In Search of the Holy Grail of Epistemology

Pritchard calls his epistemological disjunctivism ‘the holy grail of epistemology’. What this metaphor means is that the acceptance of this thesis puts the internalism-externalism debate to an end, thanks to satisfaction of intuitions standing behind both competing views. Simultaneously, Pritchard strongly emphasizes that the endorsement of epistemological disjunctivism does not commit one to metaphysical disjunctivism. In this paper I analyze the formulations and motivations of epistemological disjunctivism presented by Pritchard and McDowell. Then I consider the most common argument for the claim that epistemological disjunctivism can be held without the support of metaphysical disjunctivism. I conclude that the plausibility of epistemological disjunctivism depends on the plausibility of metaphysical disjunctivism. If the latter is false, the former postulates a set of conditions for epistemic justiﬁcation that are impossible to be fulﬁlled.

In this paper I analyze the formulations and motivations of epistemological disjunctivism presented by Pritchard and McDowell. Then I consider the most common argument for the claim that epistemological disjunctivism can be held without the support of metaphysical disjunctivism.
I conclude that the plausibility of epistemological disjunctivism depends on the plausibility of metaphysical disjunctivism. If the latter is false, the former postulates a set of conditions for epistemic justification that are impossible to be fulfilled.
"So it would be odd if knowledge could be had in virtue of having a true belief based on evidence that merely made it more or less probable that p." 1 1 Introduction: epistemic justification and the holy grail of epistemology.
One of few things that everyone in the contemporary internalism-externalism debate regarding epistemic justification seems to agree about is its significance. This debate is commonly seen as metaepistemological, in that it is not only crucial for our understanding of knowledge and truth, but also for establishing what is 'the nature and goals of epistemological theorizing' 2 .
What makes the dispute in question appear impossible to be resolved is the fact that intuitions standing behind the competing views lead to seemingly incompatible conclusions. According to epistemic internalism, justification is fully determined by conditions internal to the subject. Epistemic externalism, on the contrary, claims that justification is dependent on factors external to the subject. Thus, if we compare my present perceptual belief and the corresponding belief of my counterfactual counterpart which is a brain in a vat, the justificatory status of our beliefs will be the same from the internalist perspective, yet different regarding the externalist standpoint. This is because our beliefs are based on subjectively indistinguishable experiences (i.e. our internal states are identical), but I am perceiving whereas my counterpart is hallucinating (i.e. the relevant external factors are different).
The above example, known as The New Evil Demon problem, was primarily developed against externalism 3 . However, rather than making one view more plausible than the other, it simply illustrates what the bone of contention is. Of course, granting the externalist position seems to deprive justification of its subjective accessibility, since the subjective perspectives of mine and my BIV-counterpart are identical. Nevertheless, we certainly do not want to say that beliefs based on perceptual and hallucinatory experiences can be equally justified, as the internalist position seems to suggest. In effect, we find ourselves caught between a rock and a hard place: either the warrant for knowledge delivered by sense experience is strong but inaccessible, or it is accessible but very weak (perhaps too weak).
Although reconciling internalist and externalist intuitions would lead to potentially enormous theoretical benefits, such reconciliation, given the above considerations, seems to be like eating a cake and still having it 4 . This is precisely why Duncan Pritchard describes his attempt to end the internalism-externalism impass -a theory called epistemological disjunctivism -as 'the holy grail of epistemology, in that it is offering a bona fide internalist conception of knowledge which is able to nonetheless allow that the rational support that one's belief enjoys can be genuinely truth-connected and thus skeptic-proof' 5 . What the 'holy grail ' metaphor means here is that the acceptance of epistemological disjunctivism would bring the internalism-externalism debate to an end, thanks to satisfaction of the intuitions standing behind both sides.
The general aim of this paper is to investigate the tenability of epis-temological disjunctivism. I agree with Pritchard that this thesis really deserves the name 'holy grail of epistemology'. However, from the fact that a given standpoint is worth defending it does not necessarily follow that it is defendable. I think that epistemological disjunctivism cannot stand on its own and needs some additional argumentation. In particular, I argue that the plausibility of epistemological disjunctivism depends on the plausibility of another kind of disjunctivism, namely metaphysical disjunctivism. My argument has a form of abduction: there is no better explanation for epistemological disjunctivism than its metaphysical counterpart. Despite the fact that both views are embraced at different levels of philosophical inquiry and endorsing one does not commit one to holding the other, only the truth of both claims puts the holy grail of epistemology in our hands.
In consecutive parts of this essay, I explain what epistemological disjunctivism is and analyze Pritchard's argumentation in favour of this claim (Part 2), compare Pritchard's epistemological disjunctivism with the original formulation of this view presented by John McDowell (Part 3) and finally give a rejoinder to the most common argument for the claim that epistemological disjunctivism can be held without the support of metaphysical disjunctivism (Part 4). I conclude that such support is indispensable (Part 5).

Epistemological disjunctivism according to Pritchard
Pritchard identifies the holy grail of epistemology in the following way: 'Epistemological Disjunctivism: The Core Thesis: In paradigmatic cases of perceptual knowledge an agent, S, has perceptual knowledge that f in virtue of being in possession of rational support, R, for her belief that f which is both factive (i.e., R's obtaining entails f) and reflectively accessible to S. ' 6 Although disjunctivism comes in many varieties, what is common to all of them is the rejection of the so-called Common Factor Principle. This principle, in turn, is a consequence of the well-known argument from hallucination (and an analogous argument from illusion), according to which for every perceptual experience one might have, there is a possibility of having a subjectively (phenomenally) indistinguishable experience which is not a perception, but an illusion or hallucination. The Common Factor Principle says that such illusions and hallucinations are experiences of the same fundamental kind as perceptions from which they are indiscernible 7 . Examples of theories that follow this principle are intentionalism and the causal theory of perception. According to these theories, the difference between perception and hallucination lies not in experiences themselves, but in their etiologies 8 .
Disjunctivism denies the Common Factor Principle by claiming that there is a fundamental difference between perceptual and hallucinatory or illusory experiences, even in cases where they appear identical from the first person perspective 9 . What particular versions of disjunctivism differ about is the fundamental property that makes the difference in question. Now, epistemological disjunctivism is a very peculiar variety of disjunctivism. It sees the discussed difference in the epistemic significance of experience: perception endues the subject with the kind of epistemic justification that hallucination and illusion cannot provide 10 . Does it mean that endorsing epistemological disjunctivism commits one to denying the Common Factor Principle? Presumably, it would be an overstatement to say that the truth of epistemological disjunctivism logically entails the falsity of the Common Factor Principle. This is because epistemological disjunctivism is a theory of epistemic justification, and epistemic justification is a normative notion. It is up to us how we set the conditions a perceptual belief has to fulfill in order to be justified. Thus understood, an account of justification cannot have any bearing on the metaphysics of experience. On the other hand, while being normative indeed, a theory of epistemic justification cannot be formulated in a complete isolation from the relevant issues in the philosophy of perception. Consequently, I think that the truth of the Common Factor Principle makes epistemological disjunctivism unfounded. Under this principle, no perceptual belief can fulfill the high standard for justification required by epistemological disjunctivism. I develop this thought in Part 4.
Anyway, at this point, the Pritchard's definition should become clear: the 'paradigmatic case of perceptual knowledge' is a situation where S enjoys a perceptual experience, and conditions necessary for having perceptual knowledge are met (e.g. S is rational and not biased by delusions, etc.). In such a case, S is being put into a position to have knowledge through gaining a reflectively accessible and externally grounded (factive) warrant for a true perceptual belief.
Pritchard has good reasons to defend this thesis. One of them is the aforementioned prospect of putting an end to the internalism-externalism debate. The other is the riddance from the worry of radical scepticism: 'if epistemological disjunctivism were a viable theory then a potential route out of this problem becomes available, since we can now appeal to reflectively accessible elements of our epistemic standings which entail facts about the world' 11 . However, the above enumeration of epistemological disjunctivism's bright prospects is obviously not enough to justify the endorsement of this view. In fact, Pritchard's convincement of its tenability is not widely shared. The claim that it is possible to reconcile internalist and externalist intuitions is usually taken with disbelief: 'is it really the case that a subject can know through his reflective capacities alone what his reasons are for a belief if those reasons are factive?' 12 'Why suppose that my reflectively accessible evidence must be provided by the factive mental state of seeing that I have hands, rather than the non-factive mental state of seeming to see that I have hands?' 13 .
Indeed, the core thesis of epistemological disjunctivism might seem either trivial or dogmatic. On the one hand, it seems trivial because if a genuine case of perceptual knowledge has ever taken place, its justificatory element definitely was truth-connected and reflectively accessible. On the other hand, it might be taken as dogmatic because it completely ignores the sceptical threat of the aforementioned New Evil Demon scenario. After all, it is not possible to know by reflection alone that one is not a BIV, is it?
As it turns out, according to Pritchard, the impossibility to know that one is not a BIV does not lead to the grim consequences that are usually derived from it. He claims that 'it is far from obvious on closer inspection why possessing better grounds in favour of believing that one scenario obtains rather than another known to be incompatible scenario should entail that one thereby possesses the relevant discriminatory abilities to distinguish between the two scenarios.' 14 This is why he thinks that perceptual beliefs of my BIV-counterpart are blameless, yet not justified 15 . It should be noted that the idea of epistemic blamelessness was criticized by James Pryor, who pointed out that it overlooks the difference between rational and irrational subjects 16 . Nevertheless, the key question is the following: why do we have 'better grounds' for believing that the BIV scenario does not obtain? It is the response to this question that explains why should I think of my BIV-counterpart as epistemically blameless.
The response is this: we have better rational support for our everyday beliefs than for their sceptical alternatives (i.e. epistemological disjunctivism is true) because our commitments to denials of sceptical hypotheses are hinge commitments 17 . In order to substantiate the thesis of epistemological disjunctivism, Pritchard merged it with the idea of hinge commitments. This idea is the highlight of Ludwig Wittgenstein's last book -On Certainty. According to Wittgenstein, '[...] the questions that we raise and our doubts depend upon the fact that some propositions are exempt from doubt, are as it were like hinges on which those turn.' 18 Danièle Moyal-Sharrock, an expert on On Certainty and a proponent of the non-propositional reading of hinge commitments, describes them as indubitable, foundational, nonempirical, grammatical, ineffable and enacted 19 .
Pritchard endorses a similar interpretation of hinge commitments, which he calls 'non-epistemic'. On this reading, just as on the nonpropositional one, hinge commitments cannot be an object of knowledge, since 'there is no sense in which we can meaningfully talk of our hinge commitments as being beliefs' 20 . Contrary to the non-propositional reading, however, the non-epistemic interpretation still counts hinge commitments as propositional attitudes, which explains why we are able to recognize the logical relationships between hinge commitments and ordinary beliefs 21 .
Thus understood hinge commitments, along with epistemological disjunctivism, make a very efficient theoretical tool for blocking both ways radical scepticism might enter the picture, namely its closure-and underdetermination-based formulations. The former is based on the following indispensable principle: (1) Closure Principle: If S knows that p, and S competently deduces from p that q, thereby forming a belief that q on this basis while retaining her knowledge that p, then S knows that q.
From (1) and a seemingly plausible assumptions that (2) One cannot know that one is not a BIV, This generates a critical problem for epistemological disjunctivism. For on the ground of the closure principle, if one has factive and reflectively accessible support for one's everyday beliefs, then one should know that one is not a BIV. In effect, the truth of epistemological disjunctivism depends on the falsity of (2). And to show that (2) is false seems to be impossible.
The underdetermination-based scepticism is an upshot of another principle: (4) The Underdetermination Principle: If S knows that p and q describe incompatible scenarios, and yet S lacks a rational basis for preferring p over q, then S lacks knowledge that p.
Given (4) and a seemingly plausible assumption that (5) One does not have better rational support for one's everyday beliefs than for their sceptical alternatives.
we arrive at the conclusion that (6) One cannot have perceptual knowledge. 23 Because the underdetermination-based scepticism is not committed to (2), it is immune to our hinge commitments to denials of sceptical hypotheses. To put it differently, having such hinge commitments is compatible with not having better rational support for our everyday beliefs than for their sceptical alternatives. However, combining the idea of hinge commitments with epistemological disjunctivism allows Pritchard to solve both sceptical problems at once. Since '[...] all rational evaluation is essentially local, in that it takes place relative to fundamental commitments [hinge commitments -P.Z.] which are themselves immune to rational evaluation, but which need to be in place for a rational evaluation to occur' 24 , the lack of knowledge that one is not a BIV does not prevent one from knowing other things (e.g. that one has hands). The appeal to hinge commitments enables us to deny the inference from (2) to (3). And if the local rational support can be factive, as epistemological disjunctivism tells us, then (5) is false. Consequently, it is not the case that our everyday beliefs and their sceptical alternatives are equally justified 25 . This is merely a very brief outline of an extensive, complicated and, as far as I am concerned, quite convincing argument. Nevertheless, it seems that in order to reach the holy grail, Pritchard made a certain commitment that he did not wish to make -a commitment to metaphysical disjunctivism.
According to metaphysical disjunctivism (most generally understood), the fundamental difference between perception and subjectively indistinguishable hallucination (or illusion) lies at the metaphysical level. Perception is not merely an internal state of the subject -it is the whole relation between the subject and the object perceived. Hallucination, on the contrary, is a mere state of the subject, even if it is phenomenally indiscernible from perception. In other words, the object perceived is a constituent and not just a cause of a perceptual experience. This constituent is not shared by the indistinguishable hallucination 26 .
Of course, endorsing epistemological disjunctivism does not automatically commit one to holding metaphysical disjunctivism, as it would be logically coherent to endorse the former and deny the latter. The reverse configuration would be consistent as well. Therefore, Pritchard is right when he writes 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 illusion and hallucination are radically different does not in itself entail that there is no common metaphysical essence to the experiences of the agent in both cases' 27 .
On the other hand, the core of Pritchard's proposal is in the claim that 'a closure-style inference [...] on the Wittgensteinian account of reasons [...] simply has no application to hinge commitments such as our denials of sceptical hypotheses' 28 . But what is metaphysical disjunctivism if not a denial of sceptical hypotheses? After all, what this idea boils down to is precisely the claim that when I see my hands, I really do see hands (macroscopic physical objects, aggregates of elementary particulars) and not hallucinatory images presented to me by the Evil Demon. If this is right, then Pritchard makes an even stronger claim than the majority of metaphysical disjunctivism's proponents. Under his assumptions, metaphysical disjunctivism is a hinge commitment immune to rational evaluation. Exactly the same commitment is expressed by M.G.F. Martin's famous sentence concerning perceptual experience: '[n]o experience like this, no experience of fundamentally the same kind could have occurred had no appropriate candidate for awareness existed' 29 . It is quite ironic that being so reluctant towards metaphysical disjunctivism, Pritchard eventually lands in the same basket with the most radical proponent of this thesis.
An immediate objection to the above reasoning might be that Wittgenstein was hostile towards metaphysics, and the idea of hinge commitments belongs to the field of logic 30 . This objection does not help Pritchard, however. Firstly, Pritchard's considerations are not a mere exegesis of Wittgenstein's work 31 , so we should not be concerned whether Wittgenstein himself would agree with all this 32 . Secondly, understanding the appeal to hinge commitments as a mere reference to a rule of grammar (i) is clearly in tension with the non-propositional character of hinge commitments and (ii) makes this idea susceptible to the alternative (and radically different) interpretation of hinges defended by Martin Kusch 33 .
Kusch understands Wittgenstein along the following lines: however 'our very nature does not enable us to treat [epistemic relativism] as a live option', because '[it] jars with our natural, naïve attitude to epistemic disagreements', '[a]ll indications are that Wittgenstein [nevertheless] took epistemic cultural relativism to be a live option' 34 . On this interpretation, hinge commitments are culturally relative. Kusch argues for the plurality of radically different epistemic systems. The fact that the foundation of every such system consists of arational hinge commitments is supposed to explain why people tend to be so unwilling to change their epistemic systems 35 .
A particularly unwelcome consequence of Kusch's interpretation is that our hinge commitments to denials of sceptical hypotheses are culturally-relative. This conclusion strikes me as clearly false, for I have never heard of a community believing that the world started to exist five minutes ago, or that their hands are hallucinatory. Even the most sceptically-oriented philosophers have no problem with using their hands in their everyday lives. The disbelief in sceptical scenarios is clearly enacted in our everyday activity.
And yet it is hard to imagine how one could retreat from metaphysical disjunctivism by weakening the interpretation of hinge commitments described above without simultaneously making a concession to the Kusch's reading of On Certainty. As dogmatic as it may sound, the trust in our perceptual capacities entailed by metaphysical disjunctivism endues us with a strong criterion for evaluation of epistemic systems. Without such a criterion, we are doomed to fall into epistemic relativism.

Epistemological disjunctivism according to McDowell
As it was mentioned in Part 1 of this paper, McDowell is the original proponent of epistemological disjunctivism. Despite being inspired by McDowell's writings, Pritchard does not want his Epistemological Disjunctivism to be 'a mouthpiece for a McDowellian epistemology' 36 . He is interested in pursuing his own project, not in the exegesis of McDowell's work.
However, not much exegesis is needed to see that both philosophers share the core thesis of epistemological disjunctivism, and Pritchard is aware of that 37 . I take the following quote to confirm that McDowell is also in quest of the holy grail : 'in a non-defective exercise of [a perceptual] capacity its possessor is self-consciously equipped with an indefeasible, and so knowledge-constituting, warrant for belief about the environment. [...] When a belief owes its status as knowledge to a self-consciously possessed warrant, rationality is at work in the self-conscious possession of the warrant [...]' 38 .
In other words, perceptual experience endues the subject with epistemic justification that is reflectively accessible and externally grounded. Details aside, the key claim is the same.
That said, the way both philosophers go about motivating this claim is significantly different. As we have seen in Part 2, Pritchard is concerned with the kind of scepticism that can be called Cartesian scepticism, in that it feeds on the possibility that we cannot distinguish through reflection alone a situation in which we have perceptual knowledge from a situation in which we do not have such knowledge. This possibility creates a worry that even if there is nothing wrong with our perceptual capacities, there might be no world to be perceived, or this world can be radically different from how it seems to be 39 . McDowell, on the contrary, is worried about something else: '[...] since there can be deceptive cases experientially indistinguishable from non-deceptive cases, one's experiential intake -what one embraces within the scope of one's consciousness -must be the same in both kinds of case. In a deceptive case, one's experiential intake must ex hypothesi fall short of the fact itself, in the sense of being consistent with there being no such fact. So that must be true, according to the argument, in a non-deceptive case too' 40 .
What this quote describes is a different kind of scepticism, namely Humean scepticism. It is generated by theories of perception formulated in alignment with the aforementioned Common Factor Principle. If the nature of a hallucinatory experience is exactly the same as the nature of a perceptual experience, as this principle implies, then a perceptual experience is not capable to get the subject in touch with the mind-independent reality. In other words, the Common Factor Principle makes the difference between perception and hallucination irrespective of the nature of these experiences, and it does it at the expense of stripping perception of its capability to relate the subject with the world 41 .
McDowell denies the Common Factor Principle by endorsing the thesis of epistemological disjunctivism. He explicates this thesis in the following way: all of our sense experiences belong to one genus of seemings. However, seemings have two species: seeings (perceptions) and merely seemings (illusions, hallucinations). The difference between the species is that only seeings have the capacity to provide the subject with an indefeasible and consciously accessible warrant for a perceptual belief, and i.e. to put the subject into a position to gain knowledge about the world. As maintained by McDowell, this is a necessary assumption, the rejection of which would 'threaten the very idea of perceptual knowledge' 42 . Its adoption, on the contrary, 'is not well cast as an answer to sceptical challenges, it is more like a justification of a refusal to bother with them' 43 .
Being aware of all that 44 , Pritchard still complains that 'McDowell himself offers so little argumentative support for [epistemological disjunctivism -P.Z.]' 45 . I think that this conciseness of McDowell's argumentation might be explained by drawing an analogy between him and Martin. After all, Martin is also concerned with the Humean scepticism 46 . If such analogy holds, then McDowell's claim that '[t]here is no making sense of perceptual appearances -the testimony of one's senses -without making sense of the possibility that the objective world can be immediately present to the senses' 47 is also indicative of the hinge commitment to metaphysical disjunctivism. Of course, such hinge commitment would refer to the general idea of metaphysical disjunctivism introduced above, not to specific formulation of this thesis defended by Martin. This is because McDowell accepts, while Martin denies the possibility that a phenomenal character of perception and hallucination might be identical 48 .
Interpreting McDowell as a metaphysical disjunctivist, though perhaps controversial to some, is still a valid option 49 . Insofar as it is not my primary aim here to offer an exegesis of McDowell's work, the following reading of his standpoint seems perfectly natural to me: a certain species of experience (namely seeings) is capable of making the objective world open to us while other species (merely seemings) is not. Seeings endue us with the special kind of epistemic warrant (epistemological disjunctivism) precisely because they are capable of making the world open to us (metaphysical disjunctivism).

Epistemological disjunctivism and the causal theory of perception
Pritchard decidedly emphasizes that embracing epistemological disjunctivism does not commit him to metaphysical disjunctivism 50 . He refers to a group of authors 51 who argue that epistemological disjunctivism is compatible with the causal theory of perception -a theory that stands in contradiction with metaphysical disjunctivism. Indeed, both theories radically differ with regard to the Common Factor Principle: the causal theory assumes it, while metaphysical disjunctivism denies it. Under the former, one and the same experience might be a perception or a hallucination, depending on what is the causal chain that has led to it. According to the latter, the difference between perception and hallucination lies in the very essence of experience 52 . Nevertheless, from the authors cited by Pritchard, only Alan Millar provides a well-founded argument for the claim that epistemological disjunctivism is compatible with the causal theory of perception. Other authors -Alex Byrne and Heather Logue, Paul Snowdon -support this claim merely by indicating that McDowell is not a metaphysical disjunctivist 53 .
In the present context, however, it would be pointless to debate over what McDowell himself really thinks. As a matter of fact, any argument from exegesis would be irrelevant here, since the question is not what soand-so has really meant by this-and-that. Rather, the question is: can we make the epistemological disjunctivism a viable option? If giving a positive response to this question requires endorsing a view that certain philosophers would reject, so be it.
Much more attention has to be given to the argument proposed by Millar. He admits that '[t]he disjunctivist is right to think that experiences under the traditional conception [the causal theory of perception -P.Z.] do not have an intrinsic character that will suffice to explain how perception makes demonstrative thought, judgment and knowledge possible' 54 . Nonetheless, he thinks that we still can account for all these things without rejecting the causal theory of perception, provided that we 'abandon the idea that experiences can bear the explanatory burden that disjunctivists have placed on them' 55 .
Millar prefers to put this burden on the so-called perceptual-recognitional capacities. Getting in 'cognitive contact' with the facts about the world 'is a matter of being in a relational state in virtue of the exercise of those capacities' 56 . Such state is relational because a perceptualrecognitional capacity counts as exercised only if knowledge is gained 57 .
An exercise of such capacity is a successful (accurate) conceptualization of a phenomenal material that experience provides the subject with 58 . In effect, the term 'perception' is broader than the term 'experience', because 'perception' refers to experience plus the exercise of a relevant perceptual-recognitional capacity 59 . This, in turn, allows Millar to accept the Common Factor Principle and simultaneously maintain that perception is essentially relational. If this conception is viable, the endorsement of metaphysical disjunctivism is not the only method to reach its own goals.
Therefore, Millar challenges the disjunctivist 'to spell out what exactly is missing from an account of perception that explains seeing, for instance, just in terms of having a suitably caused visual experience.' 60 A similar objection is raised by Snowdon: '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?' 61 .
The response is this: under the causal theory of perception, perceptual experience cannot meet the justificatory standards, the fulfillment of which puts the holy grail of epistemology in our hands. The theoretical construction merging epistemological disjunctivism with the causal theory of perception shoves our understanding of epistemic justification from the golden mean to the externalist side of the debate, sustains the internalism-externalism chasm and hides the holy grail from us.
On the causal theory of perception, the difference between the aforementioned two subjectively indistinguishable experiences does not concern their intrinsic natures, but causal chains that have led to them. At the beginning of the causal chain that has led to a perceptual experience there is a fact about the world or a mind-independent object, whereas at the start of the causal chain that has led to a hallucination there is no such fact or object.
It can be agreed that perceptual experience thus understood can be a source of a factive epistemic justification. However, it is unlikely, to say the least, that the mentioned fact (or the object) at the beginning of the causal chain is reflectively accessible to the subject. What is in the range of my conscious access is whatever counts as my experience, and if the fact (or the object) is not present in it, then the epistemic justification constituted by my perceptual relation with this fact or object is not reflectively accessible to me.
As I understand epistemological disjunctivism, it is the very factivity of justification that is supposed to be reflectively accessible. Otherwise this view would be just another kind of reliabilism. Such accessibility is missing from the picture if we accept the Common Factor Principle. What this principle generates is the so-called screening-off problem. It starts from the assumption that a phenomenal character of a perception P supervenes on a certain brain activity B. If a hallucination H has the same phenomenal character as the perception P, then it supervenes on the same brain activity B. So whatever the difference is between P and H, be it the epistemic significance or the relation with a mind-independent object, it gets screened-off by the phenomenal character that is shared by both experiences 62 .
The aforementioned Millar's theory, given its commitment to the Common Factor Principle, is also susceptible to the screening-off problem. In spite of explaining the factivity, it does not fulfill the requirement of reflective accessibility. In fact, Millar explicitly subscribes himself to a certain sort of reliabilism 63 . This is why I argue that the only way to get the holy grail is to ground epistemological disjunctivism in its metaphysical counterpart. It is true that both claims do not imply each other. Nonetheless, only the assumption that an object perceived is a constituent of a perceptual experience can plausibly motivate the claim that perception endues the subject with an epistemic justification that is externally grounded and reflectively accessible at the same time. Perceptual knowledge is a kind of knowledge by acquaintance and it is only the direct acquaintance with a fact (or an object) that can serve as a source of an epistemic justification for such knowledge.
Admittedly, as it was mentioned in Part 1, this argument is based on an abductive reasoning, and as such it could be falsified by showing an alternative validation for epistemological disjunctivism at the level of the philosophy of perception. This is doubtful, however, given that the Common Factor Principle is accepted by all theories of perception other than metaphysical disjunctivism 64 .
One thing that we should learn from Millar though is that the truth of metaphysical disjunctivism is an insufficient condition for the possibility of having a perceptual belief that counts as justified by the standard of epistemological disjunctivism. This is because even an object-involving experience has to be accurately conceptualized in order to become a proper material for justification 65 . Being insufficient, the truth of metaphysical disjunctivism still is a necessary condition.
Of course, making metaphysical disjunctivism a viable theory is another problem and it is not at all obvious that this view will endure the tension of its numerous critics. In particular, a metaphysical disjunctivist has to deliver a convincing explanation for the possibility of indistinguishable hallucinations mentioned above. However, I think that there are plausible stories to tell here, so the holy grail of epistemology is not completely out of our reach.

Conclusion
By subscribing to epistemological disjunctivism, one sets a certain standard for epistemic justification. Put differently, one endorses a certain set of conditions that must be fulfilled in order to consider a given perceptual belief as justified. An endorsement of such conditions obliges one to show whether and how can they be fulfilled. This requires one to embrace a certain standpoint in the field of philosophy of perception.
In the case of epistemological disjunctivism, understood as described above, the possibility of fulfillment of its conditions for perceptual justification requires the truth of disjunctivism about perceptual experience (a.k.a. metaphysical disjunctivism). This is why I claim that the plausibility of epistemological disjunctivism depends on the plausibility of metaphysical disjunctivism. If the latter is false, the former will be postulating a set of conditions for justification that are impossible to be fulfilled.