Subjective Facts about Consciousness

1. Introduction

What shape must reality take for there to be a place for consciousness in it? This question presupposes that conscious mental states are hard to understand, so hard indeed, that we should be open to revising our general metaphysical picture of reality to come to a better understanding. This paper follows this more radical line of inquiry. The aim is to unpack the thought that for the world to harbor consciousness, reality must include subjects that are metaphysical standpoints and be such that some of the facts obtain only from the standpoint of certain subjects and not others.

The main thesis—that reality is constituted by subjective facts and that subjects are metaphysical standpoints—can sound hollow and unilluminating or, worse, seem the mark of confused thinking. The notions of ‘subjectivity’ and ‘having a perspective’ are used in so many ways. That there is something subjective about consciousness is widely acknowledged but that doesn’t mean that we are clear on what is required for something to be subjective.

The starting assumption of this paper is Nagel’s observation that consciousness, in particular the phenomenal appearances that accompany conscious mental states, are only there from a subjective viewpoint and are somehow involved in the subject’s being a viewpoint in the first place (1979: 174, 212–13). This is not simply the idea that there are certain special phenomenal properties, qualia, which are essential to conscious mental states. Nagel’s insight suggests more than that, namely that subjects are somehow standpoints relative to which different phenomenal facts obtain, standpoints that we can moreover ‘adopt’ in our conception of things.

Much progress has been made in the metaphysics of perspectival phenomena, often driven by work in the philosophy of time or modality, which some have used as a basis for illuminating work on the subjective character of mental states (e.g., in Fine 2005; Hare 2009; Merlo 2016; and views that make use of ‘many worlds’, such as List in press; Honderich 2014). These and other developments give us the resources to conceptually stabilize and further develop views of the metaphysical relativity involved in subjective facts, views that might have seemed unclear before.

I should stress that, although I rely on some central claims of Nagel’s influential early work, I do not intend to interpret his writings or do justice to all his views. The project is one of developing a general metaphysics of subjective phenomenal facts based on a few central insights that, despite their widely recognized appeal, I think we haven’t fully come to terms with yet.

The paper is exclusively concerned with phenomenal consciousness, and whenever I speak of ‘mental states’, this abbreviates ‘conscious mental states’.

2. Being Conscious Makes One a Standpoint

We start from a simple and plausible idea: there is something essentially subjective about conscious mental states. There is a strong tendency to understand the subjectivity of the mental in the first instance as an epistemic phenomenon. One might understand the subjectivity of the mental as the fact that, somehow, having an experience oneself is epistemically different from observing someone else having an experience, perhaps because experiences can only be known through introspection and are therefore epistemically private. There is much to say about these epistemic claims and about what they imply (or don’t) about the metaphysics of the mental (see, e.g., Lycan 1990; Crane 2003). These ideas might ultimately be part of the picture that we’re after, but it seems to me that they don’t get to the heart of the matter on their own and that, at the very least, there is a relatively neglected yet important alternative to explore.

The subjectivity of consciousness doesn’t seem in the first instance an epistemic phenomenon. Nagel observed that ‘an organism has conscious mental states if and only if there is something that it is like to be that organism—something it is like for the organism’ (1979: 166) and that ‘facts about what it is like to be a human being, or a bat, or a Martian, […] appear to be facts that embody a particular point of view’ (1979: 166). I propose that there is something it is like to be an organism if and only if the organism in question is a standpoint, where being a standpoint is not simply an epistemic matter of being in the position to know certain matters that others cannot.

Then what does one’s being a standpoint consist in? A subject is a standpoint when the subject is something relative to which certain facts can be said to obtain. Such facts obtain—are bits of reality—from the standpoint of the subject. A subject of consciousness is a standpoint in the very sense that there are matters that obtain relative to it.

The world is a certain way to you. It’s not simply experienced in a certain way by you, although that too of course. These two matters, (i) something obtaining from the perspective of a subject, and (ii) something being experienced by a subject, are easily conflated but shouldn’t be. Besides being experienced in particular ways by you, the world is itself constituted by certain facts that are only real relative to you, or from your perspective, facts that are not constitutive of the world relative to someone else. This can seem a suspicious or confused thought, but we should be open to the possibility that this is so because certain background assumptions make it hard to parse it in any other way than as the trivial claim that the world is thought about, or experienced, differently by different subjects. To make sense of the intended claim that the world is a certain way to you, that certain facts only obtain relative to you, we need a non-experiential sense of ‘relativity’ and of ‘relative facts’.

To fix the intended sense of relativity, it helps to think about temporal matters (as also noted by Fine 2005; Hare 2009; Merlo 2016). From the standpoint of different moments in time, different facts obtain. A sunflower is straight ‘relative to’ or ‘at’ one moment in time and it’s bent ‘relative to’ or ‘at’ a later moment in time. The world differs from the standpoint of different moments in time in the sense that different facts obtain from the standpoint of different moments in time. These facts, when they obtain, obtain simpliciter besides it being a further fact that they obtain at the time they do. When the sunflower in my garden is straight, it’s straight, period, even though it’s furthermore the case that the sunflower is straight at a certain time. The fact that it is straight and the fact that it is straight at moment t are plausibly assumed to be distinct facts. We can explore the thesis that this kind of metaphysical relativity applies to subjects as well: the world itself differs from the standpoint of different subjects but, to the extent that something is a certain way relative to a subject, it’s also just that way. I will discuss why we should admit such facts down below; we first need some time to understand and conceptually parse the suggestion.

It will be helpful to introduce some formalism. Let ‘@_(…)’ be an operator-forming device that combines with a referring expression to give a sentential operator, such as ‘@s(…)’. Such hybrid operators are used in hybrid tense logic (due to Prior 1967: 88–92, 187–97; 1968/2003). We can read ‘@s(…)’ as: ‘at s, …’ or ‘relative to s, …’ or ‘from the perspective of s, …’ depending on what sounds most natural. For our purposes, the referring expressions that slot into the first position of the operator will exclusively refer to subjects. So, the sentence ‘Lucy’s experience has phenomenal character from Lucy’s own perspective’ is formalized as: @Lucy(Lucy’s experience has phenomenal character).

With this formalism in place, the assumption that a subject is a metaphysical standpoint whenever it’s something relative to which matters obtain becomes (using sentential quantification):

Necessarily, any subject s is a standpoint if and only if for some p, @s(p).

If sticks and stones are not standpoints, then facts don’t obtain relative to sticks and stones.¹

Note that, although I informally speak about facts, this is loose talk that can be translated away in the official idiom: the claim that ‘the fact that p obtains relative to a subject s’, which seems to attribute a relation to a fact and a subject (and hence is ontologically committed to both) can always be translated into the claim that @s(p), which is ontologically committed only to subjects and not to facts.

Note also that accepting this notion of matters obtaining relative to subjects is not the same as accepting that different facts obtain relative to different subjects. It only states a connection between being a standpoint and being something relative to which matters can be said to obtain. So far, this is compatible with all facts obtaining relative to all subjects equally. But capturing variations in how things are across subjects is arguably the point of introducing a notion of metaphysical relativity. Not only do we assume that matters obtain relative to subjects, but also that different matters obtain relative to different subjects. There is subjective variegation:

Subjective variegation: some p and q are such that, for some subjects s₁ and s₂, @s₁(p) but not @s₁(q), and @s₂(q) but not @ s₂(p).

This is a substantive metaphysical assumption. It cannot be stressed enough that relativity is here distinguished from mere ‘experiential relativity’. Subjective variegation is not just the assumption that different subjects experience different things or have experiential access to different matters. Different subjects are standpoints on different facts in a more robustly metaphysical sense: certain facts only obtain, ‘out there’, relative to certain subjects and not relative to others. Yes, different subjects will experience the world differently but, on top of this, the world itself differs between subjects. Let us call any fact that obtains relative to some but not all subjects a ‘subjective fact’ (following Merlo 2016: §2).²

How does this bear on our conception of conscious mental states? Any mental state has some subjective character in the sense that one cannot be in a conscious mental state without this being reflected in one’s own perspective on the world (Nagel 1979: 166). Your ‘perspective on the world’, as used here, can be understood as loose talk for those facts that obtain from your standpoint (i.e., relative to you), and includes facts that obtain relative to any subject as well as subjective facts that obtain from your perspective only. There is a way that the world is like from one’s perspective when one is in pain and the world itself is not like that when one is not in pain. In general:

Perspectival consciousness: For any conscious mental state m, there is some p such that, necessarily, any subject s is in mental state m, if and only if, @s(p).

‘There is something it is like for a subject to be in pain’ is understood in terms of there being something that obtains from the perspective of a subject whenever the subject is in pain. The right-to-left direction is plausible here: if one’s perspective (i.e., the facts that obtain relative to one) includes the phenomenal facts that accompany pain, for example, then this suffices for one to be in pain (compare Kripke 1980: 154). Given this assumption, there must be subjective facts, which obtain at some but not all subjects, whenever there are two subjects that are in different conscious mental states. Differences in mental states imply differences in subjective facts (and hence imply subjective variegation).

So far, we have a metaphysically necessary biconditional. It’s plausibly part of the nature of a mental state that it affects the subject’s perspective in the way that it does. We can therefore also reformulate the claim as a claim about the nature of mental states: for any mental state m, there is some p, such that part of what it is for any subject s to be in mental state m is for it to be the case that @s(p).³ For a subject to be in a mental state is, at least partly, for it to be the case that some fact obtains relative to that subject. If it’s part of what it is for any subject s to be in conscious mental state m that @s(p) for some given p, then we can say that ‘p’ expresses m’s ‘phenomenal character’ and that the relevant fact that p is a ‘phenomenal fact’.⁴ Note that, given how we understand phenomenal facts, any phenomenal fact is a subjective fact. Phenomenal facts are subjective facts that necessarily accompany conscious mental states. The framework allows that there are subjective facts that are not also phenomenal facts (i.e., that do not only obtain from a subject’s perspective when that subject is in some mental state). To illustrate, one might think that facts about tastiness are subjective but not phenomenal facts: licorice can be tasty from my perspective even when I’m not in any particular mental state.

3. Adopting the Standpoint of a Subject

As Nagel noted, one cannot fully understand the nature of some mental state without representing ‘what it is like’ to be in that mental state, that is, without adopting the standpoint of someone in that mental state (1979: 166, 172). But how should we understand this notion of adopting the standpoint of a subject who is undergoing some mental state?

There may be more than one sense in which we talk of adopting standpoints. There is at least one simple and deflationary notion of ‘standpoint adoption’ that is theoretically valuable, or so I will try to show. The relevant sense is one where we adopt the standpoint of a particular subject just when we represent that which holds relative to the subject. To take an example, assume that licorice’s being tasty is an instance of a (subjective, yet non-phenomenal) fact that obtains relative to some but not all subjects (of course a controversial assumption—but here it’s just for illustration, see Merlo and Pravato 2020 for a defense). Licorice is tasty in the world as it is from your standpoint (i.e., @you(licorice is tasty)), but it isn’t in the world as it is relative to me. The thought is that, when I nevertheless represent that licorice is tasty, I thereby represent what things are like from your standpoint and can, in that narrow sense, be said to adopt your standpoint in thinking this.

Representing matters from other standpoints has a striking combination of features (the following is also discussed in Lipman in press). First, it is descriptive. By ‘descriptive thought’ I mean thought about how something is, attributing some property to something. Representing matters from other standpoints is not practical thought, nor is it a type of hypothetical thought about how things could be (the standpoints are actually there). It is, I assume, a type of descriptive thought aimed at reflecting what things are genuinely and really like from this other standpoint.⁵ As such, occurrences of such representational states can be true or false.

Normally, descriptive representations feed into deliberation, belief revision, and action. We act on what we believe and revise our overall system of beliefs under pressure of new evidence. In contrast, representations of how things are from other standpoints play a more insulated role in one’s overall doxastic state, they do not simply enter the ‘belief box’, as it were, but have a cognitive box of their own. It resembles suppositions in this respect: they too are not simply thrown into the mix of one’s overall doxastic state. But suppositions are not entertained as being genuinely descriptive of how things are, whereas representations of how things are from other standpoints are taken as reflecting how things are (namely, of how things are from that other standpoint)—or so I assume. There is a type of descriptive mental representation, which is like belief in being a species of theoretical thought purporting to capture how things are but unlike belief in having a more insulated position in our overall mental architecture. This is a substantial assumption of the proposed framework, to be judged on its fruits. Of course, it can be overstated: the insulation from evidence will not be an absolute, all or nothing affair, there are likely to be indirect constraints, for example, if there is evidence that there is no subject whatsoever relative to which p, this is evidence that we cannot engage in an immersive representation that p which is true.

Another possible point of contrast between ordinary belief and immersive representation concerns the extent to which the adoption of other standpoints is more intentional than ordinary belief. It’s often thought that occurrent explicit beliefs are not under much intentional control. If you believe that you are reading this, you cannot choose not to believe this. In contrast, it seems that the adoption of a standpoint is more naturally described as an intentional act. This may also fit its relative insulation from one’s further beliefs and evidence.

To sum up the qualifications so far, I assume that there is a type of doxastic mental state, that one can intentionally take up, that is descriptive and consists in representing what the world is like, is therefore evaluable for truth and falsity, but which doesn’t directly interact with one’s overall beliefs, evidence and actions. Let us call the postulated mental attitude ‘immersive representation’. One may rightfully want examples of this assumed representational state but, somewhat unhelpfully, it seems that the most natural examples are precisely cases that are our focus, namely cases of putting yourself in the shoes of someone else: one can willfully do this, it involves representational states (or acts) that can be evaluated for truth, but one won’t readily act on them as if there are ordinary beliefs. They are not imaginations or suppositions. When I immersively represent some experience of someone else I don’t undergo the same (kind of) experience as the subject in whose shoes I am putting myself. If asked what immersive representational states are, one answer is that they are the kind of mental states involved in one’s representation of what things are like from someone else’s perspective, the perspective of another moment in time, and so on. The label of ‘immersive representation’ is therefore appropriate.

One further qualification is needed, but this is just for our purposes here. Occurrent doxastic mental states, such as occurrent beliefs, are naturally taken to be thin in content compared to what one implicitly believes about the world. One doesn’t (at least, not typically) judge or think a total picture of the world. One thinks instead some specific matter, such as that it drizzles outside, that the rose smells nice, and so on. The same applies to immersive representations. Even if the content is comparably thin, one’s representations may well have some content that corresponds with how things are from some standpoint but also some content that doesn’t correspond to how things are from that standpoint. In such cases one doesn’t seem to succeed in adopting the relevant standpoint. Adoption should be determined by the entire content of an immersive representation, not parts of the content. Let us stipulate that when we say that a subject immersively represents that p, then ‘p’ specifies the entire content of what is represented in the relevant intuitive sense.

Given these qualifications, we have in general:

Necessarily, a subject s adopts standpoint x if and only if for some p, @x(p) and s immersively represents that p.

We take up or immerse ourselves in the perspective of a subject just when we represent part of that which holds relative to the subject and nothing that fails to hold relative to the subject.

Note that one’s representation can describe what things are like to many standpoints at once and hence be from multiple standpoints at once. I can adopt the standpoints of anyone who meets a certain condition, namely of being standpoints relative to which the represented matter obtains. There is such a thing as adopting the standpoint of those to whom licorice is tasty for example. This is one of various ways in which standpoints differ from ‘contexts’ as they are used in semantics: a given sentence token typically has at most one context (with the unique base context defined by local conditions that pertain to a token utterance, such as who makes it and when).

Note also that one may not always know what standpoints one is adopting, indeed, one can be mistaken about what standpoint one takes oneself to have adopted (for example, because one has mistaken beliefs about what is the case at a certain standpoint). This means that intending to adopt a certain standpoint should be distinguished from the actual adoption of a standpoint. It’s natural to assume that intending to adopt a certain standpoint is typically focused on a single target standpoint and yet, when successful, it can have the effect of adopting multiple standpoints. I can be mistaken in what standpoint I take myself to adopt, for example, if, unbeknownst to me, licorice is tasty from your standpoint, and I represent that licorice is tasty, then I adopt (amongst others) your standpoint without knowing that I’m doing so and without intending to do so.

There is undoubtedly more to be said about the notion but let us turn to its application to consciousness.⁶ With this proposal for understanding what it is to adopt standpoints, return to Nagel’s claim that one cannot fully understand the nature of some mental state without representing ‘what it is like’ to be in that mental state (1979: 166). We can now understand this to mean that one cannot fully understand the nature of some mental state without adopting the standpoint of anyone in that mental state. When I represent that @Lucy(p), where ‘p’ captures some mental state’s phenomenal character, I’m not thereby representing what it is like for Lucy to be in that mental state. The fact that some matter obtains relative to some subject is itself a (non-subjective) fact that can and plausibly does obtain from anyone’s standpoint when it obtains at all. When I represent that @Lucy(p), I do not thereby adopt the standpoint of those who are in some particular mental state and hence would not naturally represent what it’s like to be in that mental state.

If Lucy is undergoing an experience of a red tomato and if ‘p₁’ expresses the phenomenal fact that accompanies such an experience, it’s the case that @Lucy(p₁), and to represent what it’s like to Lucy to have this experience, I should immersively represent that p₁, not that @Lucy(p₁), but that p₁. By representing the subjective fact that captures the phenomenal character of a mental state, I adopt the standpoints of all and only those who are in that mental state and hence represent what it is like to be in that mental state—exactly as desired.

5. Revisiting the Argument for the Incompleteness of Objective Accounts of Consciousness

The metaphysical picture has an important consequence. If it’s right that mental states have a subjective nature of the proposed kind, then no objective representation can provide a truly complete understanding of what it is for one to have a conscious mental state—exactly as Nagel argued (1979: 167):

If physicalism is to be defended, the phenomenological features must themselves be given a physical account. But when we examine their subjective character, it seems that such a result is impossible. The reason is that every subjective phenomenon is essentially connected with a single point of view, and it seems inevitable that an objective physical theory will abandon that point of view. (Nagel 1979: 167)

The argument has force and yet it’s not evident how to understand the assumptions that are driving it. We can render the argument more precise within the framework proposed.

It seems safe to assume that when one represents some objective matter, one should not thereby adopt what things are like to only some but not all subjects. When a representation captures what things are like to some subjects but not to all subjects, the representation fails to be objective. Take facts about tastiness for illustration again. When licorice is tasty to me but not to you, then any representation of licorice as being tasty (simpliciter) describes how it is relative to me but not relative to you and this implies that it is not an objective account of licorice. Such a representation is exclusively from the perspectives of those to whom licorice is tasty and captures what the world is like to anyone who finds licorice tasty. That such a representation represents what things are like to some but not all subjects suffices for the representation to fall short of objectivity.

This suggests a straightforward necessary condition for a representation to be objective: an objective representation cannot be from the perspective of some but not all subjects. Note that this is a minimal requirement on objectivity, and compatible with different ways of understanding objectivity.

In the previous section we discussed a way of making sense of the claim that whenever we represent some phenomenal fact, we adopt the standpoint of all and only those subjects who are in the relevant mental state. Note that an immersive representation is exclusively from the perspective of those who are in a given conscious mental state m if one necessarily adopts the standpoint of anyone who is in mental state m whenever one engages in this representation. We can now formulate a simple argument against the completeness of objective accounts of consciousness (where the relevant notion of a ‘representation’ is always that of an immersive representation):

• (1)

  Subjective nature: for any mental state m, there is some p such that for s to be in conscious mental state m is, at least partly, for it to be the case that @s(p) and for it to be the case that p.

• (2)

  Minimal requirement on objectivity: if a representation is objective, then it’s not from the standpoint of only some subjects.

• (3)

  Mental differences across subjects: for any determinate conscious mental state m that some subject is in, there is some subject who is not in that mental state.

• (4)

  If a representation is from the standpoint of only those who are in some determinate conscious mental state m, then it’s not objective (by premises 2 and 3).

• (5)

  If a representation is not from the standpoint of only those who are in some conscious mental state m, then it fails to fully represent what it is like for a subject to be in that mental state m (by premise 1 and the proposed account of standpoint adoption).

• (6)

  Any objective representation of a conscious mental state m fails to fully represent what it is for a subject to be in that mental state.

Note that the assumption of sufficient mental differences in conscious mental states concerns determinate conscious mental states, a qualification intended to rule out determinable mental states such being conscious.⁸ The general idea behind the argument is that if it’s the case that, for any mental state m, there is some p such that for s to be in mental state m is, at least partly, for it to be the case that @s(p) and for it to be the case that p, then any full representation of what it is for someone to be in some mental state would have to represent (amongst other things) the phenomenal fact that p and this suffices to make the representation exclusively from the perspective of those who are in the mental state, and hence not an objective representation. No objective representation can provide a complete understanding of a mental state as it cannot fully capture what it is like to be in that mental state and hence cannot capture the subjective phenomenal facts that constitute part of the nature of that mental state.

Premise (2) is the necessary condition on objectivity discussed. As I already mentioned, it’s a rather minimal claim about conditions under which there fails to be objectivity and seems hard to reject. There seems a relevant notion of objectivity for which this is true.

The most likely place at which one might want to resist the argument is the proposed account of the subjective nature of mental states (premise 1) or the proposed account of the sufficient conditions under which one adopts a standpoint (which is presupposed in the formulation of premises 4, 5 and the inference to 6). These are both central commitments of the framework that has been discussed here; accept the framework, and there seems little room to plausibly avoid the conclusion of the argument.

There is much to say about the argument but let me only add a few brief comments. The argument showcases how the proposed account of phenomenal facts has substantive philosophical consequences and demonstrates the theoretical worth of the interlocking regimentations that have been proposed. The main philosophical interest of the reframed argument lies in drawing out an important consequence of the assumption that mental states have a nature that involves both relativization to subjects as well as the obtaining as such of that which is so relativized (something that we also find in the case of temporal matters). The argument has dialectical force to the extent that the first premise, the offered account of the subjective nature of conscious mental states, seems phenomenologically plausible.

The most influential arguments concerning the relation between the physical and the mental proceed from epistemological assumptions to a metaphysical conclusion (Chalmers 2010: 109). Note however that the reframed argument above proceeds in the opposite direction: from a certain metaphysical assumption about what it is for a subject to be in some mental state to an epistemological limitation in objective accounts. Although the key assumption is of a metaphysical character, it is not in and of itself an assumption of epiphenomenalism, dualism, or some other form of non-physicalism. If the conclusion implies any such view, it will be due to further assumptions that are not currently in play.

The conclusion of the reframed argument concerns an epistemological limitation only for objective representations, not for scientific or physical descriptions per se. There is an open question about whether the conclusion implies that mental states cannot but elude the natural sciences. This depends on what the aims of the natural sciences are, and if science is tied essentially to representing only objective facts with respect to all its aims and in the sense that is at work in the argument. This is a question for the philosophy of science—about what science is and has been so far, and about what science could be. Our conclusion concerns objective representations of the nature of conscious mental states, and not directly naturalism or physicalism (compare Crane 2003: §6).

Instead of any explicit appeal to the notion of the ‘what-it-is-likeness’ of a mental state, the argument only appeals to the regimented notions of matters obtaining relative to a subject and the notion of adopting standpoints. Some find the notion of the ‘what-it-is-likeness’ of an experience obscure (see, e.g., Snowdon 2010) but it’s not clear whether this worry applies to the notion of something being the case relative to a subject, which, we may add, is conceptually of a kind with the notion of being the case from the standpoint of a moment in time.

The assumption that phenomenal facts obtain (besides also obtaining relative to subjects) raises the question of what these facts are like. We assume that there is something it is like to be in some mental state. Now how to think of this ‘something’? Although the proposed metaphysical picture is neutral about what this phenomenal fact is like, I fear the proposed picture remains too schematic if there isn’t a working proposal for how to fill in the blanks.

6. Phenomenal Facts as Subjectless Appearances

Proposal: when you experience the world, it comes to phenomenally manifest certain qualities. These manifestations of properties, or appearances, are not intrinsically relativized to a subject, they are themselves not intrinsically manifestations of properties to a particular subject. When you undergo experience, the world out there comes to manifest itself in a certain way, period. When such manifestations (or ‘appearances’) obtain, it’s a further distinct fact that they obtain from the standpoint of certain subjects. Just as there is the fact that I sit and the fact that I sit relative to a moment in time (plausibly assumed to be distinct facts), so there is the fact that the rose appears red and the fact that it appears red relative to a subject s (here likewise assumed to be distinct facts). That there is such a thing as the fact that the rose appears red relative to a subject is uncontroversial. What is controversial is the assumption that there is also such a thing as the rose appearing red, period.

Various philosophers are drawn to some version of the so-called ‘no-subject view’—the view that there are mental states or mental events that do not intrinsically involve a subject.⁹ We can restrict this view to phenomenal facts (instead of the experiences themselves): although mental states are always only mental states of subjects, the phenomenal facts that are instantiated when a subject is in some mental state, do not intrinsically involve that subject.

We causally interact with the world and, as a result, the world comes to manifest itself in all kinds of qualitative guises. The relevant notion of ‘manifestation’ here is not simply that of an epistemic seeming, of something striking a subject as true, but of a qualitative appearance of things.¹⁰ When something appears a certain way, the object manifests certain properties in the way that they do when you experience these things as having these properties.

But what is the appearance of properties metaphysically speaking? Is it a property? If it were a property, what would it be a property of? The used notion of appearance is related (in content but not shape) to what Johnston (2007) and Hare (2009) call ‘presence’, though I do not assume that things ‘disclose some aspect of their nature’ (Johnston 2007: 233) when they appear a certain way, nor do I assume that the relevant notion of appearance is a property. The idea that experiential states are in general always the things that instantiate phenomenal properties doesn’t seem quite right to me, amongst other things because of the much-discussed transparency of experience (Harman 1990). When I see a sunset, it’s the sky that manifests itself as purple and orange, not my experience of the sky. The experience itself precisely doesn’t seem to manifest itself in any way when undergoing it, the experienced bits of world do. We should not say that appearing is only ever a property of experiential states (though we can allow that experiences are amongst the many things that manifest themselves in certain ways, and we can allow that experiences have the property of being such that certain things appear to be the case when undergoing these experience). Things out there in the world manifest themselves as being a certain way.

I want to propose that qualitative manifestation is not a property but is better expressed using a sentential operator (in the way that something being possibly the case or having been the case can be expressed using operators). Putative facts manifest themselves as obtaining. Let’s use ‘(…)’ for this. We can read a sentence of the form ‘(a is F)’ as ‘it appears that a is F’ or ‘it is phenomenally manifest that a is F’ (or simpler, as ‘a appears F’ or ‘a manifests itself as being F’ respectively). So, the sentence ‘(the tulip is red)’ says that it appears that the tulip is red.¹¹

The fact that the relevant notion of appearance is formally captured by a sentential operator assumes that the ways things appear when we interact with the world are ways things appear to be, that is, properties things appear to have. This is closely affiliated with what is known as representationalism (sometimes ‘intentionalism’) in the literature on experience. Martin (1998: 162) helpfully distinguishes between two ways of understanding ‘appearance’ in the literature, arguing that there is often equivocation between them.¹² On the one hand there is the property of being appeared to in a certain way (a property of subjects) or of making for a way of being appeared to (a property of experiences of subjects). This seems to be a common way of understanding what qualia are. On the other hand, there is the property of things appearing a certain way (when experienced). I assume that this second notion of appearance, of appearing a certain way, consists in objects appearing to have certain properties. Shoemaker calls this the ways = properties principle (Shoemaker 2006: 461).¹³ This principle underwrites the comparability of appearance and world. Things can intuitively appear to be different from the way they really are, that is, things can fail to be the way they appear to be, indeed, there may not even be such objects as there appear to be. The question of the veridicality of the way things appear is a question of whether things really have the properties they appear to have.¹⁴

There is much more to say about this proposal, but this is not the right place.¹⁵ Given the subjective nature of mental states, the account looks as follows: for you to undergo a certain mental state is for things to appear a certain way from your perspective and for those things to appear that way full stop.

Phenomenal appearances: For any mental state m, there is some p such that for any s, to be in mental state m is, at least partly, for it to be the case that @s (p) and for it to be the case that (p).¹⁶

For example, for a subject s to undergo perceptual experience of a red tulip is, at least partly, for it to be the case that @s (the tulip is red) and for it to be the case that (the tulip is red). Note the second conjunct, that the experience consists partly in the tulip appearing to be a certain way. When you represent a tulip appearing to be red, you represent what it is like to undergo the experience of a tulip’s being red. Although such a phenomenal manifestation needs to obtain—it needs to be a fact that (the tulip is red)—there is the further distinct fact that this only obtains relative to some subjects and not to others, that is, that for some s₁ and s₂, @s₁ (the tulip is red) but not @s₂ (the tulip is red). In short then, it’s assumed to be a genuine fact that (the tulip is red) and a subjective fact that this is so, one that only obtains relative to some and not relative to all subjects. When one immersively represents the phenomenal fact to be the case, one thereby adopts the standpoint of subjects relative to whom it obtains. To represent what it is like to experience the tulip as red one needs to immersively represent the tulip manifesting itself as red (not just to you, but as such).

The real question is ultimately whether this account captures the phenomenal character of experiences. Is it the case that part of the nature of mental states consists in the outside world appearing various ways from the standpoint of anyone who is in that state? Readers need to judge this for themselves. For what it is worth, I think that this is exactly what mental states are like and what is involved in taking up the perspective of subjects in the conscious mental states. The framework offers a fitting way of carving out a notion of phenomenal facts. This helps explain why representing them involves representing what things are like from the perspective of a subject in a mental state. It also helps to explain why philosophers are attracted to the claim that mental states are transparent, that intending to focus on what the experiences are like has one focusing on what the experienced world is like. When we focus on our experience in a standard case, we focus on how the experienced things manifest themselves.

7. Directions for the Metaphysics of Subjective Facts

As I stressed above, the core of the proposed metaphysical picture is neutral about what the phenomenal facts are like. I believe that thinking of them as appearances is the best way to think of them, but much of this depends on wider theorizing that can be hard to survey. Similarly, the core proposal is neutral about the wider metaphysical treatment of the facts that differ across subjects, and this too depends on wider philosophical theorizing. This final section discusses the main questions on which the further metaphysics turns.

If the proposed account of the subjective nature of mental states is correct, then any conscious mental state is accompanied by a fact of the form @s(p) and a fact of the form p. One key question here is what sort of pattern or principle underwrites the connection between the relevant fact that @s(p) and p. One could accept an unrestricted form of factivity about this relation, or more restricted forms. So, for example, the following are all in principle compatible with the regimentations and principles proposed:

Factivity restricted to appearances: for any p and any subject s, if @s (p), then (p).

Factivity restricted to oneself: for any p and any subject s, if @s(p) and s is oneself, then p.

Unrestricted factivity: for any p and any x, if @x(p), then p.

These principles lead to different pictures of the world. For example, if one restricts factivity to appearance facts, one can think of the world as the totality of objective facts with, in addition, each appearance fact that obtains relative to some subject or other. If one restricts factivity to what obtains relative to oneself, one obtains a solipsist-like or ‘metaphysically centered’ picture according to which the world is in accord with that it is like from your own standpoint only (defended by Hare 2009; Merlo 2016).

Besides the factivity question, there is a question about how to think of the variation of facts across subjects. Remember the principle that we called subjective variegation, which said that for some p, there are some subjects s₁ and s₂, such that @s₁(p) but not @s₂(p). One might think that, if it is not the case that things are a certain way relative to a subject, then, relative to the subject, things are not that way. In other words, if not @s(p), then @s(not p). This means that, if there is subjective variegation, there is also what we may call strong variation: for some p, there are some subjects s₁ and s₂, such that @s₁(p) and @s₂(not p). The variation across subjects involves contrary facts obtaining relative to different subjects.

Given these two questions (about the scope of factivity and about the variation in the facts across standpoints), one will already notice that one package of views seems problematic: if we accept that contrary facts obtain relative to different subjects, and that metaphysical relativity is factive in an unrestricted sense, then we seem to land in incoherence. If @s₁(p) and @s₂(not p), and metaphysical relativity is factive, then it would follow that p and not p.

In response, one could withdraw to one of the restricted forms of factivity (restricted to appearance facts, or restricted to subjective facts relative to oneself), or one could deny that there is a strong variation across subjects. These all lead to coherent metaphysical pictures—and I think many of them are worth exploring and broadly speaking compatible with the story given so far, although, of course, they may alter the precise formulations of the principles discussed here. I have my own preferred view on how best to navigate these questions but there is no need (and no space) to get into it here.¹⁷ What should be clear is that there is a need for a further metaphysical background that can embed and support the proposed understanding of phenomenal consciousness. My hope is that the proposed regimentations may help direct and give purpose to further work on these questions.