Content uploaded by Phila M. Msimang
Author content
All content in this area was uploaded by Phila M. Msimang on Nov 13, 2016
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=rsph20
Download by: [Phila Msimang] Date: 23 September 2015, At: 08:53
South African Journal of Philosophy
ISSN: 0258-0136 (Print) 2073-4867 (Online) Journal homepage: http://www.tandfonline.com/loi/rsph20
Problems with the ‘physical’ in physicalism
Phila Mfundo Msimang
To cite this article: Phila Mfundo Msimang (2015) Problems with the ‘physical’ in physicalism,
South African Journal of Philosophy, 34:3, 336-345, DOI: 10.1080/02580136.2015.1069662
To link to this article: http://dx.doi.org/10.1080/02580136.2015.1069662
Published online: 15 Sep 2015.
Submit your article to this journal
Article views: 5
View related articles
View Crossmark data
South African Journal of Philosophy 2015, 34(3): 336–345
Printed in South Africa — All rights reserved
Copyright © South African Journal of Philosophy
SOUTH AFRICAN JOURNAL
OF PHILOSOPHY
ISSN 0258-0136 EISSN 2073-4867
http://dx.doi.org/10.1080/02580136.2015.1069662
Problems with the ‘physical’ in physicalism1
Phila Mfundo Msimang
Philosophy, School of Religion, Philosophy and Classics, University of KwaZulu-Natal,
Howard College, Durban, South Africa
214525839@stu.ukzn.ac.za
Hempel’s Dilemma is a challenge that has to be met by any formulation of physicalism
that specifies the physical by reference to a particular physical theory. It poses the
problem that if one’s specification of the physical is ‘current’ physical theory, then the
physicalism which depends on it is false because current physics is false; and if the
specification of the physical is a future or an ideal physics, the physicalism based on it
would be trivial as it would be tautologously true, or because very little (if anything at
all) can be inferred from or about a physics that does not yet exist. I review the reasons
for thinking that the dilemma is a perpetual problem for currentist specifications of the
physical, then introduce the argument that the standard positions on the specification
question are wanting because they lack a generality which physicalism is generally
accepted to have. I end with a suggestion for a way forward for physicalism.
1. The specification of the physical and Hempel’s Dilemma
A basic claim of physicalism is that everything is physical. What we see here is that the content of
any version of physicalism is dependent, in part, on what is meant by ‘physical’. One question a
physicalist must answer is what it is to be physical. This query is known as the specification question.
To answer the specification question, many contemporary thinkers have argued that what is
meant by the term ‘physical’ is what some particular physical theory identifies as physical, meaning
that the physical is defined as whatever some particular physics says it is. In this answer to the
specification question, the term ‘physical’ is nothing more than a placeholder for a specific physics
and its terms. Two families of problems have come about from this specification of the physical by
reference to a particular physics.
I will discuss some responses to the specification question, which refers to a particular physics
as specifying ‘the physical’ as some particular physics. The first family of problems has to do with
the specification of the physical in terms of present-day physical theory, which is known as the
current physics option. The second family of problems has to do with specification in terms of
future or ideal physical theories, which is known as the future physics option. If, by having specified
the physical as physics, it is meant that current physics specifies what the physical is for us then
we know that our statements about the physical are most probably false in some respect or, at least,
incomplete and inconsistent because present-day physics is incomplete and inconsistent; and if our
specification of the physical is based on some future or ideal physics, we cannot specify what such
a physical theory would say about the nature of the universe or about anything in particular because
the content of future physics is unknown to us.
These two families of problems are called Hempel’s Dilemma. The first family of problems,
which concern the problems faced by a specification of the physical in terms of present-day physics
known as ‘current physics’, constitutes the first horn of the dilemma, and the family of problems
which concern specifications of the physical in terms of future or ideal physics constitutes the
second horn of the dilemma.
1 An earlier version of this paper was presented at the Annual Congress of the Philosophical Society of Southern Africa in Port Elizabeth,
South Africa, 12–14 January 2015. I would like to thank David Spurrett for his critical comments on drafts of this paper.
Downloaded by [Phila Msimang] at 08:53 23 September 2015
South African Journal of Philosophy 2015, 34(3): 336–345
337
I aim to show that the specification question in terms of Hempel’s Dilemma may be more
problematic than is usually realised. I will start by reporting why Hempel’s Dilemma is found to
be a devastating problem for formulations of physicalism. I expand on this by pointing out that the
‘current physics’ horn of the dilemma faces an intractable problem for its answers to the specification
of the physical. I end by introducing the argument that both the ‘current physics’ option and ‘future
or ideal physics’ option on the specification question are wanting because they lack a certain kind of
generality that physicalism is widely accepted to have. Very briefly, the view defended below is that
it is generally recognised that there is more than one physical way that physicalism could be true or
false, but no specification of what it is to be physical by reference to a single physical theory could
accommodate this generality. I show how the established strategies of answering the specification
question in the context of Hempel’s Dilemma lack this generality, which means that even successful
solutions to Hempel’s Dilemma would be unsatisfactory on other grounds.
1.1 The first horn: the currentist position
The first horn of Hempel’s Dilemma, the ‘current physics’ option taken by currentist physicalists,
concerns the charge of the incompleteness, inconsistency, and the probable falsity of the physicalist
thesis on the basis of the incompleteness, inconsistency, and probable falsity of our present-day or
current physics.
The general problem of the ‘current physics’ option in specifying the physical is that it takes on
the problems of current physics. Since current physics has not discovered all the relevant physical
laws and all fundamental physical properties that are in existence, there are physical things which
cannot be explained from physics or categorised as physical since they are not results of any known
physical laws or processes. If ‘physical’ means current physical science, then undiscovered physical
facts are not (yet) physical. Our knowledge of the physical is also incomplete in another way: we
do not know how each of our physical categories relate to one another (for instance particle physics
and gravity) and whether there is an actual relation rather than an assumed one. It is possible that
some of the sub-categories of the physical cannot be reconciled and must be accepted as separate
and fundamental.
Furthermore, there are worries that physicalism as a thesis may be uninteresting. One of these
concerns is based on the finding that some results in present-day physics seem to contradict one
another, making the case for using current physics in specification invalid (since we cannot accept
both p and not-p simultaneously). But this worry is contrasted with the view that current physics is
successful in many respects. Its present success is significant enough to warrant support (Melnyk
1997) and provides evidence enough for a strong case for the current physics option (Bokulich
2011). It is also expected to extend its successes to other problems and resolve its contradictions.
Because all physical theories we have had in the past have been false, the pessimistic worry
is that the theories we come up with are and will continue to be false (e.g. Laudan 1981); but, in
the same breath, because our theories have been improving throughout history, the optimistic view
is that we may one day have an ideal physical theory (e.g. Chomsky 1968). The question we will
deal with here is about what we can say today for or against physicalism. Specifically, the question
concerns the grounds on which we can base a claim that current physics is a viable option.
An instructive example of the current physics option is Andrew Melnyk’s (1997) account of
the dilemma and its purported overcoming. He argues that the first horn of Hempel’s Dilemma
is ‘blunt’ and leaves open a way to construct a physicalism that succeeds against the difficulties
commonly associated with the currentist option. Melnyk accepts the legitimacy of the second horn
of the dilemma because he argues that nothing can be said about future or ideal physics. He says:
[physicalism] must yield a thesis (1) that is not obviously false and (2) that possesses content
determinable by us. (A physicalism whose content was not determinable by us would
presumably be impossible for us to support empirically, and might, for all we know, not
even exclude from existence the sort of paradigmatically nonphysical items—for example,
souls, entelechies, ghosts—which physicalists have traditionally refused to countenance.)
(Melnyk 1997: p. 622).
Downloaded by [Phila Msimang] at 08:53 23 September 2015
Msimang
338
He goes on to argue that we should support a current physics physicalism because it is the
best theory that we have about the world (so far) even if we acknowledge that it is incomplete and
probably false. He notes that it is a contentious issue that current physics physicalism is the best
theory we have, but he maintains that we can do no better.
Rather, Melnyk implores us to take on the same scientific realist and anti-relativist attitude
towards physicalism that we, as naturalists, take towards the best of our current scientific hypotheses.
He says that physicalism is a scientific hypothesis, or that we should at least treat it as one because
‘if physicalism is not a scientific hypothesis, then it resembles one in every respect relevant to
current purposes’ (Melnyk 1997: p. 625). The purposes of which Melnyk speaks are for physicalism
to ‘be exhibited as being explanatory in just the way we expect scientific hypotheses to be’ (ibid.).
Since Melnyk considers physicalism the best theory we have, or the most plausible out of
all its theoretical rivals even if it is not considered to be a plausible thesis, he argues that the only
sensible thing to do under such circumstances is to be a physicalist. That is backed up by the
naturalist attitude Melnyk sketches. This attitude is
(1) to regard the hypothesis as true or false in virtue of the way the mind-independent
world is, and (2) to assign the hypothesis a higher probability than that of its relevant rivals
(Melnyk 1997: p. 625, emphasis in original).
Melnyk suggests that it does not matter if physicalism is believed to be false because it is the
most plausible theory available. To be a physicalist, he says, ‘does not require regarding physicalism
as likely to be true (let alone very likely to be true)’ (Melnyk 1997: p. 625). But it is a serious
matter if physicalism is false (or believed to be false) because then the problem would be finding an
alternative to physicalism rather than just pursuing it because it is the best theory we have available.
Melnyk recognises this early on in his account and builds up his argument based around the
premise that physicalism should be treated the same way as any scientific theory. He says this means
that it is regarded as having a higher probability of being accurate than any of its relevant rivals.
Relevant rivals of a hypothesis are explained in this way:
Hypothesis H1 is a relevant rival to H2 if and only if (a) H1 is sensibly intended to achieve
a significant number of H2’s theoretical goals; (b) the hypotheses, H1 and H2, fail to
supervene on one another; and (c) H1 has actually been formulated (Melnyk 1997: p. 626).
The argument goes that even though in the not-too-distant future we may have a physics which
is more comprehensive and accurate than the physics of today, which would mean that in that future
we may have a physicalism with a higher probability of being accurate than that of the present
day, future physics could not be said to be a relevant rival to present day physics because the two
hypotheses would presumably supervene on one another.2 Thus, according to Melnyk, currentism
would not be threatened.
Even so, fundamental developments in physical theory would still create problems for the
current physics option’s specification of the physical. An illustration may make the point:
Suppose we had a notion of the physical, x, with the content p, which is all things known to
be fundamentally physical. Imagine then that there was a discovery of a new fundamental entity
that was not implied or predicted by any of the present models of the physical associated with p. Its
nature would need to be investigated to find out if it were an entity explicable by any other entity in
the physical theory of x and, if not, if this entity is itself a fundamental physical entity. Assuming that
the new entity discovered was an addition to the known and predicted fundamental physical entities
in the universe, it would be a novel addition to the fundamental ontology of the physical theory of
x. Simplistically, let us say that all that was needed to accommodate this new entity was to change
the theory of x to p + 1. Yet the discovery of this new entity would give rise to a ‘new’ physicalism
fundamentally different to other versions of physicalism because the specification of the physical,
2 This would be the case if the new power discovered were an underlying sub-quantum force, for instance, because quantum powers would
simply supervene on this sub-quantum power, and so there would be no signicant change in the specication of the physical except to say
that the compositional structure of the physical is more ne-grained than it was at rst assumed. So long as the force is physical, it poses
no threat to physicalism so construed.
Downloaded by [Phila Msimang] at 08:53 23 September 2015
South African Journal of Philosophy 2015, 34(3): 336–345
339
taken to be that which composes all else, would have to change to say that the physical is p + 1
instead of saying that it is just p. The notion of the physical would change to p + 1 in this specification
of the physical rather than, say, just the content of x as the physical being that which changes.
Why this causes a problem for the current physicalist like Melnyk is because the fundamental
ontology that is p would not supervene on the ontology of p + 1 because the additional ontological
element + 1 is absent in the ontology of p. Although both are sensibly intended to achieve the same
theoretical goals—that is, to give content to the notion of x, the physical—they do not supervene
on each other. Once such a theory is formulated, it causes a significant problem for Melnyk’s
physicalism although it would not necessarily cause a problem for physicalism in general (unless,
for instance, the fundamental entity is mentality).
What it means to be a physicalist would not have continuity through changes in fundamental
ontology. Rather, it would be punctuated by the meaning of new discoveries and developments in
the field of fundamental physics. Anticipating this line of argument, Crook and Gillett (2001: p.
342) explain that
[c]urrentist formulations of physicalism are incapable of progress with their central theses,
namely their accounts of the ontologically fundamental entities. CURRENTISM defines
the basic physical entities as just those of current physics, no more or less. Any newly
discovered particle is clearly not amongst these entities and very often not composed
by them, thus such an entity’s discovery will force the comprehensive abandonment of
physicalismCUR [currentist physicalism] and this hypothesis consequently has no ability to
progress with its central claim. Since the possibility of progress is an obvious condition on
good standing for incomplete theories in the sciences, we can thus see…why physicalismCUR
fails to overcome Hempel’s dilemma.
This criticism may not be seen as an insurmountable problem for the current physics option
if the claim of physicalism is attached to the best theory of whatever particular time it is put to
question, and that would be the process of development in the thesis of physicalism. The current
physics of tomorrow would replace today’s physicalism not as a continuation of the thesis of
physicalism but as a replacement of it. Physicalism would be whatever the physics of the day says it
is irrespective of other physical theories that give different specifications of the physical.
The weakness of this position is its lack of generality over different versions of physicalism,
requiring that the physicalist thesis and its relation to other versions of physicalism be established
on an ad hoc basis. An associated cost of this move is not overcoming the dilemma once and for
all but having to meet it anew each time there is a change in physics, or that it experiences the
challenges of the future physics option as the future approaches. (I will look at the shortfalls of this
position later on in this paper.)
The finding here is that physicalism constructed around current physics like Melnyk’s is either
a thesis containing its own redundancy and reasons for abandonment, or it is a work in progress
which will only be completed at the end of the scientific inquiry of the physical. The first option is
devastating for the current physics option, but the second option is just another way of expressing
that the answer to the specification question would be deferred to the consequences of some future
physics. That is,
if current physics is not true, but only approximately true, then…some future physics will
do a better job of characterizing these entities—that is, will be more approximately true.
Hence, there is effectively no motivation for the physicalist to characterize the base set in
their foundationalist ontological thesis as only adverting to current physics. Even if current
physics is approximately true, reference to future (and in the supposed limit of inquiry,
ideal) physics is needed in recognition of the fact that current physics hasn’t yet gotten it
entirely right (Wilson 2006: pp. 66–67).
Although versions of currentism such as Melnyk’s meet the condition that the content of
physicalism must be known by individuals in the present, if currentist physicalism is not to be false
and abandoned, we may only answer the specification question of physicalism with some reference
Downloaded by [Phila Msimang] at 08:53 23 September 2015
Msimang
340
to the future or at the end of the scientific enquiry of the physical. Such attempts at answering the
specification question and overcoming the first horn of Hempel’s Dilemma put physicalism in the
line of fire of the criticisms which plague the future option of the dilemma (the second horn of
Hempel’s Dilemma).
1.2 The second horn: the futurist or ideal position
The second horn of Hempel’s Dilemma is a challenge to any specification of the physical based
on a future or an ideal physics. The specification of the physical based on an ideal physics has
problems distinct from those that arise from a specification of the physical based on future physics.
Nevertheless, they can be dealt with together because there are some general criticisms that affect
them both with equal seriousness.
The specification of the physical in both ideal and future physics is based upon the assumption
that we will, sometime in the future, have a physics that will be superior to current physics. One
position on the specification based on a future physics goes as far as to assume that we will have
an ‘ideal’ physical theory that will have ideal expressive or explanatory powers, whereas the other
position assumes that our future physics will be a more accurate and more complete physics.
But what can be said about this future physics? Can we state its propositions, and can we say
what problems it deals with (or how it may even do so)? Such a physics would not be available to
us for specification because we do not know what it is specifying. The point of the second horn can
be put this way:
Suppose then we don’t know what physicalism says—what follows? Well, if we do not
understand it at all, we are in no position to know it, no position to deny it, no position to
believe or disbelieve it with justification. Nor are we even in a position to speculate about
whether it is true. In fact the whole project of rationally assessing physicalism…seems to
presuppose that we know what it is, at least in outline. But if that presupposition is false,
physicalism is unworthy of assessment (Stoljar 2010: p. 100).
Jessica Wilson, suspicious of whether or not there is a point to the second horn of the dilemma,
says that
[o]n one reading, the worry is that a physicalism based (only) on future (ideal) physics does
not have determinate content, since we don’t know what entities future (ideal) physics will
treat
and that ‘[o]n another reading, the worry is that such a lack of determinate content will render
physicalism trivially true’ (Wilson 2006: p. 67).
Without an independent or systematic criteria of what would satisfy the specification question
of the physical, physicalism would become open to what Wilson aptly called the ‘inappropriate
extension’ worry. The inappropriate extension worry is drawn from Chomsky’s comment that the
notion of the physical will be extended to incorporate whatever can be explained (see Chomsky
1968). This means that if we find out that the cause of sub-atomic movements are conscious
forces of the will of quantum particles, for instance, those ‘conscious forces of the will’ would
be categorised as physical forces and incorporated into the fundamental physical ontology of
physicalism’s physical theory. Chomsky’s position, which allows such unqualified extensions of
the notion of the physical, has been challenged to be unacceptable to physicalism and has even been
dismissed as inappropriate (Wilson 2006).
Even so, taking this inappropriate extension seriously, physicalism would not necessarily be
trivially true. We could merely have a new physical theory specifying the physical for physicalism,
but the question of whether or not that physical theory explains the instantiation of all phenomena in
the universe would still remain. That is,
[e]ven if a future physics-based account of the physical placed no restrictions on what
features the relatively fundamental entities treated by future physics could have, the question
of physicalism’s truth would still depend on the entirely separate question of whether all the
Downloaded by [Phila Msimang] at 08:53 23 September 2015
South African Journal of Philosophy 2015, 34(3): 336–345
341
relatively non-fundamental entities not treated by future (ideal) physics were or were not
over and above the relatively fundamental entities treated by future (ideal) physics. Hence
a future physics-based account of the physical isn’t in danger of trivializing physicalism
(Wilson 2006: p. 68).
Wilson argues that the point of the second horn, ‘[i]f there is a point to the second horn’ (p.
68), is the concern that physics may posit mental or intentional properties for the specification of the
physical if physics develops the way Chomsky proposes it will. To answer this worry, it has been
proposed that all that needs to be done is to restrict the content of the physical which physicalism
can have, rendering it impossible for mental properties to be part of physicalism’s fundamental
ontology.
This way of facing the challenge of specification in the face of Hempel’s Dilemma, established
by Spurrett and Papineau (1999), has come to be known as the ‘via negativa’ approach, meaning
that the challenge is to be met ‘by the negative route’. Wilson (2006) discusses a general problem
with accounts specifying the physical using the via negativa approach. She explains that over and
above the constraint that there cannot be any mentality at the fundamental level of the physical,
there would have to be additional restrictions specifying what is not at the fundamental physical
level such as ‘no fundamental morality, no fundamental free will…no fundamental biology, and no
miraculous powers constraints’ (Wilson 2006: p. 75).
An account of what something is not does not necessarily tell one what something is. In fact,
the only way to get a correct or accurate characterisation of the physical in this way would be to
eliminate everything that is not physical in order to be left with only that which is physical (and then
to identify it as such). In other words, for the via negativa to work it must be an exhaustive account
of all that it is not physical and have a principled reason for its separation from the physical if it
were to escape the accusation of being unsystematic and ad hoc.
Instead of taking the negative route, Wilson argues that there is no need for the aforementioned
type of additional constraints on specification. Those categories which are to be excluded are not
physical in the first place but are physically composed, the argument goes, and so would not feature
in the specification of the physical or fundamental ontology. She argues that on analysis of notions
such as morality, aesthetics, and even chemistry and mentality, we find that they are not fundamental
and are dependent in their being on the configuration of more fundamental entities which are
physical entities (see Wilson 2006: pp. 74–76).
If the world turned out to be some other way—that there was at least one category that was not
supervenient on the physical as it has been specified by some physical theory—then physicalism
would be false because it would have been shown that not all things are ‘nothing over and above the
physical’. Not all phenomena are physically caused if something non-physical can be shown to be a
cause. That physicalism could be found to be false is part of the point because physicalism proposes
a thesis about how the world is and leaves open conditions that would demonstrate that the thesis is
mistaken.
Even though Wilson’s account may overcome the criticism of the unsystematicity of the via
negativa (in so far as it is equivalent with the ‘no fundamental mentality’ constraint), it remains
contested if this specification of the physical is specification enough as it is not particular about
what physical theories can be specified for the physical in physicalism. To be sure, the via negativa
is compatible with antiquated physical theories such as Newtonian physics without souls or
immaterial forces. The compatibility is not, in itself, problematic; rather, what is problematic is that
the via negativa is expected to specify a physical theory for physicalism. It does well to reinforce
one aspect of our notion of the physical as being a fundamental and (consequentially, by Wilson’s
account) non-mental, non-moral, non-aesthetic, etc., category but does not tell us what this category
that is the physical is by specifying any particular theory of the physical. That question remains.
Thus, in its failure to specify a physical theory, the via negativa, even if it does overcome the future
horn of Hempel’s Dilemma, is insufficient to answer the specification question.
Downloaded by [Phila Msimang] at 08:53 23 September 2015
Msimang
342
2. The problems of recurrence and generality
In the earlier sections of this paper I discussed the challenge of Hempel’s Dilemma and explored
some responses to it. In the last two sections below I give an outline of two further challenges for
the specification question.
The essence of these problems may be seen as directing our attention to questioning the
legitimacy of Hempel’s Dilemma. Stoljar (2010) has taken the view that the problem is the
formulation of the dilemma itself, and argues that the dilemma requires restructuring for it to have
the relevance and force it is customarily assumed to have. Although I agree with his analysis, the
issues I am raising here are independent of his concern.
The first of these challenges strengthens the force of the first horn of Hempel’s Dilemma. This
challenge is based on the observation that revisions in physics—which result in a revision in what
is specified for the physical—can open up the questions facing the current physics option of the
dilemma at each instance of fundamental revision. I illustrated this point by example (see Section
1.1) and reiterate the reasons for this being a problem for the specification question in the section
below.
I follow this by outlining a general challenge to the specification question of which Hempel’s
Dilemma is a part. I argue that it is misguided to say that the thesis of physicalism is determined
by what a true or ideal physics says is a true physical theory, following Stoljar (2010), but continue
on to point out that the understanding of physicalism implicit in the thesis is more general than
what a true physics could specify. Because the thesis of physicalism is satisfied by a number of
physical theories, true or false, I have argued that our notion of physicalism should make an explicit
recognition of this fact by identifying general conditions to which all satisfactory formulations of
physicalism meet (even if our interest—or the version of physicalism we wish to formulate—is the
one which is supposedly true in this world).
There has been anticipation of this second kind of concern for physicalism (e.g. Stoljar 2001,
2010; Crook and Gillett 2001; Wilson 2006). But the criticism developed here is not towards
individual specifications of the physical (to which Hempel’s Dilemma is the standard challenge). The
argument here is that there is a problem in the lack of generality over answers to the specification
question even though there is a consensus that those answers to the specification question are
themselves physicalist in nature. In answering Hempel’s Dilemma, physicalists define physicalism
in terms of a single physical theory despite the fact that there is a consensus amongst physicalists
that physicalism is a general theory (Stoljar 2010).
2.1 The recurrence of the dilemma for currentism to the limit of inquiry
The problem of progress for the currentist position has been noted in the literature (e.g. Crook and
Gillett 2001). If the inability of currentism to continue with physicalism’s central claim means that
the thesis of physicalism has to be made anew every time there is progress in fundamental physics,
what does this mean for physicalism in the context of Hempel’s Dilemma?
As the first horn of Hempel’s Dilemma would dictate, it means that we would continually
have to ask: ‘Is this new physics we have come up with false, inconsistent, or incomplete so that
the physical theory it specifies for physicalism is false, inconsistent, or incomplete?’ The only time
these kinds of questions would be answered once and for all is at the limit of inquiry when current
physics would also be the last physics. Melnyk argued that his current-horn physicalism would not
face these problems because future versions of physicalism would supervene on one another. If the
physicalist’s ontology is incomplete, then it is not guaranteed that present and future versions of
physicalism will supervene on one another. (A thought experiment illustrating this point was given
in Section 1.1.)
But, in the future, physical ontology will be more complete than it is today. Physics is making
new discoveries and the physical is becoming better understood. It is rational to believe that the
physicalism of the future would be closer to the truth than current physicalism. That is
Downloaded by [Phila Msimang] at 08:53 23 September 2015
South African Journal of Philosophy 2015, 34(3): 336–345
343
the appeal to future physics is inspired by the fact that current physics is to some extent
inaccurate and/or incomplete; but similar considerations would apply to any version of
future physics antecedent to ideal physics (Wilson 2006: p. 71).
This is the problem of recurrence. The current physicalist has a defective ontological doctrine
that is incomplete and that does not necessarily supervene on more complete physicalist ontological
doctrines. Even so, the physicalist may want to be able to appeal to the complete ontology of
physics as what is true, having the current theory a rough approximation of that truth.
These considerations may lead one to believe that defining what physicalism is, is the same as
describing the conditions under which physicalism is true. If we were to accept that argument, the
physical in physicalism could only be what is specified by the true and final physics. One reason
this is an undesirable outcome is because it requires that it be a true physics that specifies a physical
theory for physicalism in order for us to say what physicalism is in the first place. It would imply
that we do not know what physicalism is today, and will only ever know what physicalism is once
we have an ideal physics to provide the thesis with content. There are two short comments I will
make extending this point.
Firstly, for physicalism to be defeasible, which is generally considered a requirement for a
satisfactory construal of physicalism (see Stoljar 2010: pp. 101–103), we must be able to say what
physicalism is today (as was suggested by Melnyk 1997) so we are able to judge its development
against things it is not (see Crook and Gillett 2001; Wilson 2006), or we may make the thesis of
physicalism defeasible by saying what physicalism cannot have as a part of its fundamental ontology
(as suggested by Spurrett and Papineau 1999) rather than attaching the meaning of physicalism to
whatever a true physical theory would necessitate physicalism would need to be to be true (in
contrast to Chomsky’s 1968 suggestion).
Secondly, having to wait for a true physics to formulate physicalism would mean that there is
no physical notion for which physics provides confirmatory or disconfirmatory evidence or a notion
of the physical which the developments of science informs. But physicalism has been formulated
before in the absence of a true and ideal physics, even though some of those formulations of
physicalism were found to be false. Historical examples are materialist theories. They are surely
false in what they posited fundamental ontology was, and were superseded by the new findings in
physics about the physical. But the ontological doctrine that the physical—which contemporary
physicalists determine by appealing to the findings of physics—is fundamental is retained through
all the developments of physicalism until a final physics ‘proves’ it to be true or a development in
physics shows the physicalist thesis to be false.
The basic issue with this position highlighted here is that every time there is a change in
fundamental physics, it necessitates that the physical’s fundamental ontology would have to be
said to be something new as its specification with regard to physicalism would have changed to
something relatively new (e.g. the fundamental ontology of x changing from p to p + 1). Because
of this change in specification, the currentist horn of Hempel’s Dilemma is never truly overcome.
The danger here echoes the inappropriate extension worry in so far as it is noted that it is
not only the content of our notion of the physical that could change if we purport there to be an
equivalence between physics and the physical of physicalism. It is also the conflation of the two that
allows the notion of the physical to become whatever physics becomes even if the kinds of notions
physics postulates change. Not only does this way assume that physicalism is a scientific hypothesis
(e.g. Melnyk 1997), it seems to be making the assertion that physicalism is making the same claims
as physics through having physics specify the physical. Whereas Hempel’s Dilemma is not relevant
to science, it is relevant to the specification question and every specification of the physical given
for formulations of physical.3 And, within this context, we find that every specification of the
physical prior to the attainment of an ideal physics will not be able to overcome the currentist horn
of the dilemma because it recurs with each fundamental development of physical science (see
Section 1.1).
3 It should be noted that physics does not have the problem of development that physicalism does, giving us at least prima facie reason to
believe that there is some signicant difference between the two.
Downloaded by [Phila Msimang] at 08:53 23 September 2015
Msimang
344
2.2 The problem of generality for answers to the specification question referring to a single
physics
The established strategies of overcoming Hempel’s Dilemma or fulfilling the specification question
are wanting because they lack a generality, which is a consensus amongst physicalists. Both
the currentist and futurist positions work on the assumption that physicalism is what the true or
approximately true physics of a particular world says it is (Stoljar 2010: p. 107)—that is, physicalism
is defined by what a true or a currently approximately true physical theory says it is. This leaves out
the fact that there are other ways the world could be which would satisfy physicalism even though
we know them to be false.
Stoljar anticipates this kind of criticism when he highlights that the specification problem,
which Hempel’s Dilemma is meant to be dealing with, concerns what physicalism is, and not whether
or not any specific specifications of the physical for physicalism are true (ibid.). The question of
physicalism’s truth and its definition are two separate matters. While Stoljar uses this failing in
Hempel’s Dilemma to formulate his own version of the dilemma facing the specification question,
my point is to show that there is a lack of generality over all the kinds of answers that would satisfy
the specification question (even if those specifications would not necessarily overcome Hempel’s
Dilemma, be deemed as true, or overcome Stoljar’s new construal of the dilemma).
There is a consensus on the fact that there are numerous ways that the world could be that
would imply that physicalism is true—or, to put it differently, that there are a number of false
physical theories which could support physicalism. For instance, the mechanistic view of the world
in which causes are forces created by the collision of particles, such as in Hobbes’s physical theory,
could support physicalism. So could the mechanistic Newtonian view (i.e. without the spiritual and
immaterial forces) and Democritus’ atomistic view. What I mean by this is that in possible worlds
where some version of one of these physical theories were true, and everything that there was in the
world was physical, then physicalism would be true. But nobody thinks that any of these physical
theories is actually true. The content of ‘physics’ that informs our judgements that physicalism could
be true in these worlds, then, is not tied to one particular physical theory. Their satisfaction of the
specification question stands irrespective of their validity in describing the actual world we inhabit.
The question is: ‘what is it that unifies all these kinds of accounts as “physicalist” accounts, or how
is it that these accounts satisfy the conditions of a physicalist theory?’
It cannot be because one specific physics is true or that all of the other theories are good or bad
approximations of the true ideal theory. Physical theories are diverse and sometimes incompatible
with one another even though they may be compatible with the thesis of physicalism. Since there
is more than one way that the physicalistic thesis could be fulfilled, just the physicalist theory
that could be true in its description of this world we inhabit would be an incomplete definition
of physicalism at best but defective because of its failure to recognise other physical theories as
sufficient for a coherent physicalist position.
As Stoljar pointed out, ‘to generate a problem for the formulation of physicalism from within
a temporal or historical framework’ (ibid.), as is done by Hempel’s Dilemma, is a mistake. The
physics of the past is wrong because the physical theories it specifies for explaining the instantiation
of phenomena are flawed and because those theories cannot account for some physical phenomena.
We know that this trend in the development of physics and the physical theories thereof is most
likely to continue. Even though other more viable options for fulfilling the content of the physical
became available as we moved further into modernity, and more viable notions of physicalism
may arise with the progress of physics today, we should have a paradigmatic understanding of this
progress in relation to physicalism: we need a criteria which categorises all these specifications
of the physical as (versions of) physicalism. One way in which this may be achieved is through
a different kind of specification of the physical from that which is usually given as an answer to
Hempel’s Dilemma.
3. Concluding remarks
I have presented Hempel’s Dilemma and some of its standard challenges for physicalism. I then
revealed that whether or not the first horn of the dilemma is surmountable is a problem subordinate
Downloaded by [Phila Msimang] at 08:53 23 September 2015
South African Journal of Philosophy 2015, 34(3): 336–345
345
to the concern that the current physics option cannot give a conclusive answer to the specification
question (until the end of scientific enquiry). This connects to the issue of physicalism not being
able to continue with its central claim (Crook and Gillett 2001). I noted that even if we had an ideal
physical theory to specify, it would not solve the problem of stating what physicalism is (Stoljar
2010: p. 107) because even false physical theories satisfy physicalism. In this regard, I have argued
that the root of the problem is that there is no generality over satisfactory answers to the specification
question and suggested that answers that go the way of Hempel’s Dilemma can be misguiding in
their attempt to answer the specification question because they purport to be definitive stances on
the definition of the physical and not just historically situated instances of a more general stance on
what physicality is.
References
Chomsky, N. 1968. Language and Mind. New York: Harcourt Brace and World.
Bokulich, P. 2011. Hempel’s Dilemma and domains of physics. Analysis 71(4), pp. 646–651.
Crook, S. & Gillett, C. 2001. Why Physics alone cannot define the physical: Materialism
metaphysics and the formulation of physicalism. Canadian Journal of Philosophy 31(3), pp.
333–359.
Laudan, L. 1981. A Confutation of Convergent Realism. Philosophy of Science 48, pp. 19–49.
Melnyk, A. 1997. How to Keep the ‘Physical’ in Physicalism. The Journal of Philosophy 94(12),
pp. 622–637.
Spurrett, D. & Papineau, D. 1999. A note on the completeness of ‘physics’. Analysis 59(1), pp.
25–29.
Stoljar, D. 2001. Physicalism. Stanford Encyclopedia of Philosophy. Zalta, E.N. (ed.). http://plato.
stanford.edu/archives/win2014/entries/physicalism [Accessed 17 April 2014].
Stoljar, D. 2010. Physicalism. London & New York: Routledge.
Wilson, J. 2006. On characterizing the physical. Philosophical Studies 131, pp. 61–99.
Received 12 February 2015, revised 19 June 2015, accepted 2 July 2015
Downloaded by [Phila Msimang] at 08:53 23 September 2015
