Overdetermination and the Autonomy of the Mental

The problem of mental causation from the overdetermination argument could possibly be solved by rejecting the premise of physical completeness. I begin by elaborating the problem, give a priori reasons to reject the completeness of physics and examine the resulting problems associated with emergentism. I shall then move on to a posteriori reasons for rejecting the afore mentioned premise from the fields of quantum-electrodynamics (QED) and quantum chemistry (QC). I will conclude that although these reasons are justified, a rejection of the completeness of physics is but the first step to a satisfying account of mental causation.

One of the problems that the non-reductive materialists encounters in maintaining that mental events are irreducible to physical events is that of mental causation. This problem arises out of the inconsistency of five premises: 1) Mental properties are distinct from physical properties, 2) These mental properties can cause physical properties, 3) They furthermore supervene on their physical realisers, i.e. any mental property necessarily has a physical property and whatever has that physical property has the mental property, 4) The completeness of physics, i.e. every physical property is sufficiently caused by a prior physical property, and 5) There can be no two sufficient causes for a property (Kallestrup, 2006: 459, 466; Papineau, 2001; Kim, 1996). How this is accomplished was repeatedly argued by Kim and became known as Descartes’ Revenge since it showed how a similar line of reasoning that brought down substance dualism can be applied to non-reductive materialism (Kim, 1992; 1996; 1997, et cetera): If 1), 2) and 3) are true, then we get a picture of mental causation wherein a mental property M1 can cause changes in its physical realiser P1 as well as the realisers of other mental properties, say P2. However, if 4) is true, then P1 is already a sufficient cause for P2, which means that P2 is overdetermined, i.e. it has two sufficient causes, M1 and P1. This however is forbidden by 5). See above illustration where the dashed lines indicate supervenience and the unbroken causation.

The only way out of this mess is to reject one of its premises. This essay shall attempt to reject 4), the completeness of physics. Let me elaborate on the meaning of this premise: First, we need a conception of the physical, which for simplicity this essay will define as everything non-mental under the assumption that there are only these two kinds. (for a more elaborate definition, see Papineau, 2002) Examples of the physical are brain states and quantum states. What the completeness of physics suggests is a bottom-up explanation of the world, i.e. the standard model of particle physics, that defines all existing matter and forces, predicts quantum mechanics (QM), which in turn can predict all behaviour of molecules in chemistry. The same is true for bio-chemistry and eventually neurology. What this means is that brain states are “nothing more” than, and reducible to, the aggregates of quantum states (Morrison, 2006: 876; Kim, 1997: 279-286). The reason we start with QM is that it is more fundamental and closer to the “source” of higher level explanations (Weinberg, 1987: 437; also Papineau, 2002; Hendry, 2006: 154).

Rejecting completeness would mean that explanatory gaps arise within the physical world. The laws of QM would be unable to predict chemistry, which could not predict neurobiology. If a sufficiently large amount of particles come together, new properties would emerge and kept emerging whenever a system becomes sufficiently more complex. This means that there are things about chemistry that are unpredictable from QM and so forth. This position, let us call it emergentism (see elaborations on this definition in Crane, 2006: 3ff) would help the non-reductive materialist in the following way: If the physical world were a “layered world” (Kallestrup, 2006: 467), then there is room for downward causation. Higher level properties, although causally connected to lower levels, have their own fundamental emergent laws. These “unexplained explainers” (Horgan, 1993: 557-558) also have causal powers that they can downward exercise on their realisers (Crane, 2006: 4). If this image were abstracted to mental properties, it means that M1 can directly cause P2 since P1 is not necessarily a sufficient cause for P2. To give a mental example: “The macro-property of me, my decision […][an M1] affects the micro-property of my arm […][a P2]” (Crane, 2006: 15).

This seems a wee wild! How can a physical property not be sufficiently caused by another physical property? Does that not require some magical power of M1 and M2 that renders the interaction of their realisers unexplainable? In fact the very point of premise 4 was to forbid this kind of downward causation from higher level properties to lower level ones (Hendry, 2006: 154, 156; McLaughlin, 1992; Kim, 1997; Horgan, 1993), and indeed, not too many non-reductive materialists would be willing to come along without premise 4) (Horgan 1993: 560; Crane, 2006: 4).

Let me summarise where we are now: We started with Descartes’ Revenge and noted that a response to it would require to give up one of the premises of non-reductive materialism. We went for the completeness of physics. This led us to emergentism since if higher level systems are not predictable from lower level systems, there must be emergent properties and laws about them. Finally we saw that this solution is not necessarily too attractive since it allows for wild claims about downward causation. Therefore, prima facie materialists are not emergentists and whether our emergentists are still non-reductive materialists is not certain (see Crane, 2006: 4, 10, 22). Now that the scene is set, the next step in defending our rejection of 4) needs to be a posteriori. If we can answer the question whether there are emergent properties in the natural sciences with yes, rejecting the premise would seem reasonable and a solution to the problem of mental causation attainable. Before we start though, let us straighten out two definitions that we will use to test the following examples: Supervenience shall describe a situation where any higher level property change is necessarily accompanied by a lower level property change (Hendry, 2006: 153). Emergentism shall mean that these higher level properties are furthermore not predictable from lower level properties, i.e. if complete knowledge of all lower level properties exists, then the next higher level properties emerge unpredictably (for simplicity, this essay will ignore the nuances between different forms of emergentism, for instance British).

A classic example of a higher level chemical property that is inexplicable in terms of its lower level realisers that was used by the British emergentists around Broad is the transparency of water (Broad, 1925: 70ff; McLaughlin, 1992). A supervenience relation exists since any chemical change (e.g. transparency) is accompanied by physical change in the particles that make up water. Furthermore the particles that make up water could not explain how transparency came about. Yet, McLaughlin uses this as an example of why emergentism failed. He argues that since the discoveries of QM there is no more mystery about transparency of fluids (McLaughlin, 1992). Indeed, QED can give bottom-up explanations and predictions of how photons interact with the electrons around water molecules to extremely high levels of accuracy (Feynman, 1979). Therefore there are no emergent laws in QC (McLaughlin, 1992; Papineau, 2001: 19).

I disagree with McLaughlin about the implications of this example. I propose that it was merely an unfortunate choice of examples that brought about the British emergentists’ downfall so quickly since there are many other instances of possible emergent properties in QC that cannot be dismissed as quickly. Let me examine but a few:

Some physicists describe the formation of crystals as emergent phenomenon as it occurs only in sufficiently large system once enough particles are cooled to certain temperatures and start exhibiting “rigidity, elasticity and regularity” all of which cannot be found in fewer particles (Morrison, 2006: 880). This is an example of symmetry breaking where a small change to a system results in it taking a whole new orderly symmetry. What this means is that little cooling of a system or adding a few particles can result in matter “undergo[ing] mathematically sharp phase transitions to states where the microscopic symmetries and equations are in a sense violated.” (Morrison, 2006: 881). For instance sound wave particles emerge and start behaving according to simple rules independent of the underlying quantum laws (Laughlin, 1999). Furthermore we ought to note that in this emergent system, the constituents have not disappeared, rather the emergent phenomenon would, once the constituents are separated (Morrison, 2006: 883). Sound has no meaning for single particles, only systems.

How does this differ from Broad’s case for emergentism? Let us find out by testing this example against our definitions: If satisfies the supervenience condition since any change to a crystal necessitates a change to its constituents. The question whether this satisfies the emergence condition is more tricky. Are the laws of QM able to predict at which point and how a crystal will come about if more particles are added below as certain temperature? Let me try to answer with another example:

The Schrödinger equation is an example of a quantum law at Weinberg’s “source” of explanation. It extremely accurately describes the charge and mass of all matter in the universe (except radioactivity and gravitational curvatures) (Morrison, 2006: 879; Feynman, 1979). However, it produces very inaccurate results for any more that 10 particles, not because of the calculation abilities of the formula, but because larger dimensions require experimental input to calculate since the formula contains approximates (Laughlin and Pines, 2000). Therefore the physicist is unable to mathematically derive the behaviour of larger systems from quantum laws (Morrison, 2006: 879). This now suits our emergence condition, but more that that, the Schrödinger equation must solve for the individual particles within the framework of a larger system or molecule since they are somewhat “constrained” by the overall structure (Hendry, 2006: 165). The emergentist will justifiably claim that this is an example of downward causation! (For further more complex examples see Bose-Einstein condensates in Healey, forthcoming: 5 and Laughlin and Pines, 2000; and superconductivity in Weinberg, 1986.) An advocate of this conclusion is Anderson who states that at “each level of complexity entirely new properties appear [that] require research […] as fundamental […] as any other” (1972: 393) In a sense, the whole becomes “not only more but very different from the sum of its parts” (Anderson, 1972: 395).

Now we seem to have found a some candidates for empirical emergent properties in physics that fit both of our definitions. However, the physicalist has a compelling way of responding to these: Physics may not seem complete prima facie, i.e. there is no full account of the interaction of QM and chemistry, yet we ought to look for potential future explanations (Papineau, 2001; Broad, 1925: 70). She may argue that full bottom-up predictions are not made for “practicle reasons” (Hendry, 2006: 165), and that eventually this explanatory gap will be closed, just as in the example of Broad and transparency. This is a very difficult critique to respond to, but suffice it to say that unlike Broad, the modern physicist has very subtle mathematical ways of testing the validity of emergent properties in QM, that are unfortunately beyond the scope of this essay (see one-to-one mapping in Morrison, 2006: 885).

This brings us back to where we started. Anderson launches a direct assault on Kim’s “nothing more” position here. We now find ourselves asking to what extend our understanding of higher-level properties comes in terms of their constituent lower-level properties (Morrison, 2006: 879). How does this tie in with our original question? In order to find a solution to Kim’s overdetermination argument I proposed to reject the premise of the completeness of physics and hope to have demonstrated with a posteriori evidence that such a proposal is not unreasonable. However, this can only be the very first step in a revival of emergentism since the rejection of a premise does not yet bring about a satisfying account for mental causation (see Kallestrup, 2006: 468). The next thing to do for the philosopher would be to find such an account in light of this a posteriori evidence and retest modern physics examples against such an account. However, that task shall be left for another post.


Anderson, P. W. 1972. More is Different. Science 177: 393-396.

Broad, C. D. 1925. The Mind and its Place in Nature. Routledge.

Crane, T. 2006. The Significance of Emergence. in Gillett, C. and B. Loewer, eds. Physicalism and Its Discontents.

Feynman, R. 1979. Quantum Electrodynamics. Series of 4 lectures given at the Douglas Robb Memorial Lectures. University of Auckland. Vega Science Trust.

Healey, R. forthcoming. Reduction and Emergence in Bose- Einstein Condensates. Foundations of Physics.

Hendry, R. F. 2006. Is there Downward Causation in Chemistry? in Philosophy of Chemistry: Synthesis of a New Discipline 153-179. Springer.

Horgan, T. 1993. From Supervenience to Superdupervenience: Meeting the Demands of a Material World. Mind 102: 555-586.

Kallestrup, J. 2006. The Causal Exclusion Argument. Philosophical Studies 131(2): 459-85.

Kim, J. 1992. Downward Causation. in Beckerman, A., Flohr, H., and J. Kim, eds. Emergentism and Reduction.

Kim, J. 2006. Mental Causation. in Philosophy of Mind 173-204. Westview.

Kim, J. 1997. Supervenience, Emergence and Realization in the Philosophy of Mind. in Carrier, M. and P. K. Machamer, eds. Mindscapes: Philosophy, Science, and the Mind 271–293. Universitätsverlag Konstanz.

Kim, J. 1999. Making Sense of Emergence. Philosophical Studies 95: 3-36.

Laughlin, R. B. 1999. Nobel Lecture: Fractional Quantization. Reviews of Modern Physics 71: 863-874.

Laughlin, R. B. and D. Pines. 2000. The Theory of Everything. Proceedings of the National Academy of Science 97: 28-31.

Lewis, D. 1973. Counterfactuals. Blackwell Publishing.

McLaughlin, B. 1992. The Rise and Fall of British Emergentism. in Beckerman, A., Flohr, H., and J. Kim, eds. Emergentism and Reduction.

Morrison, M. 2006. Emergence, Reduction and Theoretical Principles: Rethinking Fundamentalism. Philosophy of Science 73: 876-887.

Papineau, D. 2001. The Rise of Physicalism. in Gillett, C. and B. Loewer, eds. Physicalism and Its Discontents.

Papineau, D. 2002. A Case for Materialism. in Thinking about Consciousness. Oxford University Press.

Weinberg, S. 1986. Superconductivity for Particular Theorists. Progress of Theoretical Physics Supplement 86: 43-53.