he Necessity of Idealism
Aa
S
a a
Ty
G
In Idealism: New Essays in Metaphysics,
Kenneth L. Pearce and Tyron Goldschmidt (eds.),
Oxford University Press,
.
Introduction
We formulate a version of idealism and argue for it. Sections and explicate
this version of idealism: the world is mental through-and-through. Section spells
this out precisely and contrasts it with rival views. Section draws a consequence
from this formulation of idealism: idealism is necessarily true if true at all. Sections and make the case for idealism. Section is defensive: it draws from the
conclusion of Section to reply to a central, perhaps the central, anti-idealist argument. Section is on the ofense: it develops a new argument for idealism based
on the contemporary debate in philosophy of mind. he contemporary debate in
philosophy of mind has been dominated by physicalism and dualism. While idealism has been historically inluential, the contemporary debate has neglected it.
If the reason is that defenses of idealism have not engaged so much with the rest
of philosophy of mind, we hope to rectify that a litle.
Idealism
Idealism is the view that the world is in some sense mental. But not just mental: thoroughly mental, wholly mental, mental through-and-through. It is inconsistent with, e.g. the contemporary panpsychist view that while everything, from
humans to amoebas to quarks, has thoughts, those things also stand at some distance from one another and move over time and do other things that are not to be
understood mentalistically. he panpsychist world isn’t thoroughly mental. On
the other hand, Idealism is consistent with there being physical things; indeed, it
is consistent with the world being thoroughly physical, wholly physical, physical
through-and-through. Or, at least, it’s not trivially inconsistent with those claims.
It’s not obviously incoherent to suppose that the world is both thoroughly mental
and thoroughly physical. hat would be the case if every mental aspect of the
universe were itself physical, and vice versa.
To ix ideas, we introduce some terminology. A thing is concrete if and only if
it is spatiotemporal. A thing is mental if and only if it has some mental feature,
and purely mental if and only if it has only mental features. Likewise, a thing is
physical if and only if it has some physical feature, and purely physicalif and only
if it has only physical features.
Idealism is the view that all concrete things are purely mental. Physicalism is
the view that all concrete things are purely physical. Impurism is the view that no
We use feature to cover both (monadic) properties and (polyadic) relations. hus, one can
say “John has the feature of being six feet tall” and “Harry and Sally have the feature of being in
love”. hus, a thing is mental if and only if it has some mental property or stands in some mental
relation, a thing is purely mental if and only if it has only mental properties and stands in only
mental relations, and so on.
concrete thing is purely mental and no concrete thing is purely physical. While
impurism isn’t very well known, property dualism is, and the later is just a speciication of the former. We will soon explain why it is that those who describe themselves as ‘property dualists’ are indeed commited to impurism, and how their view
is a speciication of impurism. We will also soon explain why we have grouped
just these views together with idealism, and not, say, substance dualism.
If idealism is true, if all things are purely mental, then either eliminative idealism is true, according to which there are no physical things, or non-eliminative
idealism is true, according to which there are physical things, but they are purely
mental. Of course, according to the later, all physical features are mental features.
We do not deine mental feature and physical feature, and the debate about
these terms is too big to cover here (see Stoljar
). But we do assume that if a
feature entails being a mental thing, then it is a mental feature, and if a feature
entails being a physical thing, then it is a physical feature, where one feature entails another if it is not possible for something to instantiate the irst and not the
second. he converses of these twin assumptions are entailed by our deinitions
of mental thing and physical thing. If a feature is mental (physical), then necessarily anything that has it is a mental (physical) thing. So it entails being a mental
(physical) thing. he assumptions themselves, while not so entailed, are nevertheless very natural to make. At any rate, we will use the terms mental feature and
physical feature in such a way that this constraint holds.
Idealism is Necessarily True (If True At All)
While it might appear that idealism, physicalism, and impurism are contingent
claims about concrete reality, they are stronger than that. Each is a claim about
how all of reality must be (even in non-modal ways). Or, more cautiously, each is
equivalent to such a claim.
Berkeley oten emphasises that idealism does not just happen to be true. Idealism is necessarily true, and any opposing view is necessarily false—or worse,
unintelligible. Indeed, for Berkeley, “the very notion of what is called mater or
corporeal substance, involves a contradiction in it” (Section ). He’s even prepared
to give up idealism if anyone could show the possibility of anything existing outside the mind; he sets the bar for his opponents very low: “I say, the bare possibility
of your opinion’s being true, shall pass for an argument that it is so.” (Section )
here’s much to be said for Berkeley’s arguments. But ours is new. Here is a
quick argument for the conclusion that idealism is necessarily true if true at all.
You have the feature existence. If Idealism is true, then existence is a mental feature, since you are concrete and thus have only mental features. Now, necessarily
everything has existence. Since it is a mental feature, necessarily everything is a
mental thing. So then any feature whatsoever entails being a mental thing. Given
the assumption to which we drew atention at the end of the previous section—that
if a feature entails being a mental thing, then it is a mental feature—any feature
whatsoever is a mental feature. If that’s true, then necessarily, everything has
only mental features. So: if Idealism is true, then necessarily, everything is purely
mental.
You might try to escape this implication by denying, with Kant, that there is
any such feature as existence. Or you might accept that existence is a feature but
reformulate idealism as the claim that for every concrete thing, each of its nontrivial features is a mental feature. But neither of these maneuvers will work. he
authors of this paper, despite sharing much in common, are also diferent in many
respects. hus there is some feature that Aaron has that Ty lacks. Call it Bob.
We can give the following argument: if Idealism is true—even the reformulated
Idealism—then Bob is a mental feature, since Aaron is concrete, and yet he has Bob.
But so too is Bob’s negation. For Ty is concrete, and yet has Bob’s negation. Now,
necessarily everything has either Bob or Bob’s negation. Since both are mental
features, necessarily everything is a mental thing. So then any feature whatsoever
entails being a mental thing. Given the assumption to which we drew atention
at the end of the previous section—that if a feature entails being a mental thing,
then it is a mental feature—any feature whatsoever is a mental feature. If that’s
true, then necessarily, everything has only mental features. So: if Idealism is true,
then necessarily, everything is purely mental.
You might try to escape this implication yet again by denying the existence of
Bob or Bob’s negation. Perhaps there are no “negative features,” and one or the
other is negative. Or you might accept their existence but reformulate Idealism
as the claim that for every concrete thing, each of its positive non-trivial features
is a mental feature. But even if we accept this added bit of ideology, neither of
these maneuvers will work. here are sets of features all of which are positive and
whose disjunction is trivial. Consider, for example, these features: having exactly
one (proper or improper) part, having exactly two (proper or improper) parts…,
having exactly aleph-null (proper or improper) parts, having exactly aleph-one
(proper or improper) parts…, having exactly aleph-omega (proper or improper)
parts… If there are any positive features, these are presumably all positive. None
is trivial. And if one of these is a mental feature, then all are. hey all, ater all,
“say the same sort of thing”. Or if you think that one of those—perhaps the irst—is
not positive, consider these features: numbering exactly one, numbering exactly
two…, numbering exactly aleph-null, numbering exactly aleph-one…, numbering exactly aleph-omega… Some xs have the irst when there is just one of them,
We assume throughout that what features there are is not a mater that varies from possible
world to possible world—just which features are instantiated so varies—that necessarily all features
exist necessarily. We could assume otherwise without sacriicing any of the essential points in our
arguments. But doing so would complicate some of our formulations.
they have the second when there is exactly two of them, and so on. Even more
clearly now, if there are any positive features, these are presumably all positive.
None is trivial. And if one of these is a mental feature, then all are. hey all, ater
all, “say the same sort of thing”.
So we can give the following argument: if idealism is true—even the doubly
reformulated idealism—then (at least) one of them is a mental feature, since you
are a concrete thing and you have one of them. So all of them are mental features.
Now, necessarily everything has one of these features. Since they are all mental
features, everything is a mental thing. So then any feature whatsoever entails
being a mental thing. Given the assumption to which we drew atention at the
end of the previous section—that, if a feature entails being a mental thing, then
it is a mental feature— any feature whatsoever is a mental feature. If that’s true,
then necessarily, everything has only mental features. So: if idealism is true, then
necessarily, everything is purely mental. Indeed, our argument shows something
even stronger: if anything is purely mental, then necessarily everything is purely
mental.
We do not see room for any further escape maneuvers. Every further reformulation of idealism that avoids our argument collapses into some other view from
which it ought to be distinguished. If we reformulate idealism as the view that every concrete thing is a (perhaps impurely) mental thing, then idealism is too weak:
it collapses into panpsychism, and is consistent with the denial of all traditional
forms of idealism (such as Berkeley’s). If we reformulate idealism as the view that
every concrete thing is a mental thing and not a physical thing, then idealism is
too strong: it collapses into eliminative idealism, according to which there are no
physical things, and is inconsistent with all traditional forms of idealism (such as
Berkeley’s).
So idealism is a necessary truth, if true at all. Indeed, since each of the premises
in our argument is necessarily true, we can conclude that necessarily, if idealism is
true, it is necessarily true. And so idealism is either necessarily true or necessarily
Originally, we deined idealism as the view that all concrete things are purely mental–leaving
open the possibility of non-mental abstract things. his possibility is now foreclosed. We have
shown that idealism entails that absolutely all things are purely mental.
Given the ideology of perfect naturalness, you might reformulate idealism as the claim that
for every concrete thing, each of its perfectly natural non-trivial features is a mental feature. And
then you might go on to claim that the “numbering features”—numbering one, numbering two,
etc.—are not all perfectly natural; and, moreover, that none of those that is instantiated—which is
presumably every one of them—is perfectly natural (without that assumption, our argument still
goes through, as we continue to assume that all are mental if one is). We fail to see why we should
think that none of these features is perfectly natural. What is more, idealism so formulated is
again too weak. Any view according to which none of the numbering features is mental is not one
according to which the world is mental through-and-through, and is consistent with the denial of
traditional forms of idealism.
Our premises are necessarily true only if the empty world—a world in which there are no
concrete things—is impossible. his is inessential to what follows, but dispensing with it would
false. Let us therefore call it a necessary proposition. It is in this respect like many,
although by no means all, metaphysical views.
Our arguments for the claim that idealism is a necessary proposition also makes
clear why it is. It is because idealism, like physicalism, is equivalent to a view
wholly about what possible features there are—not just about what features are
instantiated, but about what features are available to be instantiated. Idealism is
equivalent to the view that all features are mental. Physicalism is equivalent to
the view that all features are physical. A third view says some features are not
mental and some features are not physical. hese views exhaust logical space.
And the third view is equivalent to impurism. If there is some feature that is
not mental, then possibly something is not purely mental. But then, as we argued
earlier, nothing is purely mental. And, if there is some feature that is not physical, then possibly something is not purely physical. But then, in light of what we
argued earlier, nothing is purely physical. Impurism, recall, is just the view that
no concrete thing is purely mental and no concrete thing is purely physical. So
impurism follows from the third view. he other direction is straightforward. Impurism is thus equivalent to a view about what features there are: the view that
some features are not mental and some features are not physical. As we noted
earlier, property dualists are commited to impurism. hey are commited to impurism because they are commited to there being some features that are not mental and some that are not physical. And as we noted earlier, property dualism is a
further speciication of impurism: it speciies that there are physical features, some
of which are not mental, and mental features, some of which are not physical.
he upshot: idealism, physicalism, and impurism exhaust logical space. Each is
necessarily true if true at all. And each is at least equivalent to a claim wholly about
what features there are. hey are to be contrasted with other claims in the vicinity
that are about what features actual things have and that have no implications for
the (non-modal) character of possible worlds other than our own.
Panpsychism is the view that every concrete thing is a mental thing. Weak
physicalism, sometimes called just physicalism, is the view that every concrete
introduce signiicant complications into our formulations.
his follows assuming the characteristic axiom of the modal system S .
Every impossible feature, if such there be, is mental (and physical), given the assumption that
any feature that entails being a mental thing is mental (and that any feature that entails being a
physical thing is physical). So a non-mental feature is a possible feature.
One might take issue with this claim about property dualists, at least taken in conjunction
with our assumption that if a feature entails being a physical thing, then it is a physical feature.
For some property dualists might hold that necessarily everything is a physical thing, which together with our assumption implies that every feature is a physical feature. We think property
dualists, properly so-called, are commited to the possibility of spirits along with the possibility of
zombies. But we invite so-called property dualists who deny the possibility of spirits to (a) state
what property dualism amounts to and (b) characterize physical property in such a way that the
conjunction of that statement and that characterization does not commit them to the possibility of
spirits. he task seems to us very diicult.
thing is a physical thing. Strong impurism is the view is that there are concrete
things that are not physical and there are concrete things that are not mental.
While strong impurism isn’t very well known, substance dualism is, and the later
is just a speciication of the former: it speciies that there are physical things, some
of which are not mental, and mental things, some of which are not physical.
Panpsychism, weak physicalism, and strong impurism exhaust logical space.
But none of them is equivalent to any claim wholly about what features there are.
And none has any implications for the (non-modal) character of possible worlds
other than our own.
In a nutshell, there are two debates: one about what kinds of things there are
(featuring panpsychism, weak physicalism, and strong impurism), and another
about what kinds there are for things to be (featuring idealism, physicalism, and
impurism). Since our topic is idealism, our focus is the later debate.
A Wrongheaded Argument against Idealism
here are various arguments against idealism; Berkeley dealt with more than
a dozen. But what we have shown—that idealism is a necessary proposition, that
it is equivalent to a proposition about the whole of modal space—undercuts a particular argument against idealism: the argument from parsimony. his argument
is plausibly the strongest anti-idealism argument.
he argument from parsimony appeals to the alleged fact that idealism—at least
Berkeleyan non-eliminative idealism—asks us to believe in more entities, or more
complicated laws, than does physicalism. Ater all, Berkeleyan non-eliminative
idealism asks us to believe in all the physical things and a supreme mind whose
thoughts constitute all those physical things, while physicalism asks us to believe
in just the physical things. And the laws that physicalism asks us to believe are
much simpler, at least when formulated in the language of mathematics, than those
that idealism will have us believe. But, goes the argument, we ought rationally to
assign greater prior probabilities to more parsimonious theories. So, seting aside
other arguments on behalf of Berkeleyan non-eliminative idealism, we should believe physicalism, rather than Berkeleyan non-eliminative idealism.
his argument is dubious. First, the premise that we ought rationally to assign
greater prior probabilities to more parsimonious theories is controversial, especially since the usual rationales for preferring parsimony do not so obviously apply
to grand metaphysical theories (see Huemer
). Secondly, one might suggest
that it’s parsimony at the fundamental level that ought to be correlated with our
priors, and Berkeleyan non-eliminative idealism is more parsimonious at the funMost, if not all, substance dualists would accept the stronger claim that there are physical
things and mental things and no physical thing is a mental thing (equivalently, no mental thing is
a physical thing).
damental level than physicalism, since fundamentally there is one simple supreme
mind. hirdly, the argument, even if successful, doesn’t tell against idealism, period, just against a speciic version of idealism. Fourthly and inally, the argument
doesn’t tell against even that speciic version, since Berkeleyan non-eliminative
idealism is consistent with physicalism. It makes litle sense to suggest that (ceteris paribus) we should believe A rather than B on the grounds that we ought
rationally to assign a higher prior probability to A than to B, when A is consistent
with B.
here’s much to be said about these objections: for example, we might skirt
the inal two objections by reformulating the argument from parsimony so as to
compare the degrees of parsimony of idealism and its negation. he argument still
fails for there are two more objections, each based on idealism being a necessary
proposition.
he irst is that, if parsimony is ever a guide to rational priors, it seems to us
to be so only with regard to contingent propositions, not necessary ones. What
bearing could parsimony have on the prior probability we ought assign to a theory
about the extent and character of the whole of modal space? We have reason to
accept that for a given contingent proposition, if it is more parsimonious than its
negation, we ought to assign it a higher prior probability than its negation (and
hence a prior greater than . ). he reason we have is that (a) more parsimonious
contingent propositions have a higher intrinsic probability—they “take up” more
of modal space—and (b) one’s priors ought to align with intrinsic probabilities.
But we see no reason to accept that for a given necessary proposition, if it is more
parsimonious than its negation, we ought to assign it a higher prior probability
(and hence a prior greater than . ).
An analogue of our reason about contingent propositions would be that (a’)
more parsimonious necessary propositions have a higher intrinsic probability and
(b’) one’s priors ought to align with intrinsic probabilities. But necessary propositions have an intrinsic probability of or , so (a’) would amount to the claim that
more parsimonious necessary propositions have an intrinsic probability of and
less parsimonious necessary propositions have an intrinsic probability of . But
that’s incoherent. Consider three necessary propositions, each more parsimonious
than the next. We would have the contradictory verdict that the middling one has
an intrinsic probability of and an intrinsic probability of : the later because it
is less parsimonious than the most parsimonious of the three, and the former because it is more parsimonious than the least parsimonious of the three. Moreover,
given that necessary propositions have an intrinsic probability of or , (a’) and
(b’) jointly imply that we ought to assign a prior probability of to the simplest
theory about the extent and character of modal space. hat is, not only ought we
assign a prior of greater than . to whatever the simplest theory of modal space
is, we ought to be certain of it. hat strikes us as highly implausible.
his objection runs against the current of contemporary metaphysics, where
rival necessary theories are oten preferred on grounds of parsimony. David Lewis
(
), for one prominent example, prefers his modal theory partially on grounds
of parsimony. However, the preference for parsimony is as oten had as it is let
unjustiied.
he second objection is that, even if we grant that parsimony is a guide to rational priors with regard to necessary propositions, idealism is a more parsimonious
theory than its denial. he denial of idealism says nothing (non-modal) about how
the actual world is. One could consistently deny idealism and accept that all physical things are constituted by the thoughts of a supreme mind and governed by
the very same laws the idealist believes. Where idealism difers from its denial is
with regard to what’s possible. But here the idealist is more parsimonious. Let a
stripped-down possible world be a set of propositions such that for every non-modal
proposition—every proposition wholly about how things in fact stand—either it or
its negation is in the set, and such that the conjunction of all the propositions in
the set is possible. Every set that the idealist takes to be a stripped-down possible
world, the denier of idealism does as well, but not vice versa. Intuitively speaking,
the non-idealist modal space outsrips the idealist one. his is because idealism is
or is equivalent to the view that every property is a mental property; and that is
equivalent to the view that every property entails being a mental thing. So idealism is the view that such-and-such an entailment holds. Its denial is therefore the
denial of such-and-such an entailment. It says there are possible worlds in which
something has some property, in which something exists, and isn’t a mental thing.
Idealism, therefore, is stingier with possibilities than its denial. On parsimony, it
wins.
Arguing for Idealism
here are many arguments for idealism; a few can be distilled from Berkeley, and there are many more throughout this volume. We present an argument
drawing on the debate in philosophy of mind; the argument piggybacks on the
anti-physicalist and anti-dualist arguments there. his is similar to the argumentative strategy Chalmers (
) employs for panpsychism or protopanpsychism:
the problems facing standard forms of physicalism and property dualism put some
pressure on us to ind a third option in panpsychism or protopanpsychism. Chalmers,
At least with respect to how simple reality is if it’s true. hat is the sort of parsimony at issue
in the anti-idealist argument from parsimony. here is another sort of parsimony that has to do
with the simplicity of the statement of the view (when stated in some canonical or metaphysically
perspicuous language): thus Lewis’s Principle of Plenitude, for example, might be more parsimonious than its denial, with respect to this second sort of parsimony, even though it guarantees more
possibilities. But that was not the sort of parsimony at issue in the anti-idealist argument. here
was never any reason to think that a statement of idealism (in some canonical or metaphysically
perspicuous language) need be more complex than that of its denial.
however, neglects the idealist alternative. he argument for idealism is negative:
idealism is let standing ater the alternatives of physicalism and impurism are
eliminated. he argument runs as follows: Either physicalism or impurism or idealism is true; if impurism is true, then property dualism is true; so either physicalism or property dualism or idealism is true; physicalism is false; property dualism
is false; therefore, idealism is true.
To be sure, there are other versions of impurism besides property dualism.
Ater property dualism, we ind the most plausible versions to be the view that
every mental feature is a physical feature, but not vice versa; that every mental
feature is a physical feature, and vice versa; and that every physical feature is
a mental feature, but not vice versa—and in each case there is some other nonmental, non-physical feature. he irst and second views would be eliminated by
the argument against physicalism addressed below. he third view we could live
with: it would require a complication of the argument of Section . For example, if
we take the other features to be moral features, then we could let idealism be the
view that everything is purely mental-or-moral, and conclude that, necessarily,
everything is mental-or-moral. he only other impurist possibilities eliminate the
mental or the physical. We ind both uterly implausible, and we could in any case
live with the later version of impurism.
Generally, we’ve framed the whole argument as a very deinitive elimination
of alternatives: various versions of physicalism and property dualism won’t do,
and so only idealism remains. hat’s neat. But more realistically and for most
readers, the argument should be a more modest cost-beneit analysis: the various versions of physicalism and impurism face certain big, even if not deinitive,
problems whereas idealism doesn’t. hat’s powerful reason to favor idealism.
A a
P y
a
here are two main contemporary arguments against physicalism. he irst is
the knowledge argument. he argument describes a subject knowing everything
physical about a domain without knowing everything about the domain. It concludes that there are mental facts beyond the physical facts. Jackson (
) portrays the subject as a future neuroscientist and the domain as color vision: Mary
learns all the physical facts about color vision, but she learns them in a black-andwhite environment. When she leaves the environment and has the sensation of
color for the irst time, she learns something new. Since she already knew all the
physical facts, what she learns must be some non-physical fact. Assuming there
could be such a subject as Mary and the basic physical features in her environment are the same as in ours, the physical features in our environment are not all
the features of concrete things; in particular, the mental features of color vision
are not physical features. hus, not all concrete things are purely physical. hus,
physicalism is false.
he second argument is the zombie argument. he argument describes the
possibility of beings that are physically identical to us, but mentally short of us:
they lack certain mental features altogether. Chalmers (
) has them lacking
the same kinds of mental features of sensation Mary learns about. he argument
concludes that there are mental facts beyond the physical facts. he possibility of
zombies is supported by conceivability arguments: roughly, zombies are conceivable; if they are conceivable, then they are possible; thus, they are possible. here
could have been beings with the same basic physical features as us but without
all the mental features. here must be something to us over and above the physical features. Certain mental features of concrete things are not physical features.
hus, not all concrete things are purely physical. hus, physicalism is false.
W a
A
S
w
he arguments against physicalism are usually taken to be arguments for property
dualism, and by eliminating an alternative view, the arguments can indeed be a
part of a case for property dualism. However, if they help support dualism, then
they also help support idealism. For idealism is as much an alternative to physicalism as is dualism. he arguments are supposed to show that property dualism
is true, but all they really show is that physicalism is false.
he arguments leaves three options open. he irst two are versions of idealism. he irst is that absolutely no concrete things are physical, that all concrete
things are purely mental. his is an eliminative idealism which denies the existence of the physical. he second option is that concrete things have both physical
and non-physical features, and that the physical and the non-physical features are
all mental. his is a non-eliminative idealism that countenances the existence of
the physical. Just as non-eliminative physicalism takes thoughts and feelings to
be physical, so non-eliminative idealism takes mass and charge to be mental.
he third option is that some things are mental and physical, but not purely
mental or purely physical. his property dualism is indeed inconsistent with idealism. However, property dualism in no way follows from the arguments as stated.
For property dualism to follow the arguments would have to assume both that
there are physical things and that they are not purely mental. hat is a mere assumption. In taking the arguments to show that property dualism is true, idealism
is assumed to be false. Most readers make that assumption. But, so far as the arguments go, there’s nothing to support this. he same point holds for arguments
that are supposed to support substance dualism: then there will be the mere assumption that some things are not purely mental. Simply: the arguments in each
case try to show that the mental cannot be reduced to the physical. hey then just
assume that the physical cannot be reduced to the mental. he later claim is just
as crucial for the dualist conclusion. However, it is never defended.
Property dualists devote no time to arguing that some things are not purely
mental; as we have just seen, this is taken for granted. In contrast, they devote all
their time to arguing that some things are not purely physical. If it is ever because
the property dualist is most interested in the fundamentality of the mental, then
idealism gives the property dualist what they most want.
A a
P
yD a
he main contemporary argument against property dualism is the causal closure
argument. he basic idea here is that the physical causal nexus leaves no room for
distinct mental causes; if the mental does any causal work, it must be physical. he
instantiation of mental features of all kinds—experiencing, feeling, believing, desiring, both what Chalmers (
: - ) calls the phenomenal and what he calls
the psychological—have physical efects. hat’s supposedly common sense. But
next comes the closure premise: every physical efect has as a suicient cause the
instantiation of some physical features. hat’s supposedly what science tells us.
So if some mental features were distinct from the relevant physical features, then
the physical efects would be systematically overdetermined, having distinct yet
suicient mental and physical causes. And that would be a crazy cosmic coincidence. So all the mental goings on are purely physical; that is, every mental feature
must be identical to some physical feature (see Papineau
).
We’re oten told that there is a problem about understanding how diverse kinds
of things could causally interact. Berkeley himself appeals to the problems of causation in defense of idealism: “But how mater should operate on a spirit, or produce any idea in it, is what no philosopher will pretend to explain. It is therefore
evident, there can be no use of mater in natural philosophy” (Section ). he
problem usually comes down to sheer puzzlement about how distinct mental and
physical features could interact. his doesn’t make for much of an argument. Ater
all, diverse kinds of things oten causally interact; if you think about it, the moon
and the tides are quite diferent. here are, however, more worked out arguments,
and the closure argument locates the puzzle and does so with scientiic support.
W a
A
S
w
he knowledge and zombie arguments shows that physicalism is false, without
showing that property dualism is true. In contrast, the closure argument, if successful, shows not only show that property dualism is false, but that physicalism is
true. hus, the closure argument is not consistent with the knowledge and zombie
arguments; we cannot consistently believe the premises of the closure argument
along with those of the knowledge or zombie arguments. However, our strategy
does not require as much. We require merely the truth of the closure argument’s
premises conditional on property dualism.
he idealist need not accept the closure argument. Ater all, if idealism is true,
then the closure premise is unmotivated. If idealism is true, then every physical
feature is a mental feature; the physical is reducible to the mental. And we then
have no reason to believe that being a physical thing is fundamental, or a natural
kind, or carves nature at the joints. We have no reason to believe that the set of
physical things is any less miscellaneous than the set of things God like to imagine.
And that means that our total empirical evidence does not support the closure
premise; the evidence supports only the weaker claim that every mental efect has
as a suicient cause the instantiation of some mental features.
Compare: a second century sage realizes that all the tongs he has ever seen
that are brought into existence, are brought into existence with the aid of tongs.
But suppose he has no reason to believe that being tongs is fundamental, or a
natural kind, or carves nature at the joints: the set of tongs, ater all, are a relatively miscellaneous set of things. hen it seems that his tongs-related empirical
evidence does not support the claim that all tongs that are brought into existence
by something are brought into existence by tongs. He might reasonably conclude
that some tongs, perhaps the irst ones, were brought into existence directly by
God.
In both of these cases—the case of being a physical thing according to the
idealist and being tongs according to everyone—the properties (or relevant predicates) are not ‘projectible’. And they are not projectible because they do not carve
nature at the joints (uine
:
- , Sider
: ) Or at least we have no
reason to think they do.
However, the property dualist does think being a physical thing is fundamental, that it is a natural kind, that it carves nature at the joints. And so she cannot
bypass the causal closure argument in this way. If property dualism is true, then
the causal closure argument succeeds, and physicalism is true. But physicalism entails that property dualism is false. hus, if property dualism is true, then property
dualism is false. hus, property dualism is false.
What of a property dualist who thinks that physical features are reducible to some third kind
of x-ish feature, so that being x-ish is fundamental while being a physical thing is not? his
property dualist faces a dilemma. Either some mental feature is not reducible to an x-ish feature,
or all mental features, like physical features, reduce to x-ish features. If some mental feature is not
reducible to an x-ish feature, then the closure argument tells against the view. For our empirical
evidence supports the claim that every x-ish efect has as a suicient cause the instantiation of
some x-ish features, since being x-ish is fundamental. If, on the other hand, every mental feature
is reducible to an x-ish feature—but, per property dualism, there is some physical feature that is
not a mental feature—then extended versions of the knowledge and zombie arguments tell against
the view. For those arguments suggest not just that there are mental features that are not physical
features, but that there is no natural kind, other than the kind, being a mental thing, such that
every mental feature is a feature of that kind. (See our discussion of panprotopsychism at the end
of this section.) So being x-ish is just being a mental thing, and physical features are reducible to
mental features ater all.
U
a
C a
As noted, David Chalmers deploys a strategy very similar to ours, in order to argue
for panpsychism or panprotopsychism. Panpsychism is the view “that some fundamental physical entities have mental states” (
:
) and panprotopsychism
is the view that “fundamental entities… have certain special properties that are
precursors to consciousness and that can collectively constitute consciousness in
larger systems” (Ibid:
).
At the heart of his maneuver is the distinction between dispositional features
and the categorical features that confer them, or, more broadly, between roles and
realizers of those roles. Narrow physical properties (in our terms, narrow physical features) are “microphysical role properties, such as the dispositional property
associated with having a certain mass, or the second-order property of having a
property that plays the mass role” (
:
) while broadly physical features are
the microphysical role features together with the features that play those roles.
Narrow physicalism is the view that all features reduce to narrowly physical features, while broad physicalism is the view that all features reduce to broadly physical features.
Chalmers appeals to both the zombie and closure arguments. By his lights,
the zombie argument establishes the falsity of narrow physicalism, not of strong
physicalism. For when we conceive of beings that are physically identical to us, but
mentally short of us, we are conceiving of beings that are only narrowly physically
identical to us, not broadly physically identical to us. he closure argument, on the
other hand, establishes the truth of broad physicalism, not of narrow physicalism.
For we can easily reject the causal closure of the “narrowly physical domain,” but
we cannot very easily reject the causal closure of the “broadly physical domain”.
hus, we can accept the premises of both arguments, as long as the term physical
is properly disambiguated.
What emerges from this maneuver is that all features, including mental ones,
are reducible to broadly physical ones, while there are some features, mental ones
in particular, that are not narrowly physical. So there are mental features, or features to which mental features reduce, that realize the microphysical roles, and are
not themselves the roles. Chalmers takes this to establish panpsychism or panprotopsychism. But if it establishes that disjunction, then it establishes panpsychism.
And if it establishes panpsychism, then it establishes idealism—or, as explained
above, something near enough. So Chalmers should take his case to support idealism or something near enough, as we do, not just panpsychism, as he does.
If it establishes either panpsychism or panprotopsychism, then it establishes
panpsychism. Since if panprotopsychism is true, then panpsychism is true. For,
according to panprotopsychism, mental features reduce to protophenomenal features. However, mental properties do not reduce to any properties other than
mental properties. his is shown by extended versions of the knowledge and zombie arguments: Mary could know everything about the non-mental properties and
still learn something new when she has her irst sensation of color (see Nagasawa
:
- ); zombies could still be non-mentally identical to us, but mentally
short of us. Chalmers denies that such extended arguments will work: we cannot
so positively conceive of, and come to verdicts about, the relevant scenarios with
the protophenomenal features. However, so far as we can tell, once we realize
that the protophenomenal features are, taken individually, no more mental than
narrowly physical features, the extended arguments are just as plausible as the
original arguments.
For what it’s worth, idealism also has the advantage of avoiding the inscrutability of the categorical features of things panprotopsychism saddles us with. If idealism is true, we know what all feature reduce to—mental features. We have some
idea of what such features are like, and idealists have detailed views about how all
features might reduce to the mental. In contrast, if panprotopsychism is true, we
have no idea what kind of categorical features there are or how mental features
reduce to them. Whereas it was always regarded as an advantage of idealism that
it makes the nature of things more transparent, panprotopsychism lands us in total
mystery.
So we should setle with panpsychism rather than panprotopsychism. But if
Chalmers’s maneuver establishes panpsychism, then it establishes idealism. For
either narrow physical features are fundamental or they are not. If they are fundamental, then the maneuver fails. For now we cannot very easily reject the causal
closure of the narrowly physical domain, and systematic overdetermination looms.
So then the maneuver does not establish panpsychism. If, on the other hand, narrowly physical properties are not fundamental, then, if the maneuver succeeds,
they reduce to those features that are broadly physical but not narrowly physical. And if the maneuver establishes panpsychism, then the only such features
are mental features. So all narrowly physical features reduce to mental features;
and all physical features that are not narrowly physical are mental. So all physical
features, period, reduce to mental features. And that’s idealism, or something near
enough.
T
M
a
here is much to be said for and against all the knowledge and zombie arguments.
here is much to be said for and against the closure argument and, thus, for and
against our conditionalized closure argument. But we can hardly enter these debates here; a fortiori we can hardly show that the arguments against physicalism
and the arguments against dualism work. We set forth the arguments only to
show how they can be used as a part of the case for idealism that we sketched
at the beginning of this section: if the arguments against physicalism work and
he point is beyond the scope of this essay, and we’d be happy enough to leave a few readers
with the options of idealism and panprotopsychism.
the arguments against dualism work, then idealism, or something near enough,
remains.
Nevertheless, many do think the think the arguments on both sides are quite
persuasive; if you’re persuaded by the arguments on both sides, then our job is
prety much done. However, many who would otherwise ind the arguments
against physicalism and the arguments against dualism persuasive feel forced to
reject the arguments on one side. Some reject the arguments against physicalism because they ind the arguments against dualism more persuasive, and see no
alternative to physicalism, whereas others reject the arguments against dualism
because they think the arguments against physicalism are more persuasive, and
see no alternative to dualism.
Take, for example, Joseph Levine who ultimately gives up dualism for physicalism:
What makes the mind-body issue a problem is that both positions seem
to have excellent considerations in their favor. On the dualist side, one
need only point to just how distinctive [mental] features are, and how
diicult it is to see how mere mater and energy could support them…
On the other side, there are deep reasons for supposing that minds
must really be natural, physical things ater all, and [mental] features
must really be ultimately natural, physical features. (
: ; italics
in original)
Take, for another example, the earlier version of David Chalmers who ultimately
gives up physicalism for dualism:
Temperamentally, I am strongly inclined toward materialist reductive
explanation, and I have no strong spiritual or religious inclinations.
For a number of years, I hoped for a materialist theory; when I gave
up on this hope, it was quite reluctantly. It eventually seemed plain to
me that these conclusions were forced on anyone who wants to take
consciousness seriously. Materialism is a beautiful and compelling
view of the world, but to account for consciousness, we have to go
beyond the resources it provides. (
: xiv)
Take, for a inal example, Colin Mcginn, who ultimately concludes that the respective cases for dualism and physicalism leaves us with an insoluble mystery:
When we think relectively of mental phenomena we ind that we
acknowledge them to possess two sets of properties: one set which
invites us to distinguish the mental realm from the physical, the other
which irmly locates the mental within the physical world… It is impossible not to be impressed with the applicability of both sets of properties to the mind, and to admit that both must ind a place in any
account of the relation between mind and body. he problem is that
the two sets of truths seem to be in fundamental tension, since one set
makes us think the mind could not be physical while the other tells
us that it must be. It is this tension that makes it appropriate to speak
of the mind-body problem. (
: - )
hese are typical statements of the mind-body problem. hese philosophers,
along with many others, feel forced to choose between two impossible options.
However, since there is an idealist way out, they are actually free to accept the
arguments on both sides. Pointing out that there is an idealist alternative can
thus help support the arguments against physicalism and the arguments against
dualism.
Conclusion
Berkeley and others have argued that idealism provides us much of what we
want: from a problem free metaphysics to a problem free epistemology. We argue
that idealism also gives us what we want in philosophy of mind: a way out of the
problems facing physicalism and property dualism. he main arguments in contemporary philosophy of mind, when considered together, point towards idealism.
Idealism gives dualists what they want, maybe what they most want: irreducible
mentality. Idealism gives the physicalist what they want, maybe what they most
want: seamless monism. If we are right, idealism is doubly necessary then: it
could not have failed to be true, and it is indispensable in solving the mystery of
the mental.
References
Berkeley, George.
/
. A Treatise Concerning the Principles of Human
Knowledge. Ed Kenneth P. Winkler. (Indianapolis: Hacket).
Chalmers, David.
. he Conscious Mind: In Search of A Fundamental heory. (Oxford: Oxford University Press).
_______.
. ‘Panpsychism and Panprotopsychism ‘in Torin Alter & Yujin
Nagasawa, Eds. Consciousness in the Physical World: Perspectives on Russellian
Monism. Oxford: Oxford University Press,
- .
Huemer, Michael.
. ‘When Is Parsimony A Virtue?’, Philosophical uarterly, :
,
- .
hanks to Daniel Greco, Ralph Kennedy, Win-Chiat Lee, Christian Miller, Kenneth Pearce and
Helen Yeter Chappell for helpful comments.
Jackson, Frank.
. ‘Epiphenomenal ualia’, Philosophical uarterly :
,
- .
Levine, Joseph.
. Purple Haze: he Puzzle of Consciousness. (Oxford: Oxford University Press).
Lewis, David.
. On he Plurality of Worlds. (Malden: Blackwell Publishing).
McGinn, Colin.
. he Character of Mind: An Introduction to Philosophy of
Mind. (Oxford: Oxford University Press).
Nagasawa, Yujin.
. God and Phenomenal Consciousness: A Novel Approach
to Knowledge Arguments. (Cambridge: Cambridge University Press).
Papineau, David.
. hinking About Consciousness. (Oxford: Oxford University Press).
uine, Willard V.O.
. Ontological Relativity & Other Essays. (New York:
Columbia University Press).
Sider, heodore.
. Writing the Book of the World. (Oxford: Oxford University Press).
Stoljar, Daniel.
. Physicalism. (New York: Routledge).