Consciousness Makes Things Matter

Welfare

Welfare is what we have in mind when we ask what makes one’s life go best, whether one individual is better off than another, whether one has a life worth living, or whether one is doing well or badly. Oftentimes, welfare is taken to also concern whether one can be harmed or benefitted, what one wants for someone who one cares about, what is modulated in reward and punishment, how desirable it is to be in the position of a subject, and what is in a given individual’s interest. Other expressions that are used to denote the same phenomenon include ‘well-being’, ‘prudential value’, ‘personal value’, ‘quality-of-life’, and ‘good-for’.

The basic categories of welfare include welfare goods, welfare subjects, and welfare levels. A welfare good¹ is something that makes a welfare subject intrinsically better off; a welfare subject is something that has a welfare level; a welfare level is how well or badly a welfare subject is doing. As an example, you are a welfare subject, and (depending on which theory of welfare goods you endorse) your welfare level will increase if (other things being equal) you have a pleasurable experience, or have some of your desires satisfied, or acquire some new knowledge. In brief: welfare goods determine welfare levels for welfare subjects.²

Though I’ll freely talk about welfare levels, I’ll remain neutral on nearly all substantive questions about the structure of welfare. The notion of a welfare level is needed to make sense of how good or bad something is for a subject, whether one subject is better or worse off than another, and whether subject has a life worth living. But I’ll leave open whether welfare levels are absolute or relational, whether welfare levels are totally orderable, whether welfare levels are closed under addition, and whether welfare goods combine additively. I’ll later discuss some questions that arise if we assume that welfare has a zero point, marking the threshold for a life worth living. But my main arguments can be accepted even by those who reject the idea of welfare level zero.

The question of what makes an entity a welfare subject is connected to the question of what grounds moral status. While it’s controversial whether being a welfare subject is necessary for moral status, nearly everyone agrees that being a welfare subject is sufficient for moral status. Consider, for example, Crisp [2017]’s remark that “a theory which said that [welfare] just does not matter would be given no credence at all.” I’ll take this modest assumption for granted.³

In the welfare literature, there are a number of analyses of the concept of welfare. But these analyses typically focus on what it means for something to be a welfare good, leaving open which entities can be the beneficiaries of those goods. As examples, the locative analysis says that welfare goods are objective goods located in a given subject’s life, the positional analysis says that welfare goods are what determine how desirable it is to be in the position of a given subject, the suitability analysis says that welfare goods are whatever serve a given subject well, and the rational care analysis says that welfare goods are what one would desire for a subject that one rationally cares about.⁴ In each case, the notion of a subject appears in the analysans. Yet we can ask: which kinds of entities have lives, or have positions that are evaluable with respect to desirability, or can be served well or badly, or can be worthy of rational care? To answer these kinds of questions, we need a theory of welfare subjects.

The Phenomenal Theory

An entity is phenomenally conscious just in case there’s something it’s like to be that entity, just in case it has subjective experiences, just in case it feels a certain way, or just in case it has a first-person point of view.⁵ I’ll remain neutral on both metaphysical questions about the nature of consciousness and epistemological questions about how we can know which entities are conscious. Though I’ll sometimes take for granted standard assumptions about which entities are conscious (or not), this shouldn’t be taken as endorsing any particular theory of consciousness. Which theory of consciousness is correct will make a difference to which entities are welfare subjects. But I think the phenomenal theory is plausible no matter which theory of consciousness one favors.⁶

The central thesis of this paper is a metaphysical analysis: to be a welfare subject just is to be conscious. Given the definition of welfare subjects mentioned earlier, this is equivalent to saying that all and only conscious subjects have welfare levels. In other words, it’s for all and only conscious subjects that there is a fact of the matter about how well or badly that subject is doing, how good its life is, whether it’s better or worse off than another subject, and so forth. Over the rest of the paper, I’ll call this the phenomenal theory of welfare subjects:

The Phenomenal Theory

Consciousness is what makes an entity a welfare subject.

According to the capacity version of the phenomenal theory, an entity is a welfare subject just in case it has the capacity for consciousness. According to the state version of the phenomenal theory, an entity is a welfare subject just in case it’s in a conscious mental state. If you’re in a dreamless sleep, you aren’t in a conscious mental state, but you still have the capacity for consciousness. And there may be entities that have never yet been in a conscious state, but who nevertheless have the capacity for consciousness. The state version of the phenomenal theory is quite radical; the capacity version is more modest. Although I have some sympathies for the state version, I’ll focus here on the capacity version. For the rest of the paper, whenever I talk about conscious subjects without qualification, I’ll mean entities with the capacity for consciousness.⁷

On some theories of consciousness, consciousness comes in degrees, meaning that some entities are “more conscious” than others.⁸ How should we think about the phenomenal theory if such a theory is true? It’s plausible that if consciousness comes in degrees, then any entity with a positive degree of consciousness is conscious simpliciter. Compare: mass comes in degrees, and any entity with a positive value has mass simpliciter. Therefore, the phenomenal theory entails that anything with a positive degree of consciousness is a welfare subject. However, it’s a further question whether degree of consciousness is itself an ethically significant property. You could accept that consciousness comes in degrees while denying that differences in degrees of consciousness ground differences in welfare subjecthood. That’s in fact the view I favor. But the answer to that question won’t make a difference to the rest of the arguments in this paper.

The phenomenal theory leaves open a range of other questions about the relationship between consciousness and welfare. Consider, for example, the question of whether consciousness is intrinsically valuable. To say that consciousness is intrinsically valuable is to say that consciousness is a welfare good. But while I think consciousness is what makes an entity a welfare subject, I’m skeptical that consciousness is a welfare good.⁹ Consider, as another example, experientialism, or the thesis that every welfare good is experiential. Those who favor experientialism will probably also favor the phenomenal theory, but I’ll later explain why the connection is less straightforward than you might initially think.

The Dialectical Situation

How popular is the phenomenal theory? The question is hard to answer. Many authors make passing remarks that express sympathy for the view, and many discussions seem to take the phenomenal theory as a starting point. Yet the theory has received little systematic development in the contemporary philosophical literature. Furthermore—as I’ll discuss later—the phenomenal theory has some consequences that many people will feel unsure about.

Nevertheless, many authors have endorsed nearby theses. Perhaps the most common is the claim that consciousness is necessary for welfare subjecthood.¹⁰ As Lin [2020] notes, this claim is “widely endorsed, even among theorists who reject” experientialist theories of welfare goods. There are also some authors who claim that consciousness is sufficient for welfare subjecthood.¹¹ Since the phenomenal theory says that to be a welfare subject just is to be conscious, it entails both those claims.

Most prior discussions of welfare subjects have assumed that claims about welfare subjects (such as those mentioned above) ought to be derived from a theory of welfare goods. Because of this, prior discussions of the phenomenal theory (or nearby theses) have typically focused on experientialist theories of welfare goods. However, I’ll argue in §2 that the relationship between theories of welfare goods and theories of welfare subjects is less straightforward than one might initially think. Because of this, my defense of the phenomenal theory won’t directly appeal to particular theories of welfare goods.

Personally, I favor treating the phenomenal theory as a basic building block in a theory of welfare, rather than as a theorem derived from more fundamental axioms. Any theory must take some claims as starting points, and I think it’s reasonable for a theory of welfare to treat the phenomenal theory as one of its cornerstones. But my arguments won’t turn on any particular view about how the phenomenal theory fits into the rest of our theory of welfare. And in what follows, I’ll make some points about why the phenomenal theory is plausible, whether it’s treated as a foundational thesis or as a claim derivable from more fundamental principles.

In what follows, I’ll outline some motivations for the phenomenal theory. I’ll start with motivations that appeal to intuitions, and then turn to some more theoretical motivations.

Intuitions

Many people have the intuition that consciousness is ethically significant. But there’s disagreement over how to best vindicate that intuition.¹² My answer is that consciousness is ethically significant because consciousness is what makes an entity a welfare subject. This answer renders consciousness ethically significant both because the question of which things are welfare subjects is itself a part of the subject-matter of ethics, and because of plausible connections between welfare and morality. More specifically, if we grant the common idea that welfare subjecthood suffices for moral status, then the phenomenal theory has implications not only for which entities can be better or worse off, but also for which entities matter morally.

You might object that the ethical significance of consciousness could instead be explained by taking consciousness to be either necessary or sufficient for having certain kinds of welfare goods. But these alternative explanations are problematic. If consciousness is sufficient for having some welfare goods, then it follows that every conscious entity possesses some welfare good. The most natural way to justify this claim is to take consciousness itself to be a welfare good.¹³ However, while the claim that consciousness is ethically significant is widely accepted, the claim that consciousness is a welfare good is controversial. While most think that certain kinds of conscious experiences are welfare goods, few think that consciousness itself is a welfare good. Furthermore, even those who deny that consciousness is a welfare good tend to agree that consciousness is ethically significant.¹⁴

On the other hand, the claim that consciousness is necessary for some welfare goods is uncontroversial, since only conscious entities can have pleasures and nearly everyone agrees that pleasure is a welfare good. But then consider: for all (or at least most) welfare goods, being made of atoms is a necessary condition for that welfare good to be instantiated. This means that being made of atoms is necessary for all (or most) welfare goods. Yet nobody thinks that being made of atoms is an ethically significant property. This means that the mere fact that consciousness is necessary for some (or even all) welfare goods is insufficient for explaining why consciousness is ethically significant.

You might object that we could instead capture the ethical significance of consciousness by appealing to the weaker claim that consciousness is necessary (or sufficient) for welfare subjecthood. But these claims also fail to account for the ethical significance of consciousness. Since being made of atoms is necessary for welfare subjecthood (but isn’t an ethically significant property), the claim that consciousness is necessary for welfare subjecthood isn’t enough to show that consciousness is ethically significant. Since being a qualitative duplicate of me is sufficient for welfare subjecthood (but isn’t an ethically significant property), the claim that consciousness is sufficient for welfare subjecthood isn’t enough to show that consciousness is ethically significant. The phenomenal theory offers a more direct explanation: consciousness is ethically significant because consciousness is what makes an entity a welfare subject.

Consider next another common ethical intuition: if you die, then you cease to be a welfare subject. The term ‘death’ means different things in different contexts, but the most common definitions entail that if you die, then you lose the capacity for consciousness. In fact, it’s common for philosophers to simply define ‘death’ as the permanent loss of the capacity for consciousness.¹⁵ Therefore, unless we understand ‘death’ in a non-standard way, the aforementioned claim will entail that if you die, then it’s both the case that (1) you lose the capacity for consciousness, and (2) you cease to be a welfare subject. It’s easy to then see how the phenomenal theory accounts for the connection between death and loss of welfare subjecthood.

You might object that the intuition about death could instead be explained by the idea that if you die then you cease to exist, and that if you cease to exist then you thereby cease to be a welfare subject. However, even those who deny that death entails non-existence can accept that death entails cessation of welfare subjecthood. Consider a philosopher who endorses a biological theory of personal identity, according to which you’re identical to your body. If you’re identical to your body and your body still exists after death, then you still exist after death. Yet it’s obvious that those who endorse a biological theory of personal identity aren’t forced to deny that death entails the cessation of welfare subjecthood. This means that the explanation provided by the phenomenal theory is more robust than the explanation appealing to loss of existence.

If the phenomenal theory is true (and if ethical inquiry is at least somewhat truth-tracking), then we should expect debate about whether Fs are conscious to yield debate about whether Fs are welfare subjects, and ethical debates about conscious entities to differ from ethical debates about non-conscious entities.¹⁶ Both these predictions are reflective of contemporary applied ethics. In support of the first prediction, consider our uncertainty about the ethical statuses of entities such as simple organisms, embryos, artificial intelligences, persistent vegetative state patients, and cerebral organoids. In support of the second prediction, consider how there’s abundant debate about the ethics of eating animals yet little debate about the ethics of eating plants.¹⁷

Theoretical Motivations

I’ve focused thus far on identifying how the phenomenal theory fits with some common ethical intuitions. I’ll focus now on some theoretical motivations.

Towards the beginning of the paper, I identified some of the theoretical roles that characterize the concept of welfare. These included the following: welfare is what we have in mind when we ask (1) whether an individual has a life worth living, (2) whether one individual is better off than another, and (3) how desirable it is to be in the position of a given subject. The phenomenal theory enables us to conserve these (and other) theoretical roles. The kinds of questions mentioned above seem sensible whenever we’re talking about conscious entities. By contrast, such questions seem inapplicable to non-conscious entities. Consider:

# Does your houseplant have a life worth living?

# Are you doing better or worse than McDonald’s?

# How desirable is it to be in the position of ChatGPT?

This is evidence that when we talk about the “welfare” of non-conscious entities such as plants, corporations, or AIs, we aren’t invoking the same full-blooded sense of ‘welfare’ that we have in mind when talking about the welfare of conscious entities.¹⁸

A metatheoretical desideratum for a theory of welfare subjects is to identify a unified ground of welfare subjecthood. Without this desideratum, it would be easy to construct a theory that generates intuitive predictions but that has a disjunctive structure (where what it is to be a welfare subject is to be either an F, or a G, or … a Z). But a disjunctive theory would be unsatisfying, at least if there’s no explanation of what F and G and so forth have in common that makes them all grounds of welfare subjecthood. Furthermore, it’s hard enough to find one property that can reasonably accommodate all the theoretical roles associated with welfare; it’s harder to find multiple properties.¹⁹

If we accept this metatheoretical desideratum, then we constrain the space of possible theories. It then becomes a substantive challenge to develop a theory of welfare subjects that satisfies this desideratum while still faring well with respect to other factors, such as extensional adequacy, fit with other components of our ethical theories, and (as I’ll discuss later) satisfactory answers to challenges. My arguments thus far (and that will follow) aim to illustrate how the phenomenal theory does well on all of these factors.

I’ll end with a thought that’s speculative, but that I take to capture a core motivation for the phenomenal theory. Both consciousness and welfare are subject-relative properties: consciousness concerns how things feel for a subject, whereas welfare concerns how good things are for a subject. A natural thought is that to be the kind of entity for which things can be going better or worse, one must be a subject in a more basic sense. I don’t know whether this idea can be developed in more systematic terms, but similar thoughts have been invoked by other philosophers. Sumner [1996: 43] says that a “welfare subject…must also be a subject in a more robust sense—the locus of a…unified mental life”; Kahane & Savulescu [2009: 13], when addressing the question of how it is that “certain states of affairs matter, not impersonally, but in relation to someone,” say that “possession of consciousness—of a subjective standpoint—might be a general condition for an entity’s having interests”; and Rosati [2009: 225] says that “we regard as welfare subjects…only those beings who…have a point of view” and that “we do not talk in terms of the welfare of a living thing unless there is a way things can be for it.” I think the common thread in these lines of thought captures a core motivation for the phenomenal theory.²⁰

2. Experientialism and Anti-Experientialism

Let’s say a welfare good is experiential just in case it’s identical to or partly constituted by a kind of conscious experience. I’ll now discuss how the phenomenal theory relates to the following views about welfare goods:²¹

experientialism

Every welfare good is experiential.

anti-experientialism

Some welfare goods aren’t experiential.

A number of recent discussions about welfare and consciousness have assumed that the plausibility of the phenomenal theory (or at least the view that consciousness is necessary for welfare subjecthood) turns on the plausibility of experientialism. Van der Deijl [2021] takes the phenomenal theory for granted and argues that it’s best explained by accepting that all differences in welfare are due to differences in experiences. Lin [2020] argues that the thesis that consciousness is necessary for welfare subjecthood is best explained by taking all welfare goods to have experiential components. Bradford [2022], by contrast, appeals to anti-experientialism to argue against the necessity claim of the phenomenal theory. And Kriegel [forthcoming], in discussing why zombies aren’t welfare subjects, says that “it’s very natural to derive one’s view on who the wellbeing subjects are from one’s view on what wellbeing consists in.”

Although these authors argue for different positions, a common thread is that they all assume that the phenomenal theory of welfare subjects stands or falls with experientialist theories of welfare goods. Personally, I’m sympathetic to experientialism. But I think the dialectical picture is more complex than what the authors above have assumed.

I’ll argue that the phenomenal theory is logically independent of both experientialism and anti-experientialism. That is, I’ll argue that (1) experientialism doesn’t automatically entail the phenomenal theory, and (2) anti-experientialism doesn’t automatically entail the negation of the phenomenal theory. I won’t weigh in on the plausibility of different theories of welfare goods—instead, the aim is to elucidate some of the complexities that arise when thinking about the relationship between the phenomenal theory of welfare subjects and various theories of welfare goods.

To be clear, I think there are systematic connections between theories of welfare subjects and theories of welfare goods. My point is merely that it’s not obvious that either theory is explanatorily prior. As an analogy, consider the relationship between desire and belief. Few think that desire is reducible to belief (or belief to desire), but nearly everyone thinks there are systematic connections between belief and desire. And just as it’s possible to theorize about desire in its own right (independent of its connections to belief), it’s also possible to theorize about welfare subjects in its own right (independent of its connections to welfare goods).

Experientialism

Some have thought that the truth of the phenomenal theory turns on the truth of experientialism. Behind this line of reasoning is the idea that theories of welfare goods are explanatorily prior to theories of welfare subjects. To defend that idea, one might appeal to the following asymmetry: while a theory of welfare subjects can be derived from a theory of welfare goods, a theory of welfare goods cannot be derived from a theory of welfare subjects. But, I’ll argue, this apparent asymmetry is illusory.

To derive a theory of welfare subjects from a theory of welfare goods, one needs an auxiliary premise connecting welfare subjects and welfare goods. A natural move is to appeal to the following metaphysical analysis: for x to be a welfare subject just is for x to be an entity that can accrue welfare goods (or bads). But once we permit appeals to auxiliary premises, it’s likewise possible to derive a theory of welfare goods from a theory of welfare subjects.

Suppose, for example, that you accept (1) some theory of which things are good simpliciter, and (2) the thesis that what it is for g to be a welfare good is for g to be a good simpliciter and possessed by a welfare subject. With these auxiliary premises, one could instead derive a theory of welfare goods from a theory of welfare subjects. Now, you could deny either of those auxiliary premises. But a proponent of the phenomenal theory could likewise deny the aforementioned analysis of welfare subjects.

I think it’s an open question how exactly theories of welfare subjects ought to relate to theories of welfare goods. But even if we were to take for granted that the concept of a welfare subject is best analyzed in terms of the concept of a welfare good, there would still be complications in deriving the phenomenal theory from experientialism.

Consider a conscious subject—Zero—who necessarily cannot accrue any welfare goods (or bads). Suppose, for example, that Zero’s only possible conscious experiences are experiences of gray (where those experiences are neither pleasant nor unpleasant), and that Zero entirely lacks capacities for desire, knowledge, or any other standard candidates for welfare goods. Since Zero is conscious, the phenomenal theory entails that Zero is a welfare subject. For the moment, let’s set aside questions about whether Zero is metaphysically possible and whether this consequence of the phenomenal theory is plausible—I’ll address those questions in §3. The relevant point is that even experientialism supplemented with the aforementioned analysis of welfare subjects doesn’t entail the phenomenal theory, since the latter counts Zero as a welfare subject while the former doesn’t.

This issue is anticipated by Van der Deijl [2021], who accepts experientialism and the phenomenal theory, and who contends that subjects like Zero count as welfare subjects because they possess welfare neutrals.²² Van der Deijl says little about the concept of a welfare neutral. But let’s assume that a welfare neutral is like a welfare good, except that instead of increasing one’s welfare level it leaves one’s welfare level the same: for example, anhedonic experiences are candidates for welfare neutrals. Is adding a category of welfare neutrals a credible move in developing a theory of welfare? While I’m sympathetic to the idea, it’s not obvious that taking the notion of a welfare neutral as basic is ultimately better than taking the notion of a welfare subject as basic. And strictly speaking, this approach entails that the analysis of welfare subjects I articulated earlier is false: it requires instead adopting the more speculative thesis that what it is to be a welfare subject is to be an entity that can accrue welfare goods or bads or neutrals.

Here’s another way of illustrating why it’s unobvious how to understand the explanatory relationship between the phenomenal theory and experientialism. Let pluralistic experientialism be any experientialist theory that posits more than one basic welfare good. If you favor pluralistic experientialism, then it’s natural to ask why every welfare good is possessable only by conscious subjects. Let impure experientialism be any experientialist theory that takes some welfare goods to have non-experiential components. If you favor impure experientialism, then it’s natural to ask why every welfare good has an experiential component. Now, obviously explanations must end somewhere: any first-order theory must take some claims as basic. But if it’s fair to ask for an explanation of the phenomenal theory, then it seems likewise fair to ask for an explanation of experientialism.

Let me offer a speculative hypothesis, if only to exemplify why I think the philosophical space is underexplored. Consider the following principle: in order for g to count as a welfare good for a welfare subject x, g must affect whatever property makes x a welfare subject. This principle isn’t trivial, since there are plenty of things that can affect an entity without affecting the property that makes that entity a welfare subject: for example, consider the metabolic processes occurring in your liver. Moreover, different theories of welfare subjects generate different predictions given this principle: for example, a desire-based theory of welfare subjects may predict that in order for g to count as a welfare good, g must affect your desires (or their satisfaction).

If we accept this condition, and if we also suppose that what it is to be a welfare subject is to be a conscious subject, then we acquire an answer to the explanatory unity question mentioned above. The answer is that experientialism is true because the existence of non-experiential welfare goods would violate the principle outlined above. The reader may decide for themselves whether the principle is plausible, but I think it at least merits consideration. Yet, notice that this answer reverses the order of explanation: the phenomenal theory is used to explain experientialism, rather than the other way around!

Anti-Experientialism

I defined anti-experientialism as the view that some welfare goods are non-experiential. It may be tempting to assume that if anti-experientialism is true, then it automatically follows that the phenomenal theory is false. I’ll argue that anti-experientialism is compatible with the phenomenal theory.

As a preliminary point, it’s worth noting that anti-experientialism itself is more controversial than some might initially think. Obviously, those sympathetic to hedonism (or other views that take only experiences to be welfare goods) will reject anti-experientialism. For those sympathetic to desire-satisfactionism or an objective list theory, the question of anti-experientialism turns on questions about whether desire-satisfaction and (say) knowledge and friendship can be attained without consciousness. The answers to these questions aren’t obvious. A number of recent works challenge the idea that desire / knowledge / friendship is independent of consciousness.²³ And even if we were to agree that non-conscious entities can have desires / knowledge / friendship in some sense, we would have to ensure that those senses of ‘desire-satisfaction’ / ‘knowledge’ / ‘friendship’ are the ones that genuinely denote welfare goods.

But suppose that anti-experientialism is true. Since anti-experientialism is a claim about welfare goods, an argument from anti-experientialism to the negation of the phenomenal theory needs an auxiliary premise that connects welfare subjects to welfare goods. As before, you could argue that what it is to be a welfare subject is to be an entity that can accrue welfare goods. But once we recognize the role of this auxiliary premise, it’s evident that anti-experientialism itself is compatible with the phenomenal theory. That is, you could accept both anti-experientialism and the phenomenal theory, so long as you hold that some entities that can accrue welfare goods aren’t welfare subjects.²⁴

To illustrate, consider someone who accepts (a) desire-satisfactionism (the only basic welfare good is satisfied desires), (b) a permissive, functionalist theory of desire (where even non-conscious entities, such as thermostats and bacteria, can have desires), and (c) the phenomenal theory of welfare subjects. On such a view, there are many entities that can have desires (including non-conscious entities), but it’s only conscious entities that benefit from the satisfaction of those desires. This is a consistent theory. You might disfavor this theory (as I do!), but there’s nothing incoherent about the view. Furthermore, each premise of the theory can be independently motivated.

Is there a general reason for thinking that theories with this kind of structure are untenable? To accept both anti-experientialism and the phenomenal theory, you would have to endorse the following picture: only conscious entities have welfare levels, and welfare levels are determined by welfare goods, yet even some non-conscious entities can have welfare goods. Those who think that welfare subjects are to be analyzed in terms of welfare goods might disfavor this picture, but it’s not obvious what’s problematic about it.²⁵ Consider an analogy: only entities bound by a gravitational force have weight, weight is determined by mass, yet even entities that aren’t bound by a gravitational force have mass. This case is structurally analogous to the situation described above, yet there’s obviously no pressure to reject the claim that only entities bound by a gravitational force have weight.

To justify an incompatibility between anti-experientialism and the phenomenal theory, one would have to identify a relevant disanalogy between the weight/mass case and the welfare levels/welfare goods case. One response is to say that weight is determined not only by an object’s mass, but also by whether it’s bound by a gravitational force. However, we could likewise say that an entity’s welfare level is determined not only by its welfare goods, but also by whether it’s conscious. Another response is to say that weight is a relational property whereas welfare isn’t. However, we need to be careful about the relevant sense of ‘relational’. Weight is relational in that an object’s weight is determined extrinsically, but weight is also non-relational in that weight ascriptions are monadic (individual objects have weights, rather than ordered pairs of objects and gravitational fields). Similarly, welfare may be relational in the sense of being determined extrinsically (as anyone who accepts desire-satisfaction or knowledge as welfare goods would think), but welfare is also non-relational in that welfare ascriptions are monadic (individual welfare subjects have welfare levels, rather than ordered pairs of welfare subjects and sets of welfare goods).

These considerations illustrate why the relationship between the phenomenal theory and experientialism/anti-experientialism is more complex than one might initially think. Though we have a refined understanding of the theoretical space for first-order issues about welfare goods, we have a crude understanding of the theoretical space for metatheoretical issues about how a theory of welfare goods ought to relate to a theory of welfare subjects. But making progress on the first-order questions doesn’t require settling the metatheoretical issues. Just as we can advance first-order inquiry into ethics, metaphysics, or science without settling everything in metaethics, metametaphysics, or the philosophy of science, so too we can advance first-order inquiry into welfare subjects without settling the metatheory of welfare.

The Zero Objection

Earlier, I described a conscious subject—Zero—who necessarily cannot accrue any welfare goods (or bads). If the phenomenal theory is true, then Zero is a welfare subject. The Zero Objection claims that Zero isn’t a welfare subject.

It’s not obvious that Zero is metaphysically possible. The question isn’t merely whether it’s possible for a creature to have a cognitive architecture that precludes any pleasant or unpleasant experiences. Rather, the question turns on controversial issues about the metaphysics of identity across possible worlds. Even if it’s psychologically impossible for an entity to have any pleasant or unpleasant experiences (or any other mental states that generate welfare goods or bads), it may still be metaphysically possible for that entity to have such experiences. If Zero is metaphysically impossible, then the Zero Objection doesn’t even get off the ground.²⁷

Suppose we set aside these metaphysical concerns. To figure out how to think about Zero, we need to consider the difference between having welfare level zero and lacking a welfare level altogether. As an analogy, consider electric charge, which (like welfare) has positive, negative, and zero values. Most physical objects can have either positive or negative charge, but some—such as photons—necessarily have zero charge.²⁸ But suppose we’re developing an account of which kinds of entities have charge values, and imagine someone who reasons from the premise that photons cannot have either positive or negative charge to the conclusion that photons aren’t the kinds of things that have charge values. That inference would be fallacious: photons have zero charge (rather than no charge value at all). By contrast, it would be a category mistake to say that a gravitational field or the color red or the number three have zero charge—instead, their charge value is undefined. Similar points can be made using other quantities that have zero values, such as mass (gluons necessarily have zero mass, but the mass of love is undefined), height-above-sea-level (the surface of the sea is zero meters above sea level, but the height-above-sea-level of the solar system is undefined), and temperature (a universe with no kinetic energy is zero kelvin, but the temperature of time is undefined).

You might object that there are some quantities, such as number of children, where to have a zero value just is to lack a positive (or negative) value. Consider: for any x, if there doesn’t exist a y such that x bears the parent relation to y, then x has zero children. However, this observation merely indicates that we should distinguish between positive properties (which entail the possession of some feature) and negative properties (which entail the absence of some feature). As examples, having zero charge, zero mass, or zero temperature are arguably positive properties, whereas having zero children, zero prime factors, or zero moons are credible candidates for being negative properties. Whenever having a zero value is a positive property, we can make sense of the difference between an entity having value zero along that quantity versus an entity lacking a value along that quantity altogether. This enables us to respect the difference between having zero charge vs. lacking a charge value while also allowing that anything that doesn’t have a positive number of children thereby has zero children.²⁹

When we consider welfare, we can easily make sense of the difference between having welfare level zero and lacking a welfare level altogether. It’s only for entities that have welfare levels for which we can ask how well the entity is doing, whether it has a life worth living, whether it’s better off than another entity, how desirable it is to be in the position of that entity, and so forth. When we calculate the average welfare in a world, we ought to factor in entities with welfare level zero (while excluding entities that lack welfare levels).³⁰ And in decision-theoretic tasks, it’s straightforward how to assess scenarios where one ends up always having welfare level zero but unobvious how to assess scenarios where one lacks a welfare level altogether.³¹ These asymmetries indicate that welfare level zero is a positive property (rather than a negative property).

The question then is whether Zero has welfare level zero or lacks a welfare level altogether. It’s reasonable to think that Zero has a life that is neither good nor bad, that Zero’s life is on the threshold between a life worth living and a life not worth living, and that Zero’s life is worse than the life of someone in paradise but better than the life of someone in hell. Therefore, it’s reasonable to think that Zero has welfare level zero. Since anything with a welfare level is a welfare subject, it would then follow that Zero is a welfare subject. Just as photons are special cases of charge subjects and the number zero is a special case of an integer, Zero may be a special case of a welfare subject.

Are there countervailing reasons against counting Zero as a welfare subject? A first objection is that Zero isn’t a welfare subject because it cannot be better or worse off.³² But then consider Happy, who is just like Zero except that its only possible experience is a state of pleasure. Though Happy cannot be better or worse off, it’s still clearly a welfare subject: if we wish to determine the average level of welfare in the world, then Happy’s welfare should factor into that calculation. A second objection is that the fact that Zero necessarily has welfare level zero is itself a reason for denying that Zero is a welfare subject. But then consider Balanced, who is also just like Zero except that its only possible experience is a state involving both pleasure and pain (such that the goodness of the pleasure exactly balances out the badness of the pain). Though Balanced necessarily has welfare level zero, it’s likewise still a welfare subject.

If you deny that Zero is a welfare subject, then you must accept a discontinuity in the kinds of entities that are welfare subjects. Imagine a spectrum of cases, where on the left side are subjects like Happy, in the middle is Zero, on the right side are subjects like Unhappy (who is always in a state of pain), and where the ordinal position of a subject corresponds to its valence value (so that subjects further to the right have more unpleasant experiences). All subjects to the left of Zero are welfare subjects and all subjects to the right of Zero are welfare subjects. Of course, genuine discontinuities sometimes occur, but I see no compelling reason to think that Zero is such a case.

I’ll end with a dialectical point. David Lewis once said that there are some philosophical disputes where “spoils go to the victor,” meaning that “we can reasonably accept as true whatever answer comes from the analysis that does best on the clearer cases.”³³ I think cases like Zero fall within this category. It’s neither obviously true nor obviously false that Zero is a welfare subject. But the view that Zero is a welfare subject is reasonable, and the fact that the phenomenal theory classifies Zero as a welfare subject isn’t a reason for rejecting the theory.

The Plant Objection

The Plant Objection says that plants are welfare subjects (but not conscious).³⁴ Although I’ll focus on plants in this section, most of my arguments will apply also to other kinds of non-conscious entities that might be regarded as welfare subjects.

Why think that plants are welfare subjects? The most common justifications appeal to the fact that plants can flourish (e.g., by receiving sunlight) and flounder (e.g., by being uprooted) and to the idea that plants have interests (such as growing and spreading their seeds). The view that plants are welfare subjects also seems supported by ordinary language: it’s natural, for example, to talk about what’s good or bad for a plant. However, taking these features to be sufficient for welfare subjecthood will lead to a proliferation of which entities count as welfare subjects.

There are many entities—for example, corporations,³⁵ livers, and beehives—that satisfy the criteria above yet are plausibly not welfare subjects. It’s in the interest of a corporation to increase revenue and attract investors; a beehive flourishes by preserving the structural integrity of the hive and maintaining a healthy population of bees; and your liver is doing well when it’s healthy and free of toxins and badly when you consume excessive amounts of alcohol. It’s tempting to object that welfare talk about corporations, beehives, and livers is metaphorical. But anyone who favors that response must then explain why welfare talk about plants isn’t also metaphorical.

Is it possible to identify a criterion that demarcates plants and animals (including humans) from corporations, livers, and beehives? The most obvious criterion is that plants and animals are organisms whereas corporations, livers, and beehives aren’t. But imagine a conscious artificial intelligence that’s capable of feeling pleasure and pain but lacks the reproductive and metabolic capacities characteristic of organisms. Or consider bacteria, which are organisms but plausibly not welfare subjects. These cases indicate that being an organism is neither necessary nor sufficient for being a welfare subject. To develop a viable account of welfare subjecthood, the objector must then find another criterion that includes plants, animals, and sophisticated artificial intelligences yet excludes corporations, livers, beehives, and bacteria. But it’s not clear that there is a principled and unified criterion that can accommodate all such cases.

A more radical response is to endorse an extremely permissive view of welfare subjects. Consider, for example, the view that any entity that we talk about using welfare language thereby counts as a welfare subject. This sort of view risks yielding a disunified theory of welfare subjects: it’s not obvious that there’s any natural kind that applies to all and only the entities that we’re inclined to talk about using welfare language. Furthermore, there seem to be important differences between the sense of ‘welfare’ at play when we talk about plants, corporations, livers, beehives, and bacteria vs. the sense of ‘welfare’ at play when we talk about conscious entities.³⁶ To elicit the asymmetries, consider how it’s unclear whether we can meaningfully ask whether a corporation’s welfare is above the threshold for a life worth living, or how it’s plausible that animals can be harmed in ways that are ethically distinct from any ways in which plants can be harmed. Hence, this permissive view requires giving up on some of the core theoretical roles associated with the notion of a welfare subject.

Finally, it’s worth noting that those who have argued that plants are welfare subjects are often motivated by the intuition that plants can be objects of moral concern. Although one way to justify that idea is to contend that plants are welfare subjects, it’s also possible to think that plants can be objects of moral concern while endorsing the phenomenal theory. Consider, for example, someone who thinks that (1) plants are intrinsically valuable, and (2) it’s morally bad (other things being equal) to destroy intrinsically valuable things. On such a view, the justification for taking plants to be objects of moral concern doesn’t necessitate that plants are welfare subjects. Given this, I suspect that many philosophers who care about the normative status of plants have intuitions that are ultimately compatible with the phenomenal theory.

The Death Objection

The Death Objection claims that you can be made better or worse off even after death. Suppose, for example, that I slander you after you die, rendering your reputation unjustly damaged. On some theories of welfare, I’ve thereby made you worse off, even though you’re dead.³⁷ Since only welfare subjects can be made worse off, it seems to follow that consciousness isn’t necessary for welfare subjecthood.

To evaluate this objection, we need to first ask whether or not you continue to exist after death. Suppose that you cease to exist after death. Then we can ask: which entity is made worse off? Since we’re stipulating that (a) it’s you whose welfare is affected, (b) the harm occurs after your death, and (c) you cease to exist after death, there’s only one plausible answer: your past self (before death). But that past self was conscious, so this view is compatible with the phenomenal theory.

Is it possible to deny that existence is a requirement for being a welfare subject? Suppose you think that we’re morally obligated to mitigate the effects of climate change because of how it will affect the welfare levels of future people that don’t yet exist. Then it may seem that existence isn’t a requirement for being a welfare subject, since we’re ascribing welfare levels to people that don’t yet exist. However, the conclusion that we have moral obligations to future people is justified by the premise that the future people will be welfare subjects (rather than the premise that they are now welfare subjects). But the claim that the future people will be welfare subjects is compatible with the principle that existence is a requirement for welfare subjecthood. Future people aren’t yet welfare subjects (since they don’t yet exist), but they will be welfare subjects (since they will be conscious).³⁸

The other version of the Death Objection holds that you continue to exist after death. But it’s hard to know how to make sense of this if we define death as the permanent cessation of consciousness. If you continue to exist after death, then in what form do you exist? There aren’t any good candidates amongst physical entities, since your body may cease to exist after death (as in cases of cremation) and since there’s nothing special about the set of atoms that constitute your body (since the membership of that set is constantly in flux). Other wilder ideas are that you persist as a non-conscious mental entity or as an abstract object, but these options are hard to take seriously without further argument. Since no other options seem credible, the prospects for this version of the objection look bleak.

There’s an inversion of the Death Objection that we might call the ‘Life Objection’. The Life Objection says that you can be made better or worse off even before you’re ever conscious. If some embryos have the capacity for consciousness, then those embryos will count as welfare subjects under the phenomenal theory. But consider someone who thinks that embryos lack the capacity for consciousness but are nevertheless welfare subjects. This view is incompatible with the phenomenal theory. But note that an embryo may still become a welfare subject in the future even if it’s not a welfare subject now, and that denying that embryos are welfare subjects still leaves open questions about their moral status. I suspect that these observations will be enough to satisfy most who are tempted by the Life Objection.

Conclusion

I’ve argued that the phenomenal theory accounts for intuitions about the ethical significance of consciousness, explains why death entails loss of welfare subjecthood, makes correct predictions about asymmetries between various ethical debates, conserves the theoretical roles associated with welfare, identifies a unified ground of welfare subjecthood, and captures the idea that welfare subjects must have subjective points of view. I’ve also argued that the phenomenal theory is logically independent from both experientialist and anti-experientialist theories of welfare goods, and I’ve made a case for taking theories of welfare subjects to be explanatorily on a par with theories of welfare goods. Finally, I’ve defended the phenomenal theory from objections concerning Zero, plants, and death.

Along the way, I’ve addressed some more general issues about the relationship between theories of welfare subjects and theories of welfare goods. Philosophical discussions of welfare are dominated by debates about welfare goods, and theories of welfare goods are oftentimes regarded as theories of welfare simpliciter.³⁹ But without a story about welfare subjects, a theory of welfare is incomplete. I’ve argued that we should be careful about drawing conclusions about welfare subjects from premises solely about welfare goods, and I’ve given reasons to be cautious about analyzing welfare subjects in terms of welfare goods. Moreover, even those who accept that welfare goods are prior to welfare subjects must still explain what their preferred analysis of welfare subjects looks like and how their theory works in cases involving welfare level zero. These metatheoretical considerations may be fruitful even for those who favor different first-order views.

I started this paper by observing that while many philosophers think that consciousness is ethically significant, there’s no consensus on what that ethical significance amounts to. This paper has defended an answer to that question: consciousness is ethically significant because consciousness is what makes an entity a welfare subject. Maybe there are also other ways in which consciousness is ethically significant. But—in my view—the connection between consciousness and welfare subjecthood is an essential ingredient in an ethics of consciousness.⁴⁰