Epiphenomenalism is the doctrine that mental facts (properties, events, states) are causally idle. When it seems as if mental facts are causing behavior it is really correlated neural facts that are doing the causing—electrical signals sent from the brain down the efferent nerves and into the muscles. This is taken to be a singular fact about the mental: it is uniquely lacking in causal powers. Even in the case of a steam engine emitting waste steam—the standard analogy for epiphenomenalism—the steam can cause something, possibly burns in someone too close to it. The mental is thought exceptionally idle, and this is taken by many to be an objection to epiphenomenalism—too much of an anomaly. But is that true—are there no other epiphenomenal facts in the world? In fact, they seem quite plentiful upon closer examination. Obvious examples would be mathematical and ethical facts: they don’t cause anything either, but they are real. Nor is this merely contingent: it is intrinsic to such facts that causality be foreign to them. The same is true of such properties as identity, existence, necessity, and set membership—none of these feature in causal explanations or confer causal powers. Ditto for logical properties such as entailment and validity: these have no causal consequences either. Does the property of being a cause itself have causal consequences? Nope. These may be thought rather abstract properties, unlike mental properties, but we can easily find more concrete properties that likewise fail of causal potential. Colors are an obvious example: colors don’t cause anything either—only their underlying physical correlates do. If colors are projected onto the world, it is predictable that they make no difference to the causal powers of the objects onto which they are projected—subtract them and the object remains causally indistinguishabe. This is why mechanics makes no mention of colors in its theories of motion (red balls move the same way as blue balls). Acceleration and mass affect the causal powers of bodies, but not color (or taste etc.). Epiphenomenalism is all around us.
And it doesn’t end there. Do space and time have causal powers? Not inherently—not independently of material objects (even in Einstein’s General Relativity). Does infinity have causal powers? Does the infinite divisibility of space and time show up in the causal powers of things? Would it make any difference to the causal workings of the world if they were finite? It is not clear that it would. Is it even clear that geometrical forms have causal powers? Does it really make a difference to how objects behave whether they instantiate perfect circles? Causality is not fine-grained enough to care whether objects are platonically perfect. Causal powers in the physical world stem from four underlying forces: gravity, electromagnetic force, the strong force, and the weak force. But these operate independently of the geometrical and chromatic properties of objects: being a perfect circle (or an imperfect one), or being a whiter shade of pale, are not their concern. Mass matters, for sure, but not platonic forms or visual appearances. Of course, it matters whether an object is circular or rectangular in the vernacular sense, but not its fine-grained geometry. This is too abstract and ideal to be of concern to the rough and tumble of causality. In fact, objects don’t instantiate perfect platonic forms, though they do approximate to them, but it wouldn’t make any causal difference if they did.
This is all before we get to such peculiar properties as being thought of by Bertrand Russell one Saturday afternoon. Whether an object has this property makes no difference to its causal potential. Nor does the property of being round and such that 1 +1 = 2. Nor does the property of having a counterpart in another possible world. These are all far too extrinsic to count as causal powers. But mental properties are not extrinsic, so shouldn’t they have causal powers? Let’s consider functional properties, e.g. having the function of pumping the blood. The heart has this property: is it causal? It is not: the heart doesn’t pump blood in virtue of having the function of doing so. It pumps blood in virtue of its elastic and propulsive properties—the properties that enable it to discharge the function it has. The heart could have that function and not have the power to pump blood (it’s a defective heart). No more does a lawn mower cut grass in virtue of having the function of mowing lawns—for that it needs sharp blades and forward momentum. Having a certain function is not part of the causal powers of an object; having the properties that carry out that function is. So we could say that functional properties are epiphenomenal: they are facts but not causal facts. Just so, an organism has many functional properties that are causally idle, corresponding to its various organs. The color of blood is epiphenomenal, but so is its function per se. Functions are not causes. Teleology is not causality. Nor is species membership a cause: being of a certain species is not something that makes a causal difference, not in itself. True, members of a species have the power to interbreed, but not in virtue of species membership—for that they need a certain kind of physiology. Properties that sit idly by are certainly correlated with hardworking causal properties—and this may mislead us into thinking that they are less than completely indolent—but closer inspection reveals that they are not as such causally active. The pattern is that epiphenomenal properties sit atop causal properties—as functional properties are embodied in causal properties without being identical to them.
This is the picture the epiphenomenalist envisages for the mental and the cerebral. Mental facts are strictly speaking epiphenomenal facts, but they are tightly correlated with causal machinery in the brain. Is this a plausible view of mental facts? In the case of factive mental states like knowledge it certainly seems true: whether a belief counts as a case of knowledge is irrelevant to its power to affect behavior—being true is not a causally relevant property (the same for whether a perceptual experience is veridical). Intuitively, factives are too extrinsic to function causally. But much the same holds for propositional content generally: even a bit of externalism will have the result that content is epiphenomenal. And how could an abstract proposition contribute to the causal powers of a belief? The causal powers of beliefs evidently derive from underlying brain states, possibly of a syntactic nature (this is a familiar story). Beliefs as such don’t cause behavior, but only the correlates of belief; or if you like, beliefs have a causal aspect but they are not causal through and through. Thus intentionality generates epiphenomenal facts, which is not very surprising—being about something is not the kind of property that can get muscles to contract. More interesting is the property of subjectivity: is what it’s like subjectively epiphenomenal? To be a subjective state is to be such that the nature of the state can only be grasped from a single point of view, i.e. by a being that shares the state in question. Only perceivers of red can grasp what it’s like to see red. This is an epistemic property of a mental state: it can only be known in a certain way. But how could that property contribute to the causal powers of the state? How could the state cause behavior in virtue of being knowable only in a certain way? Maybe material objects can only be known about in a certain way, i.e. by perception, but that has no bearing on their causal operations—physics doesn’t care about such epistemic questions. So the subjectivity of mental states isn’t part of their causal profile—which means that what an experience is like for its possessor isn’t part of its causal profile. The fact that we can’t know what it’s like to be a bat doesn’t affect the causal powers of bat experience—it carries on causing bat behavior irrespective of what we can know. Presumably bat experiences cause bat behavior in virtue of various objective facts about the bat’s brain—electrochemical activity, as we now believe. In one clear sense, then, consciousness is epiphenomenal—it doesn’t cause behavior in virtue of there being something it is like. Of course, we can truly say of a conscious experience that it caused an episode of behavior, because it has a cerebral aspect or correlate; but we can’t infer from this that its conscious aspect is itselfa causal power. That is, given that a conscious property is not identical to a brain property, it does not have causal powers. We can loosely report that an experience of red caused someone to move in a certain way, but we can’t mean that the subjective nature of the experience acted causally—only that an underlying physical correlate did. Logically, the case is like saying that the color red caused someone to see red: it is not the color as such that does the causing but only its physical correlate in the object. Colors themselves are epiphenomenal, but that doesn’t preclude the existence of a correlated causal sequence leading from object to experience. Similarly, conscious states are epiphenomenal, but that doesn’t preclude the existence of a correlated causal sequence leading from brain to behavior. After all, epiphenomenal facts are common in nature, as we noted, and consciousness simply follows suit. It is not anomalous in this respect but typical.
This solves a difficult problem. If conscious states had intrinsic causal efficacy, then we would have causal over-determination, since brain states also cause behavior. Each would be sufficient to cause the behavior in question, which means the other cause is not necessary. It may appear that the only way out is to identify the conscious state and the brain state, thus committing ourselves to reductive materialism. But a viable epiphenomenalism allows us to avoid this result while not accepting causal over-determination: it is the brain state alone that is doing the causing—the conscious state sits idly by, serenely epiphenomenal. The conscious state is numerically distinct from the brain state and the brain state is the sole cause of the behavior—because the conscious state has no causal role to play in producing the behavior (not intrinsically and as such). Thus epiphenomenalism allows us to escape a difficult problem without succumbing to reductionism. Nor is this merely ad hoc since epiphenomenal facts are quite common in nature—the norm, we might say. The metaphysical picture is that nature has two layers: an underlying layer of causal machinery, which is quite restricted, and an overlying layer of epiphenomenal properties that coast on the first layer. A perfectly reasonable hypothesis is that the causal layer consists of nothing beyond the four fundamental forces identified by physics; the rest of nature is not in the causal line of business (save derivatively). It is thus quite wrong to think of nature as exclusively composed of causal facts. The four basic forces do all the causal work while the rest of nature stands by without lifting a finger.
 Causation is all about the induction of motion, but pure geometry lies beyond this. Which geometry holds of nature need not impinge on the causal machinery of nature.
 Reference is causally irrelevant, notoriously. Where is the energy emitted by the reference relation?
 The brain itself has various epiphenomenal properties over and above its conscious properties, so not everything about it is causally implicated in behavior—for example, its color, its taste, its precise furrowing, and its spatial location. If the brain qua brain has epiphenomenal properties with respect to the causation of behavior, then surely the mind can. Among the epiphenomenal properties of the brain are its mental properties.
 This means that if you were to remove the four forces from nature leaving everything else intact all causation would be expunged. This seems like a plausible result—even the basic atomic entities are nothing causally without their accompanying forces. They are certainly not epiphenomenal.