Are, Therefore, Not Possible. Please Note

Dear Author,
Here are the proofs of your article.
• You can submit your corrections online, via e-mail or by fax.
• For online submission please insert your corrections in the online correction form. Always
indicate the line number to which the correction refers.
• You can also insert your corrections in the proof PDF and email the annotated PDF.
• For fax submission, please ensure that your corrections are clearly legible. Use a fine black
pen and write the correction in the margin, not too close to the edge of the page.
• Remember to note the journal title, article number, and your name when sending your
response via e-mail or fax.
• Check the metadata sheet to make sure that the header information, especially author names
and the corresponding affiliations are correctly shown.
• Check the questions that may have arisen during copy editing and insert your answers/
corrections.
• Check that the text is complete and that all figures, tables and their legends are included. Also
check the accuracy of special characters, equations, and electronic supplementary material if
applicable. If necessary refer to the Edited manuscript.
• The publication of inaccurate data such as dosages and units can have serious consequences.
Please take particular care that all such details are correct.
• Please do not make changes that involve only matters of style. We have generally introduced
forms that follow the journal’s style.
Substantial changes in content, e.g., new results, corrected values, title and authorship are not
allowed without the approval of the responsible editor. In such a case, please contact the
Editorial Office and return his/her consent together with the proof.
• If we do not receive your corrections within 48 hours, we will send you a reminder.
• Your article will be published Online First approximately one week after receipt of your
corrected proofs. This is the official first publication citable with the DOI. Further changes
are, therefore, not possible.
• The printed version will follow in a forthcoming issue.
Please note
After online publication, subscribers (personal/institutional) to this journal will have access to the
complete article via the DOI using the URL: http://dx.doi.org/[DOI].
If you would like to know when your article has been published online, take advantage of our free
alert service. For registration and further information go to: http://www.link.springer.com.
Due to the electronic nature of the procedure, the manuscript and the original figures will only be
returned to you on special request. When you return your corrections, please inform us if you would
like to have these documents returned.
Metadata of the article that will be visualized in
OnlineFirst
ArticleTitle Access Denied to Zombies
Article Sub-Title
Article CopyRight Springer Science+Business Media Dordrecht
(This will be the copyright line in the final PDF)
Journal Name Topoi
Corresponding Author Family Name Piccinini
Particle
Given Name Gualtiero
Suffix
Division Department of Philosophy, Center for Neurodynamics
Organization University of Missouri – St. Louis
Address 599 Lucas Hall, 1 University Blvd., St. Louis, MO, 63121-4400, USA
Email piccininig@umsl.edu
Received
Schedule Revised
Accepted
Abstract I argue that metaphysicians of mind have not done justice to the notion of accessibility between possible
worlds. Once accessibility is given its due, physicalism must be reformulated and conceivability arguments
must be reevaluated. To reach these conclusions, I explore a novel way of assessing the zombie
conceivability argument. I accept that zombies are possible and ask whether that possibility is accessible
from our world in the sense of ‘accessible’ used in possible world semantics. It turns out that the question
whether zombie worlds are accessible from our world is equivalent to the question whether physicalism is
true at our world. By assuming that zombie worlds are accessible from our world, proponents of the
zombie conceivability argument beg the question against physicalism. In other words, it is a mistake to
assume that the metaphysical possibility of zombies entails that physicalism is false at our world. I will
then consider what happens if a proponent of the zombie conceivability argument should insist that zombie
worlds are accessible from our world. I will argue that the same ingredients used in the zombie
conceivability argument—whatever exactly they might be—can be used to construct an argument to the
opposite conclusion. At that point, we reach a stalemate between physicalism and property dualism: while
the possibility of zombies entails property dualism, the possibility of other creatures entails physicalism.
Since these two possibilities are mutually inconsistent, either one of them is not genuine or one of them is
inaccessible from the actual world. To resolve this stalemate, we need more than traditional conceivability
arguments.
Keywords (separated by '-') Possible world - Accessibility - Metaphysical possibility - Zombie - Conceivability argument - Physicalism
- Property dualism
Footnote Information This paper was written in 2005 and presented at Tucson VII—Toward a Science of Consciousness 2006,
Tucson, AZ, April 2006, and at the APA Central Division, Chicago, IL, April 2008.
Frankish (2007) independently published an argument that is similar in spirit to that of Sect. 7 below;
similar considerations are also found in Marton (1998) and Sturgeon (2000), pp. 114–116. But these
authors follow the mainstream in accepting the step from the metaphysical possibility of zombies to the
denial of physicalism (Frankish 2007, p. 653), whereas that step is precisely the main target of the present
paper. So the argument in Sect. 7 acquires a different significance within the present framework.Due to
other projects, I haven’t had time to publish on this topic until now, although a version of this paper was
posted on my website since about 2008. Given the tangled history of this paper and that this paper has been
cited (Brown 2010), I decided to publish it in a form that is close to its original 2006 version. Because of
this, I do not discuss the possibility of blockers (Hawthorne 2002) and the ceteris absentibus formulation of
physicalism it gave rise to (Leuenberger 2008), even though they raise interesting and subtle issues.
Briefly: I would recommend handling blockers in the same way that I recommend handling zombies in the
main text—that is, I would recommend that any physicalist who holds that physicalism is a contingent
thesis deem the possibility of blockers inaccessible from the actual world. I hope to address some of the
remaining issues at greater length in the future.I am grateful to those who commented on this paper in the
distant past; I forgot who they were—sorry! For recent comments, I am grateful to the anonymous referees,
Dave Chalmers, Terry Horgan, Robert Kirk, Waldemar Rohloff, Stephen Yablo, and the Washington
University in St. Louis Philosophy of Science Research Group. Thanks to Brendan Ritchie, Glenn
Carruthers, and Liz Schier for inviting me to submit this paper to Topoi. Thanks to Elliott Risch for
editorial assistance. This material is based on work supported in part by a University of Missouri research
award.
Topoi
DOI 10.1007/s11245-015-9323-6
2
1
3 Access Denied to Zombies

4 Gualtiero Piccinini1
F
OO
Author Proof
5
6 Springer Science+Business Media Dordrecht 2015
PR
7 Abstract I argue that metaphysicians of mind have not physicalism is false at our world. I will then consider what 23
8 done justice to the notion of accessibility between possible happens if a proponent of the zombie conceivability ar- 24
9 worlds. Once accessibility is given its due, physicalism gument should insist that zombie worlds are accessible 25
10 must be reformulated and conceivability arguments must from our world. I will argue that the same ingredients used 26
11 be reevaluated. To reach these conclusions, I explore a in the zombie conceivability argument—whatever exactly 27
12 novel way of assessing the zombie conceivability argu-
ED they might be—can be used to construct an argument to the 28
13 ment. I accept that zombies are possible and ask whether opposite conclusion. At that point, we reach a stalemate 29
14 that possibility is accessible from our world in the sense of between physicalism and property dualism: while the 30
15 ‘accessible’ used in possible world semantics. It turns out possibility of zombies entails property dualism, the possi- 31
16 that the question whether zombie worlds are accessible bility of other creatures entails physicalism. Since these 32
17 from our world is equivalent to the question whether two possibilities are mutually inconsistent, either one of 33
CT
18 physicalism is true at our world. By assuming that zombie them is not genuine or one of them is inaccessible from the 34
19 worlds are accessible from our world, proponents of the actual world. To resolve this stalemate, we need more than 35
20 zombie conceivability argument beg the question against traditional conceivability arguments. 37
36
21 physicalism. In other words, it is a mistake to assume that
22 the metaphysical possibility of zombies entails that
E
RR
This paper was written in 2005 and presented at Tucson VII—Toward handling zombies in the main text—that is, I would recommend that
a Science of Consciousness 2006, Tucson, AZ, April 2006, and at the any physicalist who holds that physicalism is a contingent thesis deem
APA Central Division, Chicago, IL, April 2008. the possibility of blockers inaccessible from the actual world. I hope
to address some of the remaining issues at greater length in the future.
Frankish (2007) independently published an argument that is similar I am grateful to those who commented on this paper in the distant
in spirit to that of Sect. 7 below; similar considerations are also found past; I forgot who they were—sorry! For recent comments, I am
in Marton (1998) and Sturgeon (2000), pp. 114–116. But these grateful to the anonymous referees, Dave Chalmers, Terry Horgan,
CO
authors follow the mainstream in accepting the step from the Robert Kirk, Waldemar Rohloff, Stephen Yablo, and the Washington
metaphysical possibility of zombies to the denial of physicalism University in St. Louis Philosophy of Science Research Group.
(Frankish 2007, p. 653), whereas that step is precisely the main target Thanks to Brendan Ritchie, Glenn Carruthers, and Liz Schier for
of the present paper. So the argument in Sect. 7 acquires a different inviting me to submit this paper to Topoi. Thanks to Elliott Risch for
significance within the present framework. editorial assistance. This material is based on work supported in part
Due to other projects, I haven’t had time to publish on this topic until by a University of Missouri research award.
now, although a version of this paper was posted on my website since
UN
about 2008. Given the tangled history of this paper and that this paper & Gualtiero Piccinini
has been cited (Brown 2010), I decided to publish it in a form that is piccininig@umsl.edu;
close to its original 2006 version. Because of this, I do not discuss the http://www.umsl.edu/*piccininig/
possibility of blockers (Hawthorne 2002) and the ceteris absentibus
1
formulation of physicalism it gave rise to (Leuenberger 2008), even Department of Philosophy, Center for Neurodynamics,
though they raise interesting and subtle issues. Briefly: I would University of Missouri – St. Louis, 599 Lucas Hall, 1
recommend handling blockers in the same way that I recommend University Blvd., St. Louis, MO 63121-4400, USA
123
Journal : Large 11245 Dispatch : 22-4-2015 Pages : 12
Article No. : 9323 h LE h TYPESET
MS Code : TOPO-D-14-00073 4 CP
h 4 DISK
h
G. Piccinini
38 Keywords Possible world Accessibility Metaphysical laws obtain, the possible worlds relevant to whether 87
39 possibility Zombie Conceivability argument physicalism holds at our world are those where the same 88
40 Physicalism Property dualism relevant metaphysical laws (or the same metaphysical 89
truths) obtain that obtain at our world. 90
Although I will focus on the zombie conceivability ar- 91
gument, the most significant observations and morals of 92
41 1 Introduction this paper extend to any conceivability argument that is 93
relevantly similar to the zombie conceivability argument. 94
42 In this paper, I look at some conceivability arguments from Physicalism is the view that everything is physical. So 95
F
43 a new vantage point. I argue that metaphysicians of mind long as ‘‘physical’’ can be explicated in a way that does not 96
44 have not done justice to the notion of accessibility between trivialize physicalism, my paper does not depend on the 97
45 possible worlds—a notion that is central in modal logic. details of such an explication. I assume that there is a suitable 98
OO
46 Accessibility is a relation between possible worlds that is explication of ‘‘physical.’’ We will be mostly concerned with 99
47 employed in modal logic to provide a formal semantics for phenomenal physicalism, according to which phenomenal 100
Author Proof
48 modal statements—statements involving possibility and consciousness is physical. The following is a popular and 101
49 necessity. In modal logic, a model is a triple \W, R, a[, useful explication of physicalism in terms of supervenience: 102
50 where W is a nonempty set of possible worlds, R is a
Physicalism: every property supervenes on physical 103
PR
51 binary relation on W (that is, R is the accessibility rela-
properties 104
52 tion), and a is an assignment function that assigns the
Phenomenal physicalism: phenomenal properties su- 105
53 following truth values: ‘‘possibly p’’ is true at a world w if
pervene on physical properties 106
54 and only if there is a world v accessible from w (i.e., wRv)
55 and ‘‘p’’ is true at v; ‘‘necessarily p’’ is true at world w if The relevant notion of supervenience is such that the 107
56 and only if for each world that is accessible from w (i.e., for supervening properties are conceptually entailed by the 108
57 each world v such that wRv), ‘‘p’’ is true at such a world
ED
base properties. According to physicalism, then, physical 109
58AQ1 (Garson 2012, pp. 64–66). truths entail (fix, determine) all truths, including phe- 110
59 I will argue that, once accessibility is given its due, nomenal truths. Two entities cannot differ without differing 111
60 physicalism must be reformulated and conceivability ar- physically. This formulation is somewhat imprecise, but it 112
61 guments must be reassessed. My objection differs from will remain our working formulation of physicalism. Later, 113
62 standard objections to conceivability arguments because, I will discuss one attempt at making physicalism more 114
CT
63 for present purposes, I deny neither conceivability nor that precise and its shortcomings, and in the end I will propose 115
64 conceivability entails possibility. Instead, I argue that, even an improved formulation.1 116
65 if the premises of conceivability arguments are accepted, What is relevant here is primarily the aspect of 117
66 their conclusion doesn’t follow because it remains to be physicalism that pertains to phenomenal properties. Roughly 118
67 shown that the envisaged possibilities are accessible from
E
68 our world in the sense of ‘accessible’ used in possible 1

I took the present working formulation of physicalism from
69 world semantics. By assuming that the possibilities in Chalmers (1996) and Jackson (1998), who offer extended defenses
RR
70 question are accessible from our world, this type of con- of its adequacy. Formulations of physicalism that employ weaker
71 ceivability argument begs the question. notions of supervenience suffer from difficulties canvassed in Horgan
(1993) and Wilson (1999, 2005). Chalmers and Jackson go on to offer
72 To illustrate, consider the case of physical possibility
a more precise formulation along the lines of Horgan (1982) and
73 and necessity. To evaluate whether something is physically Lewis (1983), but I will postpone discussion of that formulation until
74 possible or necessary at our world, we cannot simply later. Keeping with the more generic formulation is consistent with
75 consider arbitrary possible worlds. For there are possible what other writers do (e.g., Stoljar 2001, 2005, 2006). Thus
CO
formulated, physicalism is weak enough to encompass both reductive

76 worlds—e.g., worlds according to Newtonian physics—in
and non-reductive forms of physicalism. For instance, according to
77 which things occur that are physically impossible at our some versions of functionalism about consciousness, facts about
78 world. As a consequence, we need to consider only pos- consciousness supervene on physical facts even though the nature of
79 sible worlds that are accessible from our world, that is, consciousness is functional not physical. Zombies are also discussed
in debates specific to functionalism, where they are sometimes
80 possible worlds that are relevant to evaluating what is
defined as creatures behaviorally or functionally identical to us who
UN
81 physically possible at our world. The two most popular lack phenomenal consciousness (Searle 1992; Polger 2000). For
82 ways of doing this are to restrict accessibility either to present purposes, since functional properties are generally assumed to
83 worlds where the same physical laws obtain or to worlds be realized by physical properties, nothing hinges on the putative
distinction between physical and functional properties. This notion of
84 where the same physical truths obtain (Hermes 2004).
physicalism, either reductive or non-reductive, is intended to be in
85 By the same token, just as the relevant physically pos- opposition to property dualism, not scientific pluralism [of the kind
86 sible worlds are those where the same relevant physical defended, for instance, by Mitchell (2003)].
123
h 4 DISK
h
Access Denied to Zombies
119 speaking, these are properties instantiated by entities in is a contingent thesis. I will then ask whether these zombie 161
120 virtue of their possessing phenomenal states (i.e., possessing worlds are accessible from our world in the sense of ‘ac- 162
121 phenomenal consciousness, or being phenomenally con- cessibility’ employed in possible world semantics. As it 163
122 scious). According to this aspect of physicalism—phe- turns out, the question whether zombie worlds are acces- 164
123 nomenal physicalism—phenomenal properties supervene sible from our world is equivalent to the question whether 165
124 on physical properties. physicalism is true at our world. By assuming that zombie 166
125 In the recent philosophical literature, zombies are usu- worlds are accessible from our world, proponents of the 167
126 ally defined as creatures physically identical to us that lack zombie conceivability argument beg the question of 168
127 phenomenal consciousness. The zombie conceivability ar- physicalism. In other words, it is a mistake to assume that 169
F
128 gument goes as follows (cf. Chalmers 2002, 1996, p. 122). the metaphysical possibility of zombies entails that 170
physicalism is false at our world. 171
But suppose that zombie-philes should insist, with or 172
OO
129 2 The Zombie Conceivability Argument without argument, that zombie worlds are accessible from 173
our world. I will argue that, when relevant matters are 174
Author Proof
130 1. Zombies are conceivable properly clarified, the same ingredients used in the zombie 175
131 2. Therefore, zombies are metaphysically possible conceivability argument—whatever exactly they might 176
132 3. Therefore, the physical facts do not entail the phe- be—may be used to construct an argument to the opposite 177
PR
133 nomenal facts conclusion. If that is correct, we reach a stalemate between 178
physicalism and property dualism: while the possibility of 179
134 According to the zombie conceivability argument, zombies
some zombies entails property dualism, the possibility of 180
135 are conceivable, and hence possible, and hence physicalism
other creatures entails physicalism. Since these two pos- 181
136 is false. For zombies are physically identical to us and yet
sibilities are mutually inconsistent, either one of them is 182
137 lack consciousness. Therefore, if they are possible in an
not genuine or one of them is inaccessible from the actual 183
138 appropriate sense, what they lack—consciousness—is not,
ED world. To resolve this stalemate, we need more than tra- 184
139 and cannot be, determined by the physical facts.2
ditional conceivability arguments. 185
140 The zombie conceivability argument entails a form of
141 property dualism about consciousness, according to which
142 phenomenal facts do not supervene on physical facts but
3 Accessibility Between Worlds 186
143 are either further primitive facts or facts that supervene on
CT
144 further non-physical facts, such as laws that connect the

The step from (2) to (3) in the zombie conceivability ar- 187
145 physical and phenomenal facts.
gument is generally taken for granted, even by physicalists. 188
146 Critics of the conceivability argument have responded in
I will argue that this is a mistake, and that questioning the 189
147 two ways. Some deny that zombies are conceivable.3
inference from (2) to (3) yields insights into both 190
148 Others concede that zombies are conceivable but deny that
physicalism and conceivability arguments. So although I 191
E
149 they are possible.4 But the zombie conceivability argument

suspect zombies are both inconceivable and metaphysically 192
150 has been resilient in the face of criticism. Much of the
impossible, I will accept premise (2) for the sake of the 193
151 controversy hinges on how to establish and understand
RR
argument and explain why (3) still doesn’t follow. 194

152 what is conceivable, what is possible, and the link between
To evaluate the step from (2) to (3), it is convenient to 195
153 the two—matters that are at least as obscure and contro-
consider interpretations of (2) and (3) in terms of possible 196
154 versial as whether consciousness is physical.
world semantics. This is consistent with the way the debate 197
155 In this paper, I explore an alternative strategy to assess
is framed in the literature. Presumably, whether physical- 198
156 the zombie conceivability argument. I accept the possi-
ism holds and whether zombies are possible are meta- 199
CO
157 bility of phenomenal zombies in the sense that there are

physical statements, to be evaluated using a logical system 200
158 metaphysically possible worlds where phenomenal zom-
for metaphysical modality. 201
159 bies exist. I think this is the most that a physicalist should
To avoid begging the question, we need to be clear 202
160 grant—especially a physicalist who thinks that physicalism
about how possible world semantics works. To evaluate a 203
modal statement at a world, we look at which statements 204
UN
2
People who run conceivability arguments against physicalism (with are true at the possible worlds that are accessible from that 205
or without zombies) include Bealer (1994), Campbell (1970),
world. For instance, ‘‘P is possible’’ is true at world w if 206
Chalmers (1995, 1996, 2002, 2003, 2004), Kirk (1974), and Kripke
(1980). and only if there is a possible world w’, accessible from w, 207
3
E.g., Melnyk (2001), Worley (2003), Marcus (2004), Kim (2005). where ‘‘P’’ is true. ‘‘P is necessary’’ is true at world w if 208
4
E.g., Balog (1999), Yablo (2002), Stalnaker (2002), Lynch (2004), and only if ‘‘P’’ is true at all possible worlds that are ac- 209
Levine (2001), Rowlands (2001), McGinn (2004), Papineau (2002). cessible from w. 210
123
h 4 DISK
h
G. Piccinini
211 How do we determine which worlds are accessible from actual world for the purpose of evaluating the relevant 264
212 which? Accessibility between worlds is a relation invented modal statements at our world? This question has not been 265
213 by logicians to construct a semantic interpretation for adequately addressed to date. 266
214 modal statements belonging to different systems of modal The problem here is that, having granted (1) and (2) for 267
215 logic (Kripke 1963a, b; Hintikka 1963). When building a the sake of the argument but leaving open whether (2) 268
216 model for generic modal statements, such as ‘‘P is possi- holds at the actual world, the question of whether any of 269
217 ble’’ or ‘‘P is necessary,’’ we can freely posit arbitrary the zombie worlds posited by (2) is accessible from the 270
218 possible worlds and decide which worlds are accessible actual world is equivalent to the question of whether 271
219 from which, consistent with the axioms of the modal physicalism is true at the actual world. In other words, to 272
F
220 logical system we are working with. For instance, in sys- simply assume without independent arguments or consid- 273
221 tem K there are models such that ‘‘P is necessary’’ is true at erations that at least one of the zombie worlds posited by 274
222 world w even though there is no world accessible from w at (2) is accessible from the actual world begs the question of 275
OO
223 which ‘‘P’’ is true. In other words, in some modal logical physicalism.5 276
224 systems there are models in which ‘‘P is necessary’’ is
Author Proof
225 vacuously true. This is not the case in some stronger modal
226 systems, such as M, where ‘‘P is necessary’’ is true at w if 4 S5 to the Rescue? 277
227 and only if there is at least one world accessible from w
PR
228 where ‘‘P’’ is true—specifically, w itself. Notice that, if we There may appear to be a way around this problem of 278
229 are not careful in setting up the accessibility relation, we accessibility between worlds: S5. S5 is perhaps the most 279
230 may generate contradictions. For example, suppose that ‘‘P popular formal system for modal logic among philoso- 280
231 is necessary’’ is true at w and ‘‘*P’’ is true at w0 . We phers. The semantics for S5 is often formulated by saying 281
232 cannot posit that w0 is accessible from w on pain of con- that, in models of S5, the accessibility relation is univer- 282
233 tradiction, because our semantics is such that ‘‘P is nec- sal—that is, all worlds are accessible from one another. If 283
234
ED
essary’’ is true at w if and only if ‘‘P’’ is true at all possible all worlds are accessible from one another, then the step 284
235 worlds that are accessible from w. Therefore, if we want to from (2) to (3) becomes valid. This is because if all worlds AQ2285
236 avoid a contradiction, w0 must remain inaccessible from w. are accessible from one another, any zombie world is au- 286
237 When we shift from doing modal logic to evaluating tomatically accessible from our world, which makes 287
238 substantive modal statements, such as ‘‘zombies are ‘‘zombies are metaphysically possible’’ true at our world, 288
239 metaphysically possible,’’ and their logical consequences at which in turn falsifies physicalism at our world. 289
CT
240 our world, we cannot simply posit whatever possible If S5 were the correct system for metaphysical modality 290
241 worlds and accessibility relations suit us, on pain of beg- and if all models for S5 were such that accessibility is 291
242 ging the question. In order to posit appropriate accessibility universal—that is, such that all worlds are accessible from 292
243 relations between an appropriate set of possible worlds, we
244 need to know which relevant facts hold at each of those
E
5
I am working with the standard notion of accessibility that is
245 worlds. To suppose otherwise is to put the cart of semantic
employed in contemporary modal logic. Here is some textual
246 machinery before the horse of truth. evidence: (1) Kripke (1963a, b) does not use the expression ‘‘H2 is
RR
247 With these preliminaries in place, we can go back to accessible from H1’’ but rather ‘‘H2 is ‘possible relative to’ H1,’’
248 physicalism and conceivability arguments. We are asking which he explains by saying that ‘‘every proposition true in H2 is
possible in H1’’ (1963a, b, p. 64 in Linsky 1971). (2) Hintikka uses
249 whether (3) is true at our world. In terms of possible world
the term ‘‘alternative world’’ and explains it thus: ‘‘whatever is
250 semantics, physicalism [i.e., the denial of (3)] may be in- possible must be true in some alternative world and… whatever is
251 terpreted as follows: in any possible world that is acces- necessary must be true in all the alternative worlds… only these
252 sible from the actual world for the purpose of evaluating genuine alternatives really count’’ (Hintikka 1963, p. 67 in Loux
CO
1979). (3) Hughes and Cresswell say the accessibility of a world is a

253 the relevant modal statements, the physical facts entail the
matter of what is conceivable. This may seem to help the zombie-
254 phenomenal facts. For that to be true, none of the relevant phile, until we notice that what is conceivable is not unrestricted: it
255 worlds can be such as to be phenomenally different from must be restricted by the way the actual world is. So even under
256 our world while being physically the same. Hughes and Cresswell’s construal, the question of whether zombie
worlds are accessible from the actual world reduces to the question of
257 But (2) may be interpreted to say that there are possible
whether physicalism is true. (4) Salmon (1989, cf. also Garson 2003)
UN
258 worlds that are physically the same as our world while argues that accessibility is a matter of what is possible at a world.
259 being phenomenally different. Therefore, if any of those Again, the question of accessibility of zombie worlds reduces to the
260 worlds are accessible from our world for the purpose of question of physicalism. (5) Bigelow and Pargetter (1991) say it
would be desirable if what is accessible from a world would depend
261 evaluating the relevant modal statements, then (3) follows
on the intrinsic features of that world (cf. McDermott 1999); once
262 from (2) and physicalism is false. The question is, are any again, the question of accessibility reduces to the question of whether
263 of the zombie worlds posited by (2) accessible from the physicalism is true at our world.
123
h 4 DISK
h
293 one another—proponents of conceivability arguments relation, but not vice versa. It is easy to show that if a 319
294 would be home free. Unfortunately for them, they face model is universal, then the accessibility relation is an 320
295 serious obstacles. equivalence relation. But it is also easy to show that if the 321
296 A first obstacle is arguments that S5 is not the correct accessibility relation is an equivalence relation, then the 322
297 logic for metaphysical modality.6 Anyone who wants to
AQ3 accessibility relation need not be universal. All you have to 323
298 save conceivability arguments needs to show why, in spite do is posit multiple systems of worlds, such that within 324
299 of existing objections, S5 is the correct logic for meta- each system all worlds are accessible from one another, but 325
300 physical modality. I will not dwell on this first obstacle across systems no worlds are accessible from one another. 326
301 because there is a more serious obstacle: S5 does not For example, consider a model containing the following 327
F
302 mandate that accessibility be universal. On the contrary, worlds: w, w0 , w00 , v, v0 , v00 . The three worlds from the first 328
303 there are models of S5 that lack this universality property. system (the w’s) are all accessible from one another, and so 329
304 For present purposes, the correct way of introducing the are the three from the second system (the v’s). But none of 330
OO
305 semantics for S5 is that its accessibility relation is an the worlds from the first system are accessible from any 331
306 equivalence relation: a relation that is reflexive, transitive, member of the second system, and no member of the 332
Author Proof
307 and symmetric. In other words, accessibility is an equiva- second system is accessible from any member of the first 333
308 lence relation if and only if the following three conditions system. In this model, the accessibility relation is an 334
309 obtain: any world is accessible from itself (reflexivity), if equivalence relation, so this is a legitimate model of S5. 335
PR
310 w0 is accessible from w and w00 is accessible from w0 then Yet the accessibility relation is not universal in this model. 336
311 w00 is accessible from w (transitivity), and if w0 is acces- When we formulate the semantics of S5 for purely 337
312 sible from w then w is accessible from w0 (symmetry). logical purposes, the difference between the two methods 338
313 Both ways of formulating the semantics of S5—in terms (universality-based and equivalence-based) is often unim- 339
314 of universal accessibility and in terms of accessibility as an portant and gets ignored even in textbooks (e.g., Garson 340
315 equivalence relation—model the theorems of S5. But they 2013, p. 105). But when evaluating substantive modal 341
316 determine different classes of models, and the second type
ED claims, this difference is critical—assuming without argu- 342
317 of semantics is more permissive than the first. Universal ment that accessibility is universal runs the risk of ignoring 343
318 accessibility entails that accessibility is an equivalence the best way to explicate metaphysical possibility. For 344
consider a bunch of zombie worlds in a model of S5 based 345
6
For instance, Nathan Salmon has argued that S4 (and hence S5) on accessibility being an equivalence relation though not 346
contains an axiom schema some of whose instances are false (Salmon universal. All zombie worlds may well be confined to 347
CT
1989). The axiom schema is that necessary propositions are neces- systems of worlds whose members are inaccessible from 348
sarily necessary, or ‘‘h/.hh/’’. Salmon’s counterexample in-
our world. If so, ‘‘zombies are metaphysically possible’’ 349
volves the composition of material artifacts. According to a widely
shared intuition, any particular artifact could have originated from remains false at our world, which is consistent with 350
matter that is slightly different from its actual original matter; for physicalism being true at our world. In other words, even 351
instance, a particular statue s could have originated from a lump of within S5, if our model is not universal (3) doesn’t follow 352
E
bronze m* slightly different from the lump m, which it actually

from (2).7 353
originated from. But the same artifact could not have originated from
matter that is too different from its actual original matter; a particular
RR
7
statue could not have originated from a lump of bronze m’ that is Metaphysicians ignore models of S5 that lack a universal acces-
sufficiently different from m while remaining the same statue. If this sibility relation at their peril. For example, Chalmers (2010, p. 179),
is accepted, it follows that for some matter m’ sufficiently different following Yablo (1999, pp. 456–457), discusses whether the
from m but sufficiently overlapping with m*, s could have originated statement ‘‘it is possible that it is necessary that there is an omniscient
from m* but could not have originated from m’; but at the same time, being’’ is consistent with the statement ‘‘it is possible that it is not
if s had originated from m*, then it would have been possible for s to necessary that there is an omniscient being’’. Chalmers argues that, in
originate from m0 . Hence, although it is necessary that s did not S5, these two statements are mutually contradictory, because in S5
CO
originate from m0 , it is not necessarily necessary that s did not ‘‘possibly necessarily P’’ entails ‘‘necessarily P’’. But this argument
originate from m0 . This example violates the axiom schema implicitly assumes that all models of S5 are such that accessibility is
mentioned above. Therefore, according to Salmon, S4 (and S5 with universal, whereas there are models of S5 where accessibility is non-
it) is fallacious. Salmon believes that the correct logic for meta- universal. If the latter are allowed, as they should be in the present
physical necessity is probably T (ibid., p. 4). context, the two statements are consistent with one another, provided
If critics of S5 are right, we cannot assume that zombie worlds are that the two statements are true at worlds that belong to two mutually
accessible from the actual world even if we hold the question of inaccessible systems of possible worlds.
UN
physicalism to be a matter of straightforward metaphysical necessity. Similarly, Wedgwood discusses the relation between the meta-
Rather, we need to discuss which possible worlds are accessible from physics of mind and accessibility between possible worlds (2000,
the actual world. What is the reason to think that the zombie worlds p. 410 and Appendix). He argues that Kim’s (1993) argument from
are accessible? The answer is important, for as we have seen, even if strong supervenience to global supervenience is valid in S5 but
we assume that physicalism is a metaphysical claim of the relevant invalid in any logical system weaker than S5. On that basis,
type, determining whether zombie worlds are accessible is equivalent Wedgewood argues that non-reductive physicalism can’t hold if we
to determining whether physicalism is true. choose S5 as our logic but can hold if we choose weaker logical
123
h 4 DISK
h
G. Piccinini
354 Later I will argue that models with multiple systems of If we were to do the same thing when evaluating 393
355 worlds that are closed in the above sense—that is, no world metaphysical modality, we would have to restrict our do- 394
356 from one system is accessible from any world from any main to the set of worlds that are relevantly similar to our 395
357 other system—is an excellent way to model metaphysical world—that is, worlds with the same metaphysical laws 396
358 modality, especially if someone maintains—as many that obtain at our world. To decide whether to include 397
359 physicalists do—that the views under considerations are zombie worlds in our model, which we need to do in order 398
360AQ4 not metaphysically necessary but contingent.8 In getting to evaluate whether the zombie conceivability argument is 399
361 there, it will be helpful to consider what happens if we try valid, we would need to know whether physicalism obtains 400
362 to restrict our reasoning to models of S5 that have the at our world. But this is precisely what the zombie con- 401
F
363 universality property. ceivability argument is supposed to establish! 402
364 Suppose that a proponent of the zombie conceivability To avoid this conundrum, the proponent of the zombie 403
365 argument argued somehow that we should restrict our conceivability argument must argue that all possible worlds 404
OO
366 reasoning to models of S5 that have the universality (or at least all possible worlds that are physical duplicates 405
367 property, in which, because of the universality property, the of our world; since this restriction makes no difference in 406
Author Proof
368 step from (2) to (3) is valid. (I’m not aware of any such what follows, I will ignore this qualification from now on) 407
369 argument, but let’s ignore that.) Her challenges are not must be included in our domain, and any world that is ruled 408
370 over. What she also needs to show is that our model must out is not genuinely possible. To mount such an argument, 409
PR
371 include all possible worlds and, conversely, that any world she might recall that physicalism is the claim that all 410
372 excluded by our model must be metaphysically impossible properties supervene on physical properties as a matter of 411
373 not only relative to the metaphysical laws that obtain at our entailment. She might argue that supervenience is an un- 412
374 world, but metaphysically impossible simpliciter. In other restricted modal claim, and any counterexample in any 413
375 words, she must show that metaphysical modality is not possible world is enough to refute it. Therefore, all possible 414
376 restricted to a relevant class of worlds that are meta- worlds are relevant. If all possible worlds are relevant, we 415
377 physically similar to ours.
ED must include them all in our domain. If we do that and our 416
378 Consider again the analogy with physical modality. model has the universality property, then the zombie con- 417
379 Contemporary physicists may all agree that at our world ceivability argument goes through. 418
380 it’s physically impossible to travel faster than the speed of But this option turns physicalism into an unrestricted 419
381 light, but this is a relatively recent consensus. To model the necessity. If true, physicalism holds in all possible 420
382 transition between Newtonian physics and relativity theory, worlds—it is metaphysically necessary. If false, physical- 421
CT
383 we may posit a system of possible worlds governed by ism holds in no possible world—it is metaphysically im- 422
384 Newtonian physics and a system of possible worlds gov- possible. I like the view that physicalism is metaphysically 423
385 erned by relativity theory. When we evaluate what speeds necessary, but most physicalists are less sanguine. They 424
386 are physically possible at our world, we restrict our domain accept that physicalism might have been false. For in- 425
387 to the second set of worlds—those governed by relativity stance, Cartesian dualism might have been true instead. So 426
E
388 theory. We may still use a model with the universality physicalism is usually held to be a contingent thesis. Ac- 427
389 property (all worlds accessible from one another), but our cording to modest physicalists, there are possible worlds in 428
RR
390 domain is suitably restricted to include only the possible which physicalism fails. 429
391 worlds that are physically similar to our world in the
392 relevant respect.
5 Horganic Physicalism 430
Footnote 7 continued
The standard way to formulate physicalism as a contingent 431
CO
systems, and therefore non-reductive physicalists should choose a

logic weaker than S5. But Wedgewood’s argument implicitly assumes thesis can be traced to a paper by Terence Horgan (1982). 432
that all models of S5 are such that accessibility is universal, whereas In honor of its originator, I will call it Horganic 433
there are models of S5 where accessibility is non-universal. If the physicalism9: 434
latter kind of model of S5 is considered, Wedgewood’s refutation of
Kim’s (1993) argument applies within S5 too, so non-reductive
physicalism may hold even under S5. So, contra Wedgewood, non-
UN
reductive physicalists need not reject S5 for the reasons offered by

Wedgewood. 9
Following Horgan, a similar formulation was endorsed by Lewis
8
Cf. Lewis (1983, p. 362): ‘‘Materialism is meant to be a contingent (1983), and then by Jackson (1993, 1994, 1998) and Chalmers (1996,
thesis, a merit of our world that not all other worlds share.’’ Both p. 42). There are subtle differences between the formulations offered
Terry Horgan and Philip Pettit independently told me they hold by these authors, but they do not matter for present purposes. In the
phenomenal physicalism to be a contingent truth (personal commu- main text I adopted Jackson’s version of Horganic physicalism
nication). Similar statements are plentiful in the literature. because it is the simplest.
123
h 4 DISK
h
435 Horganic physicalism: any possible world that is a At this point, while we are still granting (2), we face two 479
436 minimal physical duplicate of our world is a duplicate options: either we accept Horganic physicalism as a good 480
437 simpliciter of our world. formulation of physicalism and reject physicalism, or we 481
438 Horganic phenomenal physicalism: any possible keep the question of physicalism open by finding a for- 482
439 world that is a minimal physical duplicate of our mulation of physicalism that is more adequate than Hor- 483
440 world is a phenomenal duplicate of our world. ganic physicalism. So far as I know, the second option has 484
not been explored explicitly. This is surprising, for there is 485
441 A minimal physical duplicate of a world is a world in
an obvious reason to think that Horganic physicalism needs 486
442 which all the same fundamental physical truths obtain but
to be replaced: Horganic physicalism is refuted by the 487
443 no extra, non-physical substances or principles obtain—
F
zombie conceivability argument.11 488
444 same microphysics, no new spooky stuff (Jackson 1998,
Notice that Horganic physicalism, when suitably gen- 489
445 p. 13). A duplicate simpliciter of a world is a world in
eralized to apply to arbitrary worlds (this is how Lewis 490
OO
446 which all facts, including all higher-level and higher-order
1983 formulates it), counts zombie worlds as worlds in 491
447 facts entailed by the fundamental facts obtain.
Author Proof
448 An important motivation for Horganic physicalism is to

449 allow for possible worlds where physicalism fails. For in- Footnote 10 continued
450 stance, Horganic physicalism allows for Cartesian worlds, At this point, Jackson might reply that I am misunderstanding the
451 where Cartesian dualism obtains, because those worlds are relevant notion of entailment. If the physical facts entail the phe-
PR
not minimal physical duplicates of our world. So, accord- nomenal facts, as physicalism maintains, there cannot be worlds with
452
the same physical facts in which the phenomenal facts do not obtain.
453 ing to Horganic physicalism, the possibility of Cartesian It would be like saying that the physical facts do not, after all, entail
454 worlds does not refute physicalism. Thus, physicalism the phenomenal facts, which, again, would show physicalism to be
455 turns out to be contingent as desired. false. But my reply is straightforward: entailments between physical
Horganic physicalism does not prohibit access to zom- facts and phenomenal facts depend in part on what metaphysical laws
456
obtain at a world—we’ve already established that on independent
457 bie worlds. Since some zombie worlds are by definition grounds. Otherwise, worlds with disembodied minds, where such
ED
458 minimal physical duplicates of our world, Horganic entailments clearly fail, would be ruled out by physicalism as well
459 physicalism counts them as relevant to establishing the and physicalism would fail to be a contingent thesis. So some non-
truth of physicalism. Thus, if Horganic physicalism cap- zombie worlds are such that physical facts do not entail phenomenal
460
facts. Zombie worlds might be like that too, i.e., they are among the
461 tures the content of physicalism correctly, the step from (2) worlds where physicalism may fail. To disprove a metaphysical thesis
462 to (3) is valid. like physicalism, we need more than a world where physicalism may
CT
463 For (2) says that zombies are metaphysically possible. fail—we need to show that that world is relevant because it’s ac-
By definition, some zombie worlds are physical duplicates cessible from ours in the relevant respect.
464
465 of our world but are not phenomenal duplicates. It follows Another way to put the point is by saying that whether the
entailment holds at all worlds depends on what kind of entailment we
466 that, if zombie worlds are metaphysically possible, there are appealing to. Different notions of entailment range over more or
467 are possible worlds that are physical duplicates of our less comprehensive sets of possible worlds. If Jackson should insist
E
468 world but contain zombies, and hence are not phenomenal that we are talking about logical entailment in the narrow sense,
469 duplicates of our world. Ergo, Horganic phenomenal ranging over all possible worlds, we would have to retreat to the
possibility that zombie worlds are inaccessible from our world [or
470 physicalism fails, and Horganic physicalism simpliciter
RR
where some aspect of classical logic doesn’t hold; perhaps they are
471 fails with it. non-standard worlds (Rescher and Brandom 1979) or impossible
472 So here is why the accessibility of zombie worlds has worlds (Salmon 1989)]. But we need not go that far, for we already
473 received so little attention: it is buried under the carpet of know there are worlds where the entailment doesn’t hold.
Jackson was almost right. In one sense, physicalism is true if and
474 Horganic physicalism, and Horganic physicalism is widely only if zombie worlds are impossible. That is the sense in which
475 accepted as an adequate formulation of physicalism. But worlds that are inaccessible from a world are impossible. They are
CO
476 the wide acceptance of Horganic physicalism as a formu- impossible relative to that world. But this doesn’t tell us whether
477 lation is a historical accident, which should not prevent us conceivability arguments are sound. One point of this paper is that
since the question of whether physicalism is true is equivalent to the
478 from carrying our investigation further.10 question of whether zombie worlds are accessible, the zombie
conceivability argument, by assuming accessibility, begs the question.
10
Jackson (1998) argues that Horganic physicalism captures the Mutatis mutandis, the same holds for analogous conceivability
content of physicalism as follows. If physicalism fails, than Horganic arguments.
UN
11
physicalism fails. So far, so good. Jackson also says that if Horganic Leuenberger (2008) offers an alternative formulation of physical-
physicalism fails, then our world must contain something non- ism (for independent reasons), but his alternative formulation of
physical that is not duplicated by duplicating the physical facts about physicalism is still in the spirit of Horganic physicalism, so
it, and hence physicalism fails. This is a non sequitur. For it may be Leuenberger ends up having to argue that zombie conceivability
that zombie worlds are possible but inaccessible from our world. arguments do not refute his formulation of physicalism because
Physicalism fails at zombie worlds, but this doesn’t show that zombie worlds without blockers are not conceivable. Chalmers (2010,
physicalism fails at our world. p. 164) insists that they are.
123
h 4 DISK
h
G. Piccinini
492 which physicalism is true. For zombie worlds certainly neither Lewis nor anyone else seems to fully appreciate is 528
493 have no minimal physical duplicates that are not also that an analogous objection applies to Horganic physical- 529
494 phenomenal duplicates. But this is clearly against the spirit ism with respect to property dualist worlds: a world w with 530
495 of physicalism. Physicalism is intended to include phe- phenomenal properties and its zombie duplicate w0 could 531
496 nomenal physicalism, and phenomenal physicalism is false indeed differ without differing physically, if at least one of 532
497 at zombie worlds, at least in its intended interpretation, them is a world where physicalism is false. Which of the 533
498 according to which the phenomenal facts of our world are two? The moral usually drawn is that either w0 is 534
499 entailed by the physical facts of our world. Therefore, we impossible or physicalism fails at w (the world with 535
500 should find a formulation of physicalism than improves on phenomenal properties). But it is equally legitimate to 536
F
501 Horganic physicalism.12 conclude that physicalism fails at w0 (the zombie world). If 537
physicalism fails at w0 , then there is no reason to deem it 538
impossible (as Horganic physicalism does). All a physical- 539
OO
502 6 The Contingency of Phenomenal Physicalism ist needs to do is let w0 be possible while deeming it just as 540
inaccessible from our world as a Cartesian world. 541
Author Proof
503 There is an independent reason to amend Horganic To put the point another way, if phenomenal physical- 542
504 physicalism: it doesn’t fully capture the putative contin- ism is contingent, it should be contingent all-around. It 543
505 gency of physicalism. To capture the contingency of should allow for the possibility of property dualist worlds 544
PR
506 physicalism, as we have seen, a good formulation of as well as Cartesian dualist worlds. By ruling out zombie 545
507 physicalism must allow for possible worlds in which worlds, Horganic phenomenal physicalism (suitably gen- 546
508 physicalism fails. Horganic physicalism is explicitly de- eralized to apply to any worlds, as Lewis 1983 formulates 547
509 signed to allow for some possible worlds in which it) rules out the possibility of property dualist worlds that 548
510 physicalism fails. For instance, it allows for worlds that are physical duplicates, even while not ruling out the 549
511 contain disembodied minds in addition to everything that is possibility of Cartesian dualist worlds that are physical 550
512 contained in our world. But it rules out other worlds where
ED duplicates. For a property dualist world that is a physical 551
513 physicalism fails, and for no principled reason. duplicate is a zombie world. By ruling out zombie dupli- 552
514 The problem may become clear if we take a step back. cates, Horganic physicalism rules out property dualist 553
515 As a first pass at a formulation of physicalism in terms of worlds that are physical duplicates. But there is no prin- 554
516 supervenience, Horgan, Lewis, and Jackson consider the cipled reason why physicalism should be consistent with 555
517 following: the possibility of Cartesian worlds that are physical du- 556
CT
plicates but inconsistent with the possibility of property 557

518 Generic supervenience physicalism. Any two worlds
dualist worlds that are physical duplicates. All of those 558
519 that are physical duplicates are duplicates simpliciter.
worlds refute a generic formulation of physicalism if they 559
520 Generic phenomenal supervenience physicalism. Any
are accessible, so a consistent physicalist should deny ac- 560
521 two worlds that are physical duplicates are phe-
cess to all of them. 561
E
522 nomenal duplicates.

The flip side of this observation is that, given Horganic 562
523 The natural objection to this formulation is that it turns physicalism, the zombie conceivability argument proves 563
RR
524 physicalism into a necessary thesis. But, as Lewis puts it, too much. For the zombie conceivability argument, if 564
525 ‘‘two worlds could indeed differ without differing physical- sound, would prove that phenomenal physicalism is a 565
526 ly, if at least one of them is a world where Materialism necessary falsehood. The zombie conceivability argument 566
527 [i.e., physicalism] is false’’ (Lewis 1983, p. 35). What is a perfectly general argument; it holds at all worlds (ex- 567
cept for zombie worlds, which, according to Horganic 568
12
It may seem that rejecting Horganic physicalism as a good physicalism conjoined with the possibility of zombies, are 569
CO
formulation of physicalism on these grounds begs the question against

the only physicalistic worlds—what irony!). Consider an 570
the anti-physicalist. From the present perspective, it’s the other way
around. Horganic physicalism begs the question of which zombie arbitrary world w. Suppose (hypothesis for the reductio) 571
worlds are accessible from which worlds in favor of the anti- that physicalism holds at w (thus there are no non-physical 572
physicalist. That is, by accepting Horganic physicalism, the physical- things and properties there to distract us). Now consider a 573
ist took up a huge burden of proof, which makes the task of the dualist
physical duplicate of that world where there are no phe- 574
easier by comparison. If we want to redress the balance, it is better to
UN
accept the possibility of zombie worlds and ask whether they are nomenal facts. By the logic of the zombie conceivability 575
accessible from our world. Because of this, Horganic physicalism argument, the second world exists and shows that 576
should be amended. physicalism is false at w. Hence, if the zombie conceiv- 577
At any rate, Horganic physicalism was a tentative formulation,
ability argument is sound and Horganic physicalism is a 578
which may be superseded by better formulations. If we find out that it
was flawed—and being refuted is quite a flaw—this is reason enough correct formulation of physicalism, phenomenal physical- 579
to look for a better formulation. ism is a necessary falsehood. But phenomenal physicalism 580
123
h 4 DISK
h
581 strikes modest physicalists as no more necessary than Of course it would be desirable to have an independent 613
582 physicalism in general. argument for the necessity of phenomenal physicalism and 614
583 The moral is simple. If we want phenomenal physical- property dualism—an argument not relying on the sound- 615
584 ism to be contingent, we need to reject Horganic ness of conceivability arguments. Either way, we can still 616
585 physicalism as a formulation of physicalism.13 Horganic show that no such argument is going to save the zombie 617
586 physicalism needs to be supplemented with a further re- conceivability argument. 618
587 striction on which worlds are accessible, so as to rule out If property dualism is necessary, it is true if and only if it 619
588 zombie worlds as irrelevant. is true at all worlds where phenomenal properties are in- 620
589 Therefore, not all possible worlds—not even all possible stantiated. It cannot have counterexamples. It must be un- 621
F
590 worlds that are minimal physical duplicates of our world— derstood along the following lines. According to property 622
591 are relevant to determining whether physicalism is true. dualism, the phenomenal facts are not fixed by the physical 623
592 The relevant ones are those that are accessible from the facts alone. They are fixed by some other facts, such as 624
OO
593 actual world. And the worlds that are accessible from world laws correlating the physical to the phenomenal facts 625
594 w are those that are possible given the facts at world w. If (possibly in combination with the physical facts). Now call 626
Author Proof
595 the facts at world w include Cartesian or property dualism, whatever non-physical facts are needed to fix the phe- 627
596 then zombie worlds are accessible from w. But if the facts nomenal facts according to property dualism basic facts. 628
597 at w include physicalism, then zombie worlds are inac- Here is how property dualism needs to be formulated 629
PR
598 cessible from w, regardless of how great a degree of pos- (analogously to Horganic physicalism): 630
599 sibility they otherwise enjoy.14
Property dualism: any possible world that is a mini- 631
mal physical duplicate of our world but does not 632
duplicate the basic facts is not a phenomenal dupli- 633
600 7 Conceiving Against Dualism
cate of our world. 634
601 At this point, a committed property dualist might reply
ED In other words, you can’t duplicate phenomenal facts 635
602 that—contrary to what most physicalists maintain— without also duplicating the basic facts. Given this 636
603 physicalism is not a contingent thesis. At most, (phe- formulation, it is easy to show that, by the logic of the 637
604 nomenal) physicalism is true in zombie worlds, though of zombie conceivability argument, property dualism is as 638
605 course it’s only vacuously true for lack of phenomenal vulnerable to refutation by conceivability arguments as 639
606 properties. In light of the zombie conceivability argument, physicalism is. This is instructive because it allows us to 640
CT
607 (phenomenal) physicalism is necessarily false (or neutralize the zombie conceivability argument by its own 641
608 vacuously true only at zombie worlds).15 As a conse- logic, without appealing to any special considerations 642
609 quence, according to this ambitious property dualist, the involving concepts, modality, or reference.16 Instead of 643
610 denial of physicalism—property dualism—is necessarily attempting to determine who is right about those compli- 644
611 true (or at least true at all worlds where phenomenal cated matters, we can construct a conceivability argument 645
E
612 properties are instantiated). that, while following the same logic of the zombie 646
conceivability argument, reaches the opposite conclusion: 647
RR
The Z Conceivability Argument 648

13
The tension between Horganic physicalism and the zombie (1*) Z’s are conceivable 649
conceivability argument was already noted by Neil Tennant (1994).
Tennant correctly pointed out that another option open to the
(2*) Therefore, Z’s are metaphysically possible 650
Horganic physicalist is to conclude that we are zombies. Tennant took (3*) Therefore, the basic facts do not entail the 651
this to be a reductio of the view that Horganic physicalism is a correct phenomenal facts 652
CO
formulation of physicalism, and so do I.

14
This qualification is especially germane in light of the dispute, The conclusion of the argument is that property dualism 653
related to the debate about conceivability arguments, about kinds of is false, which is the exact opposite of the conclusion of the 654
possibility and necessity. For if, as some allege (Jackson 1998, zombie conceivability argument. But the logic is the same. 655
Chalmers 1996), there is no important distinction to be drawn
This happens because Z’s are different from ordinary 656
between logical and metaphysical necessity, and therefore the
metaphysical possibility of a world may be established on purely phenomenal zombies in just the right way. By definition, 657
UN
logical grounds, then it becomes very important to distinguish Z’s are such that the phenomenal facts about them super- 658
between those possible worlds that are accessible from a world and vene on the physical facts alone and the basic facts are not 659
those that aren’t. For if physicalism and other metaphysical doctrines
are to remain contingent, as many people think they should, the set of
possible worlds that are relevant to establishing their truth at a given 16
Those considerations are the bread and butter of the current debate
world must be appropriately restricted. about conceivability arguments. The problem is, it’s difficult to reach
15
David Chalmers professes such a view (personal correspondence). consensus on any of the relevant notions.
123
h 4 DISK
h
G. Piccinini
660 needed to entail them. Thus, given the logic of the zombie inconsistent with (3) under the plausible assumption that 704
661 conceivability argument, the possibility of Z’s is incom- the phenomenal facts are entailed either by the physical 705
662 patible with property dualism as defined above. The step facts or by the basic facts, but not both. The grounds for (3) 706
663 from conceivability to metaphysical possibility is also the is precisely (1), namely, a premise with no more a priori 707
664 same as in the zombie conceivability argument. What re- ground than (1*) itself. Therefore, the dualist cannot use 708
665 mains to be shown is that Z’s are conceivable, or at least no the intuitive plausibility of (1) to rule out (1*) any more 709
666 less conceivable than ordinary phenomenal zombies. than the physicalist can use the intuitive plausibility of (1*) 710
667 A good example of a Z is a human being as the to rule out (1). 711
668 physicalist conceives of her. According to the physicalist, The dialectical situation may be described as follows. 712
F
669 the phenomenal facts about us are entailed by the physical Supporters of the two conceivability arguments are in 713
670 facts alone. In fact, according to many physicalists, the epistemic divergence, which means their disagreement 714
671 physical facts explain the phenomenal facts—a condition cannot be resolved by the methods they employed in 715
OO
672 stronger than entailment, for the physical facts might entail reaching their conclusions. Epistemic divergence can be 716
673 the phenomenal facts without there being any way to ex- defined more precisely as follows. A method of investi- 717
Author Proof
674 plain the latter in terms of the former. But to make the Z gation is public just in case it can be applied to answer the 718
675 conceivability argument go through, we need not assume same questions by all investigators and, when it is so ap- 719
676 that physicalists are right. All we need is physicalism to be plied, different investigators obtain the same answer. For 720
PR
677 conceivable and hence possible in the same way that the instance, ordinary scientific observational and experimental 721
678 dualist claims zombies to be conceivable and hence pos- methods are public in this sense, because different inves- 722
679 sible. In other words, all we need to assume is that there is tigators can replicate each other’s results by following the 723
680 no incoherence hidden in the physicalist position. same methods. By contrast, receiving divine revelations, 724
681 The result of running the two arguments in parallel—the introspecting one’s mental states, and reading tea leaves 725
682 zombie conceivability argument and the Z conceivability are not public methods. The reason is that different in- 726
683 argument—is paradoxical. For now we have two conceiv-
ED vestigators cannot apply the same methods (e.g., divine 727
684 ability arguments that start with premises that are intu- revelations occur only sporadically and unpredictably, to 728
685 itively plausible, follow the same logic, but reach mutually say the least), different investigators cannot apply the same 729
686 inconsistent conclusions. methods to the same questions (e.g., we can only introspect 730
687 To avoid this paradox, a defender of property dualism our own mental states), or different investigators that apply 731
688 cannot consistently reject the logic of the Z conceivability the same methods to the same questions obtain incom- 732
CT
689 argument on pain of undermining the zombie conceiv- patible answers (e.g., different people looking at the same 733
690 ability argument, which is the justification for her view. tea leaves make different predictions). Epistemic diver- 734
691 The only way out is for the property dualist to reject (1*). gence occurs when people answer questions by means of 735
692 But on what grounds? (1*) is supported by the same kind of methods that, for one reason or another, are not public. And 736
693 a priori intuitions that support (1). The proponent of the since their methods are not public, they cannot use those 737
E
694 zombie conceivability argument maintains that the only same methods to resolve their disagreement.18 738
695 legitimate ground for rejecting (1) would be that it contains In our case, the question is whether physicalism or 739
RR
696 a hidden contradiction, and the burden to find that con- property dualism is true, and the method being applied is 740
697 tradiction lies on the physicalist. So now, in the face of the the method of possible cases. This method consists of 741
698 Z conceivability argument, the consistent dualist must consulting one’s a priori intuitions about what is conceiv- 742
699 maintain that (1*) contains a hidden contradiction. But the able and hence possible (Jackson 1998). Since different 743
700 burden is now on the dualist to find the contradiction possibilities entail different answers to the question of 744
701 hidden in (1*).17 physicalism, it would be inconsistent to accept both (1) and 745
CO
702 Notice that it’s not open to the dualist to claim that (1*) (1*) while accepting the method of possible cases. So, one 746
703 is contradictory because it leads to (3*), which is or the other premise must be rejected. But which one? This 747
question cannot be answered by the method of possible 748
17
cases. 749
Chalmers asserts that physicalism ‘‘is not obviously conceivable in
The advantage of the present argumentative strategy is 750
any stronger sense’’ than prima facie negative conceivability (2010,
UN
p. 180) and seems to imply that physicalism is inconceivable but he that it need not resolve existing disputes about conceiv- 751
doesn’t explicitly assert that physicalism is inconceivable and does ability, possibility, and whether the former is a guide to the 752
not identify any contradiction within physicalism. Chalmers is in the latter. As a consequence, it need not locate the flaw of 753
same epistemic situation as a physicalist who asserts that zombies are
not obviously conceivable in any stronger sense than prima facie
18
negative conceivability, without being able to identify what contra- For a more explicit definition and detailed discussion of method
diction lies within property dualism. publicity and epistemic divergence, see Piccinini (2003).
123
h 4 DISK
h
754 conceivability arguments in any particular place. This is an properties conceptually entail the phenomenal properties, 803
755 advantage because the relevant notions of conceivability, without needing any further properties or facts (such as the 804
756 possibility, and inference from the former to the latter are brute inter-level necessitation relations linking physical 805
757 at least as contentious as the notion of zombie and the properties to phenomenal properties that are discussed by 806
758 possibility of physicalism. Horgan 2006, 2010). 807
759 The upshot is that conceivability arguments lead to ei- Proper Physicalism and Proper Phenomenal Physicalism 808
760 ther inconsistency or epistemic divergence, neither of correctly rule out any property dualist world, Cartesian 809
761 which is desirable. Therefore, this kind of conceivability dualist world, or zombie world as irrelevant to the 810
762 argument is inconclusive. evaluation of physicalism. Those non-physicalistic worlds 811
F
are ruled out because, according to physicalism, their 812
metaphysical laws are different from those that hold at our 813
763 8 Towards a More Adequate Formulation world.19 Incidentally, an analogous point applies to the way 814
OO
764 of Physicalism we should formulate property dualism, Cartesian dualism, 815
or any other metaphysics of mind: if we hold them to be 816
Author Proof
765 In this paper, I focused on the underappreciated link be- metaphysically contingent, they should all be formulated so 817
766 tween accessibility between possible worlds and conceiv- as to be compatible with other possibilities, such as 818
767 ability arguments, focusing on the zombie conceivability physicalism, holding at other possible though inaccessible 819
PR
768 argument. I pointed out that for the zombie conceivability worlds that are physical duplicates. What goes for 820
769 argument to go through, we need to insure that zombie physicalism goes for Cartesian dualism and property 821
770 worlds—possible worlds that are physical duplicates of our dualism. 822
771 world but at which there are no phenomenal facts—are When we do the semantics, we can express the restric- 823
772 accessible from our world. The more carefully we examine tion on relevant possible worlds that is required by proper 824
773 this issue, the clearer it becomes that whether zombie physicalism in a number of ways: we may choose a logical 825
774 worlds are accessible from our world is logically equiva-
ED
system weaker than S5, we may posit that all non- 826
775 lent to whether physicalism holds at our world. Therefore, physicalistic worlds are inaccessible from our world, we 827
776 zombie conceivability arguments as currently formulated may posit that all non-physicalistic worlds are irrelevant 828
777 beg the question. and therefore excluded from our possible world model, or 829
778 We also learned that the standard formulation of we may posit that all non-physicalistic worlds are not 830
779 physicalism in the literature—Horganic physicalism—is genuinely possible and therefore excluded from the set of 831
CT
780 inadequate. We cannot simply say that, according to possible worlds. If we choose the first, second, or third 832
781 physicalism, any possible world that is a minimal physical option, we preserve the contingency of physicalism. If we 833
782 duplicate of our world is a duplicate simpliciter of our choose the last option, we turn physicalism into a meta- 834
783 world. If we do this, we automatically assume that zombie physical necessity after all. However we choose to do it, 835
784 worlds, if they are possible, are accessible from our world. zombie conceivability arguments do not undermine 836
E
785 By doing so, we beg the question of physicalism. physicalism. Mutatis mutandis, the same applies to other 837
786 To avoid begging the question, we should refine Hor- similar conceivability argument. 838
RR
787 ganic physicalism by analogy with the way we formulate Much work remains to be done to do justice to the many 839
788 the notion of physical possibility: subtle issues that are raised by the conceivability argu- 840
ments literature—issues having to do with kinds of con- 841
789 Proper Physicalism: in any possible world with the
ceivability, kinds of necessity, kinds of entailment, kinds of 842
790 same metaphysical laws that hold at our world, every
supervenience, semantic content, etc. I hope the present 843
791 property supervenes on physical properties.
framework will help us make progress on those issues, but 844
CO
792 Proper Phenomenal physicalism: in any possible

that will have to wait for future work. 845
793 world with the same metaphysical laws that hold at
846
794 our world, phenomenal properties supervene on
795 physical properties.
796 As I said earlier, I use ‘‘metaphysical law’’ for the
UN
797 metaphysical analogue of physical laws. Examples include

798 principles governing composition, principles governing the
799 relation between physical and phenomenal properties, or 19
To rule out zombie worlds, we also need to assume that there are
800 any other metaphysical principles governing any other
phenomenal properties at our world. If we are zombies, then
801 contentious metaphysical issue. As before, the kind of obviously physicalism does not rule out zombie worlds. I am taking
802 supervenience posited here is such that the physical for granted that we are not zombies.
123
h 4 DISK
h
G. Piccinini
847 References Lewis DK (1983) New work for a theory of universals. Australas J 911
Philos 61:343–377 912
848AQ5 Balog K (1999) Conceivability, Possibility, and the mind-body Linsky L (ed) (1971) Reference and modality. Oxford University 913
849 problem. Philos Rev 108(4) Press, Oxford 914
850 Bealer G (1994) Mental properties. J Philos 91:185–208 Loux MJ (ed) (1979) The possible and the actual. Cornell University 915
851 Bigelow J, Pargetter R (1991) Science and necessity. Cambridge Press, Ithaca 916
852 University Press, Cambridge Lynch MP (2004) Zombies and the case of the phenomenal AQ7917
853 Brown R (2010) Deprioritizing the a priori arguments against pickpocket. Synthese 918
854 physicalism. J Conscious Stud 17(3–4):47–69
Marcus E (2004) Why zombies are inconceivable. Australas J Philos 919
855AQ6 Campbell KK (1970) Body and mind. Doubleday, Garden City 82(3):477–490 920
856 Chalmers DJ (1995) Facing up to the Problem of Consciousness. Marton P (1998) Zombies vs. materialists: the battle for conceiv- 921
922
F
857 J Conscious Stud 2 ability. Southwest Philos Rev 14:131–138
858 Chalmers DJ (1996) The conscious mind: in search of a fundamental McDermott M (1999) Counterfactuals and access points. Mind 923
859 theory. Oxford University Press, Oxford 108(430):291–334 924
OO
860 Chalmers DJ (2002) Does conceivability entail possibility? In: Szabó McGinn C (2004) Consciousness and its objects. Oxford University 925
861 Gendler T, Hawthorne J (eds) Conceivability and possibility. Press, Oxford 926
862 Clarendon Press, Oxford, pp 145–200 Melnyk A (2001) Physicalism unfalsified: chalmers’s inconclusive 927
Author Proof
863 Chalmers DJ (2003) Consciousness and Its place in nature. In: Stich conceivability argument. In: Gillett C, Loewer B (eds) Physical- 928
864 SP, Warfield TA (eds) The Blackwell guide to the philosophy of ism and its discontents. Cambridge University Press, Cambridge 929
865 mind. Blackwell, Madden, pp 102–142
Mitchell SD (2003) Biological complexity and integrative pluralism. 930
866 Chalmers DJ (2004) Imagination, indexicality, and intensions. Philos Cambridge University Press, Cambridge 931
Papineau D (2002) Thinking about consciousness. Oxford University 932
PR
867 Phenomenol Res 58(1):182–190
868 Chalmers DJ (2010) The character of consciousness. Oxford Univer- Press, Oxford 933
869 sity Press, Oxford Piccinini G (2003) Epistemic divergence and the publicity of 934
870 Frankish K (2007) The anti-zombie argument. Philos Q scientific methods. Stud Hist Philos Sci 34(3):597–612 935
871 57(229):650–666 Polger TW (2000) Zombies explained. In: Ross D, Brook A, 936
872 Garson J (2003) Modal logic. In: Zalta EN (ed) The stanford Thompson D (eds) Dennett’s philosophy: a comprehensive 937
873 encyclopedia of philosophy (summer 2005 edition). http://plato. assessment. MIT Press, Cambridge, pp 259–286 938
874 stanford.edu/archives/sum2005/entries/logic-modal/ Rescher N, Brandom RB (1979) The logic of inconsistency: a study in 939
ED non-standard possible-world semantics and ontology. Rowman 940
875 Garson J (2013) Modal logic for philosophers, 2nd edn. Cambridge
876 University Press, Cambridge
and Littlefield, Totowa 941
877 Hawthorne JP (2002) Blocking definitions of materialism. Philos Stud Rowlands M (2001) The nature of consciousness. Cambridge 942
878 110:103–113 University Press, Cambridge 943
879 Hermes CM (2004) Two concepts of nomic accessibility. Southwest Salmon N (1989) The logic of what might have been. Philos Rev 944
880 Philos Rev 20(2):87–94
98(1):3–34 945
Searle JR (1992) The rediscovery of the mind. MIT Press, Cambridge 946
CT
881 Hintikka J (1963) The modes of modality. Acta Philos Fenn 16:65–79
882 Horgan T (1982) Supervenience and microphysics. Pac Philos Q Stalnaker R (2002) What is it like to be a Zombie? In: Szabó Gendler 947
883 63:29–43 T, Hawthorne J (eds) Conceivability and possibility. Clarendon 948
884 Horgan T (1993) From supervenience to superdupervenience: meet- Press, Oxford, pp 385–400 949
885 ing the demands of a material world. Mind 102:555–586 Stoljar D (2001) The conceivability argument and two conceptions of 950
886 Horgan T (2006) Materialism: matters of definition, defense, and the physical. Philos Perspect 13:393–413 951
887 deconstruction. Philos Stud 131:157–183 Stoljar D (2005) Physicalism and phenomenal concepts. Mind Lang 952
E
888 Horgan T (2010) Materialism, minimal emergentism, and the hard 20(5):469–494 953
889 problem of consciousness. In: Bealer G, Koons R (eds) The waning Stoljar D (2006) Zombies and actors. In: Byrne A, Jarvis Thomson J 954
890 of materialism. Oxford University Press, Oxford, pp 309–329 (eds) Content and modality: themes from the philosophy of 955
RR
891 Jackson F (1998) From metaphysics to ethics: a defence of conceptual Robert Stalnaker. Oxford University Press, Oxford 956
892 analysis. Clarendon Press, Oxford Sturgeon S (2000) Matters of mind: consciousness, reason and nature. 957
893 Kim J (1993) Supervenience and mind. Cambridge University Press, Routledge, London 958
894 Cambridge Tennant N, Jackson F (1994) Logic and physicalism. In: Casati R, 959
895 Kim J (2005) Physicalism or something near enough. Princeton Smith B, White G (eds) Philosophy and the cognitive sciences: 960
896 University Press, Princeton proceedings of the 16th international wittgenstein symposium. 961
897 Kirk R (1974) Zombies vs materialists. Proc Aristot Soc (Suppl Vol) Hölder-Pichler-Tempsky, Vienna, pp 113–126 962
CO
898 48:135–152 Wedgewood R (2000) The price of non-reductive physicalism. Nous AQ8963
899 Kripke SA (1963a) Semantical analysis of modal logic i: normal 34(3):400–421 964
900 modal propositional calculi. Zeitschrift für mathematische Logik Wilson J (1999) How superduper does a physicalist supervenience 965
901 und Grundlagen der Mathematik 9:67–96 need to be? Philos Q 49:33–52 966
902 Kripke SA (1963b) Semantical considerations on modal logic. Acta Wilson J (2005) Supervenience-based formulations of physicalism. 967
903 Philos Fenn 16:83–94 Nous 29:426–459 968
Worley S (2003) Conceivability, possibility, and physicalism. 969
UN
904 Kripke SA (1980) Naming and necessity. Harvard University Press,

905 Cambridge Analysis 63(1):15–23 970
906 Leuenberger S (2008) Ceteris absentibus physicalism. In: Zimmer- Yablo S (1999) Concepts and consciousness. Philos Phenomenol Res 971
907 man DW (ed) Oxford studies in metaphysics. Oxford University 59:455–463 972
908 Press, Oxford, pp 145–170 Yablo S (2002) Coulda, woulda, shoulda. In: Szabó Gendler T, 973
909 Levine J (2001) Purple haze: the puzzle of consciousness. Oxford Hawthorne J (eds) Conceivability and possibility. Clarendon 974
910 University Press, Oxford Press, Oxford, pp 441–492 975
123
h 4 DISK
h
Journal : 11245
Article : 9323 123
the language of science
Author Query Form
Please ensure you fill out your response to the queries raised below and return this form along
with your corrections
Dear Author
Author Proof
During the process of typesetting your article, the following queries have arisen. Please check your typeset proof carefully
against the queries listed below and mark the necessary changes either directly on the proof/online grid or in the ‘Author’s
response’ area provided below
Query Details Required Author’s Response

AQ1 References Garson (2012), Jackson (1993, 1994) are cited in text but not provided in the reference list.
Please provide references in the list or delete these citations.
AQ2 Bigelow and Pargetter (1990) has been changed to Bigelow and Pargetter (1991) so that this citation
matches the list.
AQ3 Please provide appropriate value in place where ‘box’ present in footnote 6.
AQ4 Lewis (1982) has been changed to Lewis (1983) so that this citation matches the list.
AQ5 Please provide the page range for References Balog (1999) and Chalmers (1995).
AQ6 Please check and confirm the inserted publisher location is correct for Reference Campbell (1970).
AQ7 Please provide the volume number and page range for Reference Lynch (2004).
AQ8 Reference Wedgewood (2000) is given in list but not cited in text. Please cite in text or delete from list.

Are, Therefore, Not Possible. Please Note

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Are, Therefore, Not Possible. Please Note

Uploaded by

Copyright:

Available Formats

Dear Author,

Here are the proofs of your article.

• You can submit your corrections online, via e-mail or by fax.

• The printed version will follow in a forthcoming issue.

3 Access Denied to Zombies

68 our world in the sense of ‘accessible’ used in possible 1

formulated, physicalism is weak enough to encompass both reductive

144 further non-physical facts, such as laws that connect the

149 they are possible.4 But the zombie conceivability argument

argument and explain why (3) still doesn’t follow. 194

157 bility of phenomenal zombies in the sense that there are

1979). (3) Hughes and Cresswell say the accessibility of a world is a

bronze m* slightly different from the lump m, which it actually

systems, and therefore non-reductive physicalists should choose a

reductive physicalists need not reject S5 for the reasons offered by

448 An important motivation for Horganic physicalism is to

plicates but inconsistent with the possibility of property 557

522 nomenal duplicates.

formulation of physicalism on these grounds begs the question against

The Z Conceivability Argument 648

formulation of physicalism, and so do I.

792 Proper Phenomenal physicalism: in any possible

797 metaphysical analogue of physical laws. Examples include

904 Kripke SA (1980) Naming and necessity. Harvard University Press,

Query Details Required Author’s Response

You might also like