Unfortunately, he is talking outside his domain, because he is talking about metaphysics (which is not religion), and his lack of knowledge shows glaringly
Haven't read the book, I take it. No, he makes no arguments from a non-empirical basis.
Metaphysics is not about empiricism, at least not directly. It’s a branch of philosophy, not science. He is indeed talking outside his domain of expertise.
The book in question is not about metaphysics, except insofar as metaphysical ideas become the basis for pseudoscience.
The book in question is INADVERTENTLY about metaphysics, because free will v. determinism is a metaphysical problem. It cannot be a scientific problem because the whole issue is unempirical. Sopolsky can muster all the scientific data he wants, but he can never conduct one crucial experiment: Imagine I pick Pepsi over Coke at time T. If we played back the whole history of the universe with the EXACT SAME ANTECEDENT EVENTS, would I pick Pepsi again, or would I choose Coke this time? DBT claims I would choose Pepsi again because antecedent deterministic events NECESSITATE my picking Pepsi, and that I would ALWAYS pick Pepsi no matter how many times the experiment were run.
I don’t, in fact, disagree with him that under these circumstances, I would indeed ALWAYS pick Pepsi. Where I disagree is that I MUST pick Pepsi — that my action is necessary, as he would have it, entailed by the identical antecedents. That is his argument. HIs argument is logically incorrect, because it commits a modal scope fallacy, as I have explained numberless times. I hold that if we could run this experiment, I would indeed always pick Pepsi, because that is what I WANT to do, not because I HAVE TO DO IT.
As I have explained many times, his fallacy is:
Given antecedents w, x, and y, necessarily z.
That IS the modal scope fallacy.
The corrected argument is:
Necessarily (w, x, y —>z), where z remains now and forever contingent (could have been otherwise).
The corrected modal argument removes the fallacy DBT commits, but it also destroys his hard determinism.
In any case, the experiment above cannot be run. We cannot back up, so to say, and replay the history of the universe, to see whether I would pick Pepsi or Coke. And even if we could, and I picked Pepsi every time, it would never prove his claim, that I MUST pick Pepsi, because his claim commits a logical fallacy, as explained.
Hence, the debate Sopolsky has entered upon is indeed METAPHYSICAL, even if he doesn’t know it (and indeed, also involves logic, another branch of philosophy). That’s his problem — if he wrote the book, he ought to be able to address its metaphysical aspects, but he doesn’t, because I think he is unaware of them. In much the same way, in one of his later books (possibly his last), Stephen Hawking wrote, on the very first page, “philosophy is dead.” He then proceeded to write an entire work of … philosophy. Without even knowing it.