Hi Derek, 
The post was from me not Brent. 

So you acknowledge that you do not need to know what death was like to understand what an imagining of it was.

I can understand the distinction you are now making with an atheist assumption of a lack of an afterlife from there being an afterlife to imagine. Especially when it comes to imagining it in terms of what-it-is-like. Since for a comparison, there must be at least two to compare (a comparison is not a unitary operator so to speak) and if one is "nada" then it can be thought of as an absence of something which can be compared. Thus I think the idea that it-is-like-something to consciously experience (the "nada" being eliminated). But since the "nada" is not ambiguous neither is other than "nada" (thus if you were aware of anything at all). 

Jackson's idea that if there were features other than those observable from a third person perspective meant that reality was not physical entailed the assertion that all physical features were observable from a third person perspective. A flaw in his argument. 

A zombie should not be thought of as a human minus certain features (as people like Dennett thought of it). Instead it should be thought of as where one only imagines those features described (those observable from a third person perspective), and does not add the features which one knows the human has but which were not mentioned in the description, but instead imagines that the features in the description were all that it had. 

Yours sincerely,

Glenn

P.S.

Regarding thought experiments, in scientific experiments to do with theory (as opposed to just see what happens type experiments, in which it is easier to check the result rather than work it out given current understanding ) there can be a hypothesis, and a null hypothesis, and the experiment result can distinguish between which it was correct. The thought experiment can allow you to express your conclusions regarding whether it was result was as expected given the hypothesis or not without knowing what the result actually was. In other words the conclusion you would draw if the experiment showed the hypothesis to be correct, and the conclusion you would draw it the hypothesis was not correct. The EPR experiment that I mentioned (plato.stanford.edu/entries/qt-epr/ ) allowed the realisation that quantum randomness implied "spooky action at a distance" without doing the experiment, but considering the conclusions that would be drawn, or perhaps as is more likely encouraged the thinking of an experiment which highlighted the implications of the idea (the implications being logically derived), which could then be tested. Though with the latter, one could still consider the conclusion drawn from the hypothesis being confirmed or rejected.