But I kept the doubt in mind: is it not a matter of semantics even then? Because in epiphenomenalism, the mental changing the material is impossible within that metaphysics. But in epiphenomenalism, isn't the inability to change the material part of the definition of what is mental? And thus the mental changing the material becomes a logical contradiction within that metaphysics? — Lionino
For what it’s worth, my current thought process on the matter is along these lines:
Logical possibility—all of it pivoting on laws of thought—pivots on what is (taken to be) ubiquitous to all subjects of awareness. This irrespective of whether subjects of awareness might hold a comprehension of what these laws of thought might be; e.g., a preadolescent child will think via the laws of thought (however imperfectly and, hence, at times, illogically) although not holding a comprehension of them.
Physical possibility—when divorced from any metaphysics regarding what the physical entails, e.g. materialism, idealism, or substance dualism, etc.—also pivots on what is (taken to be) ubiquitously applicable to all subjects of awareness. Here, though, without a metaphysics there cannot occur a comprehension by which to make sense of physicality.
So both the logically possible and the physically possible will at root address ubiquitous actualities, actualities that are thereby universal and, in this sense, singular.
Metaphysics, on the other hand, will always make use of the logical and of the physical—at least in part, to which experiences, i.e. subjective actualities, can be added as well—to arrive at understandings regarding that which is in any way actual (including, for example, that which is actually possible). There are multiple ways metaphysics could be derived via physicality-bound (as well as, at times, experience-bound) logic. Thus resulting in multiple, often enough contradicting, metaphysical models of what is.
Each metaphysical system will then galvanize its own semantics; most of the time the validity of these metaphysics-specific semantics will be evaluated by their individual explanatory power—this in explaining what is actual (be it laws of thought, be it the physical, or be it our sometimes discordant and sometimes commonly held experiences, which could then extend into things such as cultures, languages, etc.). And, by extension, these individual explanatory-power-endowed semantics that together form the given metaphysics then grants the given metaphysics as a whole its explanatory power.
So the individual understandings, or semantics, imbedded within a metaphysical system (such as that of epiphenomenalism’s impossibility of mind affecting matter) is tied into, and is justified via, a webbing of ideally fully self-consistent semantics—all minimally conforming to what is known of logic and of physicality—that work together to explain all that is actual. To deprive epiphenomenalism of the impossibility of mind affecting matter is to then nullify the entirety of the metaphysical webbing of understandings which epiphenomenalism is. This, were it to occur, would then leave a vacuum of explanatory power and, hence, of general understanding, for all those that previously upheld the metaphysics of epiphenomenalism.
This being a longer path toward saying that I fully agree metaphysical differences can be said to boil down to semantics. I’d only add that, for one example, the particular semantic of “mind” in the case of epiphenomenalism appears to me inextricably bound into the entire webbing of semantics—of logic- and physicality-bound understandings—which this one metaphysics in fact is, if not merely being a webbing of understandings from which this metaphysics is constituted.
[This, to my mind, could get deep into epistemological issues of justification: which, as per the above, I currently perceive to involve some variation of
foundherentism. This being a crossbreed of foundationalism (in conforming to the laws of thought and to physicality, if not also to some aspects of experience) and coherentism (in relation to a particular metaphysics' ideal lack of self-contradiction in the understandings it holds). Likely a different issue, though.]
At any rate, this is only a rough sketch of a general idea. Still, while its likely incomplete, I nevertheless so far find it to, well … to hold a fair share of explanatory power—this in terms of the different types of modalities addressed in this thread.