Introduction
David Chalmers’s (1996; 1999; 2002; 2004; 2006; 2009; 2010a,b) two-dimensional (2D) argument against physicalism—the thesis that consciousness is wholly physical—stands as one of the most extensively debated arguments in contemporary analytic philosophy. Notoriously, in its most general form the argument can be stated as follows:
[P1] It is conceivable that there exist phenomenal zombies: beings that are physically and behaviourally indistinguishable from us yet are devoid of phenomenal consciousness.
[P2] If zombies are conceivable, then zombies are possible.
[P3] If zombies are possible, then physicalism is false.
[C] Physicalism is false.
Despite the extensive scholarship generated by this argument (see Kirk 2023), one possible response to it has received comparatively little attention. Some argue that considerations of conceivability cut both ways: just as the conceivability of zombies is taken to undermine physicalism, so too the conceivability of exact physical duplicates of ourselves that do have phenomenal consciousness in just the way we do—variously called anti-zombies (Frankish 2007), bright illuminati (Balog MS) or shombies (Brown 2010; 2025)—provides equally compelling grounds for rejecting dualism (see Brown 2010; 2025; Marton 1998; Sturgeon 2000; Frankish 2007; Piccinini 2017; Balog MS; Forthcoming). Again, in its most simple and general form the 2D argument against anti-physicalism can be stated as follows:
[P1*] It is conceivable that each one of us is a shombie.1
[P2*] If it is conceivable that each one of us is a shombie, then it is possible that each one of us is a shombie.
[P3*] If each of us is a shombie, then anti-physicalism is false.
[C*] Anti-physicalism is possibly false.
The strength of this argument lies in its promise to engage anti-physicalists on their own terms (Brown 2010, fn. 2; 2025, 225). Its proponents aim to grant as many assumptions as possible to the anti-physicalist: they can accept the two-dimensional framework within which Chalmers originally formulated the zombie argument; they can also accept the formulation of physicalism in terms of metaphysical supervenience, the target of Chalmers’s original argument. According to this formulation, physicalism is true if and only if every metaphysically possible world that is a minimal exact physical duplicate of the actual world is also a duplicate of the actual world with respect to consciousness. A minimal exact physical duplicate of the actual world is a metaphysically possible world that duplicates all physical entities and features of the actual world without adding anything else. Thus, the version of physicalism targeted by Chalmers’s zombie argument holds that it is metaphysically impossible for there to exist all the entities and properties posited by an ideal theory of fundamental physics and nothing else without consciousness. This implies that a complete physical description of the universe—i.e., the conjunction of all physical truths about the universe—will entail necessarily that there exists consciousness. By contrast, on this framework the dualist will claim that it is metaphysically possible for there to exist all physical entities and properties of our world and nothing else without consciousness. Thus, if the dualist is correct then consciousness is something over and above the physical, hence a conjunction of all the physical truths about the universe will not necessarily entail its existence.
The dialectic is crucial here. Proponents of the shombie argument do not aim to deliver a knock-down refutation of dualism. Their claim is more modest: prima facie, phenomenal consciousness is conceivably physical, just as it is conceivably non-physical. There does not seem to be anything intrinsically contradictory in the supposition that we might be wholly physical beings that enjoy conscious experience. Yet if shombies are metaphysically possible, zombies cannot be, and vice versa. This incompatibility stems from the nature of a shombie world: a minimal exact physical duplicate of the actual world with the exact same conscious features that are instantiated in it. As said, by definition a minimal exact physical duplicate of our world is a world with all and only the physical features of our world. Thus, if a shombie world is possible, then the physical features of our world are metaphysically sufficient for consciousness. Consequently, it would be impossible for there to exist a minimal exact physical duplicate of our world that lacks the very conscious features instantiated in actuality, as the physical features of our world cannot be simultaneously metaphysically sufficient and insufficient for consciousness. Zombies would therefore turn out to be impossible. This, in turn, implies that if zombies are possible, shombies cannot be. Furthermore, assuming that metaphysical possibility is implied by ideal conceivability—a premise necessary for both the zombie argument and the shombie arguments to proceed—it follows that while zombies and shombies may both be prima facie conceivable, their ideal conceivability is mutually exclusive: the ideal conceivability of one precludes the ideal conceivability of the other, where we might preliminarily understand ideal conceivability as a kind of conceivability that cannot be defeated by improved reasoning (see Frankish 2007, 9).
The moral drawn from the resulting stalemate will largely depend on one’s antecedent commitments. The first, most radical option is to deny both the prima facie and ideal conceivability of either zombies or shombies. These positions are known as type-A physicalism and type-A dualism, respectively (see Chalmers 2003; Brown 2010, 54). Alternatively, one may maintain that both zombies and shombies are both prima facie and ideally conceivable. To do so, however, one must deny that ideal conceivability entails metaphysical possibility. From this denial, some alternative conclusions may be derived, depending on one’s independent theoretical leanings. If one possesses reasons to favour physicalism independent of matters of conceivability, one will conclude that while zombies are ideally conceivable, they are not metaphysically possible; this yields type-B physicalism. Conversely, if one has independent reasons to favour dualism, one will conclude that while shombies are ideally conceivable, they are not metaphysically possible, resulting in type-B dualism (Brown 2010, 54). If one lacks independent reasons to favour either physicalism or dualism, and believes that neither a priori nor a posteriori means exist to adjudicate between them, one will be driven toward a form of agnosticism.
Finally, one may accept that ideal conceivability implies metaphysical possibility, thereby granting that while both shombies and zombies may be prima facie conceivable, one of them must ultimately be ideally inconceivable. This approach posits that the prima facie conceivability of both zombies and shombies stems from the incompleteness of the current sciences of the mind. Future scientific progress will eventually either reveal phenomenal states as entirely physical configurations—without any lingering sense of puzzlement—or force acceptance that a fully physical account cannot exhaustively explain consciousness.
Fans of this view typically urge that the debate over the truth of physicalism ought to be “de-prioritized” (Brown 2010), as it cannot be resolved on purely a priori, conceptual grounds alone. If empirical evidence (or independent future theoretical developments) will favour physicalism, then the balance will tilt toward it; vice versa, it will tilt toward non-physicalism. The crucial point is that matters of prima facie conceivability alone cannot settle the issue.
Depending on whether one bets that future evidence will favour physicalism or non-physicalism the resulting view will be type-C physicalism and type-C dualism respectively.2 At a minimum, the shombie strategy, so understood, is meant to exhort physicalists not to give up on the project of explaining consciousness within a fully, robustly physicalist framework, emphasizing that acknowledging the reality of consciousness3 does not, ipso facto, commit one to its irreducible non-physicality (Brown 2025, 227).4
In this paper, I aim to critically examine the shombie argument and do a number of things. First, I will clarify the commitments that “shombists” must accept if they really wish to engage “zombists” on the battlefield of their choice. After outlining the two-dimensional framework employed by Chalmers for formulating the zombie argument, I will present four distinct versions of such an argument. Two versions assume that phenomenal concepts—i.e., concepts that refer to experiences via their phenomenally manifest qualities—have identical primary and secondary intensions (I elaborate on this below). Drawing on textual evidence, I will argue that Chalmers likely regarded this claim as highly plausible when formulating the zombie argument, thereby favouring the first two versions over the latter two, which do not assume it.
Furthermore, I contend that the claim that phenomenal concepts have identical primary and secondary intensions is equivalent to a thesis I label phenomenal essentialism (following Nida-Rümelin, e.g., 2007), or at least it entails such a thesis. According to phenomenal essentialism, by referring to the way an experience phenomenally appears—i.e., its phenomenal character—a subject is thereby in a position to grasp what that experience essentially is.5
I will also suggest, again on textual grounds, that Chalmers likely assumed an approach I label weak grounding6 (following Aleksiev 2022) in formulating the zombie argument. On this approach, grounded, metaphysically derivative entities merely re-describe their ground but ultimately amount to nothing over and above it, thereby existing only nominally, conventionally, or “in a lightweight sense”. With this groundwork in place, I will argue that the combination of phenomenal essentialism, weak grounding, and the ideal inconceivability of zombies is plausibly inconsistent with physicalism. Thus, shombists cannot challenge zombists on just any battlefield. Rather, the assumptions underpinning Chalmers’s original formulation of the zombie argument are inconsistent with at least some versions of physicalism.7 This raises the question of whether these assumptions ought to be accepted or rejected. I will suggest that indeed phenomenal essentialism is, intuitively, a highly plausible thesis. I will therefore conclude that type-A and type-C physicalists—i.e., those who maintain that zombies are ideally inconceivable—should either explain away phenomenal essentialism in a manner consistent with physicalism while accounting for its deep intuitive plausibility, or adopt non-reductive forms of physicalism that reject weak grounding in favour of strong grounding: an approach positing that grounded, derivative entities are metaphysically dependent upon fundamental ones while remaining genuinely distinct from them.
Again, a clarification is due: my aim is not to demonstrate that physicalism is incoherent or manifestly false. I grant that physicalism is both conceivably and possibly true. My aim is primarily dialectical: I aim to show that if phenomenal essentialism is true, then type-A and type-C reductive versions of physicalism are likely false. However, as said I also believe that we have strong prima facie reasons to regard phenomenal essentialism as highly plausible. If we grant that matters of intuitive plausibility bear any weight in determining the degree of credence that we ought to assign to given philosophical positions, then these forms of physicalism do not command a particularly high degree of credence, and therefore should not be assumed as our “default stance” in consciousness studies. I leave the evaluation of whether type-B physicalism—which allows both zombies and shombies to be ideally conceivable, thereby rejecting the conceivability-possibility entailment—and/or non-reductive versions of physicalism—which reject weak grounding—can be reconciled with phenomenal essentialism to another work.
The paper proceeds in three sections. Section 1 sets the stage by briefly outlining Chalmers’s two-dimensional semantics. Section 2 advances the claim that the first two of four articulated versions of the zombie argument implicitly presuppose phenomenal essentialism. I argue that this thesis is incompatible with the ideal inconceivability of zombies. Furthermore, this section introduces the notion of weak grounding to deflect a salient objection to my argument. Finally, Section 3 defends the plausibility of phenomenal essentialism itself. The paper concludes with a brief synthesis of these findings.
1. Two-dimensional Semantics
This section sets the stage by briefly outlining the two-dimensional semantic framework developed by Chalmers for his argument against physicalism.
1.1 Conceivability
Let us begin with conceivability. Within Chalmers’s (e.g., 2002, 156-9; 2010a, 143-4) two-dimensional framework, a statement is said to be primarily conceivable if it is a priori coherent considered as a hypothesis about how the actual world might turn out to be. Furthermore, a statement is ideally conceivable if and only if it is conceivable under conditions of ideal rational reflection, and negatively conceivable if it cannot be ruled out by a subject (Chalmers 2010a, 143).
Let “P” denote a complex statement formed by conjoining all truths about the nature and structure of the universe that are expressible in the vocabulary of an ideal, complete theory of fundamental physics, together with an additional clause asserting that nothing else exists beyond what is described in the other conjuncts. For simplicity, we might understand P as stating that “there exists the entire physical world and nothing else”. Let “Q” be an arbitrary true phenomenal statement, such as, say, “someone is phenomenally conscious”. The first premise of the zombie argument is meant to be read as “P∧¬Q is primarily ideally negatively conceivable”: that is, an ideal rational subject could not rule out a priori that there might be the entire physical world and nothing else without consciousness.8 In what follows, unless otherwise specified, I will use “conceivability” simpliciter to refer to primary ideal negative conceivability. The shombie argument aims to show that for any arbitrary true phenomenal statement Q, the conjunction of P∧Q is conceivable in this sense: it is primarily ideally negatively conceivable that there is the entire physical world and nothing else and Q is true.
1.2 Primary and Secondary Intensions
Let us now turn to the distinction between primary and secondary intensions. The primary intension of a concept assigns extensions to it relative to epistemically possible scenarios—namely, hypotheses complete in every detail about how the actual world might turn out to be that do not contain any contradiction (see, e.g., Chalmers 2010a, 168). The idea is that the primary intension of the concept of water will pick out the “watery stuff”—i.e., the clear, drinkable substance that fills oceans and rivers, etc.—in any scenario in which a stuff of this kind exists: XYZ in the twin-earth-scenario, H2O in the actual-world-scenario, and so forth. Furthermore, the primary intension of a concept C can be evaluated a priori: a speaker who understands C will be endowed with a certain amount of a priori knowledge enabling them to determine the primary intension of C in any scenario in which it is not empty.
As for secondary intensions, they can be conceived as functions that assign extensions to concepts relative to metaphysically possible worlds considered as counterfactual—i.e., ways the world might have been but is not. The secondary intension of our concepts typically depends on features of the actual world, where our knowledge of the character of the actual world is typically a posteriori (see Kripke 1980; Putnam 1975). When evaluating the secondary intension of a concept C we hold fixed the character of the actual world and ask what are the features that an entity could not possibly fail to have to fall under the extension of C—i.e., the essential features of the referent of C. Moreover, secondary intensions have a stable value, since most of our concepts function as Kripkean rigid designators, designating the same features in any metaphysically possible world in which they are not empty.
1.3 The Four Versions of the Zombie Argument
As a matter of fact, four distinct versions of the zombie argument can be formulated. These versions are distinguished according to the following schema (PI = primary intension; SI= secondary intension):
| Version | Phenomenal concepts | Physical concepts |
| 1 | PI = SI | PI = SI |
| 2 | PI = SI | PI ≠ SI |
| 3 | PI ≠ SI | PI = SI |
| 4 | PI ≠ SI | PI ≠ SI |
That is to say, the first version assumes that both physical and phenomenal concepts have identical primary and secondary intensions; the second assumes that while phenomenal concepts do, physical concepts do not; and so forth. Here phenomenal concepts are understood as those concepts that refer to experiences via their phenomenally manifest qualities, whereas physical concepts are defined as those deployed within an ideal, complete theory of fundamental physics to characterize the physical entities and properties posited by such theory.
It would be distracting to focus here on how Chalmers defends each premise of each version of the argument. What matters for present purposes is that shombists aim to challenge zombists on the battlefield of their choice. To do so, they must construct four symmetrical versions of the shombie argument, premised on the conceivability of P∧Q, rather than on the conceivability of P∧¬Q. In the next section, I will argue that the ideal inconceivability of P∧¬Q (implied by the ideal conceivability of P∧Q, given the entailment from conceivability to possibility) when combined with the claim that phenomenal concepts have identical primary and secondary intensions—as in the first two versions of the zombie argument, hence in the first two versions of the intended shombie argument—is plausibly inconsistent with physicalism. This arguably gives the anti-physicalist a strategic advantage. For, if it could be shown that it is plausible that phenomenal concepts have identical primary and secondary intensions, then the shombie argument would be undermined. In the third section, I will suggest that the claim that phenomenal concepts possess identical primary and secondary intensions is intuitively highly plausible.
2. Evaluating the Shombie Argument
2.1 The Shombie Argument and the Claim that Phenomenal Concepts Have Identical Primary and Secondary Intensions
In this section, I aim to argue that the conjunction of the ideal inconceivability of zombies with the claim that phenomenal concepts have identical primary and secondary intensions is plausibly inconsistent with physicalism.
Let us examine the implications of claiming that phenomenal concepts possess identical primary and secondary intensions. As we saw, within Chalmers’s two-dimensional framework, there is a robust case for the view that the primary intension of given concepts can be evaluated a priori. For instance, Chalmers (2010b, 546) mentions the following principle as one of the “core claims” of two-dimensional semantics: “A sentence token S is a priori (epistemically necessary) if and only if the primary intension of S is true at all scenarios”. Shortly thereafter, he (Ivi, 548) says that there is “a very strong and general connection between primary intensions and apriority”. As previously noted, evaluating the secondary intension of a concept, in turn, requires holding the character of the actual world fixed and determining the features that an entity could not fail to have to fall under that concept’s extension across all possible worlds. By definition, the features that an entity could not fail to have to fall under the extension of a concept in any possible world—i.e., those the referent of the concept could not possibly lack while remaining the entity it is—coincide with the essential features of that referent. Consequently, if the primary of a concept C coincides with its secondary intension, it follows that a subject would be in a position to grasp the essential features of C’s referent a priori—that is, solely in virtue of understanding C.
More generally, phenomenal concepts refer to experiences via their phenomenal character, or simply, via how those experiences feel. It seems very plausible that the phenomenal character of an experience is at least partially constitutive9 of its essence: if an experience manifests the phenomenal qualities of pain—if it hurts, or is experienced painfully—then, intuitively, that experience just is and could not possibly fail to be a real, genuine experience of pain; conversely, if nothing is hurting, there is a very intuitive sense in which one could not possibly be having a genuine, occurrent experience of pain—one could not possibly be feeling pain. If this is right, then by deploying a phenomenal concept—by referring to an experience via the way it feels—a subject can understand what that experience is and could not possibly fail to be (see Kripke 1980, 152-3).
Chalmers appears to endorse this line of reasoning. He claims that
As Kripke noted, there does not seem to be the same strong dissociation between appearance and reality in the case of consciousness as in the case[s] of water (…). While it is not the case that anything that looks like water is water (…) it is plausibly the case that anything that feels like consciousness is consciousness. (Chalmers 2010a, 149-50)
Shortly thereafter, he (ibid.) argues this fact provides strong support for the coincidence of the primary and secondary intensions of phenomenal terms. Thus, overall, I think there is a compelling case that the claim that phenomenal concepts possess identical primary and secondary intensions is equivalent to, or at least entails, the thesis that by referring to an experience via its phenomenal character, a subject is positioned to grasp its essential nature. Following Martine Nida-Rümelin (2007), I will refer to this thesis as phenomenal essentialism.
We saw that zombies and shombies cannot be both ideally conceivable. This incompatibility stems from the assumption that ideal conceivability implies possibility. Since shombies and zombies cannot be both metaphysically possible, they cannot be both ideally conceivable. Suppose now shombies were conceivable in the relevant sense. That is to say, assume that for any arbitrary phenomenal truth Q, P∧Q is conceivable (where, recall, P represents an ideal, complete physical description of the world coupled with a “that’s all” clause). In light of the foregoing, this would imply deeming P∧¬Q—i.e., zombies—inconceivable. Given how primary ideal negative conceivability is defined, the inconceivability of P∧¬Q implies that an ideal rational subject would detect a contradiction in P∧¬Q a priori. Thus, such a subject would be able to infer Q from P a priori: If, for any experience conceived in terms of its phenomenal character, an ideal rational subject deems it contradictory to suppose that all the entities and features posited by an ideal physical theory and nothing else exist without that experience, then clearly such a subject takes the existence of the experience to be entailed by a complete physical description of the world. In brief, If P∧¬Q is contradictory, P entails Q. As far as I can see, this further implies that our ideal rational subject would judge that each experience just is one of the entities posited and described by an ideal theory of fundamental physics—or a sum of such entities, or a set constructible from them, or something of that sort. It is difficult to see how it could be inconceivable that there might be the entire physical world and nothing else without consciousness unless each experience were taken to be nothing more than one of the entities figuring in an ideal theory of physics (or a sum or set thereof). If experiences were assumed to be something genuinely over and above the entities posited by an ideal physical theory, then it could not be conceivable that there might the entire physical world and nothing else and those experiences. After all, nothing else means nothing else!
I take it that if thinking about an experience in terms of its phenomenally manifest features allows a subject to grasp its essence—as seemingly entailed by the claim that phenomenal concepts have identical primary and secondary intensions—then experiences are essentially and entirely phenomenal entities: all that it is for them to be part of reality is their phenomenal character. If the essence of experiences were taken to transcend their phenomenal appearance, then one could not be in a position to understand what those experiences essentially are a priori solely by reflecting about their phenomenal character. But in that case, phenomenal essentialism would have to be rejected.
So, given the conceivability of P∧Q, each experience is assumed to be nothing more than a physical entity (or a sum or set thereof). At the same time, given phenomenal essentialism, experiences are taken to be essentially and entirely phenomenal entities—entities whose essence is exhausted by their phenomenal character.
This, I contend, places the physicalist in a rather uncomfortable position. It implies that some (sums or sets of) allegedly entirely physical entities immediately manifest their essence in introspection, allowing a subject to know what these entities essentially are simply by undergoing given experiences and reflecting about them in terms of how they feel. I take it that most physicalists would plausibly reject this claim. As Lewis (1995, 142) would put the point, making discoveries in neurophysiology cannot be so easy.
Furthermore, the view under consideration threatens to violate what appears to be an essential constraint of most of the approaches that define themselves as physicalist without qualification. It seems to be a core commitment of physicalism—at least as standardly conceived—that in order for it to be true there cannot exist fundamental entities or properties that are essentially-experience-involving or bear any sort of special relation to experiential properties (see, among others, Montero 2010; Morris 2016). Here by a fundamental entity, I mean an entity that is not metaphysically dependent on any other concrete object. By a fundamental property I mean a property whose instantiation does not occur in virtue of any other fact. A no-fundamental-mentality constraint is typically regarded as the core element distinguishing physicalism from alternative views such as dualism, panpsychism, etc.
So, shombists appear to face a dilemma. On the one hand, they aim to challenge the zombist on the battlefield of their choice. To do so, they must provide a shombie-analogue of the first two versions of the zombie argument—versions in which it is assumed that phenomenal concepts have identical primary and secondary intensions, which is plausibly equivalent with, or at least entails, phenomenal essentialism. On the other hand, they maintain that experiences are essentially wholly physical. But arguably, no matter how thoroughly one analyses their own experiences, armchair reflection based on the sole phenomenology will always and only reveal phenomenal properties: properties that appear in a certain way in introspection and specify what it’s like to undergo certain experiences. So, if experiences are assumed to be essentially wholly physical states, then it is not the case that one can grasp what experiences essentially are simply by referring to them via a phenomenal concept. Phenomenal essentialism would therefore have to be rejected. Thus, if it could be shown that phenomenal essentialism is a plausible thesis, then shombists would be under pressure. Arguing for the plausibility of phenomenal essentialism is the task of the next section. Before turning to that, however, let me address some relevant objections that may be raised against what I have argued so far.
2.2 Responding to Potential Objections
First, a shombist might respond along the lines of type-B physicalism, arguing that phenomenal concepts enable us to grasp the nature of phenomenal states, but present that nature under a distinctively phenomenal, non-physical characterization. The shombist would further insist that insofar as the relevant metaphysical relations between experiences and fundamental physical facts obtain, phenomenal states are nonetheless essentially wholly physical. On this view, “physicality” and “phenomenality” would be regarded as distinct but equally valid ways of grasping what it is for given experiences to exist—i.e., their essence.
This move appears to be unavailable to anyone who deems zombies ideally inconceivable, given their commitment to an a priori entailment between physical and phenomenal facts under conditions of ideal rational reflection. To clarify: as previously established, the ideal inconceivability of zombies entails that an ideal rational subject would be capable of inferring phenomenal facts a priori from a complete physical description of the universe. This conclusion precludes the possibility that phenomenal and physical concepts may provide two distinct but equally valid ways to grasp the nature of one and the same kind of entity, because, arguably, there cannot be an a priori entailment between robustly distinct ways of conceptualizing given entities.10 The problem is that phenomenal concepts only present entirely phenomenal features. For phenomenal concepts to be a priori connected to the concepts of a future, ideally complete theory of fundamental physics, both sets of concepts should present their referents in a recognizably similar manner. Given that phenomenal concepts present their referents under an entirely phenomenal characterization, the concepts of an ideal physical theory would likewise need to provide a phenomenal characterization—or at least one that is sufficiently close to phenomenology to guarantee an a priori entailment. But this is arguably incompatible with physicalism, as it seems to imply that the referents of the concepts of an ideal physical theory would resemble the phenomenal properties we encounter in introspection. So, either the a priori entailment must be denied, in which case zombies would remain ideally conceivable (thereby rendering shombies ideally inconceivable and collapsing the argument) or phenomenal essentialism must be rejected. I therefore conclude that type-A and type-C physicalists are committed to denying phenomenal essentialism.
Another line of response an opponent might consider is to argue that the conceivability of shombies does not imply that experiences are fundamental: they may be identical to entirely physical states, yet these states may in turn be grounded on some more fundamental—and equally entirely physical—facts or properties that are not essentially phenomenal. In this way, one may concede that some wholly physical but not fundamental facts—namely, experiences—are essentially and entirely phenomenal, while maintaining that the no-fundamental-mentality constraint is preserved. Let me note a couple of things.
First of all, I still find puzzling that a physicalist may accept that the essence of some allegedly entirely physical states—whether fundamental or not—may be accessible to subjects simply by reflecting on the experienced qualities of given experiences. Lewis’s (1995) point that making discoveries in neurophysiology cannot be so easy still applies, for it does not hinge on whether the discoveries in question concern fundamental physical facts.
That being said, let me distinguish between two conceptions of ground which, following Damian Aleksiev (2022), we may call strong ground and weak ground (but see also Fine 2012, 51-2 on strict versus weak ground). As Aleksiev puts it, an entity is a weak groundee if and only (1) if it is derivative and (2) it is nothing over and above its ground. Thus, a weak groundee contains no new information about reality beyond what is provided by its ground. The idea is that, on a weak approach to ground, “the grounding fact provides all that is essentially required for the entities contained in the grounded fact to be part of reality” (Goff 2017, 45; see also Armstrong 1997, 12, among others). On this approach, only the ground is taken to be capital-R Real: the grounded entities are taken to exist only nominally, conventionally or “in a lightweight sense”. By contrast, Aleksiev defines a strong groundee as an entity that (1) is derivative and (2) it is something over and above the ground. On this approach, the groundee contains genuinely new information about reality beyond what is contained in the ground. Thus, both the grounded entities and the ground are taken to be capital-R Real: they are both taken to exist in a robust sense of “to exist” (see also Schaffer 2017, 23).
There is evidence that Chalmers assumed a weak conception of ground when he first formulated the zombie argument. He (1996) argues that derivative facts logically supervene on their ground, and that if facts of kind B logically supervene on facts of kind A then “all there is to the B-facts being as they are is that the A-facts are as they are”11 (Ivi, 36). He (Ivi, 41) also says that macro-physical facts “merely redescribe” what is described by micro-physical, fundamental facts (see also Chalmers 2009, 120). All this strongly suggests that Chalmers rejected a strong conception of ground with respect to macro-physical entities. Accordingly, the zombie argument appears to be targeted, at least in its original version, against those forms of physicalism which posit that experiential facts can be entirely reduced to their fundamental physical ground.
If this is the case, then my arguments still hold. For, given the purported ideal inconceivability of P∧¬Q, each experience would be taken to be nothing more than one of the entities posited by an ideal theory of physics (or a sum or set thereof). Even if the relevant physical entities were regarded as non-fundamental, they would be weakly grounded on a fundamental physical base. Hence, they would amount to nothing over and above such base. Phenomenal essentialism implies that experiences are essentially and entirely phenomenal. It follows that at least some fundamental physical entities ought to be regarded as essentially and entirely phenomenal—a thesis that, I believe, most physicalist would plausibly reject.
Of course, one may reject a weak conception of ground altogether, and argue that we should adopt a strong conception instead—one in which experiences are robustly metaphysically dependent from their fundamental physical ground while remaining somehow distinct from it. By arguing along these lines, however, one would not be challenging the zombist on the battlefield of their choice anymore. For, as I have noted, there is strong evidence that Chalmers meant the zombie argument to target metaphysically reductive versions of physicalism. Thus, the zombist will have succeeded in at least one respect: showing that if phenomenal essentialism is true then reductive physicalism ought to be rejected. This is not the place to settle the question of whether reductive or non-reductive versions of physicalism are more plausible. Readers can interpret my arguments as concerned exclusively with reductive forms of physicalism.
3. Denying Phenomenal Essentialism is Implausible
So far, I have argued that type-A and type-C versions of reductive physicalism are plausibly incompatible with phenomenal essentialism. Thus, if one could show that phenomenal essentialism is, indeed, a plausible thesis, then these versions of physicalism would emerge as implausible. Arguing for the plausibility of phenomenal essentialism is the task of this section.
As we have seen, the primary intension of a concept captures the superficial features of its referent; the secondary intension, in turn, captures the essential features of the referent. In the case of phenomenal concepts, the features picked out by the primary intension are the phenomenal features of an experience.
Suppose now phenomenal concepts did not have identical primary and secondary intensions. This would imply that the superficial, phenomenal features of an experience do not coincide with what that experience necessarily is. So, for example, a given experience may present the phenomenal features of pain—i.e., being experienced painfully, hurting—without thereby necessarily being a real experience of pain, just as, in the case of water, a given substance may present the superficial features of water—being a drinkable, odourless, liquid substance—without thereby necessarily being H2O.
I take it that the latter is an extremely counter-intuitive and ultimately utterly implausible view. I invite you to pause for a moment to slowly and carefully consider what is being claimed: you might be undergoing an excruciatingly painful sensation right now, and THAT might not necessarily be a real experience of pain.
Notice, crucially, that when I say that a given experience may not appear painful without thereby necessarily being a real experience of pain, the appearance in question is phenomenal, not epistemic. That is to say, I am not concerned with how things might seem or appear to a subject in the sense that might involve a false belief. Rather, I’m concerned with what is immediately phenomenally presented to a subject in undergoing an experience:12 that is what phenomenal concepts refer to. Importantly, I am not defending any form of infallibilism about phenomenal judgments, understood as the thesis that subjects cannot err in recognizing, categorizing or describing the phenomenal character of their experiences. Nor am I endorsing self-intimation, understood as the thesis that subjects cannot undergo given experiences without thereby ipso facto forming the belief that they are having them. All I am saying is that introspective phenomenal judgments—judgments concerning the phenomenal character of given experiences—sometimes get things right: sometimes, these judgments succeed in correctly characterizing the way an experience is felt by its subject. If and when this happens, according to phenomenal essentialism, the judging subject will grasp what the experience in question essentially is.
Another reason to endorse phenomenal essentialism is that denying it risks generating an infinite regress. Even if one denies that an experience’s phenomenal appearance coincides with the essence of the experience, one must still admit that the essence of the phenomenal appearance of an experience coincides with how it, the phenomenal appearance, appears. After all, an appearance is nothing but this: an appearance. Claiming that the essence of the way an experience phenomenally appears does not coincide with its appearance would yield an infinite regress. As far as I can see, the only way to halt this regress is by positing that there is something—whether experiences tout court or their phenomenal mode of appearing—such that their appearance collapse onto their essence. By directly referring to the way those entities or properties appears, one would thereby be in a position to grasp their essence. Furthermore, if essentialism about something—be it consciousness tout court or its phenomenal mode of appearing—must be posited to avoid an infinite regress, then it is unclear whether denying that phenomenal concepts have identical primary and secondary intensions will help the shombist. For even if one denied that the essence of conscious states qua complex, multifaceted wholes is presented to subjects in introspection, one would still have that the essence of at least one aspect of these wholes—namely, the way experiences phenomenally appear—is, indeed, phenomenally manifest, being exhausted by its appearing phenomenally in a certain way. Thus, by directly referring simply to the way an experience phenomenally appears—rather than to the whole experience qua complex, multifaceted entity—a subject would be in a position to grasp its nature. Given the alleged impossibility of P∧¬Q—necessary for physicalism to be true—combined with a weak approach to ground (which, as we saw, Chalmers most likely assumed) we obtain that these entirely phenomenal aspects of experiences would result as identical to some fundamental physical entity or state (or a sum or set thereof). It would thereby follow that some (sums or sets of) fundamental physical entities or properties are essentially phenomenal.
My point may be further appreciated by looking at how the four versions of the zombie argument are formally presented by Chalmers. As one can see, the last two versions appear structurally identical to the first two ones, with the sole difference of being targeted against physicalism concerning consciousness’s mode of appearing, rather than against physicalism concerning consciousness tout court (see Chalmers 2010a,b). My conjecture is that Chalmers was implicitly relying on the—to my mind, highly plausible—idea that either in conscious states tout court or in their phenomenal mode of appearing appearance and essence collapse. If this is right, then the general rationale underlying all four versions may be synthesized as follows:
[I] There is something—whether consciousness tout court or its phenomenal mode of appearing—such that by directly referring to how it appears a subject can understand what it essentially is. Let us call “*” the entity or feature for which appearance and essence coincide.
[II] It is conceivable that there might be entire physical world and nothing else without *.
[III] Given [I], if it is conceivable that there might be the entire physical world and nothing else without *, then it is possible that there might be the entire physical world and nothing else without *.
[IV] If it is possible that there might be the entire physical world and nothing else without *, then physicalism about * is false.
[C] Therefore, physicalism about * is false.
In turn, the argument I am proposing for the implausibility of reductive physicalism, given essentialism about *, may be summarized as follows:
[i] If essentialism about * is true, then * is essentially and entirely phenomenal: its essence is exhausted by its appearing phenomenally in a certain way.
[ii] If reductive physicalism about * is true, then * is nothing but a suitable arrangement of fundamental physical entities.
[iii] Therefore, if, per absurdum, both reductive physicalism about * and essentialism about * were true, some suitable arrangements of fundamental physical entities would be entirely and essentially phenomenal.
[iv] If some suitable arrangements of fundamental physical entities are entirely and essentially phenomenal, then reductive physicalism is plausibly false.
[v] Therefore, either essentialism about * is false and reductive physicalism is true or vice versa.
[vi] It is highly plausible that essentialism about * is true.
[C] Therefore, reductive physicalism is plausibly false.
My analysis has aimed to explore the battlefield on which the shombie argument is meant to engage zombists, arguing that such battlefield is less neutral than shombists assume. While the shombie strategy promises to mirror the zombie argument, I have argued that its success depends upon commitments—most notably, the denial of phenomenal essentialism—that seem both theoretically unstable and intuitively implausible. If phenomenal essentialism is taken seriously—as, I contend, both the phenomenology of experience and Chalmers’s two-dimensional semantics suggest we should do—then the shombie argument cannot be constructed, at least in its strongest form, without collapsing into theses the physicalist plausibly ought to reject. These include, for example, the thesis that some (sums or sets of) fundamental physical entities are essentially and entirely phenomenal, or the claim whereby the concepts of a future, ideally complete physical theory would provide a (quasi-)phenomenal characterization of their referents, thereby making fundamental physical entities and properties appear too much like the phenomenal properties we encounter in experience, with the result that one slides towards seemingly (quasi-)panpsychist views. The upshot is that shombists cannot, in fact, challenge zombists on just any battlefield.
This is not to say that I have settled the debate over the metaphysics of consciousness, nor that I have definitely undermined physicalism. However, I hope to have at least highlighted some important constraints. Type-A and type-C reductively physicalist approaches should explain away essentialism about either consciousness or its phenomenal mode of appearing in a manner that (1) is consistent with physicalism and (2) makes justice to why essentialism about either consciousness or its phenomenal mode of appearing seems intuitively compelling. Alternatively, it remains to be seen whether nonreductive versions of physicalism—i.e., those rejecting weak grounding in favour of strong grounding—or type-B versions—i.e., those rejecting the entailment from ideal conceivability to metaphysical possibility, thereby allowing that both zombies and shombies may be ideally conceivable—can be rendered consistent with phenomenal essentialism while preserving a genuinely physicalist framework.
Acknowledgments
I would like to express my gratitude to the two anonymous reviewers of this article, whose comments enabled me to produce the version of the article you are reading, which is significantly improved compared to previous versions. I would also like to thank the editor of the European Journal of Analytic Philosophy for his patience and interest in my work. Versions of this article have been presented at academic events at the universities of Turin, La Laguna, and Fribourg, and at the School of Advanced Studies IUSS, in Pavia. I thank everyone who participated in these events for their feedback. Without the following people, this work would not have been possible: David Chalmers, Alfredo Tomasetta, Martine Nida-Rümelin, Gianfranco Soldati, Jacopo Pallagrosi, and Davide Bordini. Finally, I would like to thank my family—and in particular Francesco Cortesi, Giulia Cortesi, and Valerio Pepe—without whom my life would be neither conceivable nor possible.
REFERENCES
Aleksiev, Damian. 2022. “Lightweight and Heavyweight Anti–physicalism.” Synthese 200 (112): 1–23.
https://doi.org/10.1007/s11229-022-03506-3
Armstrong, David. 1997. A World of States of Affairs. Cambridge: Cambridge University Press.
Balog, Katalin. “Illuminati, Zombies and Metaphysical Gridlock.” Unpublished manuscript.
Balog, Katalin. Forthcoming. “The Rise and Fall of the Mind Body Problem.” In Meaning, Modality and Mind: Essays Commemorating the 50 Anniversary of Naming and Necessity, edited by Corine Besson, Anandi Hattiangadi, Romina Padro, and Antonella Mallozzi. Oxford: Oxford University Press.
Botin, Marcelino. 2023. “Russellian Physicalists Get Our Phenomenal Concepts Wrong.” Philosophical Studies 180 (7): 1829–48. https://doi.org/10.1007/s11098-023-01955-1
Brown, Richard. 2010. “Deprioritizing the A Priori Arguments Against Physicalism.” Journal of Consciousness Studies 17(3–4): 47–69.
Brown, Richard. 2025. Consciousness as Representing One’s Mind: The Higher–order Approach to Consciousness Explained. New York: Oxford University Press.
Chalmers, David. 1996. The Conscious Mind. Oxford: Oxford University Press.
Chalmers, David. 1999. “Materialism and the Metaphysics of Modality.” Philosophy and Phenomenological Research 59: 473–96. https://doi.org/10.2307/2653685
Chalmers, David. 2002. “Does Conceivability Entail Possibility?” In Conceivability and Possibility, edited by Tamar S. Gendler and John Hawthorne, 145–200. Oxford: Oxford University Press.
Chalmers, David. 2003. “Consciousness and its Place in Nature.” In Blackwell Guide to the Philosophy of Mind, edited by Steven P. Stich and Ted. A. Warfield, 102–42. Blackwell.
Chalmers, David. 2004. “Epistemic Two-dimensional Semantics.” Philosophical Studies 118(1/2): 153–226.
https://doi.org/10.1023/B:PHIL.0000019546.17135.e0
Chalmers, David. 2006. “The Foundations of Two-dimensional Semantics.” In Two-Dimensional Semantics, edited by Manuel Garcia-Carpintero and Josep Macià, 55–140. Oxford: Oxford University Press.
Chalmers, David. 2009. “The Two-dimensional Argument Against Materialism.” In The Oxford Handbook of Philosophy of Mind, edited by Brian McLaughlin, Angsar Beckermann, and Sven Walter. New York: Oxford University Press.
Chalmers, David. 2010a. “The Two-dimensional Argument against Materialism.” In The Character of Consciousness, 141–206. New York: Oxford University Press.
Chalmers, David. 2010b. “Two-dimensional Semantics.” In The Character of Consciousness, 541–68. New York: Oxford University Press.
Chalmers, David. 2015. “Panpsychism and Panprotopsychism.” In Consciousness in the Physical World: Perspectives on Russellian Monism, edited by Torin Alter and Yujin Nagasawa, 246–76. New York: Oxford University Press.
Cutter, Brian. 2019. “Against the Middle Ground: Why Russellian Monism is Unstable.” Analytic Philosophy 60 (2): 109–29.
https://doi.org/10.1111/phib.12149
Fine, Kit. 2012. “A Guide to Ground.” In Metaphysical Grounding, edited by Fabrice Correia and Benjamin Schneider, 37–80. Cambridge: Cambridge University Press.
Frankish, Keith. 2007. “The Anti–Zombie Argument.” Philosophical Quarterly 57 (229): 650–66. https://doi.org/10.1111/j.1467-9213.2007.510.x
Goff, Philip. 2017. Consciousness and Fundamental Reality. Oxford: Oxford University Press.
Kirk, Robert. 2023. “Zombies.” The Stanford Encyclopedia of Philosophy (Fall 2023 Edition), edited by Edward N. Zalta. https://plato.stanford.edu/archives/fall2023/entries/zombies/
Kripke, Saul. 1980. Naming and Necessity. Oxford: Blackwell.
Lewis, David. 1995. “Should a Materialist Believe in Qualia?” Australasian Journal of Philosophy, 73 (1): 140–44.
https://doi.org/10.1080/00048409512346451
Liu, Michelle. 2024. “Introspection and the Appearance/Reality Distinction.” Oxford Studies in Philosophy of Mind 4. https://doi.org/10.1093/9780198924159.003.0011
Marton, Peter. 1998. “Zombies versus Materialists: The Battle for Conceivability.” Southwest Philosophy Review 14 (1): 131–38. https://doi.org/10.5840/swphilreview199814115
Morris, Kevin. 2016. “Russellian Physicalism, Bare Structure, and Swapped Inscrutables.” Journal of Consciousness Studies 23 (9–10): 180–98.
Nida-Rümelin, Martine. 2007. “Grasping Phenomenal Properties.” In Phenomenal Concepts and Phenomenal Knowledge: New Essays on Consciousness and Physicalism, edited by Torin Alter and Sven Walter, 307–38. Oxford: Oxford University Press.
Pereboom, Derk. 2019. “Russellian Monism, Introspective Inaccuracy, and the Illusion Meta-problem of Consciousness.” Journal of Consciousness Studies, 26 (9–10): 182–93.
Piccinini, Gualtiero. 2017. “Access Denied to Zombies.” Topoi 36 (1): 81–93. https://doi.org/10.1007/s11245-015-9323-6
Putnam, Hilary. 1975. “The Meaning of ‘Meaning’.” In Philosophical Papers, volume 2: Mind, Language and Reality, 215–74. Cambridge: Cambridge University Press.
Schaffer, Jonathan. 2017. “The Ground Between the Gaps.” Philosophers’ Imprint 17 (11): 1–26.
http://hdl.handle.net/2027/spo.3521354.0017.011
Sturgeon, Scott. 2000. Matters of Mind: Consciousness, Reason and Nature. London: Routledge.
https://doi.org/10.4324/9780203424551