The Soul Hypothesis is a collection of 9 essays from authors of diverse backgrounds (i.e., philosophy of mind, philosophy of physics, brain science, linguistics). While this sort of eclecticism often leads to very uneven work, this is most definitely not true in this case. The editors and authors do an excellent job of maintaining continuity among the essays, primarily by providing ‘bridging’ commentaries between the chapters which allow for the reader to see the emerging themes. In attempting to replicate this thematic consistency, I’ve been forced to discuss the book’s contents in an ‘out of order’ fashion.[i]
Themes. It will be helpful to first sketch the two main themes of the book. The first theme is that there is no one soul hypothesis (henceforth SH).[ii] This means that falsifying one version of the SH does not entail that no version of the SH is true. [iii] This also means that the extent to which “the” SH is judged plausible will vary according to which specific SH is being proffered. Likewise, there is no single version of the materialist hypothesis (MH) [iv]. Arguably, as with the SH, this means that falsifying a version of the MH does not entail that no version of the MH is true, nor that all versions of the MH will be judged equally (im)plausible. Despite this caveat, I will for simplicity of exposition use ‘the’ SH or ‘the’ MH language, and relegate finer-grained discussion to the last footnote of this review. However, at various points in the text the authors argue against the ‘softer’ forms of materialism as well.[v]
Epistemic Criteria: What They Are, and How the MH and SH Fare. Here are the epistemic criteria. First, explanatory power: to what extent do the hypotheses explain the data? Second, simplicity: to what extent do the hypotheses explain data in a simple and elegant way? Third, to what extent are the hypotheses predictively fertile? Put differently, is the concept such that we can test it in a variety of illuminating ways?[vi]
In terms of explanatory power, a variety of strategies are pursued. These can be characterized as defensive and offensive. Defensive maneuvers involve rebutting allegations that the SH is inconsistent with known facts. Offensive maneuvers involve demonstrations that the SH fits the data better than the MH. In Chapter 3, Stewart Goetz rebuts arguments from causal closure and methodological naturalism by pointing out that observations of a neuron (N1) being activated by another neuron (N2) do nothing whatsoever to entail that N1 can never, in principle, be activated by something immaterial (i.e., a soul). He also points out that the neuroscientist Penfield adopted the SH on the basis of (not in spite of) his brain research. In a similar vein, Daniel Robinson’s essay in Chapter 3 argues that the correlational nature of the brain sciences makes it difficult (if not impossible) to determine which direction the causal relationship goes. Is it that some neural event (NE) causes some mental event (ME), or the other way around? In Chapter 3, Mark Baker argues that the flexible and creative way in which humans produce and understand language is arguably more consistent with the SH than the MH.[vii] In Chapter 5, Robin Collins rebuts arguments against mind/body interaction which draw upon the principle of the conservation of energy (CE). CE-based arguments propose first that energy can be neither created nor destroyed within a closed system, and that causation can only happen with exchanges of energy across a physical medium. Drawing upon findings from relativity theory and quantum mechanics, Collins shows that CE is not a universal principle even when it comes to matter. Why think, then, that CE would hold for causal relationships between an immaterial mind or soul and a physical body? In Chapter 1, Charles Taliaferro tackles an extreme form of the MH (eliminativism) and argues that the idealized ‘third person’ perspective of science presupposes the to-be-eliminated first-person perspective inherent in the SH.[viii] Drawing upon the authority of science to deny the SH, therefore, seems self-defeating.[ix] In Chapter 7, Dean Zimmerman argues that the vagueness of composite objects like mountains and brains is hard to fit with the precise ways in which the laws of nature link ‘simpler’ (i.e., non-composite) objects to each other—but that the SH has less problem in explaining precise links between percepts and an immaterial mind, as the soul is not an imprecise, composite object. A similar tact is taken in Chapter 6 by Hans Halvorson, in what is perhaps the most difficult essay of the bunch. Put in the most succinct way possible, it appears that if quantum mechanics (QM) is true, then there would be no believers in quantum mechanics![x] In Chapter 8 William Hasker argues that a version of the SH can do a better job than the MH in explaining the mental life, but also avoid some of the problems with Cartesian dualism.[xi]
In terms of simplicity, it might be anticipated that the MH has a leg up here. After all, the MH is not positing a different kind of entity (all things are material) whereas the SH obviously does. It after all requires that there be an immaterial mind. This does not of course settle the matter. As Robin Collins points out in Chapter 9, if simplicity was the sole epistemic criterion, then we could just stop with any given set of observations and not posit any further entities. Citing the example of atomic theory, however, he points out that in science (and everyday reasoning) we feel justified in hypothesizing further entities if there is sufficient explanatory power in doing so. Given the previous lengthy paragraph, it is unsurprising that Collins feels that any gain in simplicity for the MH is more than offset by its low explanatory power. But Collins does not stop there. He also feels that the MH, to be credible, will in effect wind up postulating something very much like an additional entity, and thus is on a par with the SH in that respect.[xii] Further, recall that simplicity is not just a matter of the numbers and types of entities posited. Complexity also arises in terms of the laws of nature which are posited to explain cause and effect. Collins argues that positing a carefully crafted version of the SH allows for simpler laws than plausible versions of the MH will.[xiii]
In terms of predictive fertility, the comments sprinkled throughout give only a hint. When they do, they seem to be more geared not so much to determining which is true (the SH or the MH)[xiv], but rather as to how to make meaningful (read: testable) predictions about which version of the SH is likely true.[xv] This is perhaps inevitable, given the clear conviction of the authors that the MH is so low in explanatory power. Nonetheless, the more specified a SH becomes in terms of what functions it does and does not become involved in, the more testable the hypothesis will be.
Summary. This was one of the most rewarding books I’ve ever read. In decreasing order of confidence, I now am convinced that (1) some version of the SH is true (2) that many of the objections levied against the SH are based on misunderstandings of science[xvi] (3) that the philosophical implications of scientific findings are clearly understood by scientists (4) that the SH may in fact have an advantage not only in terms of explanatory power but also simplicity and (5) that much of the animus against the SH is philosophical, not scientific, in nature.[xvii] This is an absolute must-read for anyone interested in philosophy of mind.
Apologetics 315 Book Reviewer Latter Day Inkling is a U.S.-based research psychologist for the military. He is especially interested in epistemology and natural theology.
[i] For example, the three epistemic criteria I list are only made explicit in the final chapter of the book. It seemed clear, however, that all of the earlier chapters could be evaluated in terms of those criteria.
[ii] Speaking loosely, the hypothesis continuum can be conceived as follows: at one end is reductive materialism, at the other end is Cartesian dualism. In between will be non-reductive (i.e., property) dualism, emergent dualism, and (perhaps) what Collins in Chapter 9 terms the ‘dual aspect soul model’.
[iii] “If a simple or naïve version of the soul hypothesis is falsified by scientific results, then that might mean that dualism is false. But it might also mean that it is time to move to a richer, more thoughtful version of the soul hypothesis. This courtesy is routinely extended to other kinds of hypotheses….”—Location 210, Kindle edition
[iv] “The apparent [materialist] consensus could be something of a mirage, with the only thing holding it together being a denial of the Soul Hypothesis. If so, it begins to look more like a shared assumption than a shared discovery.”-Location 121, Kindle edition
[v] For example, in Chapter 7 Dean Zimmerman argues that raising the plausibility of property dualism will tend to raise the plausibility of substance dualism as well. Further, in Chapter 9 Robin Collins argues that while adopting property dualism does a better job of meeting one of the epistemic criteria (explanatory power) than does reductive materialism, it does a poorer job of meeting another of the epistemic criteria (simplicity) than does a version of the SH. See paragraph 5 of this review for more detail.
[vi] All of these criteria will be familiar to readers of Richard Swinburne, especially the first two. See his Simplicity as Evidence of Truth and Chapter 4 of his Epistemic Justification.
[vii] Interestingly enough, this aspect of language use has historically attributed to the soul by various Cartesians.
[viii] “The first question to raise in response to the above radical materialist proposal is whether one can make any sense of the ‘third-person’ point of view at work in science (and in ordinary, nonscientific reflection) without there being a first-person perspective of self-aware, conscious subjects. Dennett claims to be more certain about mass, charge, and space-time than he is of experience, but how might we have any idea at all of mass and charge or any science at all unless there are scientists who have experiences of the world and can reason about those experiences?”-Location 456, Kindle edition
[ix] Put differently, science presupposes measurement, which presupposes observation, which presupposes qualia-which are true only from a first person perspective.
[x] Put a bit less succinctly, Halvorson argues that (1) QM entails that no material objects have determinate (binary, yes/no) states. This is true of both simple objects as well as composite objects. (2) Any observers of QM become ‘causally entangled’ and thus will fall under the purview of QM. (3) On the MH, all aspects of observers are physical. (4) From (1) and (3) it follows that no aspect of observers, including their beliefs, can be determinate. So (5) There can be no determinate beliefs, like QM is true (or QM is false)! What we have here is something like a weird version of Alvin Plantinga’s EAAN. Instead of P (R|N&E) is low, we have P (B|QM&M) is low, where B=determinate beliefs, QM=quantum mechanics, and M=materialism. Halvorson takes pains to note that there are other ways out of this-one can, for example, deny one or more of the postulates of QM that lead to (1) above. But each of these come with their own set of problems. His major point is that the MH appears to conflict with observations (that QM is true and that somehow there are believers in QM). However, a proponent of the SH is prima facie not presented with this contradiction.
[xi] Hasker argues that Cartesian dualism proposes too vast an ontological gulf between animals and man, and that his emergent (read: gradual) dualism does a better job of gelling with the theory of evolution.
[xii] This gets complicated, and I will do my best to clearly convey the gist of the argument here. Collins argues that reductive materialism doesn’t do the trick explanatorily, and that to be credible one must adopt some non-reductive form of materialism (like property dualism) along with ‘linking laws’ that specify what brain states give rise to what qualia states. The most fundamental linking law, however, must provide a dividing line between those complex physical composites which have consciousness from those that do not. In doing so, the nonreductive materialist is positing something very much like an additional entity-a material entity that has these subjective properties unlike all the other material entities which do not.
[xiii] Again, subtleties abound. Collins argues that reductive materialism fails explanatorily, and thus one is driven to nonreductive (property) dualism to explain the data. He then mounts a (to my mind) plausible argument as to why the laws linking physical states to subjective states would have to be unbelievably (perhaps infinitely) complex. However, by positing what he terms ‘dual aspect soul model’ dualism (that is, assigning both non-subjective and subjective properties to the soul) the linking laws can be mathematically specified and thus be simple, with qualia states arising through reducible, simple combinations of parameters in the brain states and the non-subjective states of the soul. He illustrates this helpfully through string theory and ‘brane’ discussions.
[xiv] For an exception to this, see Chapter 1 where Charles Taliaferro sketches an experiment which might help adjudicate between the MH and the SH.
[xv] See Chapter 9 by Robin Collins. In addition, some comments by Mark Baker (Chapter 3) involve the observation that many aspects of language use might not be total ‘either/or’ in terms of brain and/or mind involvement. Put differently, take three proposed aspects of language use. Call them A1, A2, and A3. Perhaps the SH explains all of A1, ½ of A2, and none of A3. A version of the SH which fits better with those percentages will, presumably, be more plausibly true than a version of the SH that does not fit those percentages as well.
[xvi] I can’t help but wonder if we really have a clear grasp of what matter is, much less mind!
[xvii] See footnote iv. Also, in the Afterword Charles Taliaferro and Mark Baker discuss the relationship between theism (T) and the SH. One of them finds SH plausible, and thus finds T plausible. The other editor moves from belief in T to belief in TH. They note that one can logically embrace one without the other—the relationship between T and the SH is much weaker than mutual entailment. Nonetheless the two are seen as inextricably linked. For a revealing and amusing example of this, listen to the linked interview between Julian Baggini and Richard Swinburne on this exact point: (Most salient portion is between the 8 minute and 9 minute mark, most especially Swinburne’s response around 8:40): http://www.microphilosophy.net/podcasts/mp001swinburne.mp3