blog.kennypearce.net

Reductivism, Eliminativism, and Berkeley's Theory of Physical Objects

In present-day metaphysical discussions it is common to distinguish between 'reductivism' and 'eliminativism' with respect to some class of objects, C. These can be thought of as two different ways of denying the (fundamental, metaphysical) existence/reality of the objects in C. Examples of classes discussed by philosophers in this way include minds, conscious experiences, and macrophysical objects. The two views may be given a linguistic formulation as follows:

Linguistic Reductivism (LR): Sentences which appear to assume the existence of the putative objects in C are strictly and literally true, although, in metaphysical rigor, the putative objects do not exist. (The appearance is misleading.)

Linguistic Eliminativism (LE): Sentences which appear to assume the existence of the putative objects in C are strictly and literally false (although they may be true in some loose and popular sense).

Someone committed to LR with respect to macrophysical objects would say that when I say there is a table in my living room I only appear to assume that there are such things as tables - in reality, the truth of this sentence does not require that there be tables. Someone committed to LE would say, instead, that when I say there is a table in my living room that sentence is false, since there is no such thing as a table. Still, an adherent of LE might admit, there are certain facts about arrangements of atoms and my perceptions and so forth such that talk about tables can be pragmatically useful, and so might be 'true' in some loose and popular sense.

Reductivism and eliminativism can also be formulated in more explicitly metaphysical terms:

Metaphysical Reductivism (MR): The facts about the objects in C are nothing over and above the facts about objects in some other class, D.

Metaphysical Eliminativism (ME): There are no objects in C, and so there are no facts about the (putative) objects in C.

In the metaphysical formulation, the eliminativist about macrophysical objects would simply and straightforwardly state there are no such things as tables. The reductivist would instead say that the facts about tables are exhausted by facts about particles. In other words, to move back in a linguistic direction, if you were to describe the whole world just in terms of particles, with no mention of tables, you wouldn't be missing anything. This might be because tables simply are collections of particles (this corresponds to the 'identitarian' position in philosophy of mind), or the relationship might be more complicated.

It seems to me that in contemporary discussions, the metaphysical and linguistic versions are usually treated simply as different formulations of the same position, and it should be easy to see why.

Now, where Berkeley's view of physical objects stands among these theses is a disputed matter. Sometimes, for instance, he is cast as a reductive phenomenalist, holding that the facts about physical objects are really about experiences. Others cast him as a reductive idealist, holding that physical objects are really collections of ideas. (Strong support for this comes from PHK 1 and a few passages in the Dialogues that I won't go hunting for right now.) It can also be argued that Berkeley is really at heart an eliminativist, on the grounds that he only talks about physical objects when he is "speaking with the vulgar." In other words, one might take his view to be that statements about physical objects are only 'true' in some loose and popular sense. ("Strictly speaking, Hylas, we do not see the same object we feel" - DHP 245.)

A further complication that I think has not been adequately appreciated is that Berkeley's theory of language causes the linguistic and metaphysical formulations to come apart. This is because, according to Berkeley, sentences don't have to be about anything in order to be true. On his view, very few words in human language correspond directly to any entity or idea. Some of them can stand 'indifferently' for any of a number of ideas, and some don't stand for anything at all, but are nevertheless meaningful. (In "The Semantics of Sense Perception" I dubbed these "referentless abstract terms" - 'abstract' because they don't correspond to ideas, and 'referentless' because they don't refer to extra-mental objects, like other minds, either.) It is possible, then, that physical object terms don't refer to anything, mental or extra-mental, at all, yet were still meaningful and could still be used in true affirmations. This view would combine LR and ME.

In point of fact, I think it is moderately clear that Berkeley holds LR, but it's hard to tell which of the metaphysical formulations would be more properly attributed to him. That is, it is moderately clear that Berkeley thinks that plain language statements about physical objects are strictly and literally true (once one correctly understand the 'strict and literal' meaning of such plain language statements), but it is unclear whether he thinks that physical object terms refer to anything.

It should be noted that many nominalists are committed to something like Berkeley's claim that sentences can be true without being about anything: many nominalists say that this is precisely the case with sentences that appear to be about abstract objects like numbers or universals. This will cause the linguistic and metaphysical formulations to come apart again.

Posted by Kenny at August 27, 2009 11:19 AM