1 Introduction
Richard Rorty is widely regarded as one of the early developers and proponents of a version of materialism known today as eliminative materialism. This characterization is based primarily on papers wrote in the midâ1960s, especially his âMindâBody Identity, Privacy and Categoriesâ (Rorty 1965, from hereon MBIPC). While in his later work, Rorty arguably abandons the view suggested in these papers by rejecting the metaphysical and linguistic assumptions they demand, nearly all commentators treat the earlier Rorty as having helped define and establish the radical thesis that the posits of commonsense psychology, especially sensations like pain, do not actually exist.
I am going to argue here that on a closer reading of MBIPC, this interpretation of Rorty as a proponent of eliminative materialism is actually a mistake. Others have noted inconsistencies and difficulties in Rortyâs position. But here I will argue that the most plausible interpretation of Rortyâs account of mental states is one in which he rejects several central features of contemporary eliminativism. As other writers have suggested, such as Bush (1974), I believe Rorty should instead be viewed as a more conventional sort of materialist who makes a misguided claim about what mindâbrain identity entails. While Rorty argues for the eliminability of terms used to pick out mental states, he does not base this claim about language on a clear commitment to the nonexistence of the states or properties that those terms denote. In the end, while his writing lends itself to different interpretations, I believe Rorty should be seen as simply a goodâoldâfashioned reductionist/identity theorist.
To show this, this chapter will have the following organization. First, I am going to spell out eliminative materialism as it is commonly understood today, highlighting key elements that are relevant to the question of whether or not Rorty was himself an eliminativist. Then, I will spend some time going over the claims and arguments that Rorty makes in MBIPC in an effort to capture his picture of the nature of mental states. I will then elaborate on the mismatch between modern eliminativism and Rortyâs account, and argue that Rorty is best viewed not as an eliminative materialist, but rather as a proponent of the very different view that mental states should be regarded as identical (in some sense) to brain states. I will also contrast my interpretation with the interpretations offered by others, in particular, Lycan and Pappas (1972) and Bush (1974), and explain why I think my interpretation is better.
2 Contemporary Eliminative Materialism
Generally, eliminativism about something is the thesis that the targeted entity, process, event, property, or whatever else assumed (at some time) to exist actually does not exist. Thus, we are all eliminativists about some things â ghosts, demons, the crystal spheres that were thought to hold the stars in their fixed position. Eliminativism is thus a form of antirealism regarding something once assumed by some to be real. The term is misleading as the allegedly nonexistent item is not actually eliminated (it canât be eliminated, since it doesnât exist). What is proposed is the elimination of the reliance on, or perhaps the positing of or explanatory appeal to, the corresponding notion or concept. So while no one has eliminated demons, as such, we (or most of us) have eliminated the practice of invoking our concept of demons to explain various phenomena. We have eliminated demons from our understanding of the furniture of the world.
Modern eliminative materialism is a form of eliminativism with regard to certain mental states or processes. Thus, modern eliminative materialists typically claim that specific sorts of mental states or mental processes do not actually exist. In recent years, the most popular target of eliminativist writings have been intentional states and, in particular, propositional attitudes, such as beliefs and desires. Thus, eliminative materialists like Paul M. Churchland (1981, 1993) and Patricia Churchland (1986) have claimed that there are no such things as beliefs, in the same sense in which there are no such things as demons. But there have also been those who have endorsed an eliminativistic outlook toward phenomenal states like pain (Dennett 1978; Hardcastle 1999), consciousness (Rey 1983), concepts (Machery 2009), and a variety of other mental states, capacities, and processes.
Most contemporary arguments for various sorts of eliminative materialism typically involve two major steps. The first step involves establishing a set of central properties that the relevant mental state or kind is assumed to possess. Often this is done by claiming that our understanding of that mental state is based on some broader theoretical framework, like our commonsense or âfolkâ psychology (Churchland 1981). In other words, our conception of the mental state in question is claimed to be rooted in a theory that is used to explain and predict behavior; as an explanatory posit, the state in question is assigned a set of defining characteristics, such as various causal properties. The second step involves arguing that such a thing cannot be identified with â at any level of analysis â anything in our most advanced scientific account of the brain, either now or down the road. Thus, eliminative materialists share with dualists a sort of antireductionist outlook. They agree that a given mental state cannot be identified with anything going on in the brain. Dualists say this is so because mental states exist as nonphysical states, whereas eliminative materialists say this is so because the mental states do not actually exist.
Some further clarification is needed to understand each of these steps. Regarding the first step, a great deal of discussion has been devoted to the question of whether or not our invoking of commonsense mental notions like âbeliefâ or âpainâ is the expression or application of some broader theoretical framework. Proponents of the soâcalled âtheoryâtheoryâ claim that commonsense psychology is a predictive and explanatory theory, albeit a nonscientific one, that we learn while growing up. Many philosophers (and some psychologists) reject the theoryâtheory, and claim instead that we make sense of one another through some other process, like simulating their decision making using our own decisionâmaking machinery (Goldman 2008). However, strictly speaking, eliminative materialism does not really require that our commonsense mental notions are embedded in a theoretical framework used for explaining and predicting. Just about any sort of embedding conceptual framework can be used to support an eliminativistic outlook. In fact, although it is seldom appreciated, the only thing eliminative materialism actually requires is the boring assumption that we have mental concepts (that is, concepts of mental states and processes) and that those concepts assign certain properties to those mental states and processes. That is an extremely weak thesis. Even the most ardent opponent of the theoryâtheory will typically grant that we possess concepts about mental states like beliefs and pains, and that those concepts (at least tacitly) assign to their corresponding mental entities a variety of intrinsic, intentional, phenomenal, causal, temporal, and other properties. To deny this would be to deny that we possess distinct concepts of mental states.
Regarding the second step, the immediate and crucial question that comes up is this: Just how mistaken does a given conception of something need to be for us to properly say that it turns out there is no such thing? Because so many eliminative materialists regard our mental notions as part of a folk theoretical framework, this question is often framed as this: Just how wrong does folk psychology need to be for its theoretical posits to fail to correspond with anything that exists? Obviously, wrongness of the theory alone is insufficient for claims of nonexistence. We have been initially wrong about many things invoked by both commonsense frameworks (stars, light, fire) and scientific theories (planets, genes, atoms), and yet we nevertheless regard these things as real, but just different than we initially thought. In response to Rortyâs MBIPC, Steven Savitt (1974) described the relevant difference as one between ontologically conservative (retentive) and ontologically radical (eliminativistic) theory change. But beyond making this helpful distinction, Savitt did not go far in explaining what motivates it. So why do we revise and retain (reduce) ontological posits of some frameworks and abandon the ontological posits of others?
As it turns out, this is not an easy question to answer. There are no clear, fixed, and acknowledged sets of criteria that tell us how much (or what sort of) mismatch is needed between a given concept and the world to motivate and justify a change that is ontologically radical. A survey of the history of science seems to reveal a hodgepodge of different factors influencing different cases. Still, as Savitt notes, one plausible thing we can say is that in many clear cases of retention, the mismatch was relatively minor, whereas in most cases of elimination, the mismatch was quite substantial. For instance, while we were wrong to think that planets orbit the Earth as Ptolemy claimed, we were right that there are spherical bodies in solar systems that orbit so...