Regarding the paper: the arguments seem reasonably persuasive to me where they differ with the prior view (with which I'm not familiar, so you have an advantage, but it makes sense). However, it occurs to me that if value is reified as neural activity patterns or levels of neurotransmitters, then at the level of the brain, knowing what you value may be very much like knowing that your leg is itchy: it is a direct perception.

How does anyone else know that your leg is itchy? They can't very well detect the firing of your itch-sensitive neurons with their eyes or fingers. It's just neural activity activated by histamine; if some other neurons are activated by dopamine and that happens to be how value is reified, why wouldn't you know that the same way?

Of course, there is all the complexity of the coupling between ideas and the qualia of valuing-a-thing, and one could very reasonably say that one is mistaken (that is, the feeling of value is mismatched with one's ideas, and there is cause to say that it is the feeling rather than the ideas which is in error). But you might also be mistaken about an itchy leg--that is, there might be no histamine or other inflammatory molecules released in the location to which your brain is ascribing an itch. Your leg is not really itchy, you just feel that it is. You don't really value that, you just feel that you do.

I don't think I am particularly disagreeing with you regarding value, though. Rather, I think itch is in the same boat (albeit a somewhat more seaworthy one, as such sensory modalities typically are). Well, and I think that it is possible that we know what we value (in an important sense) because we perceive it, not because we derive it in a declarative fashion from other beliefs.