Causal Closure of the Physical Domain

views updated


The thesis that physics is causally closed asserts that:

Every physical effect has a sufficient immediate physical cause, insofar as it has a sufficient immediate cause at all.

If this thesis is true, it distinguishes physics from all other subject domains. The biological realm is not causally closed, for example, because biological effects often have nonbiological causes, as when the impact of a meteorite precipitated the extinction of the dinosaurs. Again, meteorology is not causally closed: The burning of carbon fuelsa nonmeteorological eventis causing global warming. Nor, importantly, is the mental realm causally closed: A mental pain can be caused by sitting on a physical thumbtack, and a train of thought can be interrupted by a loud noise.

Physics, by contrast, does seem to be causally closed. If one considers any physical effect, then there will arguably always be some prior physical cause: People expect to be able to account for physical effects without leaving the physical realm itself. In particular, this seems to hold even for physical effects that take place within the bodies of conscious beings. When the muscle fibers in a person's arm contract, this is presumably due to electrochemical activity in the nerves, which is due to prior physical activity in the person's motor cortex, and so on. In principle, it would seem possible to account for this entire sequence solely in terms of the resources offered by physics itself, and without making any essential appeal to any other subject matter.

At first pass, the causal closure of physics is solely a claim about how things go within physics itself. It does not assert that everything is physical, only that everything physical that has a cause has a physical cause. As such, it does not rule out realms of reality that are distinct from the physical realm. It is entirely consistent with the causal closure of physics itself; there should be nonphysical realms that operate independently. The closure of the physical says only that within the physical realm every physical effect has a physical cause.

Even so, the causal closure of the physical does give rise to a powerful argument for reducing many prima facie nonphysical realms to physics: It suggests that anything that has a causal impact on the physical realm must itself be physical. The reason is that the causal closure of the physical seems to leave no room for anything nonphysical to make a causal difference to the physical realm, because it specifies that every physical effect already has a physical cause.

Intuitively, of course, people take it that many prima facie nonphysical events, such as biological, meteorological, and mental events, do have physical effects. A burrowing animal can dig a hole in the ground; a hurricane can destroy houses; one's current thoughts can give rise to patterns on a computer screen. However, the causal closure of the physical says that these effects already have physical causes. So it seems that the only way to respect the causal efficacy of realms such as the biological, meteorological, and mental is to conclude that they are not distinct from the physical after all. If one wants to maintain that the animal's burrowing, the hurricane, and a person's thoughts have physical effects, then apparently there is a need to identify these processes with the physical causes that their physical effects are already known to have.

Note that this line of reasoning only argues for a reduction to physics of those realms that do have physical effects. The causal closure of the physical provides no argument against the possibility of nonphysical realms that lack any physical effects. For example, it is arguable that mathematical, moral, and modal facts have no physical effects. If this is right, then the causal closure of the physical offers no reason to collapse these realms into the physical. (Of course, there may be other arguments against the possibility of such nonphysical realms of reality, such as their epistemological inaccessibility, but that is a different matter.)

The remainder of this entry contains three sections: First, a discussion of the evidence for the causal closure of physics from a historical perspective; second, a consideration of how the thesis can be made properly precise; and finally, an examination of the details of the argument that causal closure implies physicalism about the mental and similar realms.

A Historical Perspective on the Evidence for the Causal Closure of Physics

Why should one believe the causal closure of physics (which for the moment shall be regarded as the simple claim that every physical effect has a sufficient physical cause)? If this thesis is true, it is not an a priori matter, but something that follows from the findings of science. But exactly which findings? What part of science, if any, argues that physical is causally closed?

At first sight it may seem that causal closure follows from the presence of conservation laws in physics: If there are laws specifying that important physical quantities stay constant over time, does not this show that the later values of physical quantities must be determined by earlier values? However, it depends what conservation laws one has. Not just any set of physical conservation laws rule out the possibility of nonphysical causes for physical effects.

Thus consider Descartes's early seventeenth-century physics. This was based on the conservation of amount of motion, which Descartes took to be the product of the masses of all bodies by their scalar speeds. (So amount of motion is different from momentum, which is the product of mass by vectorial velocity : A car going round a bend at a constant speed conserves amount of motion but not momentum.) As Leibniz observed, Descartes's conservation of amount of motion alone leaves plenty of room for nonphysical causes to intrude on the physical realm. In particular, if mental causes (operating in the pineal gland?) cause particles of matter to change their direction (but not their speed), this would not in any way violate the conservation of amount of motion.

Descartes's physics might allow an independent mind to affect the brain, but Descartes's physics is wrong, as Leibniz further observed. Leibniz himself replaced Descartes's law of the conservation of "motion" with the two modern laws of conservation of (vectorial) momentum and of (scalar) kinetic energy, and thereby arrived at what are now regarded as the correct laws governing impacts. Leibniz's physics, unlike Descartes's, did indeed imply that the later values of all physical quantities are determined by their earlier values, and therewith the causal closure of the physical. However, Leibniz did not draw the modern physicalist conclusion that the mind must therefore be identical to the brain. Because it seemed incontrovertible to him that mind and brain must be ontologically separate, he instead inferred from the causal closure of physical that the mind in fact has no causal impact on the physical world. (It only appears to do so because of the "preestablished harmony" with which God has arranged both the mental and physical worlds.)

Whereas Leibniz's physics implies the causal closure of the physical, this is not true of the Newtonian system of physics that replaced it at the end of the seventeenth century. The crucial difference is that, where Leibniz upheld the central principle of the "mechanical philosophy" and maintained that all changes of velocity are due to impacts between material particles, Newton allowed that accelerations can also be caused by disembodied forces, such as the force of gravity. Moreover, Newton's system was open-ended about the range of different forces that existed. In addition to gravity, Newton and his followers came to recognize magnetic forces, chemical forces, and forces of adhesionand indeed vital and mental forces, which arose specifically in living bodies and sentient beings. If one counts vital and mental forces as nonphysical (and this point will be revisited in the next section), then the admission of such forces undermines the causal closure of the physical. For it means that physical effects, in the form of accelerations of particles of matter, will sometimes be due to the operation of nonphysical vital or mental causes.

Newtonian physics has its own conservation principles, but unlike Leibniz's they do not uphold the causal closure of the physical. Crucially, Newton's physics differs from Leibniz's in the way the conservation of energy must be understood. The existence of Newtonian forces means that Leibniz's conservation of kinetic energy no longer holds true; for example, two bodies receding from each other will slow down due to their mutual gravitational attraction, and so lose kinetic energy. Newtonian conservation additionally needs the notion of potential energy: the latent energy stored when bodies are "in tension" in force fields, as when two receding gravitating bodies cease to move apart and are about to accelerate together again. The notion of potential energy was not prominent in early Newtonian physics, but by the middle of the nineteenth century physicists concluded that all forces operated so as to conserve the sum of potential and kinetic energyany loss of kinetic energy would mean a rise in potential energy, and vice versa.

This emergence of the modern version of the "conservation of energy" placed strong restrictions on what kinds of forces can exist, but it by no means ruled out vital and mental forces. Provided that the fields of these forces stored in latent form any losses of kinetic energy they occasioned (consider by way of comparison the notion of "nervous energy"), their presence would be perfectly consistent with the conservation of kinetic plus potential energy. True, the conservation of kinetic plus potential energy did apparently imply that all forces must be governed by deterministic force laws (otherwise what would ensure that they always paid back any kinetic energy they borrowed?), and this greatly exercised many Victorian thinkers, especially given that nothing in early Newtonian physics had ruled out spontaneously arising mental forces. But, even so, the Newtonian conservation of energy did not stop deterministic vital and mental forces affecting the physical realm.

Nevertheless, during the late nineteenth and twentieth centuries an increasing number of scientists have come to doubt the existence of vital and mental forces. The most significant evidence seems to have come from physiology and molecular biology, rather than from physics itself. During this period a great deal has come to be known about the workings of biological systems (including brains), and there has been no indication that anything other than basic physical forces are needed to account for their operation. In particular the twentieth century has seen an explosion of knowledge about processes occurring within cells, and here too there is no evidence that this involves anything other than familiar physical chemistry. The result has been that the overwhelming majority of scientists now reject vital and mental forces, and accept the causal closure of the physical realm.

The Causal Closure Thesis Refined

Much recent discussion of the causal closure thesis has revolved around the question of exactly how "physical" should be understood in the claim that every physical effect has a physical cause. As Carl Hempel originally observed, advocates of the causal closure thesis seem to face a dilemma. On the one hand, they can equate "physical" with the category of phenomena recognized by current physical theory. But then it seems implausible that "physics" in this sense is closed; past form suggests strongly that physics will in time come to posit various new fundamental causal categories. Alternatively, advocates of causal closure might wish to equate "physical" with the ontology of some ideal future physics. But then it is hard to see how the causal closure of the "physical" could have any current philosophical significance, given that people are as yet ignorant of exactly what this "physical" includes.

However, this dilemma is by no means inescapable. True, neither current physics nor ideal future physics gives us a suitable notion of "physics" for framing the causal closure thesis. But this does not mean there are not other suitable notions of "physics." Indeed there are arguably a number of different ways of understanding "physics" that will yield a well-evidenced and contentful causal closure thesis.

For a start, one could simply define physical as "neither essentially mental nor biological." This understanding of "physical" was in effect assumed at the end of the last section, in the argument that the nonexistence of vital or mental forces establishes the causal closure of physics. Note that nothing in that argument assumed a definitive list of fundamental physical categories; rather the thought was simply that this list would not include any sui generis mental or vital entities. This is a relatively inclusive understanding of "physical"; it counts as a "physical" cause anything that is not mental or vital, and to this extent renders the causal closure of the physical a relatively weak thesis. But even so it remains a thesis of much philosophical interest, because it still argues that any mental or vital causes of physical effects must be identical to causes that can be identified without using mental or vital categories.

A rather stronger reading of "physical" would take it to cover any categories of the same general kind as are recognized by current physical theory. Now the list of fundamental "physical" categories will be taken to include not just anything nonvital or nonmental, but more specifically only items that display the same kind of spatio-temporal pervasiveness and simple mathematical characterizability as those assumed in contemporary physics. Again, there seems good reason to suppose that "physics" in this sense is casually closed, and therefore that anything that in this sense has "physical" effects must itself be "physical."

Finally, and even more specifically, there is the option of equating "physical" with microscopic. Modern physical theory characteristically operates at a level of microscopic spatiotemporal detail. Correspondingly, it is plausible that every microscopic effect can be accounted for by (a combination) of microscopic causes. This version of the causal closure thesis thus argues that anything that has microscopic effects must itself be identical to (a combination) of microscopic causes.

The remainer of this section deals with some complications in the formulation of the causal closure thesis. In the version of the thesis at the beginning of this article, every physical effect was referred to as having a "sufficient immediate physical cause," rather than just having "a physical cause." This was to ensure that the physical realm is genuinely causally closed. The specification that the physical cause be "sufficient" is needed to ensure that it causes the physical effect by itself, and not solely in virtue of its conjunction with some sui generis nonphysical causesuch a mixed cause would obviously violate the causal closure of physics. Again, the requirement that the physical cause be "immediate" is needed to ensure that it not produce the physical effect only via some intermediary nonphysical causesuch nonphysical intermediaries would again violate the causal closure of physics.

The earlier formulation of the causal closure thesis also specifies that every physical effect has a sufficient immediate physical cause "insofar as it has a sufficient immediate cause at all." The reason for this latter qualification is to accommodate the indeterminism of modern quantum mechanics, which states that certain physical effects are random, without any sufficient determining cause. It remains the case, however, that according to quantum mechanics these random physical effects still have their probabilities fixed by sufficient immediate physical causes. And this in itself will sustain the argument that anything that affects the physical realm must itself be physical. At first sight it may seem that quantum indeterminism creates room for nonphysical causes (determinations of the will, perhaps) to exert a downward influence on the physical realm, by influencing whether or not certain random physical events occur. But this in itself would violate the causal closure of the physical, understood now as including the claim that the probabilities of underdetermined quantum physical events are fixed by sufficient immediate physical causes. For, if a nonphysical cause influences whether or not random physical events occur, it must presumably make a difference to the probabilities of those events, and this itself will contradict the thesis that those probabilities are already fixed by sufficient physical causes.

The Argument from Causal Closure to Physicalism

What follows now is a closer look at the argument that moves from the causal closure of the physical to the conclusion that anything with a physical effect must itself be physical. The focus will be on the case of mental causes of physical effects, but most of the points made will apply to items with physical effects generally.

Recall the point that the argument gets no grip on realms that have no physical effects. As mentioned earlier, Leibniz used this point to evade physicalism about the mental by holding that the mental and physical realms are causally insulated from each other, albeit unfolding in "preestablished harmony." Contemporary philosophers who share Leibniz's conviction that mental states cannot possibly be physical tend to adopt a somewhat different ploy. Instead of denying any causal contact between conscious mind and brain, they allow that brain processes cause conscious mental effects but deny that these conscious states then have any converse influence on the physical realm. The contemporary philosophers of mind Frank Jackson and David Chalmers have both argued in favor of this "epiphenomenalist" position (Jackson 1982, Chalmers 1996). By viewing conscious states as "causal danglers" that exert no independent influence on the physical realm, they avoid any conflict with the thought that the causal closure leaves no room for anything nonphysical to make a difference to physical effects.

Perhaps there is another loophole in the argument from causal closure. In effect, this argument holds that a nonphysical mind cannot have physical effects because then those effects would have too many sufficient causesboth a nonphysical mental cause and the physical cause guaranteed by causal closure. However, such overdetermination of effects by two sufficient causes is not unknown. Imagine a case of a man who is simultaneously shot and struck by lightning, where either cause would have sufficed for his death on its own. Why should the physical effects of mental causes not similarly be overdetermined by two independent causes?

However, it is not clear that this is a good comparison. Overdetermination by distinct causes occasionally occurs by chance. But if a nonphysical mind has physical effects, then causal closure means that overdetermination of those effects will be routine. This calls for some explanation of why the two independent causesmental and physicalshould always be found together. If the two causes really are distinct, then will not some mechanism be needed to ensure that a sufficient physical cause is in place whenever a mental cause has a physical effect? However, no plausible such mechanism suggests itself.

So the possibility of overdetermination by both physical and mental causes does not support a distinct mental realm in the face of the causal closure of physics. However, there is another sense in which the causal closure of physics does leave it open that all behavioral effects may have both a physical cause and a nonidentical mental cause.

It has been a common theme in much recent physicalist philosophy of mind that mental properties are not type-identical to physical properties: many physicalist philosophers of mind are persuaded that, because beings with different physical constitutions can share mental properties such as pain, mental properties must be functional properties that are variably (or multiply) realized by physical properties, or disjunctions of physical properties, or some other kind of property that is metaphysically fixed by (supervenes on) physical properties, but not strictly identifiable with them.

Now, to the extent that causes involve properties, this denial of type-identity for mental and physical properties means that the mental and physical causes of behavioral effects cannot be strictly identical. However, this kind of double causation does not amount to the kind of overdetermination by distinct causes that was argued to be unacceptable above. As long as mental causes supervene metaphysically on physical causes, they are not fully distinct from them, and there is already a built-in explanation for why there should always also be a physical cause (as required by the causal closure of the physical) whenever a mental cause produces a behavioral effect. The denial of type identity creates some space between mental and physical causes, but not so much as to render it mysterious that they are always found hand in hand.

See also Causation: Philosophy of Science; Physicalism.


Chalmers, D. The Conscious Mind. Oxford: Oxford University Press, 1996.

Jackson, F. "Epiphenomenal Qualia." Philosophical Quarterly 32 (1982): 127136.

Kim, J. Mind in a Physical World. Cambridge, MA: MIT Press, 1998.

Papineau, D. Thinking about Consciousness. Oxford: Oxford University Press, 1992.

David Papineau (2005)