Response to Dr. Jacobus Erasmus on the Soul
It is always pleasing—especially these days—to see a critique that does not descend into invective. Dr. Jacobus Erasmus does me the favor of offering such a critique of my earlier blog post on the soul:
He gives his reply at:
Here I would like to offer a response.
In my OP, I gave three arguments against the existence of souls. First, though, I gave two reasons for why I consider the burden of proof to be on the defender of substance dualism, the claim that there exist substantial spiritual entities that constitute persons and which perform all mental functions. My first reason was this:
We know that certain configurations of matter—those configurations we refer to as “human beings,” for instance—are capable of performing mental functions. They think, feel, perceive, imagine, desire, will, believe, and so forth. If, then, certain configurations of matter can perform mental functions and possess mental properties, the parsimonious, spontaneous, and natural assumption would be that matter, when organized in suitable ways, can perform mental functions and possess mental properties.
It seems perverse to make the opposite assumption, namely that material beings cannot think, and that therefore their mental functions and properties must be due to the operation of something non-physical, a soul perhaps. To make this last assumption would seem to be a bizarre instance of an a priori prejudice. The proper starting point therefore appears to be that material beings are capable of doing whatever we observe them doing, not with the gratuitous and a priori assumption that they cannot.
Erasmus responds that this argument is viciously circular:
Parsons first assumes non-dualism in order to argue that one should first assume non-dualism, and this is viciously circular (I use the term ‘non-dualism’ as an umbrella term to refer to all views that claim that the soul does not exist, such as physicalism, materialism, and property dualism). When Parsons claims that it is the brain that can ‘think, feel, perceive, … and possess mental properties’, he is simply describing non-dualism. Dualism, on the other hand, states that it is the soul (or the immaterial mind or self), and not the brain, that feels, perceives, has mental properties, etc. Thus, Parsons first assumes non-dualism, and then he concludes on the basis of this assumption that one should first assume non-dualism…
But the argument is not circular at all. It makes no assumption whatsoever about souls, brains, or dualism. It simply makes the (I think) utterly uncontroversial and everyday assertion that we know that human beings think, feel, perceive, imagine, etc. In our own cases, that we are capable of doing these things is obvious to our first-person consciousness. With respect to others, we often see and hear them engaging in rational discourse and argument and displaying emotional reactions. Saying that we know that human beings think, feel, etc. is certainly no more controversial than saying that we know that chimps, cats, or dogs have a mental life. If I see my cat displaying anger, fear, or affection, then I conclude that cats are capable of feeling emotion. I further observe that, like cats, humans are physical beings, and I make no assumptions at all about anything more than physical that a human might be.
(a) Human beings think.
(b) Human beings are physical beings.
I conclude that the parsimonious, natural, and spontaneous assumption should be:
Some physical beings are capable of thinking.
I see this reasoning as no more controversial or circular than:
(a) Cats have feelings.
(b) Cats are physical beings.
Some physical beings have feelings.
Further, it would be a piece of perverse a priori prejudice to recognize that some physical objects think, feel, etc. but to assume that they cannot qua physical, think, feel, etc. and instead to assume that such capacities must be due to hidden spiritual entities. In general, if F’s are known to G, then the assumption should be that F’s are capable of G-ing not that F’s are incapable of G-ing—which capacity must be attributed to something unobserved.
Where is the circularity? Actually, Erasmus puts in “brain” where I have said “human beings” and this creates an appearance of circularity, as though I were presuming a physicalist thesis when I am not.
If, then, it is utterly uncontroversial that some physical beings think and feel, then, as I say, it seems to be a peculiar a priori prejudice to begin the discussion on the assumption that such beings cannot think or feel, but only are capable of doing so by the operation of something nonphysical. Therefore it seems eminently fair to ask that those who think that, despite appearances, physical beings cannot think should bear the burden of proof.
A further reason I give for putting the burden of proof on the dualist is this:
… surely, by now, the heuristic assumptions of neuroscience have gained some degree of authority. As I mentioned earlier, a regulative assumption of all the sciences that study mind and brain is that the brain is sufficient for all mental activity. Perhaps, as David Chalmers famously argued, we may never solve the “hard problem,” that is, to understand exactly why some physical events should cause mental events. Still, Chalmers takes for granted that physical events cause mental events, so he accepts the regulative assumption. When a program of inquiry has produced hard, reproducible, and important results, as has neuroscience, then this would warrant prima facie acceptance of the heuristic principles that have guided such research.
To which Erasmus replies:
To make such a bold claim as ‘neuroscience presupposes (A)’, one must provide some evidence for the claim. Parsons does not do this. He does not reference any neuroscientist that explicitly states that (A) is the heuristic principle of neuroscience. Nor does he reference any study that supports his argument above. Nor does Parsons interact with the recent scholarly work that shows that Parsons has got it completely wrong. For example, in their recent paper titled ‘Neuroscience: Dualism in Disguise’, Riccardo Manzotti and Paolo Moderato (2014) persuasively argue ‘that most of current neuroscientists, contrary to often-heralded physicalist credo, embrace dualism … [and, furthermore] that the implicit assumptions adopted by most neuroscientists invariably lead to some sort of dualistic framework’ (Manzotti and Moderato, 2014:81). Contra Parsons, most neuroscientists assume dualism and, thus, according to Parsons’ argument, the burden of proof should fall, not on the dualist, but on the non-dualist.
My assertions about the regulative or heuristic assertions of neuroscience follow Owen Flanagan’s conclusions in The Problem of the Soul. After quoting a passage from a compendium of brain science to the effect that all the artifacts of human culture are accomplished by the brain, Flanagan concludes:
Modern mind science regulates its inquiry by the assumption that mind is the brain in the sense that perceiving, thinking, deliberating, choosing, and feeling are brain processes…That the mind is the brain is thus a regulative assumption that guides contemporary mind science. (Flanagan, 2002, pp. 77-78.)
Flanagan is a philosopher but also a professor of neurobiology, so he should speak with some authority.
Neuroscientist David Eagleman has this to say in the preface to his primer of neuroscience, The Brain (2015, p. 1):
The strange computational machinery in our skulls is the perceptual machinery by which we navigate world, the stuff from which decisions arise, the materials from which imagination is forged. Our dreams and our waking lives emerge from its billions of zapping cells.
Francis Crick offers what he calls “the Astonishing Hypothesis”:
The Astonishing Hypothesis is that “You,” your joys and your sorrows, your memories and your ambitions, your sense of personal identity and free will, are in fact no more than the behavior of a vast assembly of nerve cells and their associated molecules.” (Crick, 1994, p. 3).
William H. Calvin, a theoretical neurophysiologist, in his book with the telling title How Brains Think, locates our unique human aptitude for linguistic creativity in the brain:
Yet we are all constructing brand-new utterances hundreds of times every day, recombining words and gestures to get across a novel message. Whenever you set out to speak a sentence you have never spoken before…you do all your trial and error in your brain, in the last second before speaking aloud. (1996, p. 2).
Paul Churchland, another philosopher who has spent his career engaged with neuroscience, confidently claims that we now have exciting results:
…[W]e are now in a position to explain how our vivid sensory experience arises in the sensory cortex of our brains: how the smell of baking bread, the sound of an oboe, the taste of a peach, and the color of a sunrise are all embodied in a vast chorus of neural activity. (1996, p. 3).
These quotes should be sufficient to indicate that I was not making an idiosyncratic or groundless claim about the assumptions of neuroscience about the efficacy and sufficiency of the brain for the mental. Given more time and space, I am sure I could adduce quite a few more such quotes. At a more basic level, William Lyons’ excellent history of the philosophy and sciences of the mind, Matters of the Mind, makes very clear the decisive break with Cartesianism that occurred with the rise of behaviorism in the early twentieth century. As he shows, though behaviorism was rejected, the commitment to a third-person, scientific, and objective understanding of the mind was permanent and there has been no move, either in science or philosophy, to return to substance dualism.
As for the piece by Manzotti and Moderato, it does not deny what I assert, namely that neuroscientists explicitly invoke assumptions about the mental arising from the physical. Their argument is that neuroscientists implicitly accept a dualism that they explicitly reject. In other words, their argument is that neuroscientists are not aware of the actual implications of their own stated assumptions. Yet if neuroscientists are united in explicitly affirming physicalism, the burden of proof has to be on those who say that they are wrong, i.e. that they fail to recognize the implications of their own assumptions. This burden Manzotti and Moderato attempt to shoulder in their article, and this is the burden that I assert must be borne by those who oppose the explicit assumptions of brain science. In other words, Manzotti and Moderato accept the burden of proof and do not try to evade it as Erasmus does. Whether they are successful in their argument or not is not my place to say here. I would like to see the responses of practicing neuroscientists.
My first argument has to do with the interaction problem. Erasmus begins by offering an argument that I never make or endorse and attacks that straw man. I will pass over these passages in silence.
Erasmus’s statement of the interaction problem is quite inadequate:
Parsons’ first move against dualism is to appeal to the interaction problem, which states that it is difficult to see how a non-physical, immaterial entity, such as a soul, can causally interact with a physical entity, such as a brain.
But the problem is not merely that it is “difficult” to see how souls affect brains. This seriously understates the problem. The problem, is that in substance dualism “mind” and “matter” are defined in mutually exclusive terms. Mind has no physical properties; it is not composed of atoms or any other physical entities; it is not bound by the laws of physics or describable in terms of physical theory; it cannot be detected by any physical means; it is perhaps not even locatable in space. Its putative interactions with matter therefore must be of a wholly different sort than the interactions of physical things. For instance, interactions between fundamental physical particles are mediated by gauge bosons. Is that how it works with souls and matter? If not, how? We have very well developed physical theories about, say, the interaction of electrons and photons (quantum electrodynamics). With putative soul/body interactions there is a lot of speculation and hand-waving, but nothing definite—certainly nothing to compare to the detailed, coherent, rigorous, testable theories of fundamental physics. It is with justification that Flanagan says that dualists believe in psychokinesis.
The standard reply of substance dualists seems to be to concede that we do not know how souls and bodies interact but to assert the tu quoque that we do not know how bodies interact with bodies. Well, then, is it a tie? Are physical theories and soul theories equal in that neither can really explain interaction? Are we each just stuck with saying “shit happens” and that is all we can do? Here is what I said in the OP:
Even if we reach a brutally factual rock bottom in physical theory, where we just have to postulate fundamental entities and forces that turn out not to be further explicable, such brute facts lie at the end of a very long chain of deeply satisfactory explanations. Hume somewhere asserts that we will probably never understand why bread nourishes. Ah, but we do. We have for quite some time understood in very considerable detail how mitochondria break down complex carbohydrates and, via the chemical pathway known as the Krebs cycle, provide energy at the cellular level. Indeed, molecular biology is full of extremely detailed explanations of physiological processes that tell us why they happen just as they do. At a more basic level, nuclear physics can explain in detail why nuclear weapons bang so prodigiously. In innumerable cases, we do not say just that shit happens, but precisely why it happens.
But contrast the “explanations” offered by soul-theory:
With soul-theory, the incomprehensibility is right up front and on top. Your inquiry immediately hits a wall. “How do I think?” “Your soul does it.” “How does it do it?” “Let me explain: Shut up!” Sorry, but I am not just being flippant here. It matters where you put your brute facts. Mystery-mongers take you right to the occult, “explaining” in terms of astrological influences, or hexes, or psi, or chakras, or qi, or whatever. Honestly, the postulation of souls just seems to be another appeal to the occult. Neuroscience thinks we can go deeper without any such paranormal postulates, and we can.
Parsons is essentially stating (without defending) the following argument:
(B1) A brute fact that lies at the end of a very long chain of deeply satisfactory explanations is acceptable.
(B2) If soul-body interaction is a brute fact, it does not lie at the end of a very long chain of deeply satisfactory explanations.
(B3) Therefore, it is unacceptable to claim that soul-body interaction is a brute fact.
As noted above, Parsons simply states these premises without defending them. But why, exactly, should it matter where a brute fact is situated in a chain of explanations? And why, if soul-body interaction is a brute fact, would it not be situated at the end of a long chain of satisfactory explanations? Surely such interaction would take place at a very fundamental (perhaps even quantum) level of physical reality, with the chain of explanations running up to a higher level, such as the brain itself. And why think that soul-body interaction must be a brute fact? Parsons does not interact with the dualistic arguments that try explain soul-body interaction (e.g., some argue that both the soul and brain have the property of being (or being able to be) conscious, and soul-body interaction occurs, not as a brute fact, but in virtue of the transference of consciousness from the soul to the brain).
Does it matter where, in an explanatory chain, we put a brute fact? Of course it does. Three centuries ago if you asked how bread nourishes the body the answer would be shrug. It just does. Nobody really knows how or why. Today we can give a very detailed answer to that question right down to the molecular level, and below. Suppose our current explanations still reach an explanatory “rock bottom,” maybe with the fundamental properties of fundamental particles. Who knows more about how bread nourishes, us or the people in the 18th century?
Erasmus also suggests that maybe soul/body interaction takes place at a very fundamental level, perhaps the quantum level, and so a long chain of satisfactory explanations would also lead up to brute soul facts as with brute physical facts. So, this suggestion is that explanations will be satisfactory as long as they are physical, but they become brute the moment they appeal to souls. In other words, the instant that soul theory starts to do any work it becomes incomprehensible. This seems to concede my point, namely that we have many excellent physical explanations of how matter interacts with matter, but as soon as we invoke souls, understanding ends. Further, with physical theory, if and when we reach something inexplicable, it is contingently and perhaps temporarily inexplicable, and the development of theory might explain the currently inexplicable. With soul-theory, no conceivable development would make it into anything other than the blank it is. Its incomprehensibility is in principle; it is occult by nature.
Erasmus says that I fail to engage with those dualists who have offered explanations of how souls and bodies might interact. I have seen a number of these “explanations.” At best, they amount to scenarios, and there is no responsibility to engage with every scenario that pops up. Give me a theory, a real theory, one that makes robustly testable, rigorous, and specific claims—and not hand-waving, speculation, or pseudoscience—and I will engage it.
My second argument against souls is that soul-theory thinks of the self as a simple, abiding, spiritual entity that constitutes our personal identity. This is the theory of the self as a Cartesian Ego. I opposed to this theory a version of what is normally called the “Bundle” theory of the self, which is traced back to Hume, but which also has roots in Buddhism. On this theory, personal identity is not constituted by a spiritual essence or entity, but is a nexus of heterogeneous experiences and traits.
Erasmus says that in this section I equivocate between two different senses of self, meaning “self”:
Parsons…is here using the term ‘self’ to mean ‘non-spatial substance that has mental properties’. Parsons then goes on to use the term ‘self’ to refer to a set of connected mental properties and personal characteristics that change over time.
Not so. Throughout I am using self in the entirely neutral sense of my personal identity—that—whatever it is—that makes me me at any given time and over time. Further, I am considering two different theories of personal identity—the Cartesian Ego theory and the Bundle theory, and giving my reasons in support of the latter. No equivocation.
My third objection to souls begins with the simple and undeniable observation, backed by enormous empirical research, that non-human animals have minds, that is, they are capable of quite sophisticated acts of cognition and intelligence and display many of the emotions that we do. Soul-theory holds that we think, feel, etc. with our souls. So, do animals have souls? If we say “no,” then we admit that the brain is sufficient for the mental life of non-human animals. At what level of cognition or consciousness, then, are brains no longer sufficient and why? How do we give a principled, non-arbitrary answer here? If brains can do that much, then why not more?
If, on the other hand, we say that some non-human animals do have souls, then the same problem arises again. Just as Darwin did with the eye, we can point to a continuum of cognitive aptitudes and levels of sentience. As I say:
Animals show a broad range of mental functions from none at all to quite sophisticated competencies. At what point do we say that here, just here is where souls are needed and brains are not enough? With which animals do we say that their mental functions are so sophisticated that they must have souls? Bonobos? Monkeys? Cats? Snakes? Frogs? Oysters? Again, any answer would seem to be arbitrary.
Again, where do we drive the golden spike to show where brains cease to be sufficient and souls are needed? So, whether or not we say that animals need souls, soul-theory looks arbitrary and groundless.
Erasmus’ bravely bites the bullet and says that animals have souls—not just chimps and dogs and cats, but all animals. Insects and oysters have souls. There is no place to drive the golden spike. It is souls all the way down. OK. Consider the sea slug, Aplysia california. The sea slug is a big snail with a few large neurons making it easy to study. Despite its paucity of neurons, the sea slug can learn. It can be conditioned by giving it a painless electrical stimulus at the same time as a painful stimulus. Soon it reacts the same way, by withdrawing its gills and siphon, just when exposed to the painless electrical stimulation. The training causes the release of proteins that cause synapses to open between the slug’s neurons, permitting easier transfer of electrical charges between neurons. The more synapses that open, the longer the conditioning lasts. So, is learning in sea slugs not sufficiently explained in terms of neurons and synapse? Do we have to invoke sea slug souls? To me, this has the air of a reductio ad absurdum.
I conclude that, though Erasmus admirably refrains from invective in criticizing my post, his critique fails comprehensively.
Calvin, William H. How Brains Think: Evolving Intelligence Then and Now. (New York: HarperCollins, 1996).
Churchland, Paul M. The Engine of Reason, the Seat of the Soul: A Philosophical Journey into the Brain. (Cambridge: MIT Press, 1996).
Crick, Francis. The Astonishing Hypothesis: The Scientific Search for the Soul. (New York: Simon & Schuster, 1994).
Eagleman, David. The Brain. (New York: Pantheon Books, 2015)
Flanagan, Owen. The Problem of the Soul: Two Visions of Mind and how to Reconcile Them. New York: Basic Books, 2002).
Lyons, William. Matters of the Mind. (New York: Routledge, 2001).