1: What am I Talking About?
Usually epiphenomenal is used to describe a theory in which mental states exist but have no casual effect on the world. However in this article I am going to accuse other philosophical views, not necessarily about the mind, of being epiphenomenal. By this I mean that the objects that they propose to talk about do not have a casual effect on the universe. Without such a casual effect however we would have no way of knowing that they exist. Thus we should reject such ideas as being meaningless and un-provable. Of course these are broad claims, which require their own justification, so read on.
2: Knowledge and Causation
The background premise is that to know about something and to be able to meaningfully discuss it that thing must lie in your casual history. (Note: for the remainder of this article when I use the word meaning it is in the sense of meaningful. It is true that something that isn’t meaningful might have some private meaning to you, but it wouldn’t be justified to base a theory on it.) Of course to make a statement about casual history we must have defined causation first, so I suggest that you read the linked post if you have any questions about what counts as a cause for this argument. Another way of stating this premise is that to have knowledge of X there must be some time in the past in which I can say that X is a cause of my current mental state (my current knowledge). And of course for our words to have meaning we must have knowledge of the objects they propose to refer to. Now I am not saying that causation is all there is to meaning and knowledge (that story will have to wait until another time), I am simply saying that it is a necessary condition for meaning and knowledge.
(By knowledge I am referring to certain beliefs, as defined here)
An important caveat to make here is that not only must this causation exist but it must be discoverable as well. For example let us consider a possible world in which the disposition of objects to fall towards the ground is not caused by gravity but by the will of Odin. However Odin just happens to will things to fall if and only if gravity is present. People living in such a world will never be able to formulate laws involving the will of Odin, and thus as follows from my knowledge of causation post they will never think that Odin is the cause of anything. Assuming that Odin’s will is limited to making things fall they will then also never have knowledge of Odin’s will. Although they might speculate “what if the real reason things fell wasn’t gravity” they would have no way of knowing what it was, nor any evidence that it wasn’t gravity, and thus no way of knowing that it is Odin’s will.
Of course the obvious objection to the view that causation is required for knowledge is the question “what about fictional things?” It seems obvious that we have knowledge about Santa even though Santa, not being real, can never be a cause of any of my mental states. To this I would agree, and I think it shows that we don’t have knowledge of Santa, all we have is knowledge about what people have told us about Santa. To have knowledge of something implies that it is real, but what is real in this case is not Santa but the legends about Santa (they are real legends, in the sense that they exist, not that they are correct).
Does this imply then that because I have never seen the Eiffel tower that I have no knowledge of it? In a way, yes. I definitely have knowledge about pictures of the Eiffel tower, and what people have said about the tower. It is possible at least that these are all false, and thus absolute certain knowledge is impossible. However I reason thus: the most likely explanation for these pictures and descriptions is that a real Eiffel tower exists and is a cause of the pictures and explanations. Thus I feel warranted in acting as though I had knowledge of the Eiffel tower (as if it lay directly in my casual past, which it does), even though there is a minuscule possibility that it does not, and what I thought was knowledge about the Eiffel tower is really only knowledge about a complicated deception people have been playing on me. The meaning of words functions in much the same way. I use the words Eiffel tower as though I mean a real tower when all I really mean is a mental construct that has been formed by my knowledge of the tower (this is all that casually influences the words I speak, the real tower is not a cause of my speech; it can’t be it is too far away). It is possible that this mental construct is in error, or does not correspond to a real object. In this case the meaning of my words is a purely private matter, but if (as I reason is probable) other people have knowledge of the Eiffel tower as well they will have similar mental constructs, and the meaning of Eiffel tower to me will be nearly identical to the meaning of Eiffel tower to them, allowing us to communicate.
Let us propose a different objection then. Suppose that 2 million light years away there is a man called Darth Vader, who dresses all in black, ect. Now suppose we have watched a popular motion picture containing many details of what, by coincidence alone, happens to be this man’s life. Do we have knowledge of the “real” Darth Vader, even though there is no way that he could be the cause of this motion picture or the mental states that resulted from the motion picture? No, I would say that the casual link correctly tells us that we only have knowledge of the motion picture and not of the hypothetical “real” Darth Vader. Because even if Darth Vader was real it is equally possible that he isn’t real, but yet our knowledge of Darth Vader and his properties (such as his favorite color) is unchanged, he could even be popping in and out of existence as we speak, and our knowledge would remain the same. Thus since our knowledge can’t be affected by this “real” Darth Vader it seems baseless to say it is about him. On the other hand it makes perfect sense to say that it is about the fictional Darth Vader, for if somehow the movies were removed from time our knowledge would vanish with them.
Alright, enough about knowledge; if you still subscribe to a non-physical or magical theory of knowledge and meaning obviously the rest of this essay will sound hollow to you, and I apologize for that, but I have spent too long on meaning and knowledge as it is.
3: Epiphenomenal Ideas
Before we go on let me assert that the complete physical description of the universe is casually closed, by definition. Of course this isn’t true for our current physical description of the universe, as surely there are laws that haven’t been discovered yet. But, it is closed in principle because of how we decide what kind of information count as “physical”. For example let us pretend that the physical description wasn’t casually closed and that there was some non-physical force, Y, that had a casual effect on the physical world. If it has a casual effect then there is some physical system, X, that will produce result X-1 in the presence of Y and X-2 in the absence of Y. This behavior however is enough to postulate some physical cause, y, of X-1. A materialist could produce a detector for y using system X, and using this detector identify all of the ways in which Y had a casual effect on the world. This casual effect would be completely captured by y, which is part of the physical description. It is true that there may be more to Y than the physical properties captured by y. In this case we might argue that Y-y still exists and is non-physical. However Y-y has no casual effect on the world, and thus is epiphenomenal (because all the casual effects have been captured by the physical y).
Using this reasoning we can demonstrate that some forms of transcendental idealism are epiphenomenal. The form of transcendental idealism that I am referring to here is the belief that there are other aspects of reality that we have no access to, and that the aspects of reality we can perceive are in fact casually closed. Of course most philosophers who hold transcendental idealism to be true agree that we can have no knowledge about these other aspects of reality. However I would argue that if you can’t have any knowledge about it then you certainly can’t know that it exists. Thus I would argue that we should reject transcendental idealism because it is epiphenomenal; even if it were factual it can’t be known to be true, it can’t be detected, and it can’t effect us, so we might as well treat it as though it were false. Likewise when we have accepted that the physical description of the universe is casually closed we should reject dualism as epiphenomenal for the same reasons
The more interesting case though is ontology. There are two basic kinds of ontological claims. The weaker claim is that there are certain categories that we can impose on the world or that we can’t help but impose on the world. (By impose I mean that we can describe the world consistently with these divisions, and possibly argue from them.) On the other hand we have the stronger kind of ontological claim which states that reality is really divided into categories, in the sense that these divisions are part of the structure of reality, and are not in any way dependant upon people’s descriptions or thoughts. However, even if there were “real divisions” they would be epiphenomenal. For example you might think that there is a “real division” between objects under acceleration and objects in constant motion. Certainly an object cannot both be in constant motion and under acceleration. The category itself however does not have a casual effect on the world. Acceleration does, but the category is separate from acceleration, we simply use our observations on acceleration to determine which objects belong to the category. To look at it another way we might ask “what would change if the category of accelerated objects existed only in our minds?” The answer is that nothing would change. Individual objects would still accelerate and be subject to relativity even if the category dividing them from objects that were not under acceleration was only in our minds; the properties are real, but the category is not. From this I conclude that strong ontological claims are epiphenomenal, we have no way of knowing about them since they don’t have a casual effect on the world.
Someone who held one of these epiphenomenal views might now say to me “Yes, you have demonstrated that we cannot have knowledge about their contents, but neither have you demonstrated them to be false.” Well, since their contents have no casual effect on the world it is impossible to demonstrate their falsehood (except possibly through inconsistency). However because they are epiphenomenal I think that they are empty claims. Since we have no reason to believe them, and they can’t add anything meaningful to our understanding of the universe, I think that they should be discarded, even if they can’t be disproved.