Gertie wrote: ↑October 3rd, 2020, 8:56 am
As I said your What If is still stuck with the Hard Problem. If that is as much explanation as we're ever going to get, then accept the consequences. Your 'predictions' are just guesses. And the results are not reliably testable. And even if they were you couldn't know if your What If is the reason the guessed prediction is correct.
Well, all predictions can be called "guesses," I suppose. But there are good guesses and bad ones. What distinguishes them is that the former are confirmed by observation. And they are reliably testable --- either the system displays the predicted behaviors or it doesn't. If it does, then the prediction was correct, and for the reasons set forth in the theory, at least until another theory comes along, offering different reasons, that makes even more correct predictions. There is no way to assess the "correctness" of any theory other than the reliability of the predictions it makes.
The problem with that position is that experiential states are --- obviously --- not redundant. They instigate most human behavior. Did not a desire on your part instigate your above comments? That certain brain states were also involved is a theory, a conceptual construct, which is another phenomenal artifact. That argument against epiphenomenalism rests on an assumption that phenomenal states and events imply the existence of another kind of "basic stuff" which, not being reducible to physical "stuff," cannot affect it, and is thus superfluous. But that implication is gratuitous; the subjectivity of phenomenal effects does not entail that they must be of a different kind of non-physical "stuff." They are just a different kind of effect. That they are only produced (as far as we know) by physical systems is ample warrant for considering them physical effects.
Nor do those effects arise independently from the physical systems producing them, any more than the negative charge on an electron arises separately from the electron. So they don't need an independent evolutionary justification. All that needs to be justified in evolutionary terms is the system as a whole, and the evidence is pretty strong that those effects confer some survival and reproductive utility on systems that manifest them.
You haven't answered the objection - If neural correlation holds, and neurons are affected by physical causality just like any other physical stuff, then experiential states are redundant, and there would be no evolutionary pressure for them to arise. When in reality, they look honed for evolutionary utility.
The above quote does answer that, Gertie. There doesn't need to be any evolutionary pressure for experiential states to "arise." There only needs to be evolutionary pressure for systems to arise which have survival advantages. Certain kinds of systems happened to have that property, which proved to confer some survival advantage. That is true of all traits which confer some survival advantage. Various traits appear in populations at random, for physical reasons. Some confer survival advantages in a given environment, some disadvantages, some neither. Cheetahs, almost alone among cats, don't have retractable claws. There was no evolutionary pressure for that trait to appear in some ancestor population. But it did appear, due to some random genetic variant, and happened to confer an advantage on cats in a certain environmental milieu (in other environments it would be a disadvantage). Biological traits appear at random, due to some random alteration in a DNA sequence somewhere. Whether a trait confers a survival advantage can only be assessed after it appears. There is no "pressure" for any particular trait to arise.
Again, your preferred What If isn't reliably testable, because experience isn't third person observable, and you don't provide an explanation which gives us something which might be - like specific necessary and sufficient conditions. Copying something isn't explanatory. And while it might at least in principle (if it was reliably testable) rule out some What Ifs, it won't identify THE correct one.
THE correct one?
I've given you a methodology for determining whether an hypothesis, or theory, is "correct." You can't speak of "THE correct one," unless you have some methodology in mind for discovering it. The correct theory or explanation will always be, and can only be, the one which generates the most most reliable predictions. Asking how things "really are" in some transcendental sense, "from God's point of view," is a vacuous exercise. Meaningless.
No, I can't give the specific necessary and sufficient conditions for a physical system to manifest consciousness. We know that they exist, however, since we have physical systems that do manifest that property. Whether we can fully elucidate them remains to be seen; we will know when we have succeeded when we have constructed a system whose behavior warrants calling it "conscious." We'll then impute phenomenal states to it, just as we do when we deem certain animals (and other humans) to be conscious.
Then you're just defining whatever mechanism results in a sense of being an 'Experiencer-Self' in humans as a homunculus. This isn't how the term is used.
How do you think it is used? How do you understand it?
If we constructed a machine we were convinced was experiencing based on similarity to humans, we wouldn't know what particular key aspect of similarity (nec and sufficient conditions) we'd captured. So we wouldn't know if it proved your What If, or Identity Theory or Panpsychism, or something we hadn't thought of.
Behaviors we deem sufficient for imputing consciousness to other humans IS the sufficient condition, the only one we have, being unable (as third parties) to observe those internal states directly. That is the only similarity empirically accessible. We can't ask whether the machine's experiences are similar to ours; I can't even ask whether your phenomenal experiences are similar to mine. Those are unanswerable questions.
Fair play for actually having thought your position through and being able to defend it in detail, but there's really nothing wrong in saying We Don't Know, when we don't know.
I'm saying more than that --- not only do we not know precisely how phenomenal states are generated by physical systems, or whether a machine's (imputed) phenomenal states, or yours, are similar to mine, we can never know that --- because those states are not available for analysis by scientific methods and are not derivable from known scientific laws. They are, however, found only in connection with certain physical systems, which warrants considering them physical effects. We can rule out identity theories because phenomenal states are obviously not identical to brain states, per the common definitions of "identical." We can rule out panpsychism on Popperian grounds --- because it imputes a property to things which is in principle unconfirmable and unfalsifiable, to things which exhibit no behaviors that warrant imputing that property, and those behaviors are the only warrant we have for imputing it to anything.