It is now common to distinguish between two varieties of Disjunctivism: Metaphysical and Epistemological. It is also common to suggest that the epistemological variety doesn't clearly require a commitment to the metaphysical variety.Footnote 1 In this paper I distinguish between the two varieties of Disjunctivism (§1) and I describe a version of Epistemological Disjunctivism which I call Reflective Epistemological Disjunctivism, or (RED) (§2). I attempt to show that there is a route from that variety of Epistemological Disjunctivism to Metaphysical Disjunctivism (§3). Thus, we have the beginnings of an epistemological argument for Metaphysical Disjunctivism that turns on the truth of (RED).
I go on to explore, however, a certain modification that could be made to the Internalist component of (RED). It seems that the modification would, if made, threaten (RED)'s supposed commitment to Metaphysical Disjunctivism (§4). I argue that, contrary to appearances, making the modification would not undermine the reasons for thinking that (RED) is committed to Metaphysical Disjunctivism (§5). The upshot is that the commitment on the part of (RED) to Metaphysical Disjunctivism is shown to be robust enough to withstand the relevant modification. That is in turn significant because the modification in question should be made, although I do not attempt to prove that here.
1. TWO TYPES OF DISJUNCTIVISM
In this section I distinguish Metaphysical from Epistemological Disjunctivism. Both varieties of Disjunctivism will be formulated by appeal to the distinction between good cases and bad cases.
In the good case, the agent is a normally functioning inhabitant of the world, enjoying a rich set of perceptual experiences that give rise to a rich body of knowledge about the perceived environment. In the bad case, the agent is subject to deception with respect to their sensory experiences and corresponding beliefs. Instead of enjoying a series of perceptual experiences of their environment, the agent suffers a series of hallucinatory experiences that seem exactly the same from their own point of view as the perceptual experiences enjoyed by the corresponding subject in the good case, and the beliefs about the external world which arise out of those hallucinations are false.
The notion of seeming the same employed here is the epistemic notion, according to which for an experience E 1 to seem the same as a second experience E 2 is just for the subject of E 1 not to be able to know by reflection alone that E 1 is not E 2. The notion of seeming the same employed is not the phenomenological notion, according to which for E 1 to seem the same as E 2 is for E 1 and E 2 to have the same phenomenal character, for, as we shall see, the claim that hallucinatory experiences seem the same as corresponding perceptual experiences in the phenomenological sense begs the question against the Metaphysical Disjunctivist.Footnote 2
With this, albeit rough, characterisation of the good case/bad case distinction in the background, we can now examine the distinction between Metaphysical and Epistemological Disjunctivism.
1.1 Metaphysical Disjunctivism
Metaphysical Disjunctivism should be thought of as a way of defending a Relational conception of perceptual experienceFootnote 3 against the Argument from Hallucination.
When one has a perceptual experience, there is something that it is like for one to be undergoing the experience. The perceptual experience is phenomenally conscious. It is also the case that one stands in a relation of perceptual awareness to some entity in one's mind-independent environment. The Relationalist says that we should think of the perceiving relation as a primitive relation of awareness that holds between the subject and the perceived entity. Moreover, the Relationalist adds to this that we should understand the phenomenal character of the perceptual state as being constituted, at least partly, by that primitive relation of perceptual awareness. If Relationalism is true then what it is like for me to perceive the mind-independent world is at least partly constituted by the mind-independent entity I perceive. That's because what it is like for me to perceive at least partly consists in the obtaining of a relation of perceptual awareness that holds between myself and that entity.
It should be noted that Relationalism as it is understood here is consistent with the idea that perceptual experience is a state of mind which represents the world to be the way it is perceived to be. Relationalism says that we should think of the state of experience one is in when one perceives as a primitive relation of perceptual awareness that holds between oneself and the world. It is consistent with that idea that the state in question is a representational state which one can be in if, and only if one counts as perceiving the world. Perhaps the state is thought of as a primitive type of factive representational state analogous to the state of knowing that p as Williamson (Reference Williamson2000) conceives of it, for example.Footnote 4
Of course it is also consistent with Relationalism that we need to think of the perceptual state of mind by appeal to the presence of a primitive relation of awareness which is non-representational. That thesis is Naïve Realism, and it is properly categorised as a version of Relationalism.Footnote 5
The Argument from Hallucination poses a problem for Relationalism. It says that in the bad case, one's experience has the same phenomenal character as a corresponding perceptual experience. That can be argued for either by appeal to the claim that one's experience in the bad case has the same proximate causes as one's experience in the good case, or by appeal to the thought that one's experience in the bad case is introspectively indistinguishable from one's experience in the good case. Either way, the thought is that what it is like for one to perceive is the same as what it is like for one to suffer a corresponding perfect hallucination. But one doesn't stand in a relation of perceiving to anything in the bad case, even though it seems to one, in the epistemic sense, as if one does. So it can't be that the phenomenal character of one's experience in the bad case partly consists in a relation of perceptual awareness. Hence, the phenomenal character of one's experience in the good case cannot so consist. Relationalism is mistaken.Footnote 6
Metaphysical Disjunctivists defend Relationalism by rejecting the claim that the experience in the bad case has the same phenomenal character as the experience in the good case.Footnote 7 This involves rejecting the claim that sameness of phenomenal character follows from introspective indistinguishably and the claim that sameness of phenomenal character follows from sameness of proximate cause.
Metaphysical Disjunctivism is entailed by Relationalism, given the truism that the bad case doesn't involve a phenomenal character constituted by a relation of perceptual awareness. Although Metaphysical Disjunctivism might not entail Relationalism, because the thought that the two types of state in each case have different phenomenal characters even though neither of them involves a relation to a mind-independent object is coherent, it is difficult to see how one might motivate Metaphysical Disjunctivism without appealing to Relationalism.
Metaphysical Disjunctivism is sometimes formulated in a different way, as the claim that the most fundamental kind to which one's experience in the good case belongs is different from the most fundamental kind to which one's experience in the bad case belongs.Footnote 8 Controversially, I think that the most fundamental kind to which one's experience belongs in the good and bad cases is to be specified by appeal to its phenomenal character, and so I think this way of understanding the distinction is logically equivalent to mine. Having said that, I stick to the formulation offered above in what follows because it enables us to avoid getting into debates about what ‘fundamental kind’ means in this context and about whether the fundamental kind to which an experience belongs is captured by its phenomenal character.
1.2 Epistemological Disjunctivism
Epistemological Disjunctivism is best understood as a way of defending a perhaps common sense view about the nature of perceptual knowledge against an epistemic analogue of the Argument from Hallucination. The view in question combines the following two theses:
Reasons Priority. Perceptual knowledge is belief held on the basis of a reason supplied by one's perceptual experience.
Decisiveness. The only reasons that are good enough for knowledge are decisive reasons: reasons that favour one's believing that p, which can obtain only if p is true.
Reasons Priority says that perceptual knowledge is an instance of belief held on the basis of a reason provided to one by perception.Footnote 9 Perceptual knowledge doesn't consist in belief formed or sustained by the occurrence of a reliable belief forming mechanism. Nor is the propensity of perceptual experience to generate knowledge to be treated as a primitive affair.Footnote 10 Rather, we should explain why perceptual experience has the propensity to provide us with knowledge by appeal to the idea that it provides us with reasons for belief.
In order for a belief held for a reason to be an instance of knowledge, the reason must be good enough for knowledge. Decisiveness says that only decisive reasons are good enough for knowledge. Decisive reasons are facts that make it so that the propositions they are reasons to believe are true. That Hesperus is in the sky is a decisive reason to believe that Venus is in the sky, that the pig is standing in the sty before me is a decisive reason to believe that there's a pig in this field, that the note the man is striking on the piano is a C # is a decisive reason to believe that the note isn't an F, and so on.
The upshot of Decisiveness for the proponent of Reasons Priority is that they have to say that the reason one's perceptual experience provides one with is a fact which necessitates the truth of the proposition believed on the basis of perception. It is that claim which the proponent of Epistemological Disjunctivism attempts to defend against the epistemic analogue of the Argument from Hallucination.
The argument in question says that for any reason supplied to one by one's experience in the good case, one's experience supplies one with that reason in the bad case. In the bad case, one's experience fails to provide one with a decisive reason to believe some external world proposition. After all, the relevant external world propositions are false. So in the good case one's experience fails to provide one with decisive reasons too. The upshot is that experience can't provide one with knowledge by supplying one with decisive reasons for belief. Either Reasons Priority or Decisiveness must go.
Epistemological Disjunctivism says that we should protect Reasons Priority and Decisiveness by rejecting the claim that one's perceptual experience supplies one with the same reasons as one's corresponding hallucinatory experience. One could motivate the claim that one's experience in the good case supplies one with the same set of reasons as one's experience in the bad case by appeal to the fact that one's experience in the bad case is introspectively indistinguishable from one's experience in the good case. We will be looking in more detail in §2 at how that inference might work and how the proponent of a certain type of Epistemological Disjunctivism might reject it.
2. REFLECTIVE EPISTEMOLOGICAL DISJUNCTIVISM
Here is one question we can ask the Epistemological Disjunctivist: with what should we identify one's decisive reason in the good case? Disagreement on that score is one way in which different versions of Epistemological Disjunctivism are generated. Here I focus on one variety of the view: Reflective Epistemological Disjunctivism, or (RED).
(RED) has three core commitments. Here is the first:
Reasons as Perceivings. One's perceptual reason for believing that p in the good case is that one is perceiving x, where one can perceive x only if p.
According to Reasons as Perceivings, when one believes that p on the basis of a perceptual reason, one's reason should be identified with a certain fact. The fact in question is one which records the presence of one's perceptual state: the fact that one is perceiving x. That fact can obtain only if p, because the x in question is such that one can perceive x only if p.
What we replace ‘x’ with depends on which version of (RED) we're talking about. On one version of the view, we're to identify one's perceptual reason with the fact that one is perceiving that p: we replace ‘x’ in the formulation of Reasons as Perceivings above with a that-clause that denotes the very fact known. On this view, I know that there's a red cube before me because I believe that there's a red cube before me on the basis of the reason that I'm seeing that there’s a red cube before me.Footnote 11 On another version of the view, we're to identify one's perceptual reason with the fact that one is perceiving the relevant object to have the property one comes to know it to have: we replace ‘x’ as it appears in the formulation of Reasons as Perceivings with a noun phrase of the form ‘the O to be F’. On this view, I know that there's a red cube before me because I believe that there's a red cube before me on the basis of the reason that I’m seeing the cube to be red.Footnote 12 I aim to remain neutral between these two versions of the view, and indeed any others,Footnote 13 in what follows, and so I continue to talk neutrally of S perceiving x.
In order for a certain fact to function as the reason for which one believes (or indeed acts) the fact in question must be something of which one is aware. A fact which obtains in one's environment and which is a reason in favour of one's believing that p is not one which can function as the reason for which one believes as one does unless one is aware of the fact. When one is aware of the fact in the appropriate way we can say that one possesses or has the reason in question.Footnote 14
The proponent of (RED) identifies one's perceptual reason with the fact that one is perceiving x. How does one count as possessing that reason in cases of perceptual knowledge? Pritchard (Reference Pritchard2012) does not address that question, so far as I can tell. But I think McDowell, a proponent of (RED), gives us an answer to it here:
Our topic is, or at least should be, how experience figures in the warrant for knowledge of a sort that is distinctive of rational subjects: knowledge of a sort we can conceive, following Sellars, as a standing in the space of reasons … Knowledge of this sort is an act of its subject's rationality. Someone who has knowledge of this sort must be in a position to know the warrant by virtue of which her state counts as knowledge. No doubt one does not typically make explicit, even to oneself, how one's knowledgeable beliefs are warranted, or even what it is that one knowledgeably believes. But if an experience warrants one in knowledge of the sort that is an act of rationality, the warrant-constituting status of the experience must be part of the content of an at least implicit self-consciousness that belongs to one's cognitive state in knowing what one does. In the terms I have been using, the experience's epistemic significance must be part of its subjective character. (McDowell Reference McDowell2013: 148–9)
One of McDowell's thoughts here is that perceiving is an intrinsically self-conscious state. Being perceptually aware of one's environment involves an awareness of the fact that one is perceptually aware of the relevant portion of the environment in the way that one is. There is an awareness of the fact that is one's reason which is constituted by one's state of perceiving. So in answer to the question: what constitutes one's possession of one's perceptual reason in cases of perceptual knowledge, for the proponent of (RED)? We can say: the fact that is one's reason is a fact that records the existence of a perceptual state, where that very state counts as an awareness of the fact that one is in that state. Here we have the second core thesis of (RED):
The Self-Consciousness Thesis. One possesses one's reason for believing that p – that one is perceiving x – because one's state of perceiving x is an intrinsically self-conscious state.
The type of self-conscious awareness at issue is not itself a state of knowing, because one can perceive x without believing, and hence without knowing, that one is doing so. Nor is the relation of awareness perceptual – one does not perceive one's being in a state of perceiving, when one has a perceptual experience. Rather, it is a primitive type of non-doxastic awareness that comes along with the perceiving relation. This non-doxastic awareness grounds the knowledge one can have of the fact that one is perceiving x. There are, of course, numerous further questions which the Self-Consciousness Thesis raises, but the rather crude characterisation of it just offered will suffice for my purposes.Footnote 15
We are already in a position to see that for the proponent of (RED), believing for a perceptual reason is quite different from believing on the basis of a reason supplied to one by some other belief one has. In the latter case, one's reason is typically a fact about the external world and one is aware of it in the way required for believing for a reason just because one believes it and one's belief has the right epistemic credentials. Thus, the answer to the question: what is S's reason? in inferential cases can be different from the answer to the question: how is S aware of her reason? In the perceptual case by contrast, as the proponent of (RED) conceives of it, one's perceptual reason is the fact that one is perceiving x and the very state of perceiving x is what constitutes one's possession of that reason. One's reason is a fact that records the presence of one's perceptual state, it is not a fact about the external world – even if it might entail one. And the answer to the question: what is S’s perceptual reason? will be the same – or at least could be answerable using the same form of words – as the answer to the question: how is S aware of her perceptual reason? The answer in both cases is: S is perceiving x.
The final component of (RED) is a commitment to the thought that one's perceptual reasons are, in a certain way, reflectively accessible to one. A commitment to the sort of reflective access theses in question is sufficient for being a member the Access Internalist tradition in Epistemology. Thus, (RED) belongs to that tradition. Here I think we can distinguish two theses:
Reflective Access I. Whenever one believes that p for a reason, one is in a position to know by reflection alone that one does so.
Reflective Access II. Whenever one possesses a reason, one is in a position to know by reflection alone that one does so.
Reflective Access I is the thesis that whenever one believes that p on the basis of a reason, R, one is able to know just by reflection on one's situation that one believes that p on the basis of R. Believing for a reason is a luminous condition, to use terminology borrowed from Williamson (Reference Williamson2000). If I believe that the Labour party will win the by-election on the basis of the reason that the exit-poll predicts so then I am in a position to know by reflection alone that I believe on that basis. What applies in that case also applies in the case of perception too. Whatever my perceptual reason is, when I believe on the basis of it I am in a position to know by reflection alone that I do so.
The proponent of (RED) identifies one's perceptual reason with the fact that one is perceiving x. When one knows that p by perception, one holds one's belief that p on the basis of the reason that one is perceiving x. Thus, in conjunction with (RED), Reflective Access I predicts that when one knows that p by perception one is in a position to know by reflection alone that one has based one's belief on one's perceiving x. Notice that it follows from this that when one knows that p in the perceptual case, one is in a position to know by reflection alone that one is perceiving x.
Reflective Access II is a thesis about reasons possession in general. It should be read as the claim that for any state which can constitute one's possession of a reason, when that state constitutes one's possession of a reason, one is in a position to know by reflection alone that one is in that state. I possess the reason that the by-election predicts a Labour victory because I know that fact. Thus, Reflective Access II predicts that in this circumstance, I am in a position to know by reflection alone that I know that the by-election predicts a Labour Victory.
The proponent of (RED) says that it is one’s intrinsically self-conscious state of perceiving x itself which constitutes one’s possession of one’s perceptual reason in cases of perceptual knowledge. Thus, in conjunction with (RED), Reflective Access II implies that when one knows that p via perception, one is in a position to know by reflection alone that one is perceiving x.
The proponent of (RED) is at least committed to Reflective Access I and I suspect they might wish to commit themselves to Reflective Access II as well, although that is not compulsory.Footnote 16 What is important to note for our purposes is that each thesis entails, in conjunction with the rest of (RED), that in cases in which one knows that p by perception, one is in a position to know by reflection alone that one is perceiving x.
One of the premises in the epistemic analogue of the Argument from Hallucination is that the reasons one's experience provides one with in the good case are the same as the reasons one's experience provides one with in the bad case. That premise of the argument can be motivated by appeal to either Reflective Access I or II, in conjunction with the further thought that what one can know by reflection alone is the same in the good case as in the bad case. Let's take a closer look at how these derivations are supposed to work.
First, let's take Reflective Access I. Suppose that Reflective Access I is true, and suppose that one's experience in the good case provides one with a reason, R, which it doesn't provide one with in the bad case. Now let's suppose that one believes that p on the basis of R in the good case. Given Reflective Access I, in the good case one can know by reflection alone that one believes that p on the basis of R. But if what one can know by reflection alone in the good case matches what one can know by reflection alone in the bad case then that cannot be, for ex hypothesi one cannot believe that p on the basis of R in the bad case, and so one cannot know by reflection alone that one believes that p on the basis of R in the bad case.
Now let's look at Reflective Access II. Suppose that Reflective Access II is true, and suppose that one's experience in the good case constitutes one's possession of a reason, R, which one doesn't possess in the bad case. Thus, by Reflective Access II, one is in a position to know by reflection alone that one possesses R. But if what one can know by reflection alone in the good case matches what one can know by reflection alone in the bad case, then that cannot be, for ex hypothesi one doesn't possess R in the bad case, and so one cannot know by reflection alone that one does so in the bad case either.
The proponent of (RED) rejects those lines of reasoning, whilst holding fast to Reflective Access I and possibly also to Reflective Access II, by rejecting the claim that what one can know by reflection alone in the bad case is the same as what one can know by reflection alone in the good case.Footnote 17 It is no more plausible, they will say, that what one can know by reflection alone is the highest common factor of the good case and bad as it is to suppose that what one can know about the external world is so.
3. FROM (RED) TO METAPHYSICAL DISJUNCTIVISM
Having distinguished Metaphysical from Epistemological Disjunctivism I now want to raise the issue of how the two theses are related. A number of writers in this area have suggested that a commitment to Epistemological Disjunctivism doesn't obviously require a commitment to Metaphysical Disjunctivism. Here are two representative expressions of that idea, the first from Paul Snowdon, the second from Duncan Pritchard:
there is no reason to ascribe to McDowell acceptance of the claim about perceptual experience that I have been suggesting is the distinctive claim of [metaphysical] disjunctivism. For why cannot a single basic sort of (inner) experience have quite different epistemological significance in different cases, depending, say, on the context and on facts about causation? (Snowdon Reference Snowdon2005: 140)
It is reasonably clear that Epistemological Disjunctivism does not in itself entail Metaphysical Disjunctivism. For that the rational standing available to the agent in normal veridical perceptual experiences and corresponding (introspectively indistinguishable) cases of … hallucination are radically different does not in itself entail that there is no common metaphysical essence to the experience of the agent in both cases … one would need to defend further claims in order to supply the relevant ‘bridge’ between these two theses. (Pritchard Reference Pritchard2012: 24)
Both Snowdon and Pritchard suggest that Epistemological Disjunctivism doesn't obviously require a commitment to Metaphysical Disjunctivism. One would have to defend some further principle that takes us from the former to the latter, if it's to be demonstrated that Epistemological Disjunctivism does engender such metaphysical commitments. What I want to do here is supply that bridge from one version of Epistemological Disjunctivism – (RED) – to Metaphysical Disjunctivism.
Pritchard and Snowdon might well not wish to cash out Metaphysical Disjunctivism in the way that I do, of course. In which case, I might not succeed in providing a bridge from Epistemological Disjunctivism to Metaphysical Disjunctivism in their sense. However, I'll still have provided a bridge from the former to the latter as I am understanding it, which remains a significant result.
The proponent of (RED) subscribes to Reflective Access I and perhaps also to Reflective Access II. I've already noted that a commitment to either of those theses requires a further commitment to the claim that whenever one knows that p via perception, one is in a position to know by reflection alone that one is perceiving x.
But what is it to know that one is perceiving x by reflection alone? ‘Reflection’ is an umbrella term used to refer to a variety of mental activities, some of which constitute ways of knowing. Mental activities that fall under the category of reflection include: introspective attention involving one's phenomenally conscious states of mind, whatever activity one engages in in order to find out what standing states one is in, engaging in episodic recall, memory retrieval, self-critical reflection – by which I mean the sort of activity involved in answering questions such as ‘did I act appropriately?’ and ‘is my belief that p adequately justified?’, conscious inference, deliberating about whether a certain proposition is true or about how to act, whatever activity is involved in discovering a priori truths, and whatever activity is involved in discovering analytic truths.Footnote 18 Of those activities, most of them – introspective attention, the activity that yields knowledge of one's standing states, the activity involved in discovering a priori truths, the activity involved in discovering analytic truths, conscious inference, and episodic recall – constitute ways of knowing. One can know truths off the back of engaging in one of those activities.
Why do we regard each of those activities as forming a unified category which we pick out using the term ‘reflection’? The standard suggestion, which we can go along with here, is that each type of activity is not constitutively dependent upon perception. That is, for each type of activity, one can count as engaging in that very type of activity even when one is not perceiving the world. This is not to deny that there are tokens of the relevant types which are so dependent, nor that the success cases of such activities are in some way privileged.Footnote 19
How, then, might one count as being in a position to know by reflection alone that one is perceiving in the good case? I think there are two options here. The first is to say that one knows that one is perceiving by inference from other facts one knows by reflection alone. The second is to say that one knows that one is perceiving by introspection alone. I want to argue that neither of those options is acceptable for the proponent of (RED), as long as they reject Metaphysical Disjunctivism. The only further move available to them is to accept Relationalism and it's associated Metaphysical Disjunctivism. That way, the second option becomes viable after all. I'll examine each option in turn.
3.1 Option I: Inference from Reflectively Known Truths
Let's begin with the first option: that one knows that one is perceiving x in the good case by reflection alone because one counts as knowing it by inference from other propositions one knows by reflection alone. The idea here is that we come to know that we are perceiving when we are by reflecting on our situation, thus coming to know various propositions which, when taken together, entail or at least strongly indicate, that we are perceiving. We then perform a further act of reflection – deductive or abductive inference from those propositions – to arrive at knowledge that we are perceiving. That knowledge is knowledge by reflection alone because it is knowledge by inference from facts one already knows by reflection alone.
What might those reflectively known propositions be? They would be propositions such as: I am having an experience as of x, the lighting conditions seem appropriate for perceiving, my experience as of x is not a dream or a hallucination, my experience as of x is likely caused by x itself. The thought is that one can know propositions like those by reflection alone, and once one knows them one can infer from them that one is perceiving, where the inference would either be abductive or deductive depending on which propositions from the list are chosen.
The problem with this option is simply that it is too difficult to get off the ground. One would have to demonstrate that one can know the relevant propositions off the back of reflection alone and then one would have to show that one can come to have knowledge that one is perceiving just off the back of one's knowledge of those propositions. It is an onerous task indeed to demonstrate that that's possible. I'm not going to demonstrate that the task cannot be fulfilled. But I think a survey of the history of anti-sceptical epistemology licenses us in giving the idea short shrift.
3.2 Option II: Introspection
Now let's turn to the second option: that one knows that one is perceiving x in the good case by reflection alone because one knows it by introspection alone.
Let's suppose that Metaphysical Disjunctivism and hence Relationalism are false. If that's so, then when I perceive the red cube before me, I'm in a conscious state which has a phenomenal character exactly the same as the phenomenal character of the corresponding introspectively indistinguishable state of the subject in the bad case. What it is like for me to be in that state is not constituted wholly or partly by my perceiving the cube to be red.
If that is my situation then the phenomenal character of my state of perceiving could be held fixed whilst it is varied whether or not the state is one of perceiving or one of perfect hallucination. But if that's correct, then it seems plausible that I can't know just by introspection that I'm perceiving the cube to be red when I am so. I take it to be plausible, that is, that if what it is like for one to be having a certain experience E 1 is the same as what it is like for one to be having experience E 2, then what one can know just by introspective attention to E 1 matches what one can know just by introspective attention to E 2.
What's at issue here is the following principle:
The Constitution Principle. For any truth, p, about a phenomenally conscious state one is in, E, if p is knowable by introspection alone, then p is a mode of presentation of some factor which at least partly constitutes the phenomenal character of E.
According to the Constitution Principle, the only truths about phenomenally conscious states which one can know just by introspective attention to such states are those truths concerning features which partly constitute the phenomenal character of the state. It follows from the Constitution Principle that if there is some factor which does not constitute what it is like for you to be in a conscious state, then introspection alone is impotent with respect to delivering one knowledge that the factor obtains. It also follows from the principle that if there is some difference between two experiential states with respect to what one can know about them via introspection alone, then there must be a phenomenological difference between them too.
The Constitution Principle doesn't imply that one cannot come to know truths about what non-conscious standing states one is in by introspection. That's because it makes a claim about how introspection is operative with respect to phenomenally conscious states alone. So if one were to allow that the activity of introspection is the activity involved in our coming to know truths about what non-conscious standing states we're in, that would be no objection to the Constitution Principle.
Moreover, the Constitution Principle does not imply that being a constituent of the phenomenal character of a conscious state is sufficient for the subject to be in a position to know some truth that corresponds to the relevant factor. The principle allows that there are elements of phenomenal character that are not knowable via introspection in some contexts, or across all contexts. What it says is just that it is a necessary condition of being knowable by introspection alone that the relevant factor at least partly constitutes the phenomenal character of the state.
The Constitution Principle acquires some degree of inductive confirmation from typical examples of self-knowledge of conscious states of mind. When one comes to know that one has judged that p solely via introspective attention to the mental event of judging, it's plausible that the status of the event as an event of judging, as well as the particular content of the event, are partly constitutive of what it is like to judge that p. Similarly, when one comes to know that one has a pain in one's knee via introspection alone, the status of the event as a pain as well as the location of the pain are constituents of the phenomenal character of the event to which one is attending. And when one comes to know that it seems to one as if one is hearing an engine revving off the back of introspective attention to one's auditory experience alone, what one is thinking about is the phenomenal character of an event of auditory experience one is currently undergoing.
Further support is added to the principle by its explanatory power. For example, it can be invoked to explain why we cannot come to know the truths of a great many propositions about our experiences solely off the back of introspection. We cannot come to know, just by introspection, what the proximate neural causes of our states of experience are, for example. Nor can one come to know, just by introspection, the date at which one's experience occurs. The Constitution Principle explains why: neither the proximate neural causes of experience, nor the date at which one's experience occurs, partly constitute what it is like for one to be having the experience.
The constitution principle might be thought implausible because one can know the truth of propositions about the temporal profile of one's experience by introspection alone which do not seem to correspond to phenomenology constituting features of one's experience. In particular, one can know by introspection alone that one is having one's current experience now. But that one is having one's experience now is not an aspect of the phenomenal character of one's experience, for the phenomenal character of one's current experience is repeatable across time. One's past and future experiences could well have the same phenomenal character as one's present experience. Similarly, one can know by introspection alone that one's experience is one which one is having oneself – that one's experience belongs to one in the distinctive way in which one's particular mental states in general are not the sorts of things which could be shared with other people. But the fact that one's conscious states of mind belong to one in the way in question is not an aspect of the phenomenal character of one's experience, for the phenomenal character of one's experience can be exemplified by the phenomenally conscious states of other people.
In response to both of these worries I want to suggest that there is no good reason to accept the conception of phenomenal character operated with by the proponent of each objection. Why must it be the case that every aspect of the phenomenal character of experience is repeatable either across time or across numerically distinct subjects of experience? To the extent that we can know by introspection alone that one's experience is occurring now and that one's experience belongs to one, that simply puts pressure on the conception of phenomenal character in question.
I've suggested that there's a prima facie plausible principle which implies that as long as the proponent of (RED) denies Metaphysical Disjunctivism and its attendant Relationalism, they cannot identify the way in which one comes to know by reflection alone that one is perceiving x in the good case with introspection of one's perceptual state. So option II is ruled out, as long as that denial is in place. But option I should be rejected, as we have seen, and it seems to me that those options are exhaustive.
How should the proponent of (RED) respond to this quandary? What I suggest is that they need to accept Relationalism and the Metaphysical Disjunctivism which comes along with it. That way, they are committed to saying that the perceiving relation partly constitutes the phenomenal character of one's experiential state in the good case. What it's like for one to perceive is partly constituted by its status as a perception. With that claim in place, the proponent of (RED) can identify, consistently with the Constitution Principle, one's knowing by reflection alone that one is perceiving in the good case with one's knowing by introspection alone that one is perceiving.Footnote 20
It seems, then, that there is one variety of Epistemological Disjunctivism which is committed to Metaphysical Disjunctivism after all. Aside from whatever intrinsic interest this result has, it is significant because it makes it possible for one to argue epistemically for the truth of Relationalism and its attendant Metaphysical Disjunctivism by independently motivating (RED).Footnote 21
I cannot claim to have offered a decisive case for the Constitution Principle here, so I cannot claim to have decisively proven that (RED) is committed to Metaphysical Disjunctivism. However, there are two comments worth making in this connection.
First, even if I have not proven the Constitution Principle, what has been said here should still carry significant interest. For what I have said amounts to the identification of one way of bridging the gap between (RED) and Metaphysical Disjunctivism, a way of bridging the gap which is at least viable and which should be the topic for further research.
Second, in order to prove that the proponent of option II cannot deny Relationalism, there is an alternative strategy available which doesn't require appealing to anything quite so strong as the Constitution Principle. For if one can know that one is perceiving x via introspection alone, then how could that be unless perceiving x is a constituent of the phenomenal character of the experiential episode in question? What could explain why such a fact is knowable via introspection other than that it is a constituent of phenomenal character? Even if we remain neutral on the Constitution Principle, there will still be an explanatory challenge that looks difficult to meet for those proponents of (RED) wishing to pursue option II. The challenge is to explain how knowledge by introspection alone that one is perceiving is possible without recourse to the idea that it is possible in the same way in which knowledge of other phenomenology constituting conditions is possible: by dint of being an aspect of the phenomenal character of the relevant state of mind.
In any case, in the next section I want to digress in order to consider a certain modification that could be made to the reflective access theses associated with (RED). Were the proponent of (RED) to accept the modification, it seems as if that would enable the proponent of (RED) to avoid a commitment to Metaphysical Disjunctivism after all. Ultimately, I wish to show that even if the modification were made, the proponent of (RED) would still be committed to Metaphysical Disjunctivism anyway.
4. SOME REFLECTIONS ON REFLECTIVE ACCESS
Earlier, we encountered the following reflective access theses:
Reflective Access I. Whenever one believes that p for a reason, one is in a position to know by reflection alone that one does so.
Reflective Access II. Whenever one possesses a reason, one is in a position to know by reflection alone that one does so.
Reflective Access I and II are not obviously true, however much they have been taken to be by their proponents.Footnote 22 Why should it be that whenever one one believes for a reason, one is also able to know an additional fact about the way in which one's belief is based? And why should it be that whenever a mental state one is in constitutes one's possession of a reason, one is also in a position to know by reflecting on one's situation that one is in that state? These theses might be true, but they are not obviously so.Footnote 23
To buttress this point somewhat, we can note that the sort of connection between reasons and self-consciousness at issue is not obvious in the case of acting for reasons or in the case of possessing reasons for action. Indeed, so unobvious is it in the practical domain that the connection is rarely discussed by philosophers of action or meta-ethicists, if at all. Why, then, should we expect the connection to be obvious in the epistemic case?
What this brings out is that we'll have to be supplied with arguments for Reflective Access I and II, they cannot just be treated as obvious. Here I want to introduce some alternative reflective access theses (§4.1). For each of our original reflective access theses I then consider a natural way of motivating it and I suggest that each argument is undermined by the availability of the alternative thesis (§4.2). This digression is of significance for the main focus of the paper: the relationship between (RED) and Metaphysical Disjunctivism, as we shall see.
4.1 The Alternative Theses
The alternative reflective access theses I have in mind are these:
Reflective Access I*. Whenever one believes that p on the basis of a reason, R, one is in a position to engage in some activity of reflection that constitutes bringing R to mind as a reason to believe that p.
Reflective Access II*. Whenever one possesses a reason, R, in favour of believing that p, one is in a position to engage in some activity of reflection that constitutes bringing R to mind as a reason to believe that p.
The consequent of Reflective Access I* is identical to the consequent of Reflective Access II*, it is only with respect to their antecedents that the two principles differ. Thus, I'll start my elucidation of each thesis by saying something about what the consequent of each amounts to.
I intend bringing a fact to mind by reflection in the sense at issue to be a familiar phenomenon. Recalling a fact one already knows in order to answer a question asked to one by an interlocutor – for example, recalling that it will rain later when someone asks what the weather will be like this afternoon – is a paradigm case of bringing a certain fact to mind via an activity of reflection. Recalling a fact during the course of deliberation about what to do or what to believe, where the fact in question is relevant to the issue, is another paradigm example of bringing a fact to mind by reflection.
The examples just given of the phenomenon at issue both involve recalling a fact one already knows. But the phenomenon in question can sometimes involve coming to know the relevant fact for the first time. For example, knowledgeably judging that one is hearing an aeroplane fly overhead off the back of introspective attention to one's auditory experience is an instance of bringing a fact to mind by reflection – in this case, a fact about one's auditory experience.
In general, we can think of the sort of phenomenon talked about by the consequents of our new reflective access theses as the phenomenon of making it explicit to oneself that p, where one is already aware of p or some entity which suitably corresponds to p. This can take the form of recalling a known fact, in which case no new knowledge is generated. But it can take the form of transitioning from an awareness that constitutes a way of knowing that p to a knowledgeable judgement that p and thus coming to know that p for the first time.
But what is it to bring a certain fact to mind? What exactly is it to do that by reflection? And, finally, what is it to bring a fact to mind as a reason for belief? I want now to sketch an account of the type of activity in question which goes some way towards answering those questions.
To bring a certain fact, p, to mind in the sense at issue is to have an occurrent thought with p as its content. The occurrent thought in question is the manifestation of a state of mind one is already in, prior to having the occurrent thought. Such awareness constitutes a certain type of awareness of either the fact that p itself, or else of an entity which suitably corresponds to p. Bringing a certain fact to mind, then, involves three elements: (i) an occurrent thought with p as its content, (ii) a state of mind one is in prior to the occurrent thought which constitutes an awareness of the fact that p, or some other suitable entity, and (iii) some relation of manifestation that holds between (i) and (ii).
Consider, first, element (i). Occurrent thoughts are here understood to be a certain type of phenomenally conscious mental event which are individuated partly by their representational contents. Such mental items should be distinguished from standing states such as states of believing, desiring, intending and so forth because those items are states, as opposed to events, and because those items needn't be, and I think never are, phenomenally conscious.Footnote 24 Amongst occurrent thoughts we can include mental events like: consciously deciding to ϕ, feeling a desire to ϕ, wondering whether it is the case that p, imagining a certain scene, supposing that p for the sake of argument and judging that p.
To bring a certain fact to mind in the sense intended is not to have just any type of occurrent thought with p as its content, but an occurrent thought which involves affirming the truth of p in a way analogous to the way in which one does so in so far as one believes that p. It is, in other words, to judge that p. For our purposes we can think of judgement as a mental analogue of asserting that p. To judge that p is to engage in a conscious mental activity which involves avowing to oneself that p is the case.
Consider, now, element (ii). Bringing the fact that p to mind in the sense at issue is to make a certain transition from a suitable state of awareness of the fact that p one is already in, to the judgement that p. The state of awareness in question could be a standing state of knowing that p. That's typically the sort of state involved in bringing a fact to mind when one recalls that p. But the state of awareness needn't be a doxastic state, as the example of introspection given above demonstrates – it could be a state of perceptual or introspective awareness.
Finally, consider (iii): that the activity of judgement engaged in by the subject is a manifestation of the state of prior awareness the subject is in of the fact judged, or some entity suitably correspondent to that fact. I do not want to offer an account of what this manifestation relation consists in, although it will surely involve some type of counter-factual dependence of the judgement on the state of awareness.
What is it to bring R to mind as a reason to believe that p? Roughly, it is to make a judgement which is the manifestation of some suitable state of awareness one is already in, but where the judgement in question involves consciously treating R as a reason in favour of believing that p. What this amounts to is a matter for further debate. But the basic idea is that the status of R as a reason for the relevant belief is itself present to consciousness. This might involve the subject simultaneously making a further judgement, to the effect that R, for example, entails or probabilifies the relevant proposition. Or it could be that judging that R suffices for consciously treating R as a reason to believe that p, relative to background knowledge of the supporting relation R stands in to p.
Finally, to bring a certain fact to mind as a reason to believe that p via an activity of reflection is just for an activity that is of the type reflection to constitute the transition from the prior state of awareness to the judgement that R which involves consciously treating R as a reason for belief. So some act of memory recall, introspection, self-critical reflection, deliberation about what to do or think, and so on, is the act of mind that constitutes moving from some appropriate state of awareness to the judgement that involves consciously treating R as a reason for belief.
We're now in a position to understand what Reflective Access I* and II* amount to. Let's suppose that one believes that the Labour Party will win the by-election for the reason that the exit-poll predicts that that's so – a fact which one knows. What Reflective Access I* implies is that one is thus in a position, by engaging in an act of recall, to judge that the exit-poll predicts a Labour victory at the by-election, where that judgement involves consciously treating the fact about the exit-poll as a reason to believe that Labour will win and is a manifestation of one's knowledge that the exit-poll predicts a Labour victory. Likewise, let's suppose that one's knowledge that the murder weapon was found in the butler's sock drawer constitutes one's possession of that fact as a reason to believe that the butler is the murderer. It follows from Reflective Access II* that one is thus in a position to engage in some act of reflection which brings about a judgement that the murder weapon was found in the butler's sock drawer. The act of judgement involves consciously treating that fact as a reason to believe that the butler is the murderer and is a manifestation of one's knowledge about the whereabouts of the murder weapon.
4.2 Motivating Reflective Access I and II
What might motivate one to accept either Reflective Access I or Reflective Access II? As I've suggested, I don't think either thesis is obviously true – we'll need arguments for them. Here I want to examine an argument for each. In each case, I'll describe the argument before suggesting in turn that the argument is undermined by the possibility of the relevant alternative reflective access thesis.
The way of motivating Reflective Access I I want to examine says that whenever one believes that p on the basis of a reason, R, one is in a position to successfully mount a defence of one's belief in the face of a certain challenge, raised either by oneself or by someone else, that typically comes in the form of the question: ‘why do you/I believe that p?’. The challenge is to the epistemic credentials of one's belief that p. The thought is that whenever one believes that p for the reason, R, one is in a position to successfully answer that challenge by reflecting on one's situation alone.
The type of challenge is in turn construed as a request to demonstrate that one's belief that p is an instance of knowledge.Footnote 25 One can successfully answer the challenge only if one is in a position to demonstrate that one's state of belief has a certain epistemic status – the status of knowledge. This is supposed to lend support to Reflective Access I because being in a position to demonstrate, just by reflecting, that one's belief that p is an instance of knowledge requires that one is in a position to know just by reflection that one's belief is held on the basis of reasons that are adequate for knowledge. And that involves satisfying the condition laid down by Reflective Access I.
But it is not obvious that the only way to interpret the type of challenge at issue is as a challenge to demonstrate that one's state of belief is an instance of knowledge. An alternative construal of the challenge is as a challenge to demonstrate that what one believes is a proposition that there is good reason to believe. That is, an alternative construal of the challenge is to think of it as asking to demonstrate that one's belief that p, understood as the proposition believed, as opposed to the state of believing, is well-supported by some fact that constitutes a reason in favour of believing it.
The alternative construal of the challenge motivates Reflective Access I*, not Reflective Access I. Being in a position to demonstrate, by reflection, that one's belief is well-supported by reasons in favour of it would only require that one is in possession of the relevant reasons and that one is in a position to call those reasons to mind as reasons to believe that p via some act of reflection available to one. It doesn't obviously require one to be in a position to know that one's belief that p is based on those reasons.
The further point is that there is no reason for us to think it necessary for believing for a reason that one is able to answer the challenge on the first construal instead of thinking that what is necessary is that one is able to answer the challenge on the second construal. If believing for a reason requires being able to answer the challenge on some construal at all, and I don't want to deny that it does, then only on the second construal of the challenge is answering it a requirement on believing for reasons. We can conclude that this way of motivating Reflective Access I fails.
Now I want to move on to discussing the motivation for Reflective Access II. The way of motivating Reflective Access II I have in mind says that one can be in a state which constitutes one's possession of a reason to believe that p only if one is able to use the reason in episodes of deliberation about whether p. One possesses a reason only if the psychological state which constitutes one's possession of the reason enables one to use the reason as a reason to believe what it is a reason to believe in cases where one is deliberating about the truth of that proposition. According to that idea, my knowledge that the exit-poll predicts a victory for the Labour Party constitutes my possession of a reason to believe that Labour will win only if I am, during an episode of deliberation about whether or not Labour will win, able to use that fact as a reason to believe that proposition.
The further thought is that one can be in a position to use one's reason in episodes of deliberation only if one can, by reflection alone, know that one possesses the reason in question. I can use the fact that the exit-poll predicts a Labour victory as a reason during the course of deliberation about whether Labour will win only if I can know by reflection alone that I know that the exit-poll predicts a Labour victory.
The worry here is analogous to that raised above concerning the motivation for Reflective Access I. It is not clear that in order to be in a position to use a reason one possesses in a process of deliberation one must be in a position to know by reflection that one is in the state that constitutes one's possession of one's reason. All that seems required is that one is able to bring one's reason to mind as a reason to believe that p, by some reflective activity available to one during the course of one's deliberation.
What we have here is a challenge to the proponents of Reflective Access I and II. Those theses are not obviously true. Moreover, we can contrast them with two alternatives. I've considered a way of motivating each thesis. But what we really have is, at best, a motivation for the corresponding alternative thesis. The challenge is to find some alternative source of motivation for the original reflective access theses and demonstrate that we really do have a motivation for those theses there, and not simply a motivation for one of the alternative theses. I doubt that there is any such additional source of motivation available.
There is an obvious response to the challenge that I wish to consider, and it brings up an issue that has been lurking in the background throughout this digression. The response is that I've effectively been taking it for granted that Reflective Access I* and II* don't entail Reflective Access I and II respectively. If those entailments do hold, however, then whatever motivates Reflective Access I* also motivates Reflective Access I and likewise Reflective Access II* and II. So in order to show that the lines of reasoning just rehearsed only deliver us our alternative reflective access theses and not the originals I'll have to prove that the entailments don't hold.
Do Reflective Access I* and II* entail Reflective Access I and II respectively, then? I think it's unobvious that that's so. With respect to Reflective Access I* and I, a proponent of the entailment would have to show that it is impossible for there to be cases in which a subject has grounds for doubting that they hold their belief for a reason, and so they are not in a position to know that they do so, but nevertheless they do in fact hold their belief for a reason and they are in fact able to call that reason to mind. And with respect to Reflective Access II* and II, the proponent of the entailment would have to show that it is impossible for there to be cases in which a subject is not in a position to know that they know that p, even though they do know that p, their state of knowledge constitutes their possession of p as a reason for belief, and they are in fact able to call p to mind as a reason for belief.
I think that such cases are possible, and hence that the relevant entailments don't hold. However, I do not want to pursue that issue here, for that would take me too far afield. Instead, I want to concentrate on an issue more directly related to my concerns in this paper. Assuming that Reflective Access I* and II* don't entail Reflective Access I and II respectively this raises the prospect of the proponent of (RED) modifying the Internalist component of their theory by replacing Reflective Access I and II with Reflective Access I* and II* respectively. This raises the question of whether they would still have a commitment to Metaphysical Disjunctivism in those circumstances.
I'm going to argue in the next and final section for a positive answer to that question. However, one might wonder why the issue is a particularly important one to address and I want to say something about that before moving on. The issue is important because, as I have said, I don't think our alternative reflective access theses do entail our original reflective access theses. Moreover, I think anyone wishing to subscribe to an Access Internalist view would be better off accepting the alternative theses instead of the originals. I have not tried to prove either of those claims here. But if I am correct then the proponent of (RED) should commit themselves to Reflective Access I* and II* instead of to Reflective Access I and II. And if that's so then for anyone who is interested in whether (RED) requires a commitment to Metaphysical Disjunctivism, the issue of whether they would still be so committed were they to modify their Internalism in the way in question is an important one. For anyone who is interested in that issue should surely be interested in the issue of whether the strongest version of (RED) requires a commitment to Metaphysical Disjunctivism, not just the version currently on the market.
5. A CONSEQUENCE FOR (RED)
Let's suppose, then, that the proponent of (RED) decides to accept the alternative reflective access theses instead of the originals. Would that mean that they are no longer committed to Relationalism and its associated Metaphysical Disjunctivism?
On the one hand, one might reasonably think that it does. For the proof that (RED) engenders such metaphysical commitments partly relied on the thought that it is committed to the claim that when one knows that p via perception, one is in a position to know by reflection alone that one is perceiving. But that claim is something the proponent of (RED) is committed to by dint of being committed to either Reflective Access I or II. But now, supposing that they operate with Reflective Access I* and II* instead, don't we get the result that they are no longer committed to saying that one can know by reflection alone that one is perceiving in the good case? And don't we thereby get the result that they are not committed to Relationalism after all?
In fact, a commitment to either Reflective Access I* or Reflective Access II* on the part of the proponent of (RED) still requires that they are committed in turn to saying that when one knows that p in the good case, one is in a position to know by reflection alone that one is perceiving. Thus, even if they reject the original reflective access theses in favour of the alternatives, they'll still end up having to endorse Relationalism and Metaphysical Disjunctivism, by the reasoning of §3. But why do the alternative theses have this result?
I'll start with Reflective Access I*. Let's suppose that one is in the good case and that one knows that p via perception. Thus, by (RED), one believes that p on the basis of the decisive reason that one is perceiving x. Now let's suppose that the proponent of (RED) subscribes to Reflective Access I*. Thus, they are committed to saying that one is in a position, just by reflection, to come to judge that one is perceiving x, where the judgement in question involves consciously treating that one is perceiving x as a reason to believe that p, and the judgement in question is the manifestation of some prior state of awareness of one's perceiving x that one is in.
Now let's ask: what is the prior state of awareness of which one's judgement that one is perceiving x, arrived at by reflection, is a manifestation? The state of prior awareness in question will have to be the state of self-conscious awareness that is constituted by one's perceptual awareness of the mind-independent environment and which, as we've seen, constitutes one's possession of the fact that one is perceiving x as one's reason to believe that p in the first place.
Now let's ask: what is the activity of reflection which constitutes bringing the fact that one is perceiving x to mind? The answer to this is presumably: an act of introspection.
So the situation for the proponent of (RED) who commits themselves to Reflective Access I* is this: they will have to say that when one knows that p via perception, one is in a position, by introspection alone, to come to judge that one is perceiving x in a way that manifests their state of self-conscious awareness. But, given that they are in the good case, that suffices for them to know that they are perceiving x. So the proponent of (RED) who subscribes to Reflective Access I* is committed to claiming that when one is in the good case, one can come to know by reflection alone that one is perceiving x anyway.
The proof that the proponent of (RED) who subscribes to Reflective Access II*, instead of to Reflective Access II, proceeds in much the same way. They'll have to say that the subject is in a position to know by reflection alone that they are perceiving x, for the state of perceiving x is what constitutes their possession of their decisive reason to believe that p. But then, given that bringing that fact to mind by reflection is a matter of judging via introspective attention focused on one's self-conscious state of perceiving, and that that suffices, given that the subject is in the good case, for them to know that they are perceiving x, it follows that the subject must be in a position to know that they are perceiving x when they know that p in the good case, after all.
Whether the proponent of (RED) goes for the alternative or original Reflective Access Theses, then, they'll still end up having to be Relationalists and hence Disjunctivists at the metaphysical level. This is significant because it shows that the commitment to Metaphysical Disjunctivism on the part of (RED) is robust enough to survive a certain modification to the Internalist component of their theory. I think that the modification constitutes a fresh variety of Internalism in epistemology, for Reflective Access I* and II* don't entail Reflective Access I and II respectively. Moreover, I think the modification should be made. But those are matters for further investigation.Footnote 26
