This will be the reading for June 29. A pdf of the essay is available here.
As usual, we will meet at 4:00 in the Dunkin Donuts meeting room outside exit 6 of Gangnam station.
Bob punched Tom because he believed that Tom deserved it. Sally went to medical school because she desired to become a doctor. John walked into the store because he intended to buy water. I am currently conscious of myself as thinking about this paper. These propositional attitudes – beliefs, desires and intentions – are the tools we use to explain the behavior of both others and ourselves, and we are conscious of ourselves as thinking. All of this is rather obvious. It is common sense to say that John did something because he intended a particular effect, and that Sally did something because she desired something. In light of that obviousness, the first paragraph of Paul Churchland’s essay is startling:
“Eliminative materialism is the thesis that our common sense conception of psychological phenomena constitutes a radically false theory, a theory so fundamentally defective that both the principles and the ontology of that theory will eventually be displaced, rather than smoothly reduced, by completed neuroscience, a theory that we may expect to be more powerful by far than the common-sense psychology it displaces, and more substantially integrated within physical science generally. (EM, 2)
Common Sense Psychology As Empirical Theory
The materialist tradition has always been split down the middle: on one hand, there has been the tradition most obviously exemplified by Marx, in which the primary thesis is that thinking is founded upon and at least partially determined by the material circumstances it finds itself within. Economics determines psychology. The other materialist tradition is concerned with some version of atomism: the manifest world is, in some way, entirely founded upon or reducible to its constitutive parts. This tradition is most easily identified with what we now think of as science.
Eliminative materialism (EM, for short) is good news for this second brand of materialism. Thought has always been difficult for materialists to deal with; when Churchland was writing in the early 1980s, intentionality was seen as the primary barrier against a materialistic outlook. Thoughts are about things, and it is hard to see how this aboutness – aka intentionality – could be broken down into smaller pieces.
Churchland begins by pointing out that the common sense psychology that ascribes propositional attitudes (PA) to us is an empirical theory, “with all the functions, virtues and perils entailed by that status.” (EM, 3) Rather than being the fundamental nature of thought, it is one possible description of thinking, and as such is refutable.
The persistent defense of common sense psychology is odd. Common sense has said many things: that space has a preferred direction for things to fall in, that weight is intrinsic, and that the sphere of the heavens turns daily. It is easy to see how these ideas are false, because they are safely in antiquity – it is hindsight. As Churchland says, “Let us aspire to some foresight for a change.”
Seeing common sense as a theory lets us unify a range of topics: the explanation and prediction of behavior, the problem of other minds, the nature of introspection and the relation between mind and body. Explanations presuppose laws or principles, and Churchland calls common sense psychology’s network of principles folk psychology.
We do not know that another person has a particular mental state from their behavior, or from the single example of our own minds. Rather, it comes from an explanatory hypothesis which is suggested by and supported by folk psychology (FP). I observe a person’s behavior, and then put forward the explanatory hypothesis that they intended to perform that action.
FP also helps explain what is going on when I think sentences of the sort “I think I like ice cream” and “I think I love this person”. “I think” is an introspective judgment; I look inside myself and explain what I experience there by assigning a proposition attitude (PA) to it. The problem for FP is that introspective judgements are not all that reliable. They are acquired habits of conceptual response to one’s internal states, and the integrity of the response is dependent on the integrity of the framework; “Accordingly, one’s introspective certainty that one’s mind is the seat of beliefs and desires may be as badly misplaced as was the classical man’s visual certainty that the star-flecked sphere of the heavens turns daily.” (EM, 5)
The PAs are a great explanatory tool because they actually look like science. Consider the statements “the air’s temperature is 20 C” and “the table’s weight is 30 kg”; these “numerical attitudes” fix a predicate to a subject. PAs function the same way: Bob believed X, Sally desired Y, and Sam fears Z. The predicates allow the generalization of law-like generalities. Structural features of FP parallel physics quite closely. The only difference is that one deals in numbers, and the other deals in propositions.
Why Folk Psychology Might (Really) be False
As an empirical theory, it is at least possible that FP is false – but only EM really takes that possibility seriously. Most philosophies of mind see a future in which FP is at least partly preserved. Of course, FP works really well, so isn’t that good evidence for it? Churchland thinks this argument is simply tunnel vision. He discusses three criticisms of FP.
First, there is a lot FP fails to explain or address: mental illness, creative imagination, the ground of intelligence differences between individuals. We are ignorant of the physiological function of sleep. The construction of a 3D image from 2D stimulations of the retina. There are a wide variety of perceptual illusions. Memory is a sticky subject for FP. Even the ability to catch a ball is something of a mystery. Another great mystery is the learning process itself, especially “large scale conceptual change,” or pre- or non-linguistic learning like with babies or animals. FP is stuck seeing learning as the storing of PAs – but the ability to store is itself acquired.
Second, the history of FP of is one of “retreat, infertility, and decadence.” The domain of FP used to be larger; recall that the elements of nature were once understood in FP terms. That view was only restricted to higher animals in the last 2000 years. Even when restricted to humans, it has not really advanced. We have the same set of tools the Greeks did, and we are “negligibly better” at explaining behavior than Sophocles. That’s a long time for a theory to not advance, especially with all the anomalies. Perfect theories have no need to evolve, but FP is not perfect. “. . . FP is a stagnant or degenerating research program, and has been for millennia.” (EM, 10)
Third, explanatory success is not the only virtue a theory can have. We also want them to integrate well with other relevant theories. FP does not fair well here. Take natural history, from the perspective of natural science: “we can tell a coherent story of [man’s] constitution, development, and behavioral capacities which encompasses particle physics, atomic and molecular theory, organic chemistry, evolutionary theory, biology, physiology and materialistic neuroscience.” (EM, 10)
The story is incomplete, but it radically outperforms FP in many ways. And it is coherent with the rest of our world picture. The “greatest theoretical synthesis in the history of the human race” is unfolding before us, and FP has no part in it. “Its intentional categories stand magnificently alone, without visible prospect of reduction to that larger corpus.” (EM, 10) FP is in the same place as alchemy and Aristotelian cosmology as chemistry and classical mechanics were forming.
Is Folk Psychology a Necessary Normative Framework?
EM says FP is an empirical theory, and probably false. Some defenders say it is not an empirical theory, and so cannot be refuted or transcended like an empirical theory. They say FP is a normative framework. It is a description of an ideal type of internal behavior: the rational use of PAs. Language is a central category of a wide range of 20th century philosophies, and Churchland offers a sharp criticism of that centrality:
“Third, even if our current conception of rationality – and more generally, of cognitive virtue – is largely constituted within the sentential/propositional framework of FP, there is no guarantee that this framework is adequate to the deeper and more accurate account of cognitive virtue which is clearly needed. Even if we concede the categorical integrity of FP, at least as applied to language using humans, it remains far from clear that the basic parameters of intellectual virtue are to be found at the categorical level comprehended by the PAs. After all, language use is something that is learned, by a brain already capable of vigorous cognitive activity; language use is acquired as only one among a great variety of learned manipulative skills; and it is mastered by a brain that evolution has shaped for a great many functions, language only being the very latest and perhaps least of them.” (EM, 18)
As for normativity,
“EM. . . does not imply the end of our normative concerns. It implies only that they will have to be reconstituted at a more revealing level of understanding, the level that a matured neuroscience will provide.” (EM, 19)
Beyond Folk Psychology
What would the elimination of FP actually involve? This would depend on what neuroscience discovers, and how hard we follow through on it. He outlines three possible scenarios.
First, imagine a future in which research into the structure of the brain yields an entirely new concept of cognitive activity. It is uniform for terrestrial brains, humans and animals alike, and is conceptually compatible with other sciences. It may claim we have a set or configuration of complex states “which are specified within the theory as figurative ‘solids’ within a 4 or 5 dimensional phase space.” (EM, 20) A set of laws are outlined which govern the interaction of these solids. The upshot here would be better short term prediction of behavior, and a better description of long-term learning. It would also provide a new account of knowledge. An idea a person gives assent to would be a “one-dimensional projection” of the relevant parts of the brain onto the speakers language (Plato’s shadows). Being projections, they do carry worthwhile information, but they fail to reflect “the deeper reality in all its kinematically, dynamically, and even normatively relevant aspects.” However, this scenario might not have practical consequences; it might never leave the ivory tower.
His second scenario is a little more dramatic. Noam Chomsky says the brain and/or the mind is innately fitted to learn grammar. Another view is that the brain does have innate structures, but that they are built for perception, not language, which is a secondary development. Let’s assume the second view is true, and imagine another future scenario: research into neural structures shows that perceptual processing structures process more complex and a greater quantity of information than natural language. Natural language only uses part of the available information processing power. It could turn out that another linguistic system could harness that power. Learning that language would allow us to communicate at a much greater efficiency, though it would be alien in its syntactical and semantical structures.
The third scenario is crazy. We already know there is a lateralization of function between the two cerebral hemispheres, and they get information from each other via the corpus callosum. People whose callosums have been severed have a variety of behavioral problems. But people with callosal agenesis – a birth defect in which the callosum is absent – have few to no problems. This suggests “the two hemispheres have learned to exploit the information carried in other, less direct pathways connecting them through the subcortical regions.” (EM, 22)
So perhaps even in a normal case, a developing hemisphere acquires the ability to make use of the information the callosum gives it. What we see are two distinct cognitive systems “responding in a systematic and learned fashion to exchange information.” (EM, 22) An incredible amount of information is exchanged; more than 2×108 bits per second, compared with less than 500 bits per second with spoken English.
If two distinct hemispheres can learn to communicate like that, why not two distinct brains? It would require an artificial callosum, but imagine creating a workable transducer which could be implanted into some part of the body – say, the forehead – which transforms neural activity into microwaves, and which are then transformed back into neural activity for another person. Once the channel is open between 2 or more people, they can learn (learn!) to exchange information in the same way hemispheres do.
“Think what this might do for hockey teams, and ballet companies, and research teams! If the entire population were thus fitted out, spoken language of any kind of might well disappear entirely, a victim of the ‘why crawl when you can fly’ principle. Libraries become filled not with books, but with long recordings of exemplary bouts of neural activity. These constitutes a growing cultural heritage. . . . But they do not consist of sentences or arguments.”
“How will such people understand and conceive of other individuals? To this question I can only answer, ‘in roughly the same fashion that your right hemisphere ‘understands’ and ‘conceives of’ your left hemisphere – intimately and efficiently, but not propositionally.” (EM, 23)
If we want to know what the possible consequences of a thorough-going materialism entirely beholden to the scientific image of man looks like, this is it: the complete collapse of the symbolic order. The Borg! So much for Wilfred Sellars’s stereoscopic image.