is how standard phenomenalist views go. But it is not Ayer's view.

To take such a view, one would have to believe that there are material objects and sense data in much the same way that there are both tables and chairs, or subatomic particles and chairs. But that is not Ayer's view. There is one language in which one can say that there are sense data, but there is a different language in which one can say that there are chairs. Ayer wants to show how to translate between the two languages, and in particular to show how to translate the reasons we have for believing that there are certain material objects (and that there are cases of illusion) into sense-datum language. But that is not showing that material objects are clusters of sense data in the usual phenomenalist sense. Nonetheless, claims about particular material objects do get translated into claims about clusters of sense data: I am writing with a pen will get translated into some statements about pennish sense data, but not into the statement that I am writing with pennish sense data—that would be the other, more obvious phenomenalist view—compare, I am writing with a collection of subatomic particles, which is true on the standard physics.

The obvious way to make this clear would be to provide examples of translations. No one, either holding traditional phenomenalist views or holding a view like that of Ayer, has ever done so. Ayer explains why it can't be done. That sounds bad for the view, but notice that no one has ever translated I am writing with a pen into a statement solely about subatomic particles either, and that doesn't bother us. What it is to be a pen has a social definition that it is hopeless to translate into something just involving subatomic particles. That doesn't mean that there is any doubt that every pen is composed of subatomic particles.

Similarly, it is hopeless to characterize what it is to be an experience of pennish sense data solely in phenomenal terms. (That is, terms involving shape, color, ….) Even if we could describe every way every pen we can conceive of could look in every circumstance we can conceive of, that won't capture what it is to be a pen: new types of pens, not previously thought of arise all the time, and the definition won't take account of them.

Material object terms don't have translations in a direct sense into sense-datum language. That's bad enough, but the situation is even more complicated. Once again, subatomic particles exhibit the same difficulties: To describe a particular case of a use of a pen in the language of subatomic particles seems to require, at a minimum, on the order of statements: there are about that many atoms in a gram of matter. It's too complicated. Similarly, describing the full phenomenal detail of even one experience of pennish sense data, never mind that we don't even have the vocabulary, is too complicated.

There is no way to get either necessary or sufficient translations of material object language into sense datum language, or the other way around. Thus, the obvious way of getting clear on the relationship between material object language and sense datum language, providing translations, doesn't work. Nonetheless, every perception of a material object (even highly indirect cases) is a case of an awareness of sense data. Thus, every case of having empirical reason to believe that there is some array of material objects with some properties is a case of awareness of some sense data. Even if there is some question about what material objects there are or what properties they have on the basis of some empirical evidence, there is no question what the corresponding sense data are. That is what we have worked so hard to set up.

Thus, our job will be to find general principles of how sense data statements are related to beliefs and assertions about material objects and their properties, though we will not be able to descend to more specific cases. It isn't that the specific cases can't be described, but that the rules rapidly become too complex and lack general application.

The main type of criticism of phenomenalism that one sees are of the form, How could material objects, which are so unified and substantial be made of such fleeting things as sense data? For Ayer's theory this question just exhibits a misconception, for Ayer never says that material objects are made of sense data. What needs to be answered is what sorts of systems of experience of sense data give us the awareness of the unified and substantial character of material objects?

-- ShaughanLavine - 24 March 2003 - 20 Mar 2007