BOOK REVIEW: Owen Flanagan, The Problem of the Soul: Two Visions of Mind and How to Reconcile Them (New York: Basic Books, 2002). 364 pages. $27.50.
Owen Flanagan (pictured here) is a prominent philosopher of mind who exemplifies the trend in his profession of paying closer attention to psychology and neuroscience. An engaging and (for the most part) accessible writer, he does not believe in confining his efforts to the academy; indeed, he has published several books aimed at a wider audience.
In The Problem of the Soul, his mission is to promote the "scientific image" of human beings. In general terms, Flanagan is a modernist: a proponent of reason, science, naturalism, and an ethics of human flourishing. Despite some irritating concessions to academic political correctness (he employs a generic "she" throughout most of the book and routinely deplores "racism" and "sexism" without bothering to provide examples of either), Flanagan shows little regard for cultural relativism, and none at all for the postmodern dismissal of science as a mere "hegemonic discourse." Meanwhile, his rejection of premodern conceptions, particularly those based on Christian theology, is sharp and uncompromising: "Supernatural concepts have no philosophical warrant" (p. xiii).
According to Flanagan, "no scientifically minded person believes that we will need resources beyond those available to genetics, biology, psychology, neuroscience, anthropology, sociology, history, economics, political science, and naturalistic philosophy to understand the nature of persons" (p. 7)—specifically, human conscious experience, self-understanding, and moral decision-making. But what do these resources consist of? Unfortunately, Flanagan believes that the explanations provided by these sciences will have to be reductionistic. Ultimately, if he is correct, everything that happens at the mental or the psychological level will have to be explained in a "bottom-up" fashion, as a consequence of interactions among microphysical particles.
His commitments take some time to unearth, because Flanagan is so vague and wavering about the metaphysics of knowledge and mental processes. What do mental processes comprise? How do they relate to the firing of nerve cells in the brain, or to the transmission of various chemical compounds across synaptic gaps between those cells? How, in turn, do such neural processes relate to the interactions of electrons, protons, and other subatomic particles in the atoms that make up the membranes of those nerve cells? Flanagan provides no clear answer to such questions. Just as he skips around among various broadly defined views of human cognition—such as computationalism and dynamic systems, which are seen by most experts as fundamentally inconsistent with each other—he skips around among different views of the relationship between psychology and biology, or mind and brain.
On the one hand, Flanagan declares himself to be a "physicalist." He maintains that when you are shown a blue paint chip, "the activity in the blue-detection area [of your brain] is the experience of blue.… Different segments of your nervous system being activated in different ways by different color chips is all there is to your color experiences" (p. 87). The way Flanagan puts it, physicalism just means that "each and every event, each and every experience, is some physical event or other—presumably some central nervous system event" (p. 87). But physicalism, as usually understood, brings with it a far greater commitment to reductionism than Flanagan initially lets on. Physicalism doesn't just equate the experience of blue with activity in a blue-detection area of the brain; it goes on to equate the activity of that blue-detection area with an extremely complex configuration of interactions among microphysical particles. Once Flanagan has climbed aboard the Reductionist Express, he can't leave the train until it has taken him all the way down to quarks and gluons.
On the other hand, Flanagan insists that he is a nonreductive physicalist. He even claims to endorse the view that qualitatively new biological and psychological properties emerge from physical properties: "[M]ost everyone now believes that there are emergent natural properties that, despite being obedient to the laws of physics, are not reducible to physics" (p. 217). But the idea of emergent properties is consistent with both top-down causality (for instance: because you have adopted a more positive attitude, electrical activity now increases in your left prefrontal cortex) and bottom-up causality (because electrical activity has increased in your left prefrontal cortex, your attitude now becomes more positive).
The view that mental processes are real but that there is no top-down causality, and therefore less than full emergence, is called epiphenomenalism. Epiphenomenalists believe that mental processes are mere byproducts of physical processes. Flanagan says that he repudiates this view: "In the epiphenomenalist's picture, conscious thought plays no role in the execution of any act. It only seems to.... [I]t is what [William] James called an 'unwarrantable impertinence' to think that our conscious thoughts normally play no important causal role in what we choose to do" (p. 118). In other words, Flanagan seems to be saying that our conscious thoughts about the wind reaching gale force this afternoon, and gale-force winds being dangerous, can cause us to decide not to take our sailboat out on the lake. An epiphenomenalist would be required to say that some pattern of interaction among microphysical particles causes the pattern of electrical and chemical activity in nerve cells in our brains that, in turn, causes the conscious thoughts about the wind being dangerous. And the first pattern of microphysical activity causes a second pattern of microphysical activity that causes the further patterns of neural activity that cause the decision not to sail. From the epiphenomenalist's standpoint, the conscious thoughts couldn't cause the choice. The patterns of activity in various parts of the brain couldn't cause it either—only interactions among microphysical particles could.
In contrast to his stated rejection of epiphenomenalism, Flanagan comes down squarely on the side of that view in a comment on post-traumatic stress disorder in soldiers who fought in World War II. Here is how he explains their disorientation and emotional disturbances: "The terrible experiences of war cause brain changes, which cause mental changes, which cause changes in thought, motivation, and behavior.... [T]he scientific image that regulates psychology assumes 'supervenience' of the mental on the physical, that is, that any change at the level of experience is due to a change at the level of the brain" (p. 255, n. 11, emphasis added). Since he is not talking about brain tumors or head wounds, but about seeing and hearing bombs and bullets and people being killed, Flanagan is now insisting that all causal relationships are of the bottom-up variety. In fact, supervenience is a weak form of emergence that does not allow top-down causation. If Flanagan accepts the supervenience of the mental on the physical, he is stuck with the conclusion that mental processes are epiphenomenal; indeed, he will have to conclude that biological and chemical processes are epiphenomenal too.
Flanagan presumes, then, that scientific explanations of consciousness, the self, and decision-making must be reductionistic. So, when our everyday understanding or "manifest image" of ourselves contradicts reductionism—for instance, when it indicates that we could have chosen differently under the very same circumstances—it must be relying on a theological or supernatural conception of the soul. Flanagan promises to go fairly easy on the manifest image: "the scientific image can give us pretty much everything we can sensibly want from the concept of a person. Most of what we traditionally believe about the nature of persons remains in place even without the unnecessary philosophical concepts of the soul and its accompanying suite" (p. xiii). Most stays—but free will must go.
Free will bothers Flanagan so much that his writing on the subject turns uncharacteristically militant. He could have argued on epiphenomenalist grounds that there is no such thing as "libertarian" free will. But he moves less directly owing to his reluctance to own up to every consequence of reductionism, his unsureness about the kinds of causal relationships that would have to apply to conscious human decisions, and his inability to supply a detailed model of the way any human decision gets made. He argues instead that any traditional belief in free choices (choices that a person could have made differently, at the same time and under the same circumstances) requires a belief in agent causation: the whole person, an entity with certain high-level capabilities, made that choice and could have made it differently. But such a belief in agent causation is unsustainable because it is necessarily theological; agent causation would require each person to possess Godlike powers. Flanagan's entire understanding of agent causation is expressed in a single quotation from philosopher Roderick Chisholm: "[I]f we are responsible...then we have a prerogative which some would attribute only to God: each of us when we act is a prime mover unmoved. In doing what we do, we cause certain things to happen, and nothing—or no one—causes us to cause those events to happen" (p. 58). He returns to this statement over and over again.
Flanagan seems to share the outlook of the seventeenth-century philosophers whose views he ably summarizes: Human choices are just a complex pattern of matter in motion, following deterministic laws, or else they are the inscrutably mysterious doings of an immortal soul. Emergence was not seen as a live option in the 1600s. But if the psychological emerges from the biological, and the biological emerges from the physical, and systems at the higher levels can have causal powers, there is a simple and obvious alternative to Flanagan's dichotomy. If there is emergence, agent causation quits being mysterious; it is merely top-down causation by a human being. If personhood emerges via the organization of various mental processes, and mental processes emerge via the organization of processes in the brain and central nervous system, the whole person should be able to cause various mental processes, just as various mental processes ought to be able to cause changes in the biology of the brain.
Flanagan's own position, which he calls "neo-compatibilism," is that voluntary human action can be explained in a manner consistent with a reductionist worldview, but there is no free will as people usually understand it. Flanagan uses the "neo" prefix because he recognizes that calling this position "compatibilism," as is usually done, would be misleading. "[M]y view…does not in fact agree that the traditional concept of free will is compatible with determinism…some suitably naturalized conception of human agency preserves some, but not all, of what is worth preserving in the traditional concept of free will" (p. 127, n. 6).
Much more would be required for an in-depth response to Flanagan's neo-compatibilism. But a reader might ask what point there could be to Flanagan's exhortation to philosophy professors to challenge their students' religious beliefs: "If it is not good to believe what is unreasonable or false, we do our students no favor in allowing them to think they are using their minds well if they choose to believe what is false or unreasonable" (p. 23). For, from a neo-compatibilist point of view, whatever the students choose to believe is what they had to choose to believe in that context; the most that can be said is that under different circumstances they might have chosen to believe differently. But there is no need to travel down this road here. Flanagan's strategy of argument for neo-compatibilism is fatally defective. It simply fails to consider all of the alternatives.
In discussing the self, Flanagan takes a much broader view of what is consistent with the scientific image than he does in the case of consciousness and free will. Yet here again his argument against our "manifest image" of the self depends on a false alternative. His target, in two long chapters about the self, is a belief in a permanent, unchanging, indivisible, immaterial ego that will outlive the body. His arguments will indeed give trouble to those who believe in immortal souls on religious grounds. But a wide range of naturalistic, non-reductionistic accounts of the self will survive Flanagan's challenges. Basically, there are two philosophers whose views of the self must be rigorously avoided.
Descartes's conception of mind as a different substance from matter is, as Flanagan notes, a supernaturalist holdover. And so is the Cartesian insistence that the human mind can know itself earlier and more easily than it can know anything in the external world. Meanwhile, Hume's skeptical treatment of the self as something that cannot be known through introspection (because introspection just collects inner sensations) has to be rejected in favor of a view that allows the self to be known by reflecting and abstracting. And there needs to be an acknowledgment that the self develops—a mature human self has emerged in stages that began back in infancy when there was no self-knowledge yet.
Flanagan's understanding of the self and personhood, like his understanding of cognition, sometimes proves difficult to pin down. But the Cartesian view is too supernaturalistic to tempt him. The Humean view is unacceptable to Flanagan because we really can reflect and abstract about our mental processes: "The posit of a self, or better, a person makes sense if we think of mindfulness, introspection, or phenomenology as an activity performed by a changing system—a continuous and for now relatively self-contained changing system—on itself" (p. 210). Both Humeanism and postmodernism must be rejected because they would turn the self into a fiction: "Self-construction is constrained by reality: the things we have done, what we have been through as embodied beings, and the characteristic dispositions we reveal in our personal and social lives. Others will catch us if we take our stories too far afield, and we may catch ourselves" (p. 252).
Flanagan even makes a brief effort to incorporate developmental psychology into his theory (pp. 217-21), although he loses interest in the details of development after early childhood. In any event, many of the theories of the self proposed by today's developmental psychologists avoid the Cartesian and Humean traps; what's more, I believe they could be shown to address the specifics of development better than Flanagan's account does. Indeed, a theory of the self that allows for emergence, top-down causation, and agent causality would be able to make sense of remarks like the following: "[I]t would be best to say, 'the person himself is the thinker.' A person is what thinks—a whole embodied being with a history" (p. 225). Flanagan's rhetoric notwithstanding, one can easily reject his reductionism without becoming a "soulophile."
One would also have to be caught up in theology, or in an ethic of duty, to take fundamental exception to Flanagan's treatment of ethics. In its broad outlines it would pass muster with ancient non-reductionists, like Aristotle, and modern ones, like Ayn Rand: "Ethics is the normative science that studies the objective conditions that lead to [the] flourishing of persons" (p. 17). A detailed treatment of Flanagan's view of morality would need to ask why, instead of acknowledging that some form of ethical eudæmonism—Aristotelian or Epicurean or Stoic or Objectivist—best fits his definition, he sanitizes other major ethical traditions to make them fit. For instance, he declares that St. Augustine was "an exemplary moral person" (p. 58)—such minor details as believing in Original Sin or advocating persecution are airbrushed out. And if we believe Flanagan, Kant's conception of duty is really just another way of explicating what is really good for people, instead of what Kant thought it was: an effort to exclude considerations about what is good for human beings from the moral realm. When Flanagan repeats his basic formulation, he weakens it by adding that "[w]e will have moral impulses to care for nature as such," and that these will stem from an understanding that "the well-being of nature is an intrinsic good," so that human well-being is not the "beginning and end-all of ethical inquiry" (p. 267).
The reader might also raise doubts about moral responsibility left over from Flanagan's attack on free will. Flanagan says close to nothing about the differences in moral outlook that would follow from subscribing to neo-compatibilism. He suggests that, if we accepted neo-compatibilism, we would all become more forgiving toward criminals who were brought up in bad homes by abusive parents. But it would make more sense, if neo-compatibilism were true, for us to accept that people are just going to believe what they, as particular individuals in particular circumstances, have to believe: some will be more compassionate toward criminals, while others will want to lock them all up and throw away the key.
In the end, the book is signally ineffective as a defense of naturalism. Flanagan has failed to understand what naturalism is really about, or what it can really do, because he has failed to cope with emergence. (Indeed, since evolution is emergence over long spans of time, I would argue that he has failed to come to grips with the evolution of minds and cognitive capabilities, though I cannot develop that point here.) If genuine emergence is allowed in "mind science," then what is going on mentally in human beings can cause what is going on biologically, and such top-down causation will include a form of agent causality that requires no soul and no God-like powers. If genuine emergence is not allowed in mind science, then there will be no agent causation, and mental processes, conscious awareness, and reflection will all be epiphenomenal. But if epiphenomenalism is true, the "scientific image" will be frankly incompatible with the "manifest image," which will have to be scrapped in its entirety. We should not forget that Flanagan celebrates such enemies of the "manifest image" as Sigmund Freud and B. F. Skinner as champions of naturalism.
Fortunately, one can be a naturalist regarding the human mind without falling into the reductionistic trap. Ayn Rand 's compliment to Aristotle gave the issues a memorable presentation: "For Aristotle, life is not an inexplicable, supernatural mystery, but a fact of nature. And consciousness is a natural attribute of certain living entities, their natural power, their specific mode of action—not an unaccountable element in a mechanistic universe, to be explained away somehow in terms of inanimate matter, nor a mystic miracle incompatible with physical reality, to be attributed to some occult source in another dimension" (The Objectivist Newsletter, May 1963, p. 19).
Emergence will need to be part of any successful naturalistic understanding of mental processes, and plenty of work remains to be done along those lines. Meanwhile, Flanagan's attachment to reductionism keeps him off the path that researchers will need to follow to solve the problem of the soul. It is not Flanagan's commitment to modernism that defeats him, but the difficulty of the questions that he is taking on and his failure to recognize non-reductionistic ways of answering them.
Robert Campbell is a professor of psychology at Clemson University.
This review first appeared in the June 2004 issue of The Atlas Society's magazine, Navigator. (Navigator was renamed The New Individualist in 2005).