Empirical Physicalism and the Boundaries of Physics

ABSTRACT

I shall argue in this article that there are certain objectual and methodological boundaries imposed by the nature of physics that all formulations of physicalism based on physical theories should respect. Therefore, empirical physicalism—i.e. the sort of physicalism that is eager to accept all the entities included in some future, ideal, and complete physical theory and all entities dependent on them (see Jeffrey Poland and Janice Dowell)—is already committed to the exclusion of certain sorts of entities from its ontological inventory and it is far less tolerant than one might expect. After having presented my argument, I shall describe four plausible boundaries to what can be studied by physics. The boundaries will contribute to constructing a new version of physicalism, i.e. physics physicalism, whose acceptance is entailed by the acceptance of empirical physicalism. Finally, I shall briefly deal with three objections against my strategy and I shall evaluate the consequences of the acceptance of my conclusions for both physicalists and non-physicalists.

Acknowledgements

I would like to thank anonymous reviewers of this journal for their helpful comments.

Notes

1 In Paolini Paoletti (2018), I hypothesise that there are weaker versions of physicalism, according to which non-physical things can depend on physical and non-physical things and/or the former can depend on physical things in some, but not all respects. For example, some non-physical, mental property can depend on physical properties and on further mental properties. Alternatively, it can depend on physical properties only in some respect (e.g. for its being instantiated) and being independent of the latter in some other respect (e.g. for its conferring some novel causal power). However, in this article, I shall be concerned only with the stronger versions of physicalism, i.e. those versions that are incompatible with interactionist dualism, emergentism, panpsychism, and so on. Therefore, I shall assume that ‘entirely depends’ means dependence in all respects only on physical entities.

2 Following Stoljar (2010, 56–57), this constraint constitutes one of the five distinguishing features of an intuitively acceptable ‘Starting Point View’ of physicalism. I shall name it ‘Theory Constraint’. Stoljar interprets the Starting Point View of physicalism in terms of properties. Thus, the Theory Constraint turns out to assert that P is a physical property only if it is expressed by a predicate of a true physical theory. Moreover, Stoljar omits the adjective ‘true’. Yet, no physicalist would be inclined to accept properties invoked in false physical theories. Moreover, some physicalists would claim that an entity is physical if and only if (iff) it is included in some true physical theory. I prefer the weaker ‘only if’. For I wish to remain neutral (at this point) on whether further clauses (such as Wilson’s (2006) Non-Fundamental-Mentality constraint) should be ‘added’ to the entities included in true physical theories in order for them to count as physical.

3 This thesis is known as panpsychism. See, for example, Goff, Seager, and Allen-Hermanson (2017).

4 Other physicalists, who sometimes remain agnostic on the sort of physical theories to be accepted (current or future ones), assert that an entity is physical only if (or, more strongly, iff) it is treated by physics (or by fundamental physics) and it is not fundamentally mental (this is the Non-Fundamental-Mentality constraint). See Papineau (1993, 2000), Montero (1999, 2001, 2009), Montero and Papineau (2005), and Wilson (2006). For a criticism of the Non-Fundamental-Mentality constraint, see Dorsey (2011).

5 Some physicalists have a broad conception of physics and of the physical, according to which, for example, even the entities studied by chemistry, biology, etc. turn out to be physical and to be part of physics. Yet, I do think that one can make sense of physics physicalism even within this perspective. Assume that ‘narrow-physics’ stands for physics as a definite science (whose boundaries I try to characterize here) and that ‘broad-physics’ stands for the physics plus any other ‘acceptable’ science for broad physicalists (e.g. chemistry, biology, etc.). Correspondingly, assume that something is narrow-physical iff it can be studied by narrow-physics and that something is broad-physical iff it can be studied by narrow-physics and/or by any other ‘acceptable’ science. (PP) can then be rephrased as follows: everything can be studied by narrow-physics or it entirely depends on entities that can be studied by narrow-physics. For broad-physical and non-narrow-physical entities still entirely depend on narrow-physical entities. Yet, one could now object that the latter entities could have features not entirely dependent on the narrow-physical and that this would still be compatible with physicalism. For example: a biological organism could have some feature not entirely dependent on the narrow-physical and this would still be compatible with physicalism. In my perspective, the problem with such a move is that, in principle, one cannot set a border between physicalistically acceptable features/entities and physicalistically unacceptable ones. In this respect, Fiorese (2016)—whose version of physicalism seemingly implies a broad conception of the physical—is eager to invoke the Non-Fundamental-Mentality constraint.

6 On this view, physics is not just what physicists do: this is too vague and it could give arbitrary characterizations of physics. On the contrary, physics includes what physicists can do given the boundaries of their discipline.

7 For an alternative view, see Strawson’s (2008) real physicalism and Howell (2008, 2009, 2013).

8 Of course, this happens only if properties are ‘borne’ by objects. ‘Bearing’ a property can be interpreted here in different ways—not necessarily in terms of instantiation. For example, a property can be ‘borne’ by an object even when the latter is a bundle of properties and the former is included in that bundle.

9 I do not rule out that objects that cannot be exhaustively studied by physics can also have properties that can be exhaustively studied by physics. For example: a cell has a certain mass, i.e. a property that can be exhaustively studied by physics—even if it has other properties that cannot be exhaustively studied by physics. Therefore, a more appropriate, though more complex, reading of (OPP) is the following:

(OPP′) Every object can be exhaustively studied by physics or, if it cannot be exhaustively studied by physics, for what concerns the properties that cannot be exhaustively studied by physics, entirely depends on objects that can be exhaustively studied by physics.

In addition, a more appropriate ‘negative’ reading of (OPP) is the one equivalent to (OPP’), i.e., it is not the case that there are objects that cannot be exhaustively studied by physics and that have properties that (i) cannot be exhaustively studied by physics and (ii) are not entirely dependent on the properties of objects that can be exhaustively studied by physics. This still entails and is entailed by (PPP). I shall set aside such complications here.

10 For a critical overview of the options, see for example Stoljar (2017).

11 Even if Einstein showed that matter actually is energy, (ME) allows for the possibility of the former being distinct from the latter.

12 Does (ST′) also rule out non-actual models of the space-time or idealised objects, such as frictionless bodies that do not occupy space-time in the actual world? This hinges on the modal force of (ST′). Of course, a possible world where some non-actual model of the space-time is realized and/or some non-actual object occupies space-time can still be a world in which (ST′) is true. In the first case, the expression ‘space-time’ would not refer to the actual space-time, but to the space-time of that possible world.

13 Please note that (OPP*) does not make all or most physicalistically unacceptable objects—such as God and the soul—dependent* on the physical. Indeed, the reasons why the existence of such objects is postulated or accepted seem not to be exhausted by what can be studied by physics. In addition, though (OPP*) is not equivalent to (PPP)—as there could be properties of substantivalist space-time not dependent on the physical—one could try to reformulate (PPP) in order to meet the equivalence demand as follows:

(PPP*) Every property can be studied by physics or entirely depends on properties that can be studied by physics—or, if it cannot be studied by physics nor does it depend on properties that can be studied by physics, still depends* on properties that can be studied by physics or on properties that entirely depend on the former.

Dependence* for properties can be formulated as follows: some property C depends* on some property/ies D if and only if accepting C’s instantiation is only needed in order to explain something about D’s instantiation—and there is no other reason to accept C’s instantiation. On substantivalism about space-time, see for example Pooley (2013).

14 This view is at odds with the theory that Hüttemann and Papineau (2005) try to defend, i.e. that physicalists need not be microphysicalists. More specifically, they accept that non-physical entities are identical with or metaphysically supervenient on physical ones (this is the physicalist aspect of their theory). Yet, they deny that macroscopic physical entities are asymmetrically determined by their microscopic physical parts and the microscopic physical laws that apply to those parts (this is the non-microphysicalist aspect). My concern with their view is that there is no principled way to distinguish between the properties of macroscopic physical entities that are physicalistically acceptable and the ones that are unacceptable. For example: it is prima facie compatible with their view that a macroscopic physical entity has emergent properties not determined by its microscopic physical parts and laws. This seems to be in stark contrast with physicalism. On this point, see also note 5 above.

15 This methodological aspect of physics is obviously stressed in many handbooks, including Young and Freedman (2012), 2. For a discussion of the methodological aspect from a philosophical viewpoint, see Goff (2017).

16 My purpose here is not to defend Jackson’s argument. I only wish to show that a certain characterization of the mental makes the conclusions of such an argument incompatible with physics physicalism. For some responses to Jackson’s argument on behalf of physicalism, see for example Nida-Rümelin (2015).

17 For an exhaustive investigation of phenomenal properties (or qualia), see Tye (2018).

18 In principle, besides phenomenal properties, any property having irreducibly qualitative and first-personal (and/or private) aspects turns out to be at odds with (M). Phenomenal properties are not just taken here as those properties to which subjects that instantiate them have privileged access (see, for example, Swinburne 2009, 502) or that are not recognised by those subjects as physical properties. Having privileged access to a property does not imply that that property cannot be reduced to a physical property. For an interesting critique of certain form of a posteriori physicalism based on the existence of phenomenal concepts, see Goff (2011). For a reply, see Elpidorou (2016).

19 It might be objected that the Copenhagen interpretation of quantum mechanics actually introduces mental acts of measurement by the observer that produce wave function collapses. However, if this interpretation were true, two options would be open for physicalists. First, the Copenhagen interpretation would just make physicalism false. The mental acts would not entirely depend on the physical and the mental would still not be in line with some of the boundaries of physics. In particular, it would not be in line with (M). Such a consciousness would give publicly reproducible, empirically and mathematically testable results only ‘by courtesy’. Consciousness as such would not have such features: only its effects would have them. The second option consists in accepting consciousness as a dependent* object, in line with (OPP*) and (PPP*) below. On this view, consciousness—and all its features—would only be invoked for the sake of explaining what can be studied by physics. Therefore, consciousness—and all its features—would depend* on what can be studied by physics.

20 This means that a possible world with phlogiston is a world where physics physicalism can still be true. For phlogiston is something that can be studied by physics. Thus, its existence would not falsify by itself physics physicalism.