Functionalism and Epiphenomenalism

Epiphenomenalism is, rightly I think, a rather unpopular view about the nature of consciousness. If consciousness has no effect on the brain, then we should be deeply suspect of our knowledge about it. In The Conscious Mind, David Chalmers suggested that dualists can't really avoid the problems of epiphenomenalism, because the intrinsic properties that constitute consciousness must always be married to separable causal powers. It is the causal powers, not the intrinsic properties, that cause us to believe we're conscious.

The same view can, I think, be applied to basically every non-functionalist account of consciousness. Consider the view that material constitution of the brain is important. It is not just what the neurons do, but how they do it, that makes us conscious. Reproduce the same network of neural connections in another medium, (a computer chip, say) and you won't reproduce the consciousness.

However, this seems to make consciousness epiphenomenal to whatever features of our brain give us knowledge of our consciousness. If you reproduce a brain in another medium that preserves its  functional architecture, its behavior should not change. It will still claim to be conscious, speak about its consciousness, etc. It's brain will come to the same conclusions for what look to be the same reasons. It won't believe that it is conscious because of the material matter comprising its brain. It will think that it is conscious solely because of the connection strengths of the 'neurons' in its neural net. 

If epiphenomenalism is troubling for dualists, it should also trouble those physicalists who think that non-functional properties, like material constitution, are important.