Some important points about Poland's and Melnyk's versions of physicalism:
The scope of physicalism. According to Poland (p.227),
The theses of physicalism apply to all natural phenomena and all claims to truth and knowledge concerning the natural order.This leaves one wondering what, exactly, constitutes the "natural order." Poland lists
...physical, chemical, and biological phenomena, ... the psychological, the social, the moral, the aesthetic, and the commonsense world of our everyday experience.However, he excepts purely abstract realms such as mathematics.
Melnyk is (again) more specific, saying that physicalism applies to everything that is either contingent or causal. The existence of a god who is not contingent (because he is a necessary being) but who causally affects, or is affected by, the world, would refute Melnyk's physicalism. (It seems that such a god would not refute Poland's version, as god is presumably outside of the natural order.) Mathematics, being composed of necessary (i.e. not contingent) truths that have no causal efficacy, is exempt here as well.
Realization physicalism entails supervenience. That is, the physical facts about the world determine all objective facts about the world.
But supervenience by itself is not enough to guarantee the kind of physicalism that we want. And supervenience doesn't entail realization physicalism. Thus, realization physicalism is a stronger claim than supervenience.
Physicalism is reductionistic. I was surprised to learn that "reductionism" is a bad word among philosophers. I thought that reduction was the whole point of science: to find explanations of things in terms of simpler, more fundamental things. Melnyk spends a whole chapter (cutely titled "Physicalism and R*d*ct**n*sm") discussing in what sense realization physicalism is, and in what sense it is not, reductionistic. He argues that it is reductionistic in the "core sense." The core sense, briefly, is that all higher-level facts have an explanation in terms of physical facts and necessary truths (such as logical and mathematical truths).
However, realization physicalism does not satisfy what Melnyk calls the "received sense" of reductionism, namely that each higher-level type is identical with some physical type. And Melnyk's version of physicalism is stated in terms of functional, not physical, types. In particular, a functional type can be one that satisfies a purely logical condition. (An example would be the computer program considered earlier.) While realization physicalism requires that all actual instances of such a higher-level type are realized physically, it does not identify the higher-level type with those instances, or the collection of all possible such instances, or anything like that.
Other Errors To Avoid: Poland writes (p.41)
[T]here is good reason to avoid such views at that everything is completely physical, physics is the one true theory, physics describes and explains everything, the only ... legitimate methods of inquiry are those of physics, and everything is best understood from the perspective of theoretical physics.
I have to agree with him here - except for the last one, which is self-evidently true.