Saturday, 22 February 2014

What the mind isn't - abandoning epiphenomenalism

The Hard Problem of consciousness remains the subject of intense debate and discussion in science and philosophy.    In this post I will argue that one view of the way that physical brains are associated with consciousness - epiphenomenalism - isn't a coherent view.

In order to understand what epiphenomenalism is, here is a representation of what is said to happen in a brain and conscious states in epiphenomenalism:

P represents a physical state of the brain.
C represents a mental (conscious) state.
M represents a mapping between P states and C states.

Non-interacting dualism or epiphenomenalism looks like this:

C1    C2   C3
 |    |    |
 M    M    M
 |    |    |
P1 → P2 → P3 → …

Physical states cause each other and also result in conscious states.  Now it may seem to be a problem for those who support this model of the relationship between brain and mind to explain how physical brain states can result in someone talking about truly about epiphenomenalism being real? – how can a 'P' state be at all related to a 'C' state?  This can happen because if the relationship between P states and C states (the mapping 'M') doesn't vary, then each P state represents its corresponding C state, and so a physical being can talk about epiphenomenal conscious states without inconsistency – it's not absurd.

Or is it?  Does this really work?  The answer is that it doesn't work, and to see why we need to perform a thought experiment.  Imagine someone with a brain implant.  This implant affects mathematical thinking, and it can be turned on and off by someone other than the person with the implant.

With the implant off, the person with the implant does some mental arithmetic.  Suppose the sequence of physical and conscious states are as follows, as he or she thinks through a simple sum:

Step Physical Conscious

1 '1' '1'
2 'add' 'add'
3 '2' '2'
4 'equals' 'equals'
5 '3' '3'

Now, the implant is turned on, and the sequence of states is this

Step Physical Conscious

1 '1' '1'
2 'add' 'add'
3 '2' '2'
4 'equals' 'equals'
5 '4' '4'

Both the P state and the C state have changed in step 5.  This is due to signals from the implant affecting the brain.

Looking at the C states, the indirect causal connection between these states doesn't work with the implant on.

However, looking at the P states causality is still working fine, because what produced the state in step 5 was a predictable combination of state 4 and the implant signal.

So it's conceivable that the same conscious state doesn't necessarily always lead to the same subsequent conscious state.  The possibility of additional physical contributions to the physical states means that we aren't justified in insisting that the physical states represent conscious states.

Other problems for the idea that physical brain states can be thought to be about epiphenomenal conscious states is to do with the relationship between physical and conscious states – 'M':

Dancing Qualia

Changes in how conscious states arise from physical brain states have no effect on physical brain states.  Such are the subject of the thought experiment known as 'Dancing Qualia'.  Suppose you see a red rose and say 'that rose is red'.  Then, you are shown the rose again, but the physical to conscious mapping is changed so that the conscious experience of the rose is 'green'.  You will still, as a result of physical brain states being unchanged, say 'that rose is red'.  The philosopher David Chalmers says that one of these two statements from you is true and the other false.

If we assume epiphenomenalism, then, because physical brain states don't represent conscious states – the experienced relationships between conscious states don't necessarily correlate to causal relationships between physical states (as in '1 plus 2 equals 4' above) – so the statement 'the rose is red' doesn't have a causal connection to the epiphenomenal conscious experience of 'red'.  Both statements above about the colour of the rose are true, because when the word 'red' does not refer to the epiphenomenal conscious experience of red, but the physical brain state representing 'red'.

What is 'M'?

Another problem for epiphenomenal consciousness is that even if it were true that physical brain states did represent conscious states, there is no representation either in physical or conscious states of the nature of the mapping M between those states. Even if there was a possible sense of the mapping being of different kinds  M1 and M2, there is no way for a physical brain state to represent the difference between M1 and M2.  In terms of conscious experience, this means that all you get is qualia, you can't get from conscious experience why you get qualia.

But what does 'qualia' refer to?  We have seen from the Dancing Qualia thought experiment that the word 'qualia' cannot be said to refer to anything in epiphenomenal conscious experience.  The only thing 'qualia' can refer to is representations in physical brain states – there are qualities of experience, and they are physical: encoded in brains.

Epiphenomenal consciousness is incoherent, as when you try to pin down what it is it vanishes like a mirage.  The words we use to describe epiphenomenal consciousness cannot in reality refer to any actual epiphenomenal consciousness, only physical representations in brain states.  The semantics of epiphenomenal consciousness can never reach such a consciousness.


Anonymous said...

"So it's conceivable that the same *conscious state doesn't necessarily always lead to the same subsequent conscious state."


Steve Zara said...

No, it is 'conscious'.