Professor Michael Egnor has kindly responded to my post, The craniopagus twins from British Columbia: A test case for Thomistic dualism (TSZ, November 25, 2017), in a new post, titled, The Craniopagus Twins and Thomistic Dualism (ENV, December 10, 2017). In my earlier post, I had argued that “the twins’ ability to share thoughts without speaking weakens the case for Thomistic dualism, and lends support to a subtle variety of materialism which incorporates top-down causation.” In his response, Professor Michael Egnor gets to the heart of our disagreement and explains why he does not think that the experiment I proposed would serve to test whether dualism or materialism is true. Egnor proposes another test of his own, relating to mathematical abilities. In this post, I’d like to explain why I object to Professor Egnor’s test, before putting forward another one, very similar to it, which I believe could experimentally resolve whether dualism or materialism is true. Finally, I offer a few reflections on the philosophical argument which, Egnor contends, makes materialism logically untenable.
Do the Hogan twins share thoughts or just images?
In his latest article, Professor Egnor contests my claim that the Hogan twins are capable of sharing thoughts (“We talk in our heads” is how the twins describe it). Instead, he thinks, what the twins are sharing is probably mental images. Egnor also maintains that because the girls are twins who spend all their time together, it is likely that their thoughts are very much alike, which means that when they share the same image, they will both (individually) think the same thought, in parallel:
Where Torley disagrees with my view that the twins’ mental abilities are consistent with Thomistic dualism is on their ability to share thoughts. I assert that what they share are images, which are material mental things on the Thomistic view, but that they do not share abstract thought, which is immaterial. (By “images” I mean reconstruction of sensations — visual, auditory, tactile, etc. — in the absence of the object originally sensed.)…
It is clear that the Hogan twins share some thoughts — they giggle at private thoughts that they seem to share. Torley believes that these private thoughts entail some immaterial content. I believe they do not. I believe that the twins share images and perceptions, and that their reaction to the images that they share (giggling) is a manifestation that they both quite separately find the images funny.
In my view, the twins don’t share the intellectual immaterial thoughts. They do share some imaginary material thoughts (sensible species), from which they (at times) each individually extract similar immaterial thoughts (intelligible species). It is only the material sensible species that they share by virtue of their brain connection…
“Talking in their heads” can mean many things, and may just refer to shared perceptions from which they independently derive similar propositions. They may arrive at similar propositions from their shared image, in the same way that two different people may look at the same object and draw similar conclusions from it…
Personally, I think the girls’ description of how they communicate mentally without speaking aloud (“We talk in our heads”) is not at all what one would expect them to say, if they only shared mental images with one another. Professor Egnor evidently thinks differently. Fair enough. But as Huckleberry Finn famously put it, “RECKONING don’t settle nothing. You can reckon till the cows come home, but that don’t fetch you to no decision.” (Tom Sawyer Abroad, by Mark Twain, chapter 9.) What we need is an experiment.
Why Professor Egnor objects to my proposed experiment for testing dualism vs. materialism
In my earlier post, I proposed an experiment: allow each girl to (a) silently choose one statement from a list of six simple sentences containing abstract concepts, (b) decide whether she agreed or disagreed with the statement, and (c) mentally formulate an argument as to why she agreed or disagreed. If each girl could report on the other girl’s choice, agreement or disagreement, and her reasons, without the other girl telling her anything, then that, I argued, would surely show that the girls can share propositional thoughts, and not merely images. Professor Egnor is unconvinced. All it would show, he says, is that the girls think alike:
…[E]ven if they don’t share intelligible species, which is the Thomist view, they may still share sensible species/images and may separately derive the same intelligible species from it. This is particularly likely because of the close relationship between the two girls. They may routinely and individually derive the same abstract thoughts from a sensory image, yet not actually share the same abstract thought. This kind of thing is quite common, for example, with married couples over many years, who both think of abstract things at about the same time (my wife and I do this all the time). Similar things happen with non-conjoined twins and even with close siblings, for whom the actual sharing of thoughts is not an issue.
Now, this would be a perfectly legitimate criticism if the girls consistently made similar choices in the test I have proposed. But what if each girl held different opinions on some of the statements, but was nevertheless able to report accurately on the reasons underlying her sibling’s view, even if she violently disagreed with it? Now that would be a significant finding. Nor could it be explained by saying that as twins, they’re used to arguing over certain issues that regularly arise between them, such as what to eat (interestingly, the girls have different tastes in food). For in the experiment I proposed, the six sentences were entirely novel, and unlikely to have been seen or discussed by the girls previously. So I still think that the test could discredit dualism in the event of a disagreement of opinion between the girls over some of the sentences.
Professor Egnor’s new test
To his great credit, Professor Egnor then proposes a test of his own. He argues that if the girls differ in their mathematical aptitude, that would support Thomistic dualism, but if their mathematical aptitudes never diverge significantly from one another, then that would bolster the case for materialism:
I suggest another approach to testing the girls… If the twins share both material and immaterial powers of the mind, they should share mathematical ability very closely. They should have the same aptitude and the same comprehension of mathematical concepts at every stage of their education. If they share material and immaterial thought, they should share mathematical aptitude, and do so identically.
If they share perceptions (images), but not abstraction (concepts), they would be expected to differ at times in their mathematical aptitude. For example, both girls may share the perception for the symbol for a square root, but if they do not share immaterial thought, it is quite likely that one girl will understand what square roots mean before the other girl understands it… [I]f they do not share abstract thought, it is likely that in at least a few aspects of their mathematical education they will progress at different rates, because they have different comprehensions of the mathematical perceptions they share.
This can be tested rather easily. Thomistic dualism predicts that they will have at least occasional disparities in their understanding of mathematics, which would show up on standardized tests, school grades, etc. If they share intellect as well as perception, their scores and grades should be indistinguishable.
I would suggest that the mathematical test is more comprehensive and practical than the test suggested by Torley. If Thomists are right, the girls will diverge at times in their mathematical aptitude. If Thomists are wrong, and the girls share intellect and well as perception, they should not diverge at all.
I have to say that I regard this test as flawed, as it stands. My reason is very simple: the girls only share parts of their brains, not all of their brains. As I put it in my previous post, “the girls have two brains, not one, even if those brains are uniquely inter-linked.” In an article titled, Parts of the Brain Associated With Thinking Skills (Livestrong.com, August 14, 2017), neurologist Dr. Heidi Moawad writes:
Mathematical and analytical skills require a system of interaction between the temporal lobe, prefrontal region and parietal lobe, which is located near the back of the brain at the top of the head. Skills for algebraic mathematical tasks and calculations are generally concentrated in the left parietal lobe, while skills for geometric perception and manipulation of 3-dimensional figures are determined primarily by the right parietal lobe.
It is almost certain that the two girls’ brains differ in their relative proportions, in some of these areas, and in the number of neuronal inter-connections. That would likely give one girl a mathematical edge over the other, even if the materialistic hypothesis were correct.
An amended version of Professor Egnor’s test
Nevertheless, I think Professor Egnor is on the right track with his proposed test. So I’d like to propose a slight modification to it. Let’s say that one twin is having trouble grasping an abstract mathematical concept – say, the notion of congruence and how it applies to triangles, or the concept of a prime number, or for that matter, a negative number. If the more mathematically gifted twin were then able to correct her sister’s misunderstanding and enable her to grasp the new concept, without saying anything out loud, but simply by “talking in her head” to the other sister, then that would suffice to demonstrate that the two sisters can actually share abstract concepts, and not merely images.
I hope that Professor Egnor will accept my proposed modification to his test. I gather from his remarks on a recent podcast (Michael Egnor on What the Craniopagus Twins Tells Us about Mind and Brain, ID the Future, December 13, 2017) that he knows Dr. Douglas Cochrane, the neurosurgeon at B.C. Children’s Hospital who treats the Hogan twins. If that is the case, and if the girls’ mother is agreeable, then there is no reason why the test I have proposed could not go ahead. What say you, Dr. Egnor?
Why I don’t think philosophical arguments about the mind-body problem are logically compelling
Let’s suppose, for argument’s sake, that my modified version of Professor Egnor’s test was performed, and that it supported the hypothesis of materialism (the brain is capable of abstract thought) over that of Thomistic dualism (the brain stores images and sensory memories, but does not engage in abstract thought). Here’s an interesting question: would Professor Egnor change his mind? I suspect that he wouldn’t. It appears that he regards the philosophical arguments against materialism to be so powerful that no experimental finding to the contrary would cause him to alter his view. As he puts it:
I do point out, however, that in the Thomist view it is not merely empirically true that perception is material and intellection is immaterial. It is logically necessary for the intellect to be immaterial, because the intellect is that by which we contemplate universals, which by definition do not have particular existence and thus cannot be material.
If the twins were shown to share intellect as well as perception, it is not merely our theory of mind that would need revamping, but the logical and metaphysical basis for Western thought as well.
Now, I will readily acknowledge that there are some very powerful prima facie arguments in favor of dualism. Associate Professor Edward Feser has discussed some of these in a series of posts:
Some brief arguments for dualism, Part I
Some brief arguments for dualism, Part II
Some brief arguments for dualism, Part III
Some brief arguments for dualism, Part IV
Some brief arguments for dualism, Part V
But there is an ocean of difference between a highly persuasive argument and a compelling one: the former may (in theory) be mistaken, while the latter cannot. So it’s worth quoting what Dr. Feser himself has to say about the much-vaunted argument from universals:
Whatever one thinks of arguments like this, it is important to understand that (like the other arguments I’ve presented in this series) they are not the sort that might be undermined by the findings of neuroscience, or any other empirical science for that matter. They are not “soul of the gaps” arguments which purport to give a quasi-scientific explanation of some psychological phenomenon that we simply haven’t got enough empirical data to explain in a materialistic way. Rather, they purport to show that it is in principle impossible, conceptually impossible, for the intellect to be accounted for in a materialistic way. If such arguments work at all, they establish conclusively that the intellect could no more be identified with processes in the brain than two and two could make five. If they are mistaken, they would be mistaken in the way one might make a mistake in attempting to carry out a geometrical proof, and not by virtue of having failed to take account of this or that finding of brain research.
Reading between the lines, I get the sense that Dr. Feser himself isn’t 100 per cent sure that the arguments for dualism are both valid and sound. Why might that be?
(i) Universals are particularizable
Let’s look at Professor Egnor’s argument first:
1. By definition, universals do not have particular existence.
2. By definition, material things and processes have particular existence.
3. Therefore, necessarily, universals are not material.
The problem lies in the vague term “particular existence.” It is certainly true by definition that universals are not particular objects belonging to the group whose properties they generalize. The concept of a panda is not a particular panda; nor is the concept of a triangle a particular triangle. But it does not follow that these concepts do not possess particular existence of some sort or other.
How might this be so? Let’s take the common concept of a triangle: a closed two-dimensional figure having exactly three sides (and three angles). (Mathematicians will tell you that’s actually a Euclidean triangle, but let that pass.) I could represent this concept in code, if I wished: C23, where the first letter indicates whether the figure is open (O) or closed(C), the second character represents the number of dimensions (2) and the final character denotes the number of sides (3). Viewed in this way, the concept of a triangle does turn out to have particular existence, after all: each of the characters in the three-letter code has a particular value.
The same goes for the common definition of a giant panda: a large black-and-white herbivorous bearlike mammal. The panda is a mammal, and not a bird, reptile, amphibian or fish. It’s a member of the bear family (not the dog family), within the order Carnivora (animals whose teeth and claws make them specially adapted to meat-eating). Unlike other bears, it’s black-and-white in color. And most unusually, it’s a herbivore. Once again, the concept seems to be “particularizable,” to coin a term. The only vague predicate in the definition is “large,” but even here we can set particular limits by specifying a range: adults are 1.2 to 1.9 meters long, for instance. I see no reason in principle why the brain cannot store such information. I am not saying that it does, of course; only that it might.
(ii) The term “concept” is not a natural kind
Dr. Feser’s argument is somewhat more subtle (which is hardly surprising as he is, after all, an Associate Professor of Philosophy). Feser writes:
Consider that when you think about triangularity, as you might when proving a geometrical theorem, it is necessarily perfect triangularity that you are contemplating, not some mere approximation of it. Triangularity as your intellect grasps it is entirely determinate or exact… Of course, your mental image of a triangle might not be exact, but rather indeterminate and fuzzy… Any mental image of a triangle is going to have certain features, such as a particular color, that are no part of the concept of triangularity in general. A mental image is something private and subjective, while the concept of triangularity is objective and grasped by many minds at once.
Quite so; but all that proves is that the concept of a triangle is not a mental image. What it doesn’t prove is that the concept of a triangle is immaterial. But Feser is not done yet, for he continues:
Now the thought you are having about triangularity when you grasp it must be as determinate or exact as triangularity itself, otherwise it just wouldn’t be a thought about triangularity in the first place, but only a thought about some approximation of triangularity. Yet material things are never determinate or exact in this way... And in general, material symbols and representations are inherently always to some extent vague, ambiguous, or otherwise inexact, susceptible of various alternative interpretations. It follows, then, that any thought you might have about triangularity is not something material; in particular, it is not some process occurring in the brain. And what goes for triangularity goes for any thought that involves the grasp of a universal, since universals in general … are determinate and exact in a way material objects and processes cannot be.
The key premises in Feser’s argument appear to be as follows: (i) concepts are inherently determinate; (ii) our thoughts about these concepts are likewise inherently determinate; (iii) material objects and processes, on the other hand, are inherently indeterminate; therefore (iv) our thoughts are not material objects or processes.
Unfortunately, Dr. Feser does not provide us with anything like a general argument as to why he believes material things and processes are inherently ambiguous or capable of alternative interpretations. (He refers to the work of the philosopher James Ross, which is summarized here. Ross’s main argument is that the various instantiations of a mathematical concept – e.g. the concept of the square of a number – do not uniquely determine its content, as a finite number of instances cannot fix the rule: another interpretation is always possible. But what this proves is not that material processes are inherently ambiguous, but that the meaning of a mathematical concept can never be exhausted by its instances. Fine; but who said it could? And in the case of squaring, the instances aren’t even material, anyway; they’re numbers!)
Feser appears to believe that abstract objects (such as the concept of a triangle) are incapable of alternative interpretations, but again, he does not tell us why. Perhaps his thinking is that you either grasp them or you don’t. This is more promising; but the question we need to ask is: what makes them graspable? Is it their immateriality, as such? And if so, why? Feser does not tell us.
Finally, Feser neglects to mention the inconvenient fact that not all concepts are equally determinate. For instance, the biological concept of a dog [pictured above] (which is capable of hybridizing with a wolf or a fox) is much fuzzier than the mathematical concept of a triangle, the chemical concept of gold (element number 79) or for that matter, the biochemical concept of DNA (which may contain non-canonical bases). And what about the geographical concept of a mountain (arbitrary cut-off point) or for that matter, the concept of “bald” (where does one draw the line)? What about the concept of love (which some people confuse with liking), or the concept of justice (which means different things to different people)? I could go on, but I won’t belabor the point.
The real problem here is that, as philosopher Edouard Machery puts it in a brilliantly argued 2004 essay, Concepts Are Not a Natural Kind. Indeed, Machery argues that we don’t even have a single concept of “dog”: we have several concepts. He concludes:
First, the notion of concept is ill-suited to formulate scientifically relevant generalizations about the mind. Psychologists should focus instead on other classes of mental representations, particularly prototypes, exemplars, and theories (and eventually others). In other words, the notion of concept does not carve the mind at its joints. Second, the controversy between the main psychological theories of concepts is deeply misguided. Concepts are neither prototypes, nor exemplars, nor theories. Some concepts are prototypes, some concepts are sets of exemplars, some concepts are theories. The theory view of concepts, the prototype view of concepts and the exemplar view of concepts are not inconsistent theories about our concepts: instead, they characterize the main features of three basic different kinds of representations. Finally, this position raises a provocative question: if the notion of concept is ill-suited for scientific purposes, do we need it at all? But this is certainly a topic for another day.
More recently, Machery has written a provocatively titled book, Doing Without Concepts, proposing that we jettison concepts altogether (see here for a critical review). This is a very extreme move, and in my opinion, an over-reaction, but Machery has at least performed the philosophical service of forcing us to re-examine our preconceptions about what the mind does and how it works. Philosophical arguments based on the nature of our mental concepts should never be used to over-rule the findings of science, because we can never be certain that we actually think in the way we assume we do.
I would like to close with a plea for an open mind. The philosophical tradition of dualism is a venerable one, which is supported by some ingenious philosophical arguments; but as far as I can tell, the case for dualism is far from airtight. That’s why experiments are so useful. They can, at least, help to eliminate bad hypotheses about the mind, even if they can never prove any particular hypothesis to be true.
In a recent podcast (Michael Egnor on What the Craniopagus Twins Tells Us about Mind and Brain, ID the Future, December 13, 2017), Professor Egnor cited the pioneering work of Wilder Penfield, Benjamin Libet and Roger Sperry, and concluded: “Any objective person looking at the science would have to come away with the viewpoint that dualism makes the most sense here.” Penfield and Libet were indeed both dualists; but Sperry was not. He was a monist, and a strong determinist at that, although he believed in a version of downward causation. His student, Michael Gazzaniga, is a materialist who maintains that determinism is compatible with a version of free will: brains are automatic, but people are free, as he puts it. The point I wish to make here is that neither the scientific data nor the philosophical arguments for or against dualism are compelling, right now. To take one instance: the split-brain work of Sperry and Gazzaniga seemed to disprove dualism, but more recent work by cognitive psychologist and physicist Yaïr Pinto points the other way: when you split the brain, you still end up with only one person. The moral of this story is that we need to keep digging and avoid a rush to judgement.
What do readers think?