Eliminative materialism and the propositional attitudes (1981)
by Paul churchland
and
Cognitive Suicide (1987), by Lynne Rudder Baker
in: Heil, pp. 382-413
Eliminative materialism and the propositional attitudes (1981)
by Paul churchland
Introduction
A. Eliminative materialism is the theory that common-sense folk psychology is false and must be replaced by, not reduced to, neuroscience
B. As Churchland sees it, earlier objections to materialism focused on qualia – he claims “these barriers [are] dissolving”
C. More recently, however, objections to materialism have concentrated on the problem of intentionality, arguing that beliefs and desires cannot be reduced to or eliminated in favor of something material
D. To examine this position, Churchland will begin by arguing that folk psychology is in fact a theory
1. he meets a lot of resistance to this suggestion
2. it seems that people will concede that something is a folk theory only if (383)
a. it is safely in the past
b. it is clearly false
I. Why Folk Psychology is a Theory (383)
A. Churchland argues that if we regard folk psychology as theory, we can unify the most important problems in the philosophy of mind, including:
1. the explanation and prediction of behavior
2. the meanings of our terms for mental states
3. the problem of other minds
4. introspection
5. the intentionality of mental states
6. the mind-body problem
B. the explanation and prediction of behavior
1. all of us can explain and even predict the behavior of other people and even animals rather easily and successfully
2. these explanations and predictions attribute beliefs and desires to others
3. but also, explanations and predictions presuppose laws
4. Churchland, however, believes that rough and ready common-sense laws can be reconstructed from everyday folk psychology explanations -- see below
C. the problem of the meanings of our terms for mental states
1. if folk psychology is a theory, the semantics of our terms is understood in the same way as the semantics of any other theoretical terms (383)
2. that is, the meaning of a theoretical term derives from the network of laws in which it figures
D. the problem of other minds (384)
1. we cannot infer that others have minds from their behavior
2. and it's risky to generalize from our own case
3. rather, the belief that others have minds is an explanatory hypothesis that belongs to folk psychology
E. indeed, Churchland believes that one's introspective judgment about one's own case is just a special instance of the theory (q.v.)
F. the intentionality of mental states
1. for many philosophers, the intentionality of mental states marks a fundamental contrast with what is merely physical
2. From Churchland's perspective, however, the intentionality of mental states is not a mysterious feature of nature but rather a “structural feature” of the concepts of folk psychology (q.v.)
a. these structural features reveal how much folk psychology is like theories in the physical sciences (384)
b. for example, he compares the propositional attitudes – has a belief that p, a desire that p, etc. – with what he calls the "numerical attitudes:" has a mass n, a velocity n, etc. (384-85, q.v.)
1.) both are “predicate forming expressions:”
a.) in the numerical case, we substitute a number for n to form a predicate
b.) in the propositional case, we substitute a proposition for p to form a predicate (384-85)
2.) in both cases, the logical relations that hold among the "attitudes" are the same as those that hold among their contents
3.) also, in both cases we can form laws by quantifying "for all n" or "for all p" (385, q.v.)
c. the only difference between folk psychology and theories in the physical sciences is in the sort of abstract entities it utilizes: propositions instead of numbers (386)
G. regarding folk psychology as a theory sheds light on the mind-body problem (386)
1. the issue becomes one of how the ontology of one theory relates to that of another
2. the different philosophical theories then reflect different positions on the reduction of folk psychology to neuroscience
a. identity theory maintains there will be a smooth reduction that preserves the ontology of folk psychology
b. dualism says it will be irreducible, since it deals with a non-physical domain
c. functionalism says it will be irreducible, since psychology deals with an abstract set of relations among functional states that can be realized in different media
d. eliminativism says that folk psychology won't be reduced to neuroscience because it is wrong and will be replaced by neuroscience (386)
II. Why Folk Psychology Might(Really) be False
A. given that folk psychology is an empirical theory, there is at least a possibility that it is false, and that its ontology (beliefs, desires) is an illusion
B. to address this issue, Churchland says, there are three things that we must consider: (387)
1. FP's explanatory failures as well as its successes
2. FP's long-term development as a theory
3. How well FP fits in with the rest of science
C. FP's explanatory failures
1. there are many things that FP cannot explain or even fails to address, including:
a. mental illness
b. creative imagination
c. differences in intelligence
d. function of sleep
e. how we track moving objects
f. 3-D vision from 2-D retinal array
g. perceptual illusions
h. memory, and retrieval from memory
i. learning, especially in pre- or non-linguistic organisms, such as babies and animals
2. the failure to explain all these things does not prove FP false, but does raise doubts about it (387-88)
D. FP's long-term development as a theory (388)
1. it has a history of “retreat, infertility, decadence:” we used to ascribe intentionality to all sorts of things in nature
2. but even if we restrict its application just to humans, it has made no advances in thousands of years (388)
3. Churchland sees it as a “degenerating research program”
E. How well FP fits in with the rest of science
1. we may hold on to a stagnant theory in the sciences if it is the only theory that fits well with related sciences
2. however, Churchland finds that FP fairs very poorly here
a. he claims that it doesn't fit in with sciences like evolutionary theory, biology, and neuroscience
b. and that these sciences are part of a growing system of knowledge that also includes chemistry, physics (388-89)
III. Arguments Against Elimination (389)
A. these arguments consist largely of a defense of folk psychology
1. they argue that FP is not an empirical theory, that is, not a theory that is refutable by the facts
2. these arguments are raised by functionalists
B. first argument: that FP is a normative theory
1. that is, that it doesn't describe how people actually act, but characterizes an ideal: how they ought to act if they were to act rationally on the basis of beliefs and desires (389-90)
2. hence, FP could not be replaced by an description of what's going on at the neuronal level (390)
C. Second argument: that FP is an abstract theory
1. according to the functionalists, FP characterizes internal states such as beliefs and desires in terms of a network of relationships to sensory inputs, behavioral outputs, and other mental states
2. functionalists argue that this abstract network of relations could be realized in a variety of different kinds of physical systems
3. hence we cannot eliminate this functional characterization in favor of some physical one
D. As Churchland sees it,
1. both of these arguments give a stipulative character to FP (q.v.)
2. that is, they shift the onus of burden proof
a. from trying to show that FP is a good theory of the mind
b. to trying to show that certain physical systems are capable of supporting FP
3. FP is thus removed from empirical criticism (391)
E. To be fair, some functionalists, such as Fodor, do not defend FP in this way
1. Fodor wants to find the best possible functional characterization of the mind, which for him is an empirical question. It may not be our current FP (391)
2. Nevertheless,
a. Fodor still seeks some sort of abstract theory that explains our cognitive activity in terms of the manipulation of propositional attitudes
b. Fodor thinks that whatever corrections functionalist psychology makes to FP, it will not be replaced with any sort of physical theory
F. Hence, Churchland thinks that the way to defend eliminativism against functionalist objections is to attack functionalism
IV. The Conservative Nature of Functionalism
A. Churchland argues against functionalism in the philosophy of mind by drawing an analogy with alchemy (391-92)
1. alchemy explained the properties of matter in terms of four different sorts of spirits – e.g., the spirit of mercury explained the shininess of metals (392)
2. this theory was dead by the time of Lavoisier and Dalton's theories of atoms and elements
a. in other words, an elimination took place
b. the dualistic interpretation of the four elements as spirits would have been feeble and pointless
c. a reduction would have been impossible, because the old and the new gave different classifications of things
3. but suppose instead it had hung on by redefining these spirits as functional states (392)
a. so, e.g., being ensouled by mercury just is the disposition to reflect light, etc.
b. defenders of alchemy could claim that it deals with an abstract level of description that does not reduce to the atomic level (392-93)
4. for Churchland, this analogy shows that it is at least possible for the sort of moves the functionalist makes to result in an outrage against truth and reason (393)
a. one could imagine similar defenses of other outmoded theories, such as phlogiston, the four humours, vital spirits, etc.
b. the functionalist stratagem can be used as a smokescreen for error
B. in sum, given Churchland's earlier arguments for the falsity of FP, it is not an adequate defense to say that it is an abstract theory (394, q.v.)
1. the same defense could have been made for any "haywire" network of internal states
2. the defense begs the question in assuming that the intentional idiom of FP captures what's important about cognition
3. finally, eliminativism is consistent with the claim that a cognitive system must be explained in terms of the functional organization of its internal states. It's just opposed to FP (394)
C. then Churchland replies to the other argument against eliminativism, based on the normative character of FP
1. the fact that FP explanations depend on logical relations among beliefs and desires does not make FP a normative theory
a. in the same way, the mathematical relations among the quantities in the classical gas law does not imply that this law has a normative character
b. the logical relationships among propositions in FP are objective. The normative element comes in just because we value these patterns of reasoning.
2. FP ascribes only a minimal, not an ideal, rationality to human beings (395)
a. we have “no clear or finished conception of ideal rationality”
b. hence, we cannot say that the explanatory failures of FP are due to people failing to live up to its ideals (395)
3. There is no reason to think that FP is applicable to any sort of cognition other than that of adult, language-using human beings
a. to make any advance in cognitive sciences, we need to go beyond FP and give a deeper and more accurate account of cognition
b. eliminativism does not imply the rejection of all normative concerns but only that these things will have to be reconsidered at a deeper level, the level of neuroscience (395)
F. Churchland concludes that FP is nothing but a culturally entrenched theory of how humans and the higher animals work (396)
V. My comments:
A. if FP is "culturally entrenched," it's entrenched in all cultures
1. That it is universal in the species is a fact that requires some explanation at some level, either psychological or anthropological
2. FP does fit in with science, not so much as a theory but as a phenomenon for science to explain
3. as Dennett says, it’s of anthropological interest
B. FP may be regarded as the manifestation of a human cognitive capacity that can be explained by evolutionary theory
1. I would argue that FP is rooted in an evolutionary adaptation, an ability to ascribe beliefs and desires to others, that facilitates certain sorts of social interactions
2. it's tightly linked to our ability to use language
3. other species have this "theory of mind" only to a very limited degree
C. Churchland's second argument that FP is a degenerating theory could also be taken as showing that it is not a theory at all.
1. It doesn't do what theory is supposed to do, which is guide research.
2. in fact, it doesn't ever seem to have performed this function
D. as for Churchland's argument for FP being a theory that draws on analogies with physical theories, what these analogies may show is simply that such physical theories have been constructed with the same linguistic apparatus that makes FP possible
VI. Beyond Folk Psychology
A. in this section, Churchland sketches three scenarios of what the elimination of FP might look like. This is all highly speculative, of course.
B. First scenario (connectionist)
1. we arrive at a theory of all brains that is consistent with evolutionary biology that ascribes to us at any given moment a complex state that can be described as a solid within a 4- or 5-dimensional phase space
2. any sentence which a human speaks is like a one-dimensional projection of that multi-dimensional solid (396-97)
3. although this scenario may not have any effect on our ordinary lives, it is not inconceivable that at least some people may start talking about their inner states in the language of the new theory (397)
C. Second scenario – more radical (397)
1. start with the hypothesis that our ability to use language is an additional and acquired function of our apparatus for organizing perceptual experience
2. research shows that languages like English utilize only a fraction of this cognitive machinery (397)
3. the knowledge of how this machinery works allows us to invent a much more sophisticated language with a totally new sort of syntax and semantics that makes communication more efficient, and in which FP disappears (398)
D. Third scenario
1. we learn about how the two cerebral hemispheres communicate with each other
2. based on this knowledge, we invent electronic devices that we can implant in brains that allow brains to communicate with each other without spoken language (399)
3. libraries replace books with recordings of neural activity
E. given these three scenarios, the elimination of FP is not inconceivable
F. a final argument against eliminativism:
1. it is incoherent because it assumes the very things it wants to eliminate
2. after all, it's a belief, and one that eliminativists have the desire to communicate, etc.
3. hence, if eliminativism were true and there were no beliefs and desires, whatever the eliminativist says would then just be a meaningless bunch of noise (399 q.v.)
G. in reply to this argument, Churchland says that it begs the question (400)
1. it assumes that for something to have any meaning, it must express a belief
2. the argument could also be interpreted as showing that it is this theory of meaning that should be rejected (400)
3. the question-begging nature of the argument is shown through an analogy he borrows from Pat Churchland
a. she says that it would be like a 17th-century vitalist arguing that if someone denies they have a vital spirit, they must be dead, and hence not saying anything
b. this argument begs the question in the same way
H. in conclusion, the propositional attitudes of FP are not beyond the possibility of being wholly replaced by neuroscience, according to Churchland
Cognitive Suicide (1987), by Lynne Rudder Baker
I. Lynne Rudder Baker actually raises the argument that eliminitavism undermines itself (401)
A. it calls into question our common-sense framework of propositional attitudes (“that” clauses)
B. Baker maintains that if our common-sense framework is eliminated, the following are at risk:
1. rational acceptability
2. assertion
3. truth
II. Rational acceptability:
A. the Churchlands need some alternative account of things like belief and acceptance that doesn’t bring up content (402)
B. The first problem is to make sense of what it means to accept something without referring to content
C. The second problem has to do with the normative notion of justifiably accepting something
1. usually, a person is justified in accepting some thesis when there is evidence for it
2. but what sense can we make of evidence when dealing with states that don’t have content or meaning?
3. Churchlands need replacements for concepts like evidence, hypothesis, argument that don’t appeal to content (403)
D. without any account of the content, how could one be said to accept one hypothesis as opposed to another? (403-4)
III. Assertion:
A. How can there be assertion without belief? (404)
B. Churchland has tried to rebut the claim that eliminative materialism undermines itself through the analogy with vitalism (404-5)
C. Baker argues that the vitalist story is not analogous (405)
1. vitalists and anti-vitalists agree that being alive is necessary for making a claim; they disagree over the cause of life (405)
2. but eliminativists are denying that there are such things as beliefs, not giving a different account of them
D. second, the mistake in the argument against antivitalism has no analogy in the argument against eliminativism
1. whether or not antivitalism is true, it’s a mistake to charge him with being dead on account of lacking a vital spirit
a. if he’s right, he doesn’t need a vital spirit to be alive
b. if he’s wrong, then he’s got such a spirit and isn’t dead
2. but the argument against eliminativism does not beg the question by asking eliminativism to explain belief or content on the assumption that eliminativism is true (405-6)
E. also, the argument that eliminative materialism undermines itself does not, pace Churchland, assume any particular theory of meaning. It makes only the minimal assumption that people can mean what they say (406)
F. how would an alternative to the traditional view even distinguish assertion from mere noise, or assertion from lying? (407)
G. Churchland’s analogy with the way in which the two hemispheres communicate with each other is not helpful, as even Churchland himself has to put scare quotes around terms like “understand” when he says they understand each other (408)
IV. truth
A. if the thesis of eliminativism is true, there is no way to formulate it, because we can do so iff we can say what would make it true (408-9)
B. for the sake of argument, she will join the eliminativist in his contention that mental states cannot be identified with “that” clauses, that is, that thought is not like language (409)
C. but then she will argue that this contention comes to grief on the question of whether thoughts can be true or false
Case 1: If they can, then on what basis?
Case 2: If they cannot, then what makes utterances and inscriptions true or false?
D. Case 1
1. distinguishes two approaches to truth:
a. correspondence
b. causal
2. Baker sees little hope for either
a. correspondence: it is not clear at all clear how mental states without content could map onto reality (409-10)
b. causal: the problem is how to spell out the standard causal chain or what are normal conditions.
1.) cricket/locust example (410)
2.) two causal chains could be indistinguishable if described nonintentionally, yet lead to two different beliefs p and q where one is true and the other false (the muddy zebra?)
E. Case 2: that is, where mental states are not true or false. But then how can what we say or write be true or false?
1. first of all, on the supposition that mental states were neither true nor false, no one could make a mistake (411)
2. the truth or falsity of anything you say would have nothing to do with your thoughts
a. if what you say is false, it’s not your fault
b. if it’s true, it’s just a matter of luck
3. so then in what sense are the things you say true or false?
F. thus, we cannot join the eliminativist in denying that mental states have content (412)
V. in conclusion, Baker that to deny the common-sense conception of mental states as having content is a kind of cognitive suicide (413)