Consciousness
from:
Brain-Wise: Studies in Neurophilosophy (2002)
by:
Patricia Smith Churchland
Chapter 4
Part II: Dualism and the Arguments against Neuroscientific Progress
II.I Life and Conscious Experience (171)
A. at this stage of our knowledge, we cannot simply equate consciousness with any such function as being awake, attention, perceiving, etc.
B. however, we are learning more about consciousness as we learn more about these functions
C. in this way, the virtues of the indirect approach to the study of consciousness may be analogous to those of the indirect approach to the study of life
1. trying to find some micro-structure that explains being alive turned out to be the wrong way to go
2. similarly, trying to find some micro-correlate to consciousness could prove the wrong way to go (171)
D. in other words,
1. just as, if someone were to ask just what life is at the end of a biology course in which they learned all about DNA, RNA, proteins, cells, etc., he or she would be told that life is all that stuff
2. there is no “vital spirit” or “life force” (172)
3. similarly, those who pursue a scientific study of consciousness think that by studying all the details about sleep, memory, perception, self-representation and so on, we will learn about consciousness
E. as Churchland sees it, for someone to object that consciousness cannot arise out of non-conscious physical stuff would be like someone 100 years ago arguing that life cannot arise from non-living stuff (172-173)
F. this is not to say that science has ever come up with an experiment that disproves the existence either of a vital spirit to explain life or a soul to explain consciousness (172-173)
1. rather, the explanatory frameworks that included those concepts have come to be replaced by new frameworks that do not make use of them
2. people came to see that the concept of a vital spirit was no longer needed to explain life (172)
3. the concepts of impetus and caloric met a similar fate
4. in sum, for Churchland the idea of a nonphysical soul is simply an outdated theoretical notion (173)
a. the non-existence of the soul will never be shown by experiment
b. rather, psychology and the neurosciences today can explain so much more than any dualist theory and in a way that fits in with the rest of science
G. nevertheless, even some philosophers who would agree that dualism is dead continue to argue that a neurobiological approach to the study of consciousness is simply a waste of time
II.II Nine Naysaying Arguments (173)
A. here she quotes some contemporary philosophers who argue not merely that we cannot explain consciousness biologically now, but that we will never be able to (173-174)
1. Colin McGinn argues that we just are not able to form the right sort of concepts to understand this, just as a mouse cannot form the right concepts to understand calculus
2. Zeno Vendler chides neuroscientists for assuming that there is nothing science cannot understand (174)
B. Churchland will analyze such philosophers’ arguments one-by-one
1. I cannot imagine how science could explain awareness!
A. from the fact that we cannot presently explain consciousness, it does not follow that we will never be able to
B. the argument from ignorance is simply a fallacy
1. that is, from the fact that we don’t know something, nothing very interesting follows (q.v.)
2. there can be all sorts of reasons for not knowing something (175)
3. lack of evidence for something is not positive evidence for something else, especially not for something ghostly
4. the mysteriousness of a phenomenon is not a fact about the phenomenon but rather a fact about ourselves
C. some think we can make the jump from “we cannot now explain” to “we will never explain” by adding the premise, “I cannot imagine how we would ever explain”
1. but this premise, too, simply states a psychological fact about the speaker
2. analogous to someone a century ago arguing that they cannot imagine how we can get life out of matter (175)
3. another problem with this sort of argument is that it easily becomes question-begging, given that we don’t have well-defined notions of what it means for someone to be able to imagine something (176, q.v.)
a. for instance, suppose someone were to try to argue that he can imagine mechanisms for sensorimotor integration, but not for consciousness
1.) what could this mean, given that science has detailed answers for neither?
2.) if it means he can imagine this neuron connecting to that, etc., he can imagine this just as well for consciousness
b. hence, to say one can imagine mechanisms for things like attention and memory but not for consciousness, one has to do more than say that one can imagine neurons doing one and not the other – otherwise one begs the question
2. There could be zombies
A. She reads the zombie argument as claiming:
1. we can imagine a zombie
2. if it’s conceivable, then it’s logically possible (177)
3. since a zombie is logically possible, consciousness cannot be explained by brain activity (177)
4. a true explanation of consciousness would show that a zombie is not logically possible
B. Churchland replies:
1. many things, like a two-ton mouse or a spider who plays the flute, may be logically possible, but not empirically possible
2. why should we think that the logical possibility of a zombie would tell us anything about what may be fruitful areas for further research? After all, science is interested in empirical possibilities.
3. the reply to this question makes the key assumption that a scientific explanation rules out logical possibilities
a. Churchland thinks this is simply absurd
b. Holding explanations to such a standard rules out explanations of consciousness in terms of souls – or anything else – as well as brain functions
c. no scientific explanation of any phenomenon has ever met this condition
1.) explanatory reductions require that a new theory explain most of the features of the phenomenon to be reduced
2.) they do not require that the new theory entails this phenomenon in such a way that what was previously thought to be logically possible is now shown to be otherwise (177)
3.) explanations in science rule out empirical possibilities, not logical possibilities
4. facts about what people are able to conceive do not tell us anything interesting about the way things actually are
5. that something seems possible does not show that it is genuinely possible. To think otherwise is to confuse real possibility with mere grammaticality.
6. it may not even be true that we have a clear idea of what it means for something to be logically possible (178)
a. we may have some clear examples of things that are genuinely possible or impossible, but there is a large gray area
b. philosophers cannot draw conclusions from examples in this gray area about what science can or cannot achieve
C. she also suggests two similar arguments of the same form to show the problem with zombie arguments:
1. the logical possibility of “deadbies,” that are like us in every way except that they are dead, which would (supposedly) entail life does not reduce to biology (178)
2. imagine a planet where the average kinetic energy of the atmosphere increases but its temperature stays the same. What does this tell us about the relationship between mean molecular kinetic energy and temperature? Nothing.
D. the zombie argument is an exemplar of the weakness of thought experiments divorced from real science (q.v.)
3. The problem is too hard
A. Churchland asks how we can tell whether a problem is too hard before we have done much of the science on it? (179)
B. in fact, the history of science shows that we have often been wrong about what problems were harder than others
1. it was thought that the chemical composition of the stars was a hard problem, while the perihelion of Mercury was relatively easy
2. it was thought that the folding of proteins was a relatively easy problem that should be solved first before science turned to how traits were copied and transmitted from parents to offspring
C. such examples illustrate the fallacy of the argument from ignorance (179)
4. How can I know what you experience? (180)
A. the most well-known form of this argument is the inverted spectrum argument
B. in general, it says that since phenomenal experience is underdetermined by the physical facts, that phenomenal facts or qualia must be non-physical, that is, facts that can never be explained in physical terms (180-181)
C. Churchland finds these arguments to turn on questions of what is conceivable, like the zombie argument (181)
1. the key premise is that our brains could be identical yet our qualia could differ
2. but this means that it is only logically possible that the qualia could differ
3. and again, this tells us nothing about real possibilities
D. The empirical evidence goes against the premise that brains could be the same while qualia differ
1. for instance, pain from a toothache decreases as neural activity decreases
2. direct stimulation of appropriate cortical area can produce sensations in the hand
3. there is no evidence of the brain remaining the same while conscious experience changes (181)
E. other defenses of the key premise that our brains could be identical while our qualia differ
1. one is premised on the assumption that qualia are non-physical. But this is simply to assume dualism at the outset.
2. another defense is premised on the assumption that conscious experiences are not identifiable with any property of the nervous system, but this defense simply begs the question (182)
3. see table 4.1 on p. 182 for a summary of all 5 kinds of defenses of the key premise
F. the problem with the inverted spectrum argument is that it tries to establish something about the way the world is from absolutely no facts. It needs to persuade us that qualitative differences in experience cannot be detected (q.v.)
G. in fact, the strongest arguments against the reduction of consciousness to neurophysiology are those based on dualism (183)
1. that is, at least the dualist could argue that if there were no way to tell whether someone’s spectrum was inverted, that this would be evidence in favor of dualism (183)
2. for this reason, Churchland turns to the question whether an inverted spectrum is empirically possible
What Happens if we get more empirical?
A. the usual way the inverted spectrum argument is presented is much too simple
1. the spectrum we get from dividing white light with a prism includes only a fraction of all the colors we can see, leaving out brown, pink, etc.
2. our colors are better thought of as forming a 3-D solid than a 1-D spectrum (cf. plate 3, between 186 and 187)
3. in addition, this 3-D color space is not uniform (see plates 3 and 4)
a. if we make equal distances represent equal increments of color discriminability, it is not a sphere
b. this reflects that we can make finer distinctions in the green, yellow, and orange regions than in the blue
B. Churchland suggests that we imagine inverting the color solid rather than just a linear spectrum (184)
1. this would be logically possible
2. however, if it were to happen to someone, there would be ways to tell (184)
a. this person would make more discriminations in the blue area, fewer in green-yellow-orange
1.) there may be good evolutionary reasons that we make finer discriminations in the green than in the blue area
2.) birds with a 4th cone in UV area make finer discriminations among the blues
b. the boundaries among the colors would be different
C. concludes that it is not true that our color qualia could be different without there being any physical or behavioral difference
Connecting qualia and neuronal organization (185)
A. in this section, Churchland proposes to explain the shape of the color space in terms of neuronal organization
B. the story begins with the 3 kinds of cones
1. each is preferably sensitive to a narrow band of wavelengths
2. and the peaks are not evenly spaced: see figure 4.17 on p. 185 and plate 2)
C. but the next, important step is in the color opponent cells in the lateral geniculate nuclei
1. note: many of the inhibitory and excitatory arrows in fig. 4.18 on p. 186 are incorrectly drawn and do not agree with her text
2. the green vs. red cells are excited by the medium wavelength cones, inhibited by the long, so the result is a function of the relative amounts of medium and long-wave input (186)
a. notice in figure 4.17 and plate 2 that there is considerable overlap in the response curves of these two kinds of cones.
b. This means the struggle between these two kinds of cones makes possible very fine discriminations in this part of the spectrum
3. blue vs. yellow are excited by the short, inhibited by medium and long (186-187)
a. notice that here, there is very little overlap in the response curves between the two kinds of cones (187)
b. this explains why we cannot make as many fine discriminations in this part of the color spectrum
4. white vs. black are excited by all 3, inhibited by average stimulus level. Determines color brightness of a particular part of the retina, relative to rest (187, q.v.)
D. we can represent visible colors in a three-dimensional space whose axes correspond to the activity level of these three kinds of opponent cells (187, cf. plate 5)
1. when we do that, we find that the possible coding triplets do not fill the cube but form the same shape as the Munsell color solid, because of the way the cells are wired
2. the parts of the solid are even associated with the appropriate colors (187-188)
E. hence, at least to a first approximation, it looks like we have found the neuronal basis for phenomenal color space (188)
1. the coding triplets stand in the same proximity and similarity relationships as the colors
2. if we can identify phenomenal colors with coding triplets in the same way that we identify temperature with mean kinetic energy, etc., we can explain these relationships
a. however, Churchland cautions that this should be regarded as only a provisional hypothesis and that many other things could affect color perception (188-189)
b. nevertheless, the hypothesis can explain things like color blindness, color illusions, after-images, etc. (189)
F. this provisional hypothesis thus looks more plausible to her than dualism.
1. the inverted spectrum thought experiment is now harder to imagine
2. if it is to preserve the metric of similarity and dissimilarity relations, there will have to be changes in the wiring
3. An inverted spectrum of the sort proposed by the dualist may be possible, but it would show up in the behavior of the cones and in the wiring that connects them to opponent cells
5. The dualist tries again
A. the dualist may then insist that the inverted spectrum is still possible as long as we invert the similarity/dissimilarity metrics as well as the color qualia. This will require no changes at the neural level and not result in behavioral changes.
B. Churchland concedes that this could be possible (190)
1. but if the similarity relationships have changed, who could say that we’re still talking about the same set of colors?
2. But how can we say that any of the features of color we have been discussing is essential to color qualia? (190, q.v.)
3. it is the job of scientific research to tell us what is and what is not essential to color qualia (q.v.)
The dualist is hoist by his own petard
A. but this conclusion applies just as much to the dualist as to the materialist
1. indeed, it’s worse for the dualist, because the inverted spectrum argument pretends that how the colors appear to us is essential to what they are
2. but the dualist has no more right to say this than the functionalist has to say that the functional role is the essential feature of color
3. that is, no one can say what is essential and what isn’t until the science is done
B. one could turn the dualist’s argument strategy against him in the following way:
1. suppose half the population had its ability to make judgments about its qualia systematically inverted: an undetectable malady that made them always wrong in their judgments about their phenomenal qualia (refers to pp. 118-23 where she discusses mistaken judgments such as in the case of Anton’s syndrome)
2. then one could not say that our judgments about our qualia can define what they truly are (190-91)
3. Pat Churchland wouldn’t defend this argument; rather, she’s just using it to show that conceivability arguments are a two-edged sword (191)
C. one may insist that qualia by definition are those things whose appearance is their reality
D. in reply, Churchland cites examples from the history of science in which people argued that certain things were true by definition but nevertheless turned out to be wrong
1. that atoms cannot be split
2. that the Earth does not move
E. These examples raise a second point: the meaning of a word cannot be separated from beliefs about what that word applies to, and these beliefs can be wrong (q.v.)
So who is right?
A. none of this proves that the dualist is wrong about qualia
B. all that has been shown is that the inverted spectrum argument will not suffice to support the dualist position
C. what will decide the issue is which side of the debate has the better explanatory theories
1. the materialist can offer the opponent-cell activation-space theory of color coding to explain, for instance, the shape of the Munsell color solid (191-192)
2. so it’s up to the dualist to offer some explanation of the same things (192)
6. Doesn’t neuroscience leave something out?
A. one could argue that it does, since knowing all about the color space produced by color-opponent cells doesn’t help one to recognize colors
B. for instance, one could be color blind, and even if one knew the theory, not be able to recognize red
C. but, as Churchland says, it is not enough to know the theory; one must have a functioning system as described by this theory
1. but this does not mean that there’s something wrong with the theory, especially when the theory can explain color-blindess
2. having a skill and knowing a theory are entirely different things (192-193)
3. however, if the theory is true of you, then you have the skill (193)
7. It is ridiculous to expect a reduction from the behavioral level directly to the neuronal level
A. however, Churchland argues, it does not follow from this that consciousness cannot be explained neurobiologically (193)
1. there are many levels of organization in the nervous system
2. it is unlikely that high-level functions, like recognizing faces, will be directly explained at the lowest level
B. this objection turns on a misconception of reduction
8. Consciousness is not a neural effect but a subatomic effect (195)
A. Penrose and Hameroff have argued that consciousness arises from quantum effects in the microtubules of neurons
1. Penrose thinks mathematical understanding transcends what nerve cells and neuronal networks can do
2. quantum events are needed to transcend these limits and access the Platonic realm of mathematical truths (195-96)
B. But Churchland sees no hard evidence that supports this theory. Indeed, it’s hardly testable. (196-197)
9. Science cannot solve all problems (197)
A. this is Vendler’s argument, mentioned above
B. but Churchland thinks that since neuroscience has made progress in the study of the mind, there is reason to think further progress is possible (197)
II.III Conclusions
A. she has considered the inverted spectrum argument and other arguments against looking to neuroscience to understand consciousness, and found them all wanting (197-98)
B. putting a bunch of flawed arguments together doesn’t help the case, either (198)
C. of course, showing that these arguments are weak doesn’t prove that neuroscience will be successful
D. the best answer to the skeptics is to make progress in neuroscience