Imagine sitting down to play a video game called Philosophical Zombie Apocalypse that, aside from the title, you know nothing about. Upon booting it up, you surmise that it’s a first person perspective survival game set in an open world with sandbox style gameplay. What is not immediately obvious though is whether or not it’s a single-player or multiplayer game. There are other character with avatars similar to your own, and they claim to be other players, but you are aware of the trend of metaludological design in vogue at the moment – games about games, breaking the fourth wall, all that – and as such are aware of the distinct possibility that the developers might have programmed some or all of the characters other than your own to behave as indistinguishably as possible from a real player (non-player characters, i.e. NPCs). You surmise four distinct possibilities:

1. This is a single-player game. You are the only player, and all of the other characters are NPCs pre-programmed to behave the way they do and react to the events of the game world, including those instigated by you the player, either according to a logic-script or according to a random number generator, or a mix of both. 
2. This is a massively multiplayer game populated entirely by other players. There are no NPCs whatsoever, not even as quest givers. Behind every avatar is a person sitting at his computer monitor, seeing the events on screen, hearing the audio out of his speakers, and entering commands through a keyboard or controller. 
3. This is a multiplayer game partially populated by players and partially populated by Bots (NPCs programmed to behave like player characters). The sophistication with which the Bots are programmed make it virtually impossible according to all conceivable Turing tests to discriminate them from the real players. You do not know nor can you reliably guess which characters are bots and which are players. 
4. This is not a game at all but rather a movie made from computer generated images (CGI) that only appears to be a game. The inputs you attempt to deliver through your mouse or keyboard have no effect on the events on screen. Or, alternatively, you’re in Spectator Mode, seeing the game play out through what the perspective of an NPC would be. Either possibility would be functionally equivalent. 

If the fourth possibility is true, then knowing such is absolutely inconsequential insofar that the knowledge can have no impact whatsoever over the events onscreen or the outcome of the game. The character through whose perspective you’re witnessing events onscreen (one can hardly call it “your character” or “you”) will behave as he will no matter how much you pound on the keyboard or scream into your headset. Quidquid erit.

Let us therefore consider the first three possibilities on the assumption that you are actually making meaningful choices in the game, that some if not all of your inputs on the keyboard register as actions in the game. What ought to be your default assumption as to what kind of a game it is? How would you go about testing that hypothesis?

As to the first question, it has two answers. If all of the possibilities are equally likely, option three seems to conform best as a working hypothesis until more evidence can be collected. Any character could be a player or could be an NPC. I qualify that statement with an “if all the possibilities are equally likely” because it could be argued that the idea of Bots programmed to perfectly simulate players is implausible, or at least less plausible than Bots being less creative, intelligent, or otherwise as unpredictable in their behavior as players. 

The second answer is that ethically, we should assume the second possibility, that all other characters are players, and treat them accordingly until a preponderance of evidence convinces us otherwise. If we would not want to be trolled, or teabagged, or sent death threats over the in-game chat, then we ought not to do so to any character that potentially might be a player. But both of these approaches are temporary measures taken out of ignorance. To be truly ethical, we must not indefinitely assume that all characters are players, but rather pursue truth for truth’s sake and know in actuality whether they are players or NPCs. 

One interesting feature of this game is that there are temporary debuffs called “Blind” and “Deaf.” When you as the player get the “Blind” debuff, your screen goes black, even though from the audio cues you can tell that the gameworld is still proceeding along as normal (i.e. your computer hasn’t crashed). Similarly, when you get the “Deaf” debuff, your audio cuts out but the screen shows the world as normal. Because these debuffs affect how your character acts, other characters react accordingly. If, while blind, you suddenly start walking into walls, they’ll acknowledge that you’re impaired by this debuff. And, when you see other characters have an icon over their avatar indicating they’re debuffed by “Blind,” they behave similarly to how you do while debuffed, walking into walls and over cliffs, which they otherwise know not to do. If there are NPCs – whether all the characters are or only a few – they must be programmed in such a way that they behave as if they were a player whose screen has cut out – despite there being niether a player nor a screen. 

But imagine that your screen actually broke, or you closed your eyes, or you became blinded in real life. You would not have an icon over your avatar signalling a debuffed status, but your behavior would be more similar to that of a character with the debuff than those without. If other characters asked over voice chat why you were running around like such, you’d explain that you couldn’t see. So far as they could tell, there would be nothing wrong with your character, and would have to take you at your word that you were not lying or deceiving them. Similarly, vice versa. But what would not happen is that, absent the visual or audio information which you’re used to, that your character would behave like an NPC without the debuff status. Nor would it be that, if you yourself were permanently deaf and blind from birth, that your player character would behave like a non-player character from the start. 

The difference between player characters and non-player characters is not merely whether or not they can signal audio and video output devices like speakers and monitors, but also whether or not their behavior is governed (at least partially) by input methods such as keyboards or controllers. While there may be non-cancellable animations or idle animations that the player character engages in sans any input from the player – or very long cutscenes in which the character acts independently of the player input for an extended duration –  at least some of the characters actions are conditioned on button presses. Contra a non-player character, who is not governed by any input devices. 

Similarly, we can imagine cases in which your keyboard and mouse fail, and your character remains motionless as a result. And if you were to see another character inexplicably motionless, you might extrapolate that they are real players experiencing hardware malfunctions. 

None of this would definitively prove or disprove any of the hypotheses. It could always be the case that the designers programmed the bots to behave in ways which suggest that they’re real players with faulty equipment. As such, the possibility of purported “players” being bots or NPCs is unassailable and unfalsifiable. Nevertheless, behavior of other characters that is more likely explicable as caused by players or players’ hardware rather than especially clever programming would move the dial slightly towards the likelihood of there being other players, as would seem almost tautological. 

* * *

This extended analogy is exactly parallel to the situation we really do find ourselves in with respect to the uncertain existence of other minds and the possibility of being the sole solipsistic survivor of a philosophical zombie apocalypse. To paraphrase Descarte, “I think, therefore I know I am,” but I cannot be objectively certain of the existence of any other “I”s that are not me; that is, other subjective minds. As above, there are four possibilities:

1. I am the only subjective mind in existence. It may be the case that there is no objective reality whatsoever, that everything I think I see is an illusion of my mind, manifested by my subconscious. But it need not be so radical. It could also (much more likely) be the case that material reality really does exist independently, but by sheer happenstance my material brain is the only one that has an immaterial mind supervening upon it. Every other homo sapien on the planet is a meat machine – mechanically very complex, but no less an automaton than the animatronic presidents in Disney World. Similar ideas in similar though experiments have been labelled Absent Qualia, Functional Isomorphs, and Philosophical Zombies, the last of which will be my parlance going forward. The important thing to note about these Philosophical Zombies is that they are physically indistinguishable from a Person like myself (Persons being composites of bodies and minds), but they lack any perception of qualia or subjective experiences whatsoever. They are not a composite of body and mind; they are entirely body and brain, and their behaviour is governed exclusively by the physical processes that govern biology, chemistry, physics, quantum fields, and ultimately whatever is most fundamental in physical reality. Whether or not that means their actions are hard determined or random, they are not metaphysically free like I am. 
2. All humans are Persons. At minimum, every homo sapien with a fully developed and properly functioning brain likewise has a mind. At maximum, all hominids from the moment of conception and potentially other animals possess a Cartesian duality of body and mind, as could aliens and sufficiently advanced computers. Philosophical Zombies are theoretically conceivable, but among all beings with a brain or analogous information processing mechanism, an accompanying mind will always and invariably obtain, and therefore Zombies will not. The emergence of minds may or may not be necessitated by brains or their analogues in all actual worlds, but even if not, it happens to be the case in this world that all beings with brains have minds. 
3. Both Persons and Philosophical Zombies obtain at approximately similar rates, so that statistically it is no more or less probable that any given human being is a Person or a Zombie. Or even if the split is not exactly even, sufficient numbers of each obtain so that their numbers are not statistically insignificant. Regardless of rate, Zombies do not exhibit behavior that’s noticeably divergent from Persons so that one could easily distinguish between the two kinds. Thus, in this world, the problem of other minds is not an academic exercise with no real world applicability; it is a constant splinter in the mind, as one never knows whether one is dealing with a Person or a Zombie. 
4. Metaphysical Libertarianism (free will) does not exist for me, even though my perception of qualia and subjective experiences do. I am not my body, but rather a passive observer watching it from my metaphysical perch doing whatever the laws of physics will compel it to do. When my body looks at an image, the photons hit its retinas, transmit a electrochemical signal to its brain, and as it processes that material information by means of synaptic connections between neurons, my metaphysical mind involuntarily constructs a mental model of the tableau out of an arrangement of visual qualia. When my body is burned, a similar biological reaction occurs, and the metaphysical me is involuntarily subjected to the quality of intense pain. But though my mind is subjected to the behaviors of my body, it has no control over those behaviors, not even what thoughts arise. When my brain contemplates the idea of Persons and Philosophical Zombies, or writes a paper about them, my mind’s ear can hear the surface level thoughts on the subject, but cannot speak about the issues to my brain, nor reason with it, nor supervene upon it in any way. My body is a train on a fixed track, and my mind merely an observer along for the ride, powerless to derail it in any way. 

As with the game analogy above, the fourth possibility will be dismissed out of hand, not because it is inconceivable, but rather because – if true – even conceiving of it is not a knowledge of truth derived at by reason, but rather the entirely accidental and unrelated result of atoms bumping into one another in accordance with in a purely physical processes that has exactly as much truth-detecting power as billiard balls bouncing around a pool table. 

But raising the issue of free will – while not often spoken to in most explorations of the idea of Philosophical Zombies – is not incidental to the argument. As could be extrapolated from the analogy, just as we can conceive of metaphysical minds that can perceive qualia but do not possess free will, so too can we conceive of metaphysical minds that possess free will but cannot perceive qualia. (A mind that both perceives qualia and possesses free will is that of a Person, and one which does niether is no mind at all – i.e., a full Philosophical Zombie.) And as was seen, the consequence of a Person with free will but absent qualia is behavior atypical of either a full Person or a full Zombie. 

The body of a Zombie that has free will would be controlled by a metaphysical agent which is not in possession of the means by which to construct a mental model of the world necessary to navigate it. Despite being in possession of a fully functioning body, such a Zombie would behave much more like Hellen Keller than an average individual, much to the bafflement of doctors and physicians. 

A similarly strange scenario would occur if a Person could perceive certain categories of qualia but not others. In the physical domain, lightwaves and soundwaves move through entirely different mediums, and are detected by entirely different sensory organs. The ability to perceive one does not necessarily entail the ability to perceive the other, nor does the inability to perceive one impede upon perception of the other, as Matthew Murdock can attest. Similarly, nothing we know about metaphysics suggests that an inability to perceive certain categories of qualia (e.g. color) would logically necessitate an inability to perceive other categories of qualia (e.g. sounds). Moreover, given the regularity with which our physical parts are partially damaged, resulting in impairments like blindness and deafness, it’s not inconceivable that our more mysterious metaphysical parts might become partially damaged, resulting in an inability only to perceive certain categories of qualia. And even if such strange beings as the Metaphysically Deaf or the Qualitatively Blind don’t actually obtain, like Philosophical Zombies, their conceivability within the thought experiment is itself enlightening. 

Importantly, while Philosophical Zombies cannot be dismissed a priori, the evident lack of Metaphysically Deaf or Qualitatively Blind Persons can be taken as a small piece of a posteri evidence against Philosophical Zombies having yet arisen in this actual world. Were there to be a Person who was absent visual qualia, he would not act as a Zombie. A Zombie, being mindless, would have his behavior “preprogrammed.” He wouldn’t contemplate his lack of visual qualia, firstly because he’d be incapable of genuine contemplation, and secondly because he’d have no other qualitative experiences against it with which to compare. 

Not so a Qualitatively Blind man. He’d be a real “I,” capable of contemplating his subjective experiences. He’d know sound waves entering his ears would have the qualitative effect sound qualia in his mind. Against this he could compare the absence of any qualitative experiences when lightwaves hit his retinas. And because he is metaphysically free, he’d be stumbling through the dark despite possessing perfectly healthy eyes and a visual processing part of the brain. From this he would extrapolate that, if the vehicle his mind is driving is undamaged, it must be his mind itself that possesses the impairment. 

However, while we find many damaged vehicles with broken windshields driving erratically around the road (individuals with sensory disabilities), and we suspect it possible that some or all of the cars on the road could be self-driving vehicles, it’s notable that we don’t seem to find any infirmity in the drivers themselves. It may be that – to extend this metaphor poetically, if not literally – that the laws governing the interaction between the physical and metaphysical realms are such that only minds which are free and capable of perceiving a certain set of qualia are “licenced” to operate the vehicles known as human brains. If this actual world were so specifically regulated (or fine-tuned) as to narrow the types of minds that can supervene on particular parts of the physical realm (like human brains), then that ever so slightly increases the plausibility that Zombies, while possible, are not (or are no longer) permitted. 

Despite this minute move of the needle, I still suspect that Zombies are not merely conceivable, nor potentially possible, but might actually obtain. For one, I think it evident that we evolved from Zombies, and quite recently in terms of evolution, but not as a result of evolution. That is to say, some several hundred thousand or million years ago, near-human hominids were mere meat machines, no different than any other extant animal. Natural selection and mutation over time made our brains into more suitable vehicles for metaphysical agents, but nothing necessitates that the work-in-progress prototypes of, say, australopithecus africanus, even if such a species possessed a brain minimally suitable for a metaphysical mind to supervene upon, that at the exact moment such potential was realized that it was at that then also actualized. If so, for a time, Absent Qualia hunted across the African savannah while Functional Isomorphs gathered in Olduvai Gorge. The world already once experienced a Philosophical Zombie Apocalypse.

What predicated the grand miracle of – after 13.8 billions of years of a closed, purely physical system – subjective minds/metaphysical agents suddenly supervening upon the material world? That is the question. But until it is answered, we cannot know that the miracle is continuing to occur in any individual other than ourselves. 

How then ought we act towards others, not knowing whether they’re Persons or Zombies? As in the game with the question of other players or bots, the provisional answer is to treat all humans as Persons until a preponderance of evidence points us to the conclusion that an individual is indeed a Zombie. The question of animals is more open. While it may seem intuitive that higher level mammals must have subjective experience, if we admit the possibility that some homo sapiens are Zombies, our intuitions regarding other genuses and species are even more fraught. I cannot fault those who conclude animals are Persons and become vegetarians for this reason. However, it’s far more dubious that animals are metaphysically free agents. If not, they are therefore not moral actors, and while it’s conceivable – as in the fourth examples above – that they are merely moral subjects but not actors, rational minds imprisoned in irrational bodies, I find it personally improbable, and my conscious is undisturbed by eating meat (and eased over even more by just how delicious it is). 

Here and only here is where the analogy breaks down. Killing NPCs in single player games does not seem to adversely affect the player. Of the many studies conducted on the subject, violent games do not make players more violent. But I suspect that, even if we could somehow be certain that a particular human was absolutely a Zombie, we would still have ethical imperative governing our behavior towards that creature, just as we have ethical obligations not to practice needless cruelty towards animals – not because it is an agent experiencing the qualia of pain, but rather for the intrapersonal effect such cruelty has on those who engage in it. 

All of which is to say, even if it does turn out we’re living in a Philosophical Zombie Apocalypse, don’t get out your sawed-off shotgun just yet.