9

Experience and Introspection

Fabian Dorsch

Abstract

One central fact about hallucinations is that they may be subjectively indistinguishable from perceptions. Indeed, it has been argued that the hallucinatory experiences concerned cannot—and need not—be characterized in any more positive general terms. This epistemic conception of hallucinations has been advocated as the best choice for proponents of experiential (or “naive realist”) disjunctivism—the view that perceptions and hallucinations differ essentially in their introspectible subjective characters. In this chapter, I aim to formulate and defend an intentional alternative to experiential disjunctivism called experiential intentionalism. This view not only enjoys some advantages over its rival but is also compatible with the epistemic conception of hallucinations, as well as with the disjunctivist view that perceptions and hallucinations differ essentially in their third-personal structures (e.g., their causal, informational, or reason-providing links to reality). It also maintains that there are actually two aspects to the subjective indistinguishability of mental episodes: (i) we cannot distinguish their first-personal characters in introspective awareness; and (ii) we cannot distinguish their third-personal structures in experiential awareness—that is, in how they are given to consciousness. While experiential disjunctivism makes the mistake of ignoring (ii) and reducing subjective indiscriminability to (i), experiential intentionalism correctly identifies (ii) as the primary source of the subjective indistinguishability of perception-like hallucinations. Accordingly, the intentional error involved in such hallucinations is due to the fact that we consciously experience them as possessing a relational structure.

 

If I stand here, I saw him.

—William Shakespeare, The Tragedy of Macbeth, act III, scene 4

I   The Epistemic Conception of Hallucinations

1. One of the main issues in the recent debates about the nature of perception is whether it should be understood in relational or in intentional terms. While relationalist positions—such as experiential disjunctivism—are arguably more promising with respect to the elucidation of the phenomenology of perceptions and their close links to knowledge and demonstrative thought, intentionalist views are plausibly better equipped to accommodate the contentfulness of perceptions and their rational force. The two approaches compete with each other, however, only if they are concerned with the same aspect of the nature of perceptions. And, contrary to the received opinion, this need not necessarily be so. In particular, while the third-personally accessible structure of perceptions may be taken to be relational, their first-personally accessible character may be thought of as intentional. Such a view—which I aim to develop and defend here—can perhaps combine the strengths of the two more austere alternatives while avoiding the weaknesses of either.1

The contrast between relationalist and intentionalist views is most evident in their treatment of perception-like hallucinations. Accordingly, I begin with spelling out in more detail in which sense hallucinations may be subjectively indistinguishable from perceptions, and why this leads us to erroneously judge them to be perceptions (secs. IIII and VIII). Then I raise three challenges each for experiential disjunctivism and its orthodox intentionalist counterparts (secs. IV and V), notably in respect to the need to explicate why a perception-like hallucination still makes the same judgments reasonable for the subject as the corresponding perceptions. Finally, I propose my alternative both to experiential disjunctivism and to orthodox intentionalism. Experiential intentionalism takes perceptions and perception-like hallucinations to share a common character partly to be spelled out in intentional—and hence normative—terms (secs. VI and VII). The central thought is that the hallucinations concerned are intentionally—and erroneously—presented to us as perceptual relations to the world. I aim to show that the resulting view can meet all six challenges (secs. VIVIII). I end with some comments on the consequences for the nature of perceptual experiences, and on the possibility of combining experiential intentionalism with relationalism about the structure or perceptions (sec. IX).

2. Philosophy has a long and fruitful tradition of trying to clear up the nature of perceptions by shifting one’s attention to the nature of hallucinations. In particular, the epistemic conception of hallucinations helps us better to understand what is distinctive of perceptions and, more generally, of perceptual experiences.2 Hallucinations differ from perceptions in that they do not put us into contact with the world in such a way as to enable us to refer to mind-independent objects and acquire knowledge about them. In addition, certain hallucinations are special insofar as they are, in a significant sense, subjectively indiscriminable from perceptions. According to the epistemic conception of hallucinations, these two claims—one being negative and metaphysical, the other positive and epistemic—capture all that can and need be said about what these perception-like hallucinations have in common. In other words, the main tenet of this conception of such hallucinations is that their common makeup—if they share any at all—can be positively characterized only in epistemic, but not in metaphysical, terms.

Proponents of the epistemic conception typically put forward an even stronger claim, namely, that there is nothing more to having a perception-like hallucination than having an experience that is indistinguishable from that of perceiving. That is, the positive characterization of such hallucinations is taken to be exhausted by reference to their indiscriminability from perceptions. This conclusion is stronger in that it concerns not only the issue of what all perception-like hallucinations have in common but also the issue of whether there is something that differentiates them. The claim that their common makeup cannot be positively characterized in nonepistemological terms does not imply that their different individual makeups do not allow for such a description. Objects that share nothing but the feature of being perceptually indiscriminable from lemons, without being lemons, can still differ greatly in their natures. The epistemic conception (as introduced earlier) is compatible with something similar being true of perception-like hallucinations: while they do not share among each other anything but their subjective indiscriminability from perceptions, they may still have distinct individual natures that can be positively characterized in metaphysical terms (e.g., in terms of their causes or neuronal bases). By contrast, an endorsement of the stronger claim rules out this possibility: the hallucinatory experiences at issue do not possess any other positively describable feature over and above their subjective indistinguishability.3

3. The subjective (or first-personal) indiscriminability referred to by the epistemic conception is understood as indiscriminability from the inside, that is, indiscriminability relative to some form of access other than outer perception, testimony, or inference on the basis of either. There is a different and wider sense in which all kinds of access may be said to be “subjective,” given that they all involve a knowing subject with a particular perspective on what is known. But what matters for the distinction between perception-like and other hallucinations is the narrower notion of subjective indistinguishability defined in terms of nonperceptual and nontestimonial access. It becomes clear later on that this includes not only introspection but also experiential awareness—the kind of awareness that comes with conscious mental episodes and their possession of a subjective character.4

A closely related issue is which aspects of mental episodes are accessible from the inside. The object of our knowledge from the inside is the subjective (or phenomenal) character of mental episodes—that is, what the episodes are like from our conscious perspective. So whichever aspects of episodes are accessible from the inside, they have to be intimately linked to their character. The closest link possible is that of constitution. If it is assumed, for instance, that the character of perceptions is partly constituted by their relational connection to objects in the world, it follows that we have access from the inside to their relationality. But the same conclusion may be available even if the link between character and relationality is understood as something weaker than constitution—namely, as an intentional connection. As I aim to illustrate in this chapter (see esp. §47), an account along these lines can hold on to the idea that the relationality of perceptions is accessible from the inside, despite being a constituent of their third-personal structure (e.g., in the shape of their causal origin, representationality, functional role, or reason-giving power), and not of their first-personal character. The kind of access to their structure in question counts as access from the inside insofar as the intentionality involved is not perceptual or testimonial. But the accessed structure is still third-personal insofar as our canonical access to it is from the outside (e.g., a matter of empirical or metaphysical investigations).

4. These considerations have the consequence that the epistemic conception (in the sense introduced in the two previous sections) may be understood in two ways. On the one hand, it may be taken to maintain that we cannot positively characterize perception-like hallucinations in terms of a common character that they share with each other and possibly also with perceptions (Martin, 2004, 2006). On the other hand, it may instead be understood as claiming the impossibility of positively describing perception-like hallucinations by reference to a structure common to all of them, as well as perhaps also to perceptions. The main underlying issue is thereby whether the relationality of perceptions should be understood as an aspect of their character or an aspect of their structure.5 My conclusion toward the end of this chapter is that we should prefer the second way of specifying the epistemic conception of perception-like hallucinations over the first (see esp. sec. VIII).

One important motivation for the epistemic conception of hallucinations—independently of whether it is concerned with the character or the structure of perceptual experiences—is the observation that, while there is only one way in which perception can go right, there are many ways in which it can go wrong. Perception goes right when it relates us to the external world in the way just mentioned. In all other cases, it goes wrong, and it may fail to establish the required relation on different occasions for very different reasons. This observation allows us to characterize perceptions—that is, the perceptual experiences involved in successful cognition—in positive metaphysical terms. But it also suggests that we may be unable to provide more than a merely negative metaphysical description of what unifies defective perceptual experiences, given that they may vary significantly in why they fail to relate us to the world. We may therefore be able to further categorize perception-like hallucinations only by reference to the extent to which we can subjectively tell them apart from perceptions and other mental episodes. In section IV (esp. §§18–20), I say a bit more about the motivation to treat the subjective indiscriminability from perceptions as the mark of being a perceptual experience (see also Martin, 2006). There is still the possibility of perceptions and hallucinations sharing some other, nonepistemic feature—thus permitting a positive nonepistemic characterization of hallucinations, after all. But, as should become clearer later on, it is not easy to support the postulation of such a feature.

5. Experiential disjunctivism about perceptual experiences endorses the epistemic conception of hallucinations and is traditionally linked to this conception.6 The core claim of this version of disjunctivism is that the essence of perceptions includes—and presumably is also exhausted by—their introspectible property of bringing us into conscious contact with mind-independent entities. That is, perceptions are essentially instances of relational awareness or acquaintance. As a consequence, the view maintains that perceptions and hallucinations differ completely in their essences, given that hallucinations lack this kind of relationality. It also claims that although we have introspective access to these different essences, we need not always be able to tell them apart in introspection. Together with the observation about the variety of ways in which perceptual experience may go wrong, this leads naturally to a merely epistemic characterization of perception-like hallucinations: they are introspectively indiscriminable from perceptions, while lacking the link to the world that is distinctive of perceptions.

Many of the central elements of experiential disjunctivism have been well argued for (see, e.g., Martin, 2000a, 2002; Nudds, this volume; Dorsch, 2010b). Notably, the following three insights should not be readily given up: (i) perceptions, but not hallucinations, are essentially relational; (ii) the relationality of perceptions is accessible in introspection; and (iii) perception-like hallucinations may differ in their natures from each other and need not have more than their subjective indiscriminability from perceptions in common. Instead, I argue that experiential disjunctivism is problematic for other reasons. In particular, it has difficulties accounting for the nature of our error involved in taking our perception-like hallucinations to be perceptions.

Moreover, I aim to show that an alternative account of perceptual experiences can avoid the problems for experiential disjunctivism while still holding on to—or at least being compatible with—the three central claims just identified. The view in question does not fall victim to the problems because it takes perceptual experiences to be intentional. And, as illustrated toward the end of the chapter, it can accommodate the three insights by understanding the subjective indiscriminability of hallucinations from perceptions primarily in terms of experiential, rather than introspective, awareness. Accordingly, what I aim to put forward and defend is an intentionalist account of perceptual experiences that combines well both with (nonexperiential) disjunctivism about these experiences and with the epistemic conception of hallucinations. My underlying suggestion is that the error in mistaking hallucinations for perceptions should be located already at the level of experiential awareness.

6. Apart from the nature of the subjective indiscriminability of hallucinations and the distinction between introspective and experiential awareness, two other important themes—which have sometimes been neglected in the recent literature on the nature of perceptual experiences—figure prominently in what follows. One is the idea that—following the phenomenological tradition in which much of the talk of intentional phenomena originated—intentionality should be understood as a normative aspect of consciousness (Dorsch & Soldati, 2010). The other central theme is the importance for any adequate account of perceptual experiences of a satisfactory theory of our access from the inside to our own mental episodes.7 Our subjective access is our canonical way of becoming aware of, and acquiring knowledge about, our conscious experiences. It informs our ordinary opinions about them. And it enables us to notice similarities and differences among their conscious characters. Given that conscious experiences should—precisely because of their conscious status—be characterized in terms of how they are given to us in consciousness, any theory of them has to investigate our first-personal access to them.

7. As already mentioned, the epistemic conception of hallucinations refers to two important facts about hallucinations and their relationship to perceptions. The first is that the two types of experience differ from each other in their relationship to the world: perceptions relate us in a certain manner to some particular mind-independent objects and their features, while hallucinations do not. Accordingly, there is a distinctively perceptual way of being related to the world that is not realized when we are hallucinating. It may not be easy to specify the precise nature of this perceptual link. But for the current purposes, it suffices to note that it exists, and that it differentiates perceptions from hallucinations. In particular, it explains the fact that perceptions—but not hallucinations—inform us about, and enable us to demonstratively refer to, objects in our environment. The second important fact is that there can be hallucinations that are subjectively indiscriminable from perceptions. To understand better what this amounts to, it is perhaps helpful to say a bit more about the general idea of indiscriminability.8

II   The Subjective Indiscriminability of Hallucinations

8. Indiscriminability is an epistemic phenomenon. That two distinct entities—whether they are objects, events, properties, and so on—are indiscriminable means that they cannot be told apart, that is, known to be distinct. Claims about indiscriminability differ in generality relative to the extent to which they put limitations on relevant contextual features, such as the subjects, times, and sources of knowledge concerned. Inuit can tell apart more kinds of snow and ice than most of us. We may be able to visually recognize differences in shape or color if looking at the respective objects from a close range, but not if looking at them from a great distance. We may be better in discriminating certain differences in shape by vision than by touch (or vice versa). And sometimes we may able to discriminate two entities only by comparing each of them to a third entity—for instance, in cases where discriminability turns out to fail to be transitive.

In addition, indiscriminability claims may vary in scope. Most basically, indiscriminability is a relation between two distinct entities. At this basic level, it is also arguably symmetric: if one thing is indiscriminable from another, the latter is also indiscriminable from the former.9

But indiscriminability claims may concern more than two entities. They may proclaim the indiscriminability of each possible pair of entities belonging to a certain group (e.g., “all people from that country look the same to me”). Or they may assert that a certain entity is indiscriminable from each member of a certain group (e.g., “I cannot tell from his looks whether he belongs to that community”). In both cases, indiscriminability turns out to be a relation between more than two entities. And in the second example, it stops being symmetrical in any meaningful sense, for it is not necessary that any member of the comparison group is itself indiscriminable from anything other than the entity originally compared with the group. In particular, the members of the group need not be indiscriminable from each other.

9. It may help to consider a concrete example to get clearer about the possibility of hallucinations being indiscriminable from perceptions. It is normally—and perhaps even always—possible to come to know that one is hallucinating, and not perceiving. Macbeth, for example, need not have been so convinced that the perceptual character of his vision of Banquo sitting at the table was as obvious as his own presence in the room. He might just have listened properly to what the other lords present at the banquet told him. He might have inferred the hallucinatory character of his experience from his previously acquired knowledge of the murder of Banquo. Or in a more contemporary setting, cognitive scientists might have informed him that they subjected him to a treatment meant to induce guilt-related hallucinations.

However, if any such relevant evidence coming from the outside—that is, delivered by outer perception, testimony, inference, or any combination thereof—is lacking, it can be impossible for the subject concerned to notice the hallucinatory character of a current experience. This is well illustrated by the initial reaction of Macbeth, during which he seemed to be ignoring the lords’ assertions and to have forgotten about the murder of Banquo. More clearly, if he would have been uninformed about the absence or death of Banquo—that is, if he would have had no perceptual, testimonial, or related inferential evidence available to him suggesting that he was hallucinating, rather than perceiving—he might have been unable to tell that his experience was indeed hallucinatory.

In addition, perhaps no other human being in his position would have been able to come to know about the hallucinatory character of the experience. The fact that Macbeth might have been unable to discriminate his hallucination from comparable perceptions in the absence of perceptual or testimonial evidence to their distinctness need not have been due to features that distinguish him from other actual or possible human beings—such as his general cognitive or moral shortcomings, or his particular situation (e.g., the stressful guilt and anxiety that he was suffering, or the specific spatial point of view that he occupied in the banquet hall).

Finally, what is at issue is not simply whether Macbeth could have distinguished his hallucination from one or more of his other actual experiences, such as his particular perceptions of Banquo that he had had at some time before the banquet, or underwent at the same time as his hallucinatory experience, or could have experienced at some time after the banquet (assuming that Banquo would then still be alive). Macbeth could have had the very same hallucination and been unable to identify it as such, even if he would have never encountered and seen Banquo in his whole life. It would just have been for him an experience of some unknown lord.

The indiscriminability under consideration is therefore not merely a relation between the hallucination and one or more actual perceptions. What matter as well are possible perceptions of Banquo. In particular, that Macbeth—or anyone else in his position—could not have told apart his hallucination from a perception implies that, from the inside, he could not have noticed a difference if he would instead have perceived Banquo. Similarly, it also implies that he could not have noticed a difference if he would have seen Banquo at some point or another before starting to hallucinate him. That is, the indiscriminability claim about Macbeth’s hallucination compares it with the members of a larger class of actual and possible perceptions of Banquo. It is in this—and only in this—sense that, for all that Macbeth knew from the inside, his experience could have been a perception.

Not all possible perceptions are relevant, however. Macbeth would presumably have been able to distinguish his hallucination from a possible perception of his wife or, for that matter, from a possible perception of Banquo at a different location, or under a different spatial perspective, or under different lighting conditions. But this fact need not undermine the indiscriminability claim about his hallucination. The reason for this is that Macbeth would still have been able, from the inside, to tell these possible perceptions apart from those other possible perceptions just mentioned, which he could not have distinguished from his hallucination. Hence what should be said about Macbeth’s hallucination is, more precisely, that it is individually indiscriminable only from each member of a certain class of possible perceptions, which are themselves mutually indiscriminable from each other.10

10. From these considerations about Macbeth’s hallucination and its indiscriminability from certain perceptions, we can glean a more precise characterization of the subjective indistinguishability from perceptions, the possibility of which I took to be the second important fact about hallucinations. Let “us” include all possible subjects that are just like human beings as they actually are; and let a class of “corresponding” perceptions be a class of perceptions that, from the inside, are mutually indiscriminable from each other. A given hallucination is then subjectively indiscriminable from perceptions when it satisfies the following condition:

(S)  None of us could, from the inside, come to know it to be distinct from each member of some class of possible corresponding perceptions.11

The subjective indiscriminability of hallucinations is sometimes spelled out in terms of the fact that we cannot, from the inside, come to know that it does not instantiate the property of being a perception (see, e.g., Siegel, 2008). One problem with this formulation is that it remains unclear what the relata of the relation of indiscriminability are supposed to be. For instance, they cannot be the general properties of being a hallucination and of being a perception, given that even Macbeth can distinguish instances of the two. And referring instead to the more concrete properties of being a perception or hallucination with a certain content is problematic precisely because of the unclear and controversial notion of a content of perceptual experiences. It therefore seems more reasonable to take this ignorance about property instantiation to derive from the more fundamental subjective indiscriminability, as it is specified by means of (S). This also fits much better with Martin’s (2006) insistence that we understand (S) as having a plural form. Besides, the formulation proposed here stays neutral on which aspects of experiences can be introspected. It presupposes only that we can introspectively note similarities and differences among such episodes.

11. The second feature of hallucinations—their subjective indistinguishability from perceptions—gives rise to a third important fact about them and their relationship to perceptions: the priority of perceptions over hallucinations. In general, when two of our mental episodes are subjectively indistinguishable, we treat them in the same way. In particular, we take them to possess the same features and to belong to the same mental kind, and we rely on them to the same extent when forming beliefs or intentions. This is precisely what happens in the case of indistinguishable hallucinations and perceptions. We take both to be perceptions and to relate us to the mind-independent world. We endorse both in the shape of perceptual judgments and corresponding actions. And on the basis of introspection, we judge both to be perceptions. In short, we treat both as if they were perceptions—and not as if they were hallucinations. It is in this sense that perceptions enjoy priority over their indistinguishable hallucinatory counterparts. And this fact becomes manifest in the formal structure of the indiscriminability relation concerned: while some particular hallucinations are subjectively indiscriminable from all members of a group of perception, no particular perception is subjectively indiscriminable from each member of a class of hallucinations.

12. The priority of perceptions has the consequence that we end up being doubly misled when hallucinating, that is, misled not only about the world but also about the hallucinations themselves. When Macbeth is hallucinating Banquo sitting at the table and does not suspect the hallucinatory character of his experience, it is rational for him to come to believe that there is such a scene before him. But the resulting belief is surely false: Macbeth is misled about how a certain part of the world is. Similarly, when he cannot tell apart his hallucinatory experience from perceptions, it is rational for him to come to believe that he is perceiving—and not, say, merely hallucinating or visualizing—Banquo at the table. Again the resulting belief is false: this time, Macbeth is misled about how a certain part of his mind is. Thus hallucinations that are subjectively indistinguishable from perceptions may lead us to form rational but erroneous judgments or beliefs about their experiential type. And this fact is due to the priority of the perceptions over such hallucinations.

The first kind of error has already been discussed in much detail in the literature. The challenge is rather to identify the best account of the nature and source of the second kind of error linked to hallucinating. What is clear so far is that it becomes manifest in false self-ascriptions of the form “I am now perceiving.” But what still needs to be determined is what kind of judgments or beliefs give rise to these self-ascriptions, and at which stage in the epistemic process of their formation the error occurs first.

III   The Introspective Indiscriminability of Hallucinations

13. So far, our access from the inside to our perceptual experiences has been characterized in purely negative terms, namely, as a form of access different from outer perception, testimony, inference based on either, or any combination thereof. A natural way of being more positive is to identify the kind of access mentioned in (S) with introspective access. We can introspectively distinguish sensory experiences only by distinguishing some of their introspectible features (just as we can visually tell objects apart solely by recognizing a difference in their visible properties). Since the object of our knowledge from the inside is the subjective or conscious character of mental episodes, the introspectible properties of experiences—which constitute their introspective “appearance”—are all aspects of their character. Hence we can discriminate experiences introspectively only if they possess distinct characters, and only by introspectively discriminating their characters. And we can introspectively discriminate distinct characters only by recognizing a difference in one or more of the determinables that they realize. Accordingly, a given hallucination is taken to be subjectively indiscriminable from perceptions in the sense of satisfying the following specification of (S):

(I)  None of us could introspectively know its character to be distinct from the character shared by each member of some class of corresponding perceptions.

14. Since it becomes important later on to distinguish introspective indiscriminability from another aspect of subjective indistinguishability, namely, experiential indiscriminability, it is necessary to say more about the nature of introspection assumed here. In the context of (I), introspection is meant to include more than the mechanisms and products involved in the noninferential formation of judgments of the form “I am Φ-ing X, or that p,” where Φ denotes some type of mental episode or state. Indeed, any form of access from the inside to the character of mental episodes counts as introspective, as long as our access is distinct from what is accessed. When we perceive something, our perceptual access to it and the resulting episode of perceiving are distinct from what is perceived. Similarly, when we introspect a given mental episode, our introspective access to it and any resulting episodes (such as higher-order thoughts) are distinct from the episode. Introspecting an episode is therefore distinct from its subject matter—that is, from being in, or having, that episode. But otherwise the exclusion of access from the outside is the only further condition on introspection. Hence any form of reflection on the character of our mental episodes, which is not based on outer perception or testimony, counts as introspective. For instance, introspection might still involve some kind of “inner perceptions” (i.e., nonintellectual higher-order episodes).

Although more needs to be said about this issue, the idea that experiences and other mental episodes possess introspectible properties—that is, properties that we can come to know to be present by means of introspection—is not necessarily incompatible with the idea that we come to know about the presence of such properties by attending to the experienced external objects and features (Martin, 2000b). It is plausible to treat experiences not as objects of our focal attention but as determinations of our conscious focal attention to such objects. Furthermore, what kind of attentive awareness experiences constitute is partly determined by what external objects and features they present us with. Hence acquiring knowledge about experiences may require attending to these objects and features. The intentionalist view that I put forward in the second half of this chapter takes exactly this line of response.

15. Understanding subjective indiscriminability in terms of the introspection of characters permits a simple and natural explanation of why certain perceptions are mutually indiscriminable from each other in introspection and therefore form a class of corresponding perceptions: they do so because they possess the same character and character determinables. The question is now whether the same, or a different, account should be given of why (I) is true of certain hallucinations. Conjunctivism about character claims that those hallucinations satisfy (I) because they possess the same character as—that is, share all character determinables with—the corresponding perceptions. Disjunctivism about character, on the other hand, maintains that the hallucinations do not share any of their character determinables with the perceptions and therefore differ in character from the perceptions, but this difference is for us inaccessible through introspection.12

16. Introspective indiscriminability and disjunctivism about character are typically spelled out in slightly different terms from (I). Instead of maintaining that the hallucinations are indistinguishable from perceptions relative to their character (i.e., that the character of the hallucinations is indistinguishable from that of the perceptions), it is claimed that the indistinguishability of the hallucinations from perceptions is a constituent of their character (Martin, 2004, 2006; Siegel, 2008; Sturgeon, 2008). Accordingly, perceptions and hallucinations are taken to share a character determinable, namely, their subjective indiscriminability from perceptions. But this alternative characterization still presupposes that the hallucinations differ in character from the perceptions, and this difference cannot be noticed by us from the inside. Hence it comes with an endorsement of (I), too. Indeed, this should be expected, since being indiscriminable from the inside just means being indiscriminable relative to those features accessible from the inside—that is, relative to the subjective character.

The main reason for adopting the different characterization is that it can provide an account of the character of perception-like hallucinations solely by reference to their subjective indiscriminability from corresponding perceptions. Assuming (contrary to Fish, 2009) that hallucinations do possess a character, the claim that it is indistinguishable from that of perceptions, but does not share any aspects with perceptions, leaves its positive identity completely open. This issue becomes particularly pressing in the case of causally matching hallucinations—that is, hallucinations that satisfy (S) because they have exactly the same proximal causes as the corresponding perceptions. The positive aspects of character of such hallucinations cannot be due to their proximal causes, since then the character of the perceptions would involve these aspects as well—which would contradict the claim put forward. Hence the character of the hallucinations should be understood as being a matter of distal causes or certain contextual features. However, it is not clear whether, for instance, the absence of a perceived object among the causes can constitute part of the character of hallucinations. And there do not seem to be other obvious candidates for the role of relevant factor (Martin, 2004, 2006; Nudds, this volume).

Identifying the character of perception-like hallucinations with their property of being subjectively indistinguishable from corresponding perceptions—and with nothing else—solves this problem by providing a positive characterization of the character of those hallucinations. Moreover, it leads to the strengthening of the epistemic conception introduced at the beginning (see sec. 2), since it denies that the character of perception-like hallucinations involves other aspects than their subjective indistinguishability from perceptions.

But now the difficulty is that this character turns out to lack determinacy. According to disjunctivism about character, mental episodes can be subjectively indiscriminable from perceptions in at least two ways: by being a perception, or by being a perception-like hallucination. The view can assume that, in the case of perceptions, this determinable aspect of character is realized by some more determinate aspect (e.g., their special relationality), which is furthermore responsible for the difference in character between the perceptions and the hallucinations. But since the character of hallucinations is taken to be exhausted by the determinable aspect of subjective indistinguishability, it remains indeterminate. However, it is doubtful that genuine entities could instantiate determinable features without instantiating determinations of them. And we have no reason to assume that mental episodes are an exception to this rule. Hence perception-like hallucinations (in contrast to perceptions) cannot any more count as genuine entities—that is, in this case, as episodes in the stream of consciousness. Instead they should be treated as situations or states of mind in which subjects can be (see the talk of “situations” in Martin, 2004, 2006). But this is in tension with our subjective impression that we are actually undergoing an episodic experience when unknowingly hallucinating something.

For what follows, it does not matter to settle the issue of which of the two versions of disjunctivism about character is to be preferred. The subsequent considerations and objections apply equally to both. Hence I continue to assume that the subjective indiscriminability of hallucinations from corresponding perceptions is not a constituent of the character of those hallucinations but an indiscriminability in—or relative to—their character. I therefore also do not assume that the epistemic conception of perception-like hallucinations should give rise to the stronger view discussed. In accordance with this, it is compatible with that conception that such hallucinations may still possess positively describable individual natures, although their experiential kind as a whole can be positively characterized only in epistemic terms.

IV   Three Challenges for Conjunctivism about Character

The Challenge of Introspective Error

17. There is an important explanatory difference between disjunctivism and conjunctivism about character: while disjunctivism has the resources to elucidate the nature of the error involved in judging perception-like hallucinations to be perceptions, conjunctivism does not—or at least not yet. According to disjunctivism about character, the hallucinations satisfying (S) and the corresponding perceptions do not share any character determinables; but in introspection we are ignorant about the distinctness of their characters. The error in taking the hallucinations to be perceptions is therefore introspective in nature. By contrast, conjunctivism assumes that the hallucinations are subjectively indiscriminable from corresponding perceptions because both possess the same introspectible properties, that is, the same character. The error can therefore not be located at the level of introspection but has to arise either at an earlier or at a later stage in the epistemic process. But there are also other reasons why disjunctivism about character should perhaps be preferred over its conjunctivist counterpart.

The Challenge of Subjective Impact

18. Perceptual experiences are essentially conscious phenomena, that is, phenomena with a subjective character. This means that a theory that tries to capture their nature must characterize and individuate them in terms of what they are like from, or how they are given to, the subjective perspective. This has the consequence that features—such as structure, functional role, or representationality—matter for such a characterization only if they, in one way or another, make a difference for the subject. If their presence or absence has no subjective resonance, they do not have a bearing on which fundamental kind the experiences concerned belong to. This does not necessarily rule out reductionist accounts of experience. If experiences turn out to be identical with, say, certain brain states, then those brain states arguably make a difference for the subject because experiences do. They are just not given to the subject as the brain states that they essentially are. From the inside, they are simply identifiable as conscious experiences.

19. However, if the property of making a difference for the subject is understood in introspective terms, conjunctivism about character is left with a problem. What is central to this view is the claim that all perceptual experiences—that is, both perceptions and perception-like hallucinations (and illusions)—share the same perceptual character. In addition, and independently of one’s stance on the character(s) of perceptual experiences, it is natural to assume that nonperceptual experiences—such as episodes of imagining or recalling or hallucinations that are, from the inside, easily recognizable as nonperceptual—differ in character from perceptual experiences. Conjunctivism about character, together with this further assumption, entails that experiences count as perceptual by virtue of their character: sharing a character with perceptions is both sufficient and necessary for being a perceptual experiences. Mental episodes, which do not possess a perceptual character, are not perceptual experiences but belong to some other kind of experience. Hence to be able to claim that their view captures the nature of perceptual experiences, conjunctivists about character have to assume that the presence or absence of a distinctively perceptual character makes a difference for the subject.

The problem is now that there seems to be no good reason to rule out the possibility of cases in which the absence of a perceptual character does not make a difference in introspection (Martin, 2004, 2006). In such cases, an experience lacks the character distinctive of perceptions but cannot be introspectively discriminated from them. But this gives rise to a dilemma for conjunctivism about character. On the one hand, the fact that the character of the experience concerned is distinct from that of perceptions is taken to be relevant for its characterization as a nonperceptual experience. But on the other hand, this difference in character does not have any impact on how the experience is given to the subject in introspection, so it should not matter for our identification of the fundamental kind to which the experience belongs to. The only way out for the conjunctivists seems to be to deny that there can be such experiences that satisfy (I) despite not sharing a character with perceptions and hence with perceptual experiences.

What therefore needs to be shown is that introspection is infallible with respect to the detection of the absence of a perceptual character. That is, it has to be argued that each time an experience without a perceptual character occurs, we are in principle able to subjectively recognize this aspect of nonperceptual experiences’ character in which they differ from perceptual experiences. But why should it be assumed that introspection is infallible in this specific manner? Perception, for instance, does not show a comparable infallibility. There are distinct shapes or shades of color, for instance, which we—that is, any possible subject with the apparatus distinctive of humans as they actually are—cannot perceptually discriminate under suitable circumstances, though we might be able to distinguish them in different settings. So conjunctivism about character seems to be forced to present an argument of why introspection differs from perception in not allowing for a certain kind of ignorance.

20. Disjunctivism about character does not have the same problem, since it identifies the satisfaction of (I), rather than the possession of a perceptual character, as the distinctive mark of perceptual experiences.13 Accordingly, an experience that is introspectively indiscriminable from a perception counts as a perceptual experience, independently of whether it has a character in common with perceptions or not. In contrast to the possession of a perceptual character, the introspective indistinguishability from perceptions is always recognizable from the inside. This is not in conflict with the idea that the nonperceptual character of perception-like hallucinations is still to some extent introspectively accessible—if only with respect to its introspective indiscriminability from the character of perceptions. But it has the consequence that perceptual experiences do not form a natural, but rather only an epistemic, kind. This fits well with the epistemic conception of perception-like hallucinations: they are all perceptual experiences (i.e., are subjectively indiscriminable from perceptions), but each possibly for very different reasons. By contrast, perceptions are all perceptual experiences for the very same reason, namely, because of their distinctive relationality. Hence they can still be said to form a natural kind due to this shared essential feature.

The Phenomenological Challenge

21. Another motivation for adopting disjunctivism about character is the acceptance of a certain disjunctivist view about the nature of perceptual experiences. In general, disjunctivism about perceptual experiences combines two claims: (i) perceptions and hallucinations have different essences (or belong to different fundamental mental kinds); and (ii) it is essential to perceptions—but not to hallucinations—that they relate us, in the specified manner, to some mind-independent object and its features. In other words, disjunctivism treats the difference in relationality noted at the beginning as an essential difference.

Importantly, disjunctivism about perceptual experiences is distinct from disjunctivism about character, and the same is true for their conjunctivist counterparts. One difference to note is that while the two approaches to perceptual experiences make claims about the essence(s) of these experiences, the two approaches to character do not—although it is plausible to further assume, for independent reasons, that the character of an experience is essential to it. Another relevant point is that so far, disjunctivism about perceptual experiences stays neutral on whether perceptions and hallucinations share all, some, or none of their character determinables. As the view to be defended later illustrates, it is, for instance, possible to combine conjunctivism about character with disjunctivism about perceptual experiences.

22. Nonetheless there are more concrete versions of disjunctivism about perceptual experiences, which further specify or back up (ii) in such a way that they come to adopt a stance on the issue of character as well. Experiential disjunctivism about perceptual experiences is a good example of this. This version of disjunctivism understands (ii) in the following manner: (ii*) it is essential to perceptions that they establish a relation of awareness to some mind-independent object and its perceivable properties, and that each of their character determinables is determined by, or otherwise constitutively linked to, this relation of awareness. That is, the nature of perceptions consists in their establishment of a conscious contact with external entities and is accessible to introspection.14

Disjunctivism about character is a direct consequence of experiential disjunctivism, for if all character determinables of perceptions are constitutively linked to their special relationality, then the characters of the nonrelational hallucinations cannot realize any of these determinables. Accordingly, experiential disjunctivism denies that hallucinations can share any character determinables with perceptions. However, the two types of experience still have some other property in common, namely, their satisfaction of (I), a relational epistemic property. Perceptions satisfy (I) trivially, that is, simply by being perceptions. And perception-like hallucinations satisfy (I) because of one of the factors mentioned earlier—notably that introspection is insensitive to the relational nature of characters. Given that perception-like hallucinations also do not share any structural aspects with perceptions but instead are characterized by their lack of the relationality distinctive of perceptions, experiential disjunctivism embraces the epistemic conception of hallucinations. Moreover, it includes the idea that what unifies the class of perceptual experiences is precisely the property of satisfying (I).

I have no room here to properly discuss and evaluate the arguments in favor of experiential disjunctivism. But one of its advantages is that it can preserve well our ordinary conception of perceptions, according to which it is part of their nature that they bring us into contact with the external world, that is, genuinely relate us in a distinctive manner to mind-independent objects or facts (Martin, 2002; Dorsch, 2010b). By contrast, other prominent views about the nature of perceptual experiences have to adopt an error theory concerning some aspect or another of our ordinary opinions about perceptions. This is the case, for instance, if perceptions are construed as relations to mind-dependent or nonphysical entities (i.e., to some form of sense-data), or if they are construed as involving an intentional, and hence nonrelational, form of awareness of the external objects (Martin, 2000a).

23. Conjunctivism about character has to address the challenges outlined in the last few sections. Since intentionalists about perceptual experiences are typically also conjunctivists about character, they have to confront these challenges as well. I return later (see sec. VI and §45) to the issue of how experiential intentionalism—the unorthodox intentionalist version of conjunctivism to be defended here—is able to do this, namely, (i) to explain the error involved in taking perception-like hallucinations to be perceptions; (ii) to ensure that the absence of a perceptual character always makes a difference for the subject; and (iii) to accommodate our ordinary opinions about perceptions. For the time being, however, it is worthwhile to understand why disjunctivism about character and, more specifically, experiential disjunctivism are at least as problematic as conjunctivism about character.

V   Three Challenges for Disjunctivism about Character

24. That this task is perhaps less easy than might be thought is illustrated by the fact that conjunctivists about character cannot simply appeal to the satisfaction of (I) when arguing for the claim that the perceptual experiences concerned possess some common character determinables (Martin, 2006). More needs to be said to establish the claim that introspective indiscriminability should count as tracking a sameness in character. To see this more clearly, it may be helpful to compare introspection with perception.

When we are trying to determine the visible features of objects, we have a fairly good grasp of which viewing conditions are optimal for this task and which are not. Moreover, this distinction between optimal and nonoptimal viewing conditions enables us to draw a distinction between how objects seem in visual experience and how they really are. That is, it allows us to conceive of the visible features of objects as being mind independent in the sense of being instantiated independently of any particular of our visual experiences. The visible features should count as mind independent because brute error is possible in vision: we may misperceive or otherwise err about such features, despite being epistemically not at fault (e.g., despite being completely rational and possessing a well-functioning mind and brain). And brute error is possible in vision because we may view things under nonoptimal conditions (e.g., in unusual light or when immersed in water), which need not guarantee that there is no gap between how objects seem in visual experience and how they really are.

Now, the distinction between optimal and nonoptimal viewing conditions and the resulting mind independence of the features of things accessible through vision also ensure that if we cannot tell two things apart by vision alone when viewing them under optimal conditions, then they possess the same mind-independent visible features. This means that we are entitled to take visual indiscriminability under optimal conditions as an indication of a commonality on the level of mind-independent visible features.

This line of reasoning cannot, however, be easily applied to the introspective indiscriminability of hallucinations, for we do not have a similar grasp of a distinction between optimal and nonoptimal conditions in the case of introspection (Burge, 2003; Martin, 2000b, 2006). This need not mean that introspection does not allow for brute error, or that what is introspected is mind dependent. But in the absence of such a distinction, we have yet no good reason to assume that introspective indiscriminability indicates some underlying sameness in character, which occurs independently of our introspective awareness of it. Conjunctivists could maintain that all conditions are optimal for introspection, and introspection is consequently infallible. But this would give rise to the difficult challenge of identifying the mechanisms that underlie introspection and guarantee that we always correctly notice the similarities and differences among the characters of our experiences. Therefore it is better to look for other reasons to prefer the conjunctivist view over its disjunctivist counterpart.

The Challenge of Rational Sameness

25. One of the main challenges for disjunctivism about character is to explain why hallucinations possess the same rational force as corresponding perceptions when they are subjectively indiscriminable from those perceptions. That two experiences share the same rational force means that they make it reasonable for the subject to form the same judgments, beliefs, intentions, and so on. What thus needs explaining is not only that perception-like hallucinations move us to form the same judgments about the world and about themselves as the perceptions, but that it is also reasonable for us to form these judgments—such as the judgment that our current experience is a perception.

Returning to our example, if Macbeth had been rational and completely unaware of the absence and death of Banquo, as well as of his own agitated and traumatic state of mind, he would have come to believe that he really saw Banquo there before him. Moreover, it would have been reasonable for him to develop this belief. Similarly, if he had seen Banquo after having been erroneously assured by everyone else that Banquo had died, a rational Macbeth would not have formed the belief that he actually saw Banquo (but perhaps instead the belief that he “saw a ghost,” meaning that he hallucinated Banquo). In fact, it would have been unreasonable for him to believe in a perceptual encounter with Banquo, assuming that the people around him had been trustworthy. In short, it is reasonable for us to trust our hallucinations, unless we become aware of evidence about their hallucinatory character—just as it is reasonable for us to trust our perceptions, unless we believe them to be hallucinatory.

26. To get clearer about its impact and avoid potential misunderstandings, it is worthwhile to qualify the challenge raised here against disjunctivism about character in several respects.

First, that perception-like hallucinations render certain judgments and beliefs reasonable need not imply that they provide us with (access to) some epistemic reasons for those judgments and beliefs. Accordingly, the challenge stays neutral on whether the subject has epistemic reasons solely when he is perceiving, or also when he is hallucinating (assuming the absence of defeaters). Nonetheless the rational force of our perceptual experiences—whether they are perceptual or hallucinatory—remains closely linked to our subjective take on the presence of epistemic reasons for us. When we are rational, we form our judgments and beliefs in response to what we take our reasons to be. Accordingly, whether it is reasonable for us to rely on a given experience depends on whether we take the experience to be reason providing.

Second, the challenge does not impose the requirement that the rational powers of perceptions and perception-like hallucinations should receive exactly the same explanation. For the sake of argument, I assume here that perceptions render certain judgments and beliefs reasonable by virtue of the reason-providing power of their relation to the world, while the rational force of perception-like hallucinations is to be understood as deriving from the rational force of the perceptions and their priority over the perception-like hallucinations, which is part of the subjective indiscriminability of hallucinations from perceptions (for such an explanatory approach, see Williamson, 1990, 60). The present challenge thus does not presume that sameness of rational power implies sameness in the features responsible for that power. It simply asks for some satisfactory explanation of why subjectively indiscriminable hallucinations share their rational force with the relevant perceptions.

Third, it is also important to note that the rational force common to perceptions and hallucinations is accessible from the inside—perhaps not always by the subject concerned, but at least in principle by some possible human subject in the same situation (just as in the case of subjective indiscriminability). Had his judgment about the presence of Banquo been challenged by the people surrounding him, a rational Macbeth’s initial reaction might have been surprise or disbelief, since he would have taken his judgment to be perfectly reasonable until the moment of the challenge. Indeed, it would have taken extremely convincing external evidence for him to change his assessment of his judgment and its grounding in his experience. The situation would have been very different if Macbeth had merely visualized Banquo and been able to distinguish this experience from a perception. He would have recognized that it would not be reasonable for him to believe that Banquo was there before him. Perceptual and imaginative experiences differ in whether they make it reasonable for the subject to form perceptual judgments and judgments about perceiving. And this difference is accessible from the inside, even if imaginative experiences are compared with perception-like hallucinations.

Fourth, this is one reason why reasonableness is not the same as justification or entitlement (i.e., whatever is third-personally distinctive of knowledge, in addition to truth). If a given hallucination is subjectively indiscriminable from corresponding perceptions, we cannot tell from our subjective perspective that forming a judgment on its basis does not lead to knowledge (or even just true belief). If we could, this would, after all, mean that we do have access from the inside to a feature that distinguishes this hallucination from perceptions. Rational force and the power to put us into a position to know may also differ in that the latter may actually contribute to a difference in nature between perceptions and hallucinations. It may, for instance, be argued that the two kinds of perceptual experience differ essentially in whether they can be grounds for knowledge—or be veridical, for that matter (McDowell, 1998). If this is true, hallucinations cannot justify us to form judgments (other than the judgment that we are hallucinating). In particular, it would be false to assume that if a given hallucinatory experience had been veridical, it would have put us in a position to acquire knowledge (or, indeed, it would be false to assume that a given hallucination could have been veridical in the first place).

This perhaps suggests that hallucinations do not provide us with epistemic reasons, even if they sometimes seem to do so. But it does not prevent hallucinations from making the formation of certain judgments reasonable when hallucinations are subjectively indiscriminable from corresponding perceptions. Judgments based on perception-like hallucinations cease to be reasonable relative to the subject’s perspective only if the subject (rightly or wrongly) takes them to be false or lacking proper grounding—for instance, in response to recognizing the underlying experiences as hallucinatory. The mere lack of truth and justification, on the other hand, does not yet suffice to undermine the reasonableness of such judgments. Macbeth’s experience might have failed to put him in a position to acquire knowledge about the world or about his experience and, in this sense, might have been epistemically defective (e.g., by violating truth- or knowledge-related epistemic norms). But it still rendered it reasonable for him to judge that he was seeing Banquo before him, as long as he lacked evidence from the outside for the hallucinatory status of his experience. And it still enabled him to make a claim to the reasonableness of his judgment. Indeed, he would have been at fault and blameworthy (e.g., for being rationally insensitive or akratic) if he would have failed to take his experience at face value and to form the belief about Banquo’s presence in response to it.15

27. How can disjunctivism about character—or experiential disjunctivism, for that matter—explain that hallucinating something makes it reasonable for the subject concerned to judge that we are perceiving when the hallucination in question is subjectively indiscriminable from a comparable perception?

One natural answer is to say that this rational force just comes already with being perception-like. That one of our experiences is subjectively indistinguishable from a perception means in part that if we lack any opposing evidence from the outside, we are inclined to take this experience to be reason providing (assuming that we are rational). Were we lacking this inclination, we would be able to subjectively tell the experience apart from a perception. One fact in support of this conclusion is that the presence or absence of such an inclination is accessible from the inside, since we can discover by introspection which judgments our experiences move us to form (e.g., in response to asking ourselves whether our experience provides us with a reason to believe). The other fact in support is that perceptions do compel us to take them to be reason providing, as long as we are not aware of defeaters. That is, when we perceive something, we are inclined to judge not only that our experience is a perception but also that it provides us with certain epistemic reasons. We can therefore discover that an experience is not a perception by introspecting that it does not incline us to judge that it is reason giving, despite the lack of any evidence from the outside concerning its nonperceptual status.

Now, in the light of the close link between rational force and our subjective take on reason provision spelled out in the previous paragraph, it might be assumed that being inclined to take an experience to be a reason-giving perception suffices for it being reasonable for us to rely on it in the formation of respective judgments and beliefs. This assumption promises a direct route from subjective indistinguishability to sameness in rational force. The subjective indiscriminability of a hallucination comes with the inclination to take it to be reason giving, which again is assumed to ensure the reasonableness of our reliance on the hallucination when forming our beliefs.16 The assumption under consideration should, however, be rejected, since the reasonableness of relying on an experience cannot simply be a matter of what that experience inclines us to do.

One reason for this is that reasonableness is a normative feature, while inclination is not. In particular, it is no option to introduce the idea of reliability or conduciveness to truth, in the hope that this might be able to bridge the gap between the descriptive and the normative. Granted, our judgments in response to our perceptual experiences generally tend to be true because our experiences are normally perceptions. But disjunctivists about character maintain that perceptions and hallucinations form two distinct fundamental kinds of experience, which do not share any relevant aspects of their essential character or structure. This means that the inclinations coming with hallucinating something in a perception-like manner are completely independent from the inclinations coming with perceiving something. Hence the latter’s connection to truth does not extend to the former. But while the reliability or truth conduciveness of perceptual experience is limited to perception, the reasonableness of relying on experience pertains also to perception-like hallucinations.

The other reason for rejecting the assumption that inclination is sufficient for reasonableness is that being inclined to form a certain belief if prompted does not amount to forming or having that belief. More specifically, that an experience inclines us to take it to be a reason-giving perception does not imply that we actually take it to be so. Part of the explanation of this is that the kind of inclinations at issue depend solely on the occurrence of the experiences in question and our possession of the relevant concepts, and no belief is required or involved. As a consequence, the impossibility of distinguishing a hallucination from reason-giving perceptions does not amount to the positive recognition of the hallucination as reason giving. Our unavoidable ignorance cannot so easily be turned into knowledge.17 But this is problematic, since—as observed earlier—it is reasonable for us to rely on a given experience when we take the experience to be reason providing. Accordingly, the rational force of experiences is linked to our actual take on them—and not to the take on them that we would develop if prompted in a suitable way.

28. So, perhaps, the reasonableness of our reliance on our perception-like hallucinations is due not to their subjective indiscriminability from perceptions but to the fact that we actually take them to be reason providing. The suggestion cannot be that what makes it reasonable for us to trust a given hallucination is our judgment (or belief) that it is a reason-giving perception. There is no need for us to gather respective evidence from the outside to come to judge in a reasonable manner that one of our experiences is a perception and hence provides us with epistemic reasons. Forming this judgment from the inside, however, is precisely what is at issue. We are concerned with the question of what renders our introspective judgment that our experience is a reason-giving perception reasonable. This judgment therefore cannot contribute to the rational force of perception-like hallucinations.

Hence the thought should rather be that we take perception-like hallucinations to be reason giving by recognizing their subjective indiscriminability from corresponding perceptions. Experiential disjunctivism maintains that perceptions provide us with reasons for belief mainly because they acquaint us with aspects of the world or make them manifest to us.18 Since this relational aspect of perceptions constitutes part of their character, it is subjectively accessible to us. So we may perhaps be said to recognize the reason-giving power of our perceptions by becoming aware of their distinctive relationality and thus of their perceptual status. Perception-like hallucinations—the thought continues—possess the same rational force because they are subjectively indistinguishable from corresponding reason-giving perceptions (and because perceptions enjoy priority over hallucinations). Accordingly, it may be claimed that we take such hallucinations to be reason providing because we recognize them to be indiscriminable from perceptions. However, it is doubtful that this type of ignorance is often subjectively accessible to us (if at all). Acquiring knowledge of it requires taking into account the possible cases of other—and more rational and attentive—human subjects being in our current mental situation. Moreover, even if we could in principle come to know about our own necessary ignorance, this would presumably involve a considerable amount of theoretical reflection (such as engaging with Martin’s writings on the subjective indistinguishability of hallucinations). But relying on hallucinations satisfying (S)—just as relying on perceptions—is far from being intellectually demanding in this way.

29. This leaves experiential disjunctivism perhaps with the option to identify a more basic feature of perception-like hallucinations, which is responsible both for their satisfaction of (S) and for their possession of the same rational force as the corresponding perceptions. But such a feature is difficult to find. The hallucinations in question are not in any interesting way linked to the external world. They need not share among each other, or with the corresponding perceptions, any relevant aspect of their causal origins. And neither their lack of a link to the world nor their causal origins are accessible from the inside, so our awareness of them cannot ground our subjective knowledge of the rational force of the hallucinations concerned.

The best candidate for the third feature would probably be a character determinable shared by all hallucinations that satisfy (S). But experiential disjunctivists remain silent about the character determinables of those hallucinations—and for good reasons. One motivation for—and advantage of—their view has been to assume that there need not be such a common character determinable, and nothing more can be said about the hallucinations at issue, other than that they are subjectively indiscriminable from perceptions but lack the relationality of perceptions (Martin, 2004, 2006). However, if no third feature can be identified, experiential disjunctivism cannot explain why hallucinations that are subjectively indiscriminable from perceptions share their rational force with perceptions. The view might outweigh this shortcoming by its power to account for other aspects of perceptual experiences. But it may be equally beneficial to look for an alternative theory that can elucidate the sameness of rational force while perhaps not being less explanatorily powerful in other respects (for more discussion, see Dorsch & Soldati, 2010; Dorsch, 2010b).

The Challenges of Consciousness and of Rational Force

30. Before moving on to the presentation and discussion of such an alternative view, I would like to outline two other challenges to experiential disjunctivism and thus indirectly also to disjunctivism about character. So far, the objection has been primarily that experiential disjunctivism cannot account for a certain fact about perception-like hallucinations. But there are also some doubts about whether it can satisfactorily illuminate two central features of perceptions, namely, their conscious status and their rational force (as well as, relatedly, their power to provide epistemic reasons).

According to experiential disjunctivism, the essence of perceptions consists primarily in their property of relating us to mind-independent objects or facts. However, many relations between subjects and the world do not give rise to conscious states with rational powers. This is why experiential disjunctivism takes the relation in question to be a special kind of relation—a relation of awareness, or of acquaintance, by means of which aspects of the world become manifest to us. But even if we grant that such a relation exists, and that its power to give us awareness of aspects of the world is a primitive feature of our conscious minds, two questions remain.19

The first concerns the issue of how the relation of awareness can have both the power to make us aware of the world and the power to make us aware of our mind. When seeing a green tree, we are conscious not only of the tree but also of our experience of it—notably of how we are of the tree (e.g., whether we see, remember, or imagine it). Indeed, this is part of what it means that our experience of the tree is a conscious experience with a subjective character. The question is how these two types or aspects of awareness are linked to each other, and why they occur together. Perhaps the fact that our experiences of the world are conscious is a primitive aspect of our minds, too. But even so, it can be no accident that awareness of the world and awareness of the mind come together. In other words, it can be no accident that the two capacities of the relation of acquaintance are compatible with each other and, indeed, coexist. And this fact needs explaining, even if it is accepted that each of the powers on its own need—or can—not be elucidated much further.20

The second question problematizes the fact that experiential disjunctivism has to identify the obtaining relation of awareness as the source of perceptual reasons and the resulting rational force of perceptions. It asks how this is compatible with the fact that it is not intellectually demanding to come to recognize perceptions—say, when comparing them with their imaginative counterparts—as reason giving and hence as having the power to render certain judgments and beliefs reasonable (in the absence of relevant defeaters). For the experiential disjunctivist, what gives perceptions their rational force is ultimately the fact that they put us into conscious contact with the world. Hence becoming aware of the rational force of perceptions requires becoming aware of their property of establishing a relation of awareness between us and the world. The question, then, is how the disjunctivist position can ensure that we are able to recognize the relationality of our perceptions and its rational relevance with relative ease. It is not obvious how being acquainted with objects or facts in the world can account for the easy availability of our knowledge that perceptions provide us with reasons and therefore have the power to render certain judgments and beliefs reasonable.21

VI   Meeting the Challenge of Introspective Error

31. In the light of the difficulties and questions that disjunctivism about character—and, notably, experiential disjunctivism about perceptual experiences—faces, it seems worthwhile to look at conjunctivism about character as a viable alternative. Of course, any account of perceptual experiences and their character—whether disjunctivist or conjunctivist—should be able to satisfactorily address the three challenges raised in the last sections. But while it is indeed unclear how disjunctivism about character might be able to achieve this, I aim to show that conjunctivism about characters has no problems with this task—as well as with meeting the other three challenges described even earlier.

Conjunctivism about character states that the hallucinations satisfying (S) share their character with the corresponding perceptions. Perhaps it is possible to identify a character determinable common to both kinds of experience that explains their shared rational force. And perhaps reference to this character determinable can also help us to elucidate in which sense perceptual experiences count as conscious, and how they are able to provide us with reasons.22

According to conjunctivism about character, the hallucination and the corresponding perceptions satisfy (I) because they share their character. As already noted, this means that in finding the two introspectively indiscriminable, we correctly recognize their sameness in character; and the error in judging the hallucinations to be perceptions cannot be an error due to a specifically introspective failure or limitation. Instead the error has to occur either before or after introspection.

32. One way to spell out the latter option is to say that the error is inferential in nature. The idea may be that we inferentially judge perceptual experiences (whether they are perceptual or hallucinatory) to be perceptions on the basis of introspectively judging them to be perceptual experiences, plus some ancillary belief—such as the belief that perceptual experiences are normally perceptions, or that we have good reason to take them to be perceptions in the absence of evidence to the contrary. Alternatively, the introspective judgment may be that it seems that we are perceiving; and from that, we conclude that we are perceiving on the basis of the belief that what is introspectively judged as seeming to be the case is typically the case.

Both views fit very well with the fact that, once we become aware of evidence for its hallucinatory status, we stop taking a given experience to be perceptual.23 But their main problem is that they cannot easily accommodate the immediacy with which we often come to judge that we are perceiving. When challenged about his unusual words and behavior, Macbeth did not have to engage in any form of reasoning to be able to reply that he had seen Banquo.

It might be suggested that Macbeth had learned to automatize or internalize such inferences in some way or another. The thought is that while the justification for his judgments about the perceptual character of his experience was still inferential, he was able to form that warranted judgment without having to engage in any conscious inference. However, if further pressed, Macbeth would not have provided such an inferential justification but instead continued to simply point to the perceptual status of his experience. Indeed, Macbeth’s conduct would have looked strange (or, rather, even stranger than it already did) if he would have answered that he had a visual experience as if of Banquo, and that his experiences are normally perceptions. He very well realized that the others were believing that there was no Banquo to be seen. And this deeply worried and unsettled him and let him question his own sanity. Nonetheless he kept on insisting that he had seen Banquo (as illustrated, say, by the line of dialogue quoted in the chapter epigraph). If his judgment would have been inferentially justified, he would probably have dismissed it in the light of the testimony of the others. But this was not what was happening.

33. If the error in taking perception-like hallucinations to be perceptions is neither introspective nor inferential, it has to be nonjudgmental and occur at a prior level of awareness. It is natural to identify this prior level of awareness with conscious awareness that comes with having conscious experiences and occurs before, and independently of, introspective awareness of those experiences.

Consciousness comes with subjectivity. When we are conscious, things are given to us as conscious subjects. That is, we are consciously aware of them from our subjective perspective. But things can be consciously given in two different ways. Certain things—namely, episodes in the stream of consciousness—are present in (or to) consciousness. That is, they are determinations of consciousness, and we consciously experience them while being in them. By contrast, other things—such as aspects of the external world or, indeed, our mental episodes—are presented to us as being a certain way. A rose may appear to be long and red, and the respective mental image of the rose may seem to be actively produced or instead to have occurred unbidden. We are conscious of these entities and experience them as being a certain way.24 Many mental episodes involve both forms of conscious awareness. But perhaps some of them are only present in consciousness and do not come with the presentation of something as being a certain way. This may be, for instance, what is happening when we are consciously enjoying a feeling of ennui or anxiety or are undergoing an experience of vertigo.

The character of mental episodes is in fact identical with how they are present in, or determine, consciousness. To be present in consciousness just means to be conscious and to have a character. It is perhaps worthwhile to point out here that talk of what a mental episode “is like” may be understood in at least two different ways. On the one hand, it can be interpreted metaphysically, as denoting the nature of the episode (leaving it open whether this nature is first- or third-personally accessible). On the other hand, the phrase can be understood epistemically, as denoting how we consciously experience the episode. The character of mental episodes combines both elements: it is part of their nature and consists in how they seem to the subject in the sense of being present in, or a determination of, consciousness. As a result, there is no distinction between how the character of a mental episode is and how it experientially seems (Husserl, 1996). By contrast, it is certainly possible that how an episode is intentionally presented to us as being—whether in experience or in introspection—does not match how it really is.

34. The suggestion is now that hallucinatory experiences are presented to consciousness as being a certain way. More precisely, they are given to us as being perceptions, that is, as relating us to some mind-independent objects and their features in the manner characteristic of perceptions. And given that they are not perceptions, our conscious awareness of them involves some kind of error: there is a mismatch between how the experiences really are and how they are presented to consciousness. The error concerned is one about the underlying objective structure of the hallucinations: namely, their lack—rather than their possession—of the property of relating us to some mind-independent entities. And the wrongness of the resulting introspective judgment is merely a consequence of the error that occurs at the prior level of conscious awareness.

It is perhaps worthwhile to stress that the proposed type of error is not an error about the character of the hallucinations concerned. Indeed, this would be impossible, since it would mean that how these hallucinations are given to us in consciousness is wrong about itself. The character of experiences—what we have so far specified as their most determinate introspectible feature—is identical with their presence in consciousness, that is, with what it is like to consciously experience them. Having a character just means being conscious, that is, being given to consciousness. And the character of an experience cannot present itself, let alone in a mistaken manner. The only types of error possible are introspective error about the character of an experience, and experiential (or first-personal) error about its objective (or third-personal) structure (Dorsch & Soldati, 2010). The proposal here is that introspective error is a result of the experiential error.

This presupposes that we form the introspective judgment about the perceptuality of the hallucinatory experiences in direct response to our conscious awareness of them: we judge them to be perceptions because they are given to us in consciousness as perceptions, and because we introspect this feature of theirs. In accordance with this, the property of being presented to consciousness as relating us to mind-independent things or facts is to be understood as a character determinable that is common to all hallucinations that satisfy (S). But it is also shared by the corresponding perceptions, thus ensuring that the two kinds of perceptual experience end up possessing the same character, for perceptions are equally given to us as relations to mind-independent entities. In this introspectibly accessible respect, perceptions and perception-like hallucinations differ from sensory (or episodic) memories and sensory imaginings: memories and imaginings are not given to consciousness as perceptions. If Macbeth had recalled or visualized Banquo as being at the banquet, instead of hallucinating him to be there, he would not have had the conscious impression of his experience bringing him into direct contact with something that was present before him independently of his actual experience of it.

VII   Experiential Intentionalism

35. So far, it remains unclear whether, or how, the new proposal can meet the challenges to disjunctivism about character and experiential disjunctivism. And it is also left open why the fact that hallucinations are given to us as perceptions should count as involving an error, given that some kind of mismatch between two facts does not automatically manifest a genuine mistake. Both sets of issues can be resolved by understanding consciousness partly in intentional terms. More specifically, the presence of something to consciousness as being a certain way—or its appearance to the conscious subject as being a certain way—should be interpreted as a form of intentional awareness. The resulting view is experiential intentionalism—the view that we are intentionally aware not only of the world but also of our own conscious experiences.25

36. The intentionality of conscious mental states consists minimally, and centrally, in their subjection to some norm that requires the states to occur only when the world is in a certain state or meets a certain condition.26 The intentional content of the states—if one wants to introduce this notion at all—can then be understood as being determined by the nature of the relevant truths about the world. The judgment that it rains, for instance, should occur only if it rains (and this is thinkable); and it should occur if the fact that it rains is evident to the subject (e.g., when he sees that it rains). The norm for perceptual experiences, on the other hand, consists in the requirement that they should occur only when they actually relate us—in the manner characteristic of perceptions—to particular mind-independent objects with certain perceivable features. Accordingly, the visual experience of a green tree is adequate only when it relates us in the right way to a certain visible green tree before us.27

The intentionality of perceptual experiences is therefore directed both at the world and at the experiences themselves. In particular, the specific condition on the world, which is characteristic of the norm governing perceptual experiences (and determines their intentional content), concerns not only how certain external entities are like independently of the particular subject and experience concerned but also how these entities are linked to that subject and experience. This is further reflected in the fact that consciously enjoying a perceptual experience enables us to demonstratively refer not only to aspects of the mind-independent world but also to the mental experience itself. And it is in this sense that the intentionality of perceptual experiences may be said to be self-presentational or token reflexive (despite, perhaps, not necessarily involving or requiring any conceptual capacities). Indeed, their self-directed intentionality is identical with their property of being given to us—and, indeed, to themselves—as perceptions. The resulting error in the case of hallucinations is therefore intentional and self-presentational in nature. It consists in their violation of the norm constitutive of their intentionality—an intentionality that is self-directed and shared with the corresponding perceptions.

37. This raises the question of why perceptual experiences—and, especially, perception-like hallucinations—are given to consciousness as perceptions. I discuss this issue in more detail later (as well as in Dorsch, 2010b), but it may already be helpful to provide a sketch of the answer to be presented later on (§42). One of its central claims is that perceptions are intentionally given to consciousness as relations because they are relations, and because it is of epistemic—and ultimately also practical—importance for us to become aware of their relationality. Similarly, it is of value for us that hallucinations are consciously marked as hallucinatory. Perception-like hallucinations—which may occur for a variety of reasons—are worse than their nonperceptual counterparts because they disguise themselves as perceptions. They should not occur because their misleading character is counterproductive with respect to our epistemic and practical aims. Indeed, they are accidents; and not much of interest can be said about why it is possible for experiences that satisfy (S) despite not being perceptions to occur. This is part of what it means that perceptions enjoy priority over their hallucinatory counterparts.

38. Two further qualifications may help to forestall potential objections. First, the normativity at issue need not be very strong. That is, it need not put any demands on the subject and his rationality. Accordingly, that certain mental states should occur only when certain things are true of the world need not mean more than that it is better for the states to occur under those condition than to occur under all other conditions. This allows for the possibility, for instance, that the intentionality of perceptual experiences may be partially accounted for in terms of some more basic cognitive function that they have in our mental lives. However, second, intentionality is not the same as—and may not be reducible to—representationality, if representationality is understood as the possession of some teleological function concerning, or the presence of some nomological correlation with, the world. Representational states need not be conscious. By contrast, intentionality is inseparably linked to consciousness and subjectivity. This is the main point behind taking the presence of something to consciousness as being a certain way to consist in intentional awareness of it as being that way.

Meeting the Challenge of Consciousness

39. One of the challenges raised earlier against experiential disjunctivism was that assuming a relation of awareness does not shed enough light on how experience makes us aware not only of the world but also of itself, and on why the two are so intimately linked. Treating perceptual experiences as token-reflexively intentional states may do better on this count.

The twofold character of the intentionality of perceptual experiences corresponds to two ways in which we can become consciously aware of—and pay attention to—something while enjoying such an experience. On the one hand, we can become perceptually aware of something—typically some physical entities in the external world. Such awareness not only is sensory in character but also allows for the possibility of focal attention to the mind-independent objects of awareness. On the other hand, we can become experientially (or phenomenally) aware of something—this time of our mental episodes themselves. This kind of awareness is not sensory and does not involve the episodes as objects of awareness and of focal attention. Indeed, any attempts to attend to conscious experiences as objects inevitably give way to attempts to attend to the perceptually conscious external entities (Martin, 2000b).

40. What we are concerned with here is the special transparency of perceptual experiences. Having them consists in occupying a certain subjective and conscious perspective on external objects and features. But this perspective is not itself presented to us as being a certain way. We do not see our own point of view. Rather, this perspective is present to consciousness only insofar as it is the perspective from which other entities are presented to us as being a certain way. This is reflected, for instance, in the fact that perceptions inform us about one object being to the left of another relative to our point of view without explicitly presenting that point of view and its location in space. Instead they simply present the one object as being left to the other, while our perspective on them enters consciousness just implicitly, as part of how the spatial relationship between the two objects is given to us.28 Similarly, perceptual experiences as a whole are not objects of awareness and attention but rather determinations of both. We experience them as relating us to external objects, but we do not experience both relata in the same way. While the external object of awareness are given to us as being a certain way, the experiences are given to us as determinate aspects of our subjective perspective on such objects.

The intentionality of perceptual experiences is characterized by the fact that it combines the two noted ways of becoming consciously aware of things. Given that the self-presentational part of this intentionality consists in the experiential awareness of the episodes themselves, the intentional error under discussion is experiential—not perceptual, inferential, or introspective—in nature. But the two ways of becoming aware of things also have something important in common, namely, their intentional nature. Assuming that we have a fairly good grasp of how intentionality works, we are able to provide an answer to the question of how—or in which sense—we become conscious of our own perceptual experiences: they are intentionally directed at themselves and, in this way, present themselves to us.

41. Does this mean that the conscious status of experiences is a matter of their being the object of some intentional awareness? This would come dangerously close to higher-order accounts of consciousness and would in addition render the view vulnerable to objections against the thesis that mental states are conscious by being objects of awareness (for such an objection, see Martin, 1997b). But experiential intentionalists need not accept this conclusion. On the one hand, intentional presentation—like the relational presentation put forward by experiential disjunctivism—may simply be taken to presuppose a conscious subject or mind to which things are presented. So while the introduction of intentionality—or acquaintance—is meant to explain how we are linked to the objects of our awareness, it is not intended to shed light on what it means for a subject, or one of its mental episodes, to count as conscious. On the other hand, experiential intentionalism assumes that, in experiential awareness, the mental episodes do not become objects of awareness but instead are determinations of awareness. That is, while they help us to become aware of external objects as objects, they do not present themselves as objects but are instead given as subjective parts or aspects of consciousness (see the two previous sections). Accordingly, intentional awareness need not always, or not entirely, be object awareness—even in the case of visual experiences.

Meeting the Challenge of Rational Force

42. The second challenge was to account for the capacity of perceptions to render certain judgments and beliefs reasonable. The intentionalist approach explains this power in terms of the normative intentionality pertaining to perceptual experiences. Its account applies therefore not only to perceptions but also to perception-like hallucinations. The rational force of perception-like hallucinations is accounted for in the same way, and at the same time, as that of perceptions. There is no need—as with disjunctivism about character—to make the detour of referring to the priority of the perceptions and the subjective indiscriminability of the hallucinations to account for their rational force. But, as will become clear shortly, the power of perception-like hallucinations to render certain judgments reasonable is still in some sense derivative from the power of the corresponding perceptions to render those judgments reasonable.

Perceptual experiences are intentional insofar as they are subject to the requirement to occur only if they relate us, in the right way, to some perceivable and mind-independent aspects of the world. The normative status of this requirement is due—or at least inseparably linked—to the fact that it is of value for perceptual experiences to establish the required relations between us and the world, and of disvalue for them to fail to do so. The value of those perceptual relations derives from the fact that they put us into the position to acquire knowledge about the world, as well as about the respective experiences themselves (for similar ideas, see Burge, 2003; Haddock et al., 2009). When we are having perceptual experiences, both the world and the experiences appear to be certain ways. And when we are actually perceiving, taking these appearances at face value will lead to knowledge about the world and the experiences.

Now, perceptual experiences are presented to consciousness as being relational (see note 21). That is, they are given to us as possessing precisely the feature that renders them valuable with respect to the attainment of knowledge. In the absence of evidence to the contrary, it is therefore reasonable for us to trust them and to form the respective first- and higher-order judgments or beliefs—even if the result will not be knowledge. Hence what is responsible for the rational force of perceptual experiences is that they are intentional—that is, appear to satisfy a certain norm. And which judgments or beliefs they render reasonable is determined by which particular norm they purport to satisfy (or, if one prefers, which specific intentional content they possess).29

The rational force of our reliance on perceptual experiences is thus a matter of their intentionality, while their status as grounds of knowledge depends on their actual structure. The two epistemic aspects are intimately linked insofar as their intentionality partly concerns their structure. But only in the case of perceptions does the rational force correspond to their power to put us in a position to acquire knowledge. Hallucinations, by contrast, lack such power, since they are not relational in the required manner, and their rational force is not matched by a capacity to ground knowledge. So there is still a sense in which perception-like hallucinations derive their rational force from perceptions. Perceptions possess their intentional character and hence their rational force (and reason-giving power) because this reflects their intimate link to knowledge and renders it subjectively accessible to us. Hallucinations that satisfy (S), on the other hand, are mere accidents deviating from the perceptual norm: they lose the value of being grounds for knowledge but keep the rational force of perceptions by remaining subject to the relevant intentional norm.30

Meeting the Challenge of Rational Sameness

43. The third and main challenge for disjunctivism about character was to account for the fact that perception-like hallucinations share their rational force with the corresponding perceptions. What is to be explained is thus that, in situations in which we lack relevant evidence from the outside, a hallucination makes the same judgments reasonable as the respective corresponding perceptions when it satisfies (S). The intentionalist strategy for answering this challenge should be clear by now. Again, the central thought is that the rational force of perceptual experiences derives from their intentionality. Accordingly, two such experiences share their rational force when and because they share the same specific intentionality—that is, are subject to the same specific norm. In short, sameness in rational power comes with sameness in intentionality. Given that their shared intentionality is independent of whether we have any evidence from the outside about their perceptual or hallucinatory status, we can safely ignore what happens if such evidence becomes available.

Now, according to conjunctivism about character, a hallucination is indiscriminable from the inside from corresponding perceptions when and because it possesses the same character as the latter. Given that sameness in character presupposes sameness in intentionality, it follows that if a hallucination is subjectively like certain perceptions, then it also possesses the same rational force as certain perceptions. On the other hand, a hallucination possesses the same rational force as certain corresponding perceptions only if it also shares their intentionality—including the appearance of relating us perceptually to the world. And no other character is determinable in respect of which the hallucination might differ from the perceptions at issue. For instance, if the hallucination involves blur, and blur is not a matter of intentionality, there will be a comparison class of possible corresponding perceptions that equally involve blur, and likewise for any other potential character differences between hallucinations and perceptions that share the same intentionality. Accordingly, for such experiences, sameness in intentionality comes with sameness in character and hence with indiscriminability from the inside.

VIII   The Experiential Indiscriminability of Hallucinations

44. The kind of subjective indistinguishability at issue is not merely introspective in nature: it occurs already at the level of experiential awareness. That is, access from the inside is characterized by the fact that it does not involve perception, testimony, inference based on either, or any combination thereof. But introspective awareness is not the only form of access that satisfies this condition. Experiential awareness—how mental episodes are given to consciousness—is another one. While introspective awareness provides us with access to the character of independently occurring mental episodes, experiential awareness is already inherent to having such episodes. Now, perceptual experiences are given to us as relations to the world, that is, as possessing a certain third-personal structure. However, this awareness is misleading in the case of the hallucinatory experiences, given that they do not actually possess this structure. Accordingly, each of the hallucinations can be taken to be subjectively indiscriminable from perceptions in the sense of satisfying the following specification of (S):

(E)  None of us could, in experiential awareness, recognize its structure to be distinct from the structure shared by each member of some class of corresponding perceptions.

All perceptions satisfy (E) trivially because there is no distinct structure to be noticed in their case. And hallucinations satisfy (E) when and because they possess the same intentionality and therefore—as has been argued earlier—the same character as the perceptions concerned. Mental episodes, which satisfy (E), also satisfy (I), for if an experience possesses the same character as a perception, it is introspectively indistinguishable from it. Assuming that experiential awareness and introspective awareness exhaust the possibilities in which we can access something from the inside, experiential indiscriminability implies not only introspective but also subjective indiscriminability. Moreover, given that subjective indiscriminability requires both experiential and introspective indiscriminability, it turns out that an experience satisfies (E) when it satisfies (S).

But, importantly, the same does not hold with respect to (I) and (S). The reason for this is that the relationship between the introspective and the experiential indiscriminability of hallucinations (as well as of other experiences) may be asymmetric. While experiential indistinguishability entails introspective indistinguishability, the opposite is not necessarily true. Following disjunctivists about character in their assumption that we cannot always tell apart two distinct characters when introspecting them, it is possible that an experience conforms to (I) without sharing its character with perceptions and hence without conforming to (E). There is a difference in how the episodes concerned are given to consciousness, and we consciously experience this difference, though we cannot pick up on it in introspection—for instance, because we lack the required conceptual capacities. A similar gap between experience and introspection is present in other cases, too. We may, for example, start to be in and consciously experience pain while continuing to judge or believe that we do not feel any pain. Such a case may indicate some psychological problem or pathology, but there does not seem to be any good reason to deny its possibility (for a more detailed discussion, see Dorsch & Soldati, 2010).

Meeting the Challenge of Subjective Impact

45. This raises again the question of whether experiences that satisfy (I) but not (E) and (S) should count as perceptual or not—a question that relates back to the first challenge against conjunctivism about character presented earlier. Disjunctivists about perceptual experiences answer positively, since they take the introspective indiscriminability from perceptions to be necessary and sufficient for being a perceptual experience. Indeed, they are likely to equate subjective indistinguishability with introspective indiscriminability and thus to deny—or, rather, ignore—the distinct existence of experiential indiscriminability. Experiential intentionalists, on the other hand, give a negative answer, since they take the possession of a perceptual character to be the distinctive mark of a perceptual experience, which is lacking in the example case. As a consequence, they accept that there are in fact two ways in which experiences can be subjectively indistinguishable from perceptions, and that these two ways need not necessarily coincide.

The challenge raised against views of the latter kind has been to ensure that the features responsible for counting as a perceptual experience make a difference for the subject. And its prospects of success have been doubted in response to the possible case under consideration, which shows that the absence of a perceptual character cannot always be introspectively detected from the inside. But the shift of focus from introspective to experiential awareness enables experiential intentionalism to answer this challenge without having to deny the noted fallibility of introspection.

Remember the distinction between something being present in consciousness and something being presented to consciousness as being a certain way (§33). As I have argued, the latter is best understood in intentional terms. And there are different ways in which something can be intentionally given to us as being a certain way—notably in a perceptual or in an experiential way. But perceptual (and other) experiences do not simply present us with themselves or external objects as being a certain way (e.g., as having a relational structure, or as being red). They are also present in consciousness as one of its episodic determinations and, in this sense, make a difference for the subject and his or her stream of consciousness.

Experiential intentionalism can therefore respond to the challenge by pointing out that a difference in character between two experiences constitutes a difference in how they determine consciousness, that is, in how we experience them. And this remains true even if we cannot tell them apart in introspection. If the difference in character is not open to introspection, its discovery requires substantial theoretical reflection, which may very well go beyond introspective reflection. But this just illustrates that phenomenology, although being concerned with our subjective perspective and with how things are given to it, cannot always be pursued exclusively from the inside (Husserl, 1992). Moreover, presence in consciousness—that is, our conscious experience of mental episodes—may be understood as a nonintentional instance of experiential awareness. Both are characterized by the fact that they are not instances of object awareness. They consist in the awareness of mental episodes but do not present them as objects and do not allow us to focally attend to them (in contrast to any external entities that they present us with).

46. Apart from saving conjunctivism about character from the objection raised by its disjunctivist opponents, there is some independent reason to accept the distinction between experiential and introspective awareness and to prefer the former over the latter when considering the issue of why an experience counts as perceptual. The subjective perspective matters only because experiences are essentially conscious. Capturing their nature therefore means capturing how they are linked to consciousness. However, on the assumption that higher-order accounts of consciousness are to be rejected, there is no good reason to posit any significant connection between consciousness and introspection—or introspectibility, for that matter. It is true that what introspection provides us with direct access to is the character of an experience. But if higher-order accounts of consciousness are indeed inadequate, the presence of an experience in consciousness cannot be a matter of how it is, or can be, introspected. In particular, we should not doubt that there may be beings that enjoy conscious experiences but lack introspective capacities.31 Consequently, perceptual (and other) experiences should be characterized in terms of how they determine consciousness—that is, of how they are given to the subject in nonintentional and nonintrospective experiential awareness. This is why mental episodes that satisfy (I) but not (P) should not count as perceptual experiences.

Meeting the Phenomenological Challenge

47. Another important point is that the proposed intentionalist version of conjunctivism about character—in contrast to many other instances of that view—can save some of the central elements and advantages of experiential disjunctivism.32 Among the main elements preserved is the idea that we have introspective access to the relationality of perceptions. According to experiential intentionalism, we can introspect the character of perceptions. But part of that character is constituted by the token-reflexive intentionality directed at the specific relationality of perceptions. Hence we can introspect the fact that perceptions present themselves as relations to the world. And this suffices for us to come to know that they are indeed such relations. It is true that this access to their relationality is less direct than has been thought by experiential disjunctivists. But introspective access mediated by intentional awareness on the experiential level is still introspective access.

That intentionalism can preserve this element of experiential disjunctivism allows it also to uphold our ordinary conception of perceptions as genuinely and distinctively relating us to mind-independent entities. First of all, it can explain why we are of this opinion by reference to the fact that perceptions are given to us in consciousness as relational. In fact, given that experiential disjunctivism has difficulties accounting for our conscious awareness of the relationality of our perceptions (§30), experiential intentionalism seems to fare better than its disjunctivist counterpart with respect to the elucidation of why we have this specific conception of perceptions, and not another. Then intentionalism can also match disjunctivism’s ability to avoid an error theory about our ordinary views. The more straightforward option is simply to adopt structural disjunctivism about perceptual experiences—the view that perceptions and hallucinations differ essentially in their third-personal structures (Dorsch, 2010b). This permits the experiential intentionalist to embrace the ordinary opinion that perceptions are essentially relational. A slightly revisionary alternative is to argue that this opinion concerns not how perceptions are but how they are given to us in consciousness. Accordingly, what we ordinarily assume to be essential to perceptions is that they present themselves as relational, not that they are relational.

IX   The Nature of Perceptual Experiences

48. To accommodate our common views about perceptual experiences, both options take a certain stance on the nature of those experiences. The specific commitments of experiential intentionalism on this issue depend on the underlying conception of the relationship between the first-personally accessible character of experiences and their (also) third-personally accessible structure.

To get clearer about why this conception matters, consider the contrast with experiential disjunctivism. Its postulation of a relation of awareness can be interpreted as an attempt to combine or reconcile the relationality of perceptions with their character by identifying their relationality as the main (or even sole) constituent of their character. Furthermore, experiential disjunctivism maintains that their relational character is essential to perceptions. But their structure plays no role in the provided account of their nature, and the issue of how character relates to structure does not become pressing. This is neatly captured by Martin’s focus on the “phenomenal nature” of perceptions (Martin, 2006, 14).

In comparison, experiential intentionalism takes the relationality of perceptions to be part of their nonexperiential structure—along with, say, their representationality or their functional role. As already indicated, the relation in question may be understood in causal, informational, or rational terms, for instance; and natural candidates for it are the relations of reference, nomological dependence, object dependence, or reason constitution. In particular, it is natural to assume that perceptions—but not hallucinations—are relational insofar as their distinctive power to provide us with reason for belief, and to put us in a position to acquire knowledge, constitutively depends on the perceived external facts (McDowell, 1998; Dorsch, forthcoming). It is this relational aspect of the structure of perceptions that is adequately reflected by their character, which hallucinations merely seem to possess.

But the distinction between the character and the relationality of perceptions raises the issue of which of the two forms part of the nature of perceptions. Part of this question is how character and relationality—or, more generally, structure—are related to each other, given that relationality is not said to be a constituent of character. It is not easy to come up with a plausible and illuminating view about the relationship between the first-personal and the third-personal aspects of our minds. This difficulty becomes manifest when we consider the possibilities for experiential intentionalists concerning the identification of the nature of perceptual experiences.

49. A natural explanation of the fact that the satisfaction of (E) is both necessary and sufficient for being a perceptual experience is that the character of such experiences—that is, what is responsible for their satisfaction of (E)—either constitutes (and exhausts) their essence or is entirely determined by whatever constitutes their essence. This fits well with the more general idea that consciousness is central to the nature of particular mental episodes: they are not merely essentially conscious, but the specific ways in which they are present in consciousness are also part of, or determined by, their nature. However, there is still room for several different views about the nature of perceptual experiences, each of which is compatible with experiential intentionalism and with the idea that the character of an experience is essential to it.

First, it is possible—as already suggested—to adopt structural disjunctivism about perceptual experiences. Contrary to what might be thought, this view can be combined with intentionalism. The resulting position claims that perceptions and perception-like hallucinations possess different essences, despite having the same character (Dorsch, 2010b). One way of spelling this out is to maintain that the shared character of the two kinds of perceptual experience is to be understood as being realized by different underlying structural essences. The essence common to all perceptions is thereby identified with their specific relationality. By contrast, it is left open which essence(s) hallucinations possess, as well as whether they all possess the same nature—apart from the restriction that their essence(s) should imply that they necessarily lack the relationality distinctive of perceptions. One of the main problems of this view is to explain how different relational and nonrelational structures can give rise to the same intentional character.

Second, it may instead be suggested that perceptions and perception-like hallucinations possess the same nature, and that this shared nature consists in some aspects of their common third-personal structure—such as their representationality or functional role. This view denies that the relationality of perceptions is essential to them. As a consequence, it has the advantage that it is probably easier to elucidate intentionality in terms of representation or function, rather than in terms of a relational link to the world. But the close link between intentionality and consciousness still remains largely unexplained, given that representationality or functional role is not tied to consciousness in the same way. And the view is also at odds with our ordinary conception of perceptions as relational and therefore has to adopt an error theory in this respect.

Third, it might be assumed that the common nature of all perceptual experiences is exhausted by their character and does not extend to the lower structural level. This position could then be further supplemented with the idea that, in the case of perceptions, their character is in some sense realized by their relational structure, while leaving it to further investigations to discover the structural realizer(s) in the case of hallucinations. The resulting view differs from the first one only insofar as it limits the essence of perceptual experiences to the level of character and does not include the underlying structural differences. As a result, it still faces the difficulty of having to make sense of how both a relational and some nonrelational structures can give rise to (the same) intentionality. Moreover, since the view does not take the relationality of perceptions to be essential to them, it seems forced to embrace a respective error theory about our ordinary opinions.

The second option may seem to be the least problematic. In particular, it is arguable that all theories about perceptual experiences—including disjunctivism—have to take our common view about perception-like hallucinations to be erroneous, given that we judge them to be perceptions from our subjective perspective. But perhaps we should instead renounce the need—or even the possibility—of being able to account for our first-personal perspective on mental episodes in terms of our third-personal perspective on them. There is, after all, a sense in which experiences are not accessible from the outside: we can acquire knowledge about their character—what they are like as conscious episodes—only by experiencing or introspecting them. Indeed, the main source of the problems for the first and the third view outlined earlier is our expectation of being able to explain the first-personal character of perceptual experiences in terms of their third-personal structure. Giving up this expectation may perhaps lead to a more promising picture.33

In accordance with this, a fourth possible view claims that the essence of perceptions contains two relatively independent elements: their intentional character and their relational structure. This is not meant to deny that some link exists between the two. For instance, the relational structure may still determine part of the intentional character by determining part of the intentionality involved. More precisely, the specific condition on the world, the satisfaction of which is crucial for whether a given perception should or should not occur, may be fixed by the fact that the perception relates us to the world, as well as by facts about which aspects of the world it relates us to. However, that perceptions are conscious and intentional in the first place need not be a matter of their relationality. The connection between their character and their relational structure may be intentional, and nothing more. If this is the case, a complete and general account of the intentionality of perceptions should refer to more than their relationality. Perhaps it is possible to identify other aspects of their common structure—and aspects that they share with other types of mental episodes—that are responsible for their conscious and intentional status. But maybe the latter should instead be taken to be primitive aspects of subjectivity—at least in the sense that they evade explanation in terms of structural aspects, and hence explanation from a third-personal point of view.

50. The first and the fourth option have in common that they combine experiential intentionalism with structural disjunctivism, and they take both the character and the structure of perceptual experiences to be essential to them. They differ in how they conceive of the relationship between character and structure—whether it is one of mere intentionality or also one of realization. The two proposed elements of the essence of perceptual experiences correspond to our two perspectives on them: while their third-personal essence concerns how they are in objective reality, their first-personal essence concerns how they determine consciousness, that is, are given to the subjective perspective. To say, from the first-personal point of view, that an experience possesses a certain character is therefore not to say, from the third-personal point of view, that it possesses a certain structure.

One advantage of this separation of perspectives is that experiential intentionalism—when combined with structural disjunctivism—can hold on to the epistemic conception of hallucinations (see also Dorsch, 2010c). From the third-personal stance, perception-like hallucinations lack the relationality distinctive of perceptions but may otherwise differ greatly in their structural essences. From the first-personal stance, they possess the same conscious character as corresponding perceptions, that is, are experientially indiscriminable from them. And nothing more positive may perhaps be said about what these hallucinations have in common with each other, or with perceptions, but not with other mental episodes. Nonetheless the subjective perspective can still count as being concerned with the metaphysics of perceptual experiences. Indeed, the proposed view does not differ in this respect from experiential disjunctivism, which, as noted, focuses primarily—or even exclusively—on the first-personal nature of experiences.

Acknowledgments

Part of this material was presented in 2008 at the University of Crete in Rethymno during the “Hallucination on Crete” conference, in 2009 at the University of Fribourg as part of their research colloquium, and in 2009 at the University of Bremen during the seventh conference of the German Society for Analytic Philosophy. I would like to thank the respective audiences for their comments, in particular David Bain, Johannes Brandl, Paul Coates, Tim Crane, Katalin Farkas, Martine Nida-Rümelin, Matthew Nudds, Howard Robinson, and Susanna Schellenberg. Special thanks are due to the two editors of this volume, Fiona Macpherson and Dimitris Platchias, for all their work; to Mike Martin and Matthew Soteriou for the many discussions; and finally to Gianfranco Soldati, together with whom I developed many of the ideas—and probably also introduced some of the errors—present in this chapter (see also Dorsch & Soldati, 2010). Part of the research for this work was generously funded—in the form of a Fellowship for Advanced Researchers—by the Swiss National Science Foundation (grant no. PA00P1-126157).

References

Burge, T. (2003). Perceptual entitlement. Philosophy and Phenomenological Research, 67, 503–548.

Dorsch, F. (2009). Judging and the scope of mental agency. In L. O’Brien & M. Soteriou (Eds.), Mental actions (pp. 38–71). Oxford: Oxford University Press.

Dorsch, F. (2010a). Colour resemblance and colour realism. Rivista di Estetica, 43, 85–108.

Dorsch, F. (2010b). Transparency and imagining seeing. Philosophical Explorations, 13, 173–200.

Dorsch, F. (2010c). The unity of hallucinations. Phenomenology and the Cognitive Sciences, 9, 171–191.

Dorsch, F. (Forthcoming). The phenomenal presence of reasons. In F. Macpherson, M. Nida-Rümelin, & F. Dorsch (Eds.), Phenomenal presence. Oxford: Oxford University Press.

Dorsch, F., & Soldati, G. (2010). Intentionalism, experiential and phenomenal error. Manuscript.

Dretske, F. (1986). Misrepresentation. In R. Bogdan (Ed.), Belief: Form, content, and function. Oxford: Oxford University Press.

Dretske, F. (1995). Naturalizing the mind. Cambridge, MA: MIT Press.

Evans, G. (1982). The varieties of reference. Oxford: Oxford University Press.

Fish, W. (2009). Perception, hallucination, and illusion. Oxford: Oxford University Press.

Haddock, A., Millar, A., & Pritchard, D. (2009). Epistemic value. Oxford: Oxford University Press.

Husserl, E. [1911] (1996). Philosophie als Strenge Wissenschaft. [Philosophy as a strict science.] Tübingen: Klostermann.

Husserl, E. (1992). Logische Untersuchungen (Vols. 2–4). Hamburg: Meiner.

Macpherson, F., & Haddock, A. (2008). Disjunctivism: Perception, action, knowledge. Oxford: Oxford University Press.

Martin, M. (2004). The limits of self-awareness. Philosophical Studies, 120, 37–89.

Martin, M. G. F. (1997a). The reality of appearances. In M. Sainsbury (Ed.), Thought and ontology. Milan: Franco Angeli.

Martin, M. G. F. (1997b). Self-observation. European Journal of Philosophy, 5(2), 119–140.

Martin, M. G. F. (2000a). Beyond dispute: Sense-data, intentionality, and the mind-body problem. In T. Crane & S. Patterson (Eds.), History of the mind-body problem. Routledge.

Martin, M. G. F. (2000b). An eye directed outward. In C. Macdonald, B. C. Smith, & C. Wright (Eds.), Knowing our own minds (pp. 99–123). Oxford: Oxford University Press.

Martin, M. G. F. (2002). The transparency of experience. Mind and Language, 17(4), 376–425.

Martin, M. G. F. (2006). On being alienated. In T. Gendler & J. Hawthorne (Eds.), Perceptual experience (pp. 354–410). Oxford: Oxford University Press.

McDowell, J. (1998). Mind and world. Cambridge, MA: Harvard University Press.

Schellenberg, S. (2010). Perceptual content defended. Noûs, 54, 714–750.

Siegel, S. (2008). The epistemic conception of hallucination. In A. Haddock & F. Macpherson (Eds.), Disjunctivism: Perception, action, knowledge. Oxford: Oxford University Press.

Sturgeon, S. (2000). Matters of mind: Consciousness, reason, and nature. London: Routledge.

Sturgeon, S. (2008). Disjunctivism about visual experience. In A. Haddock & F. Macpherson (Eds.), Disjunctivism: Perception, action, knowledge. Oxford: Oxford University Press.

Travis, C. (2004). The silence of the senses. Mind, 113, 59–94.

Tye, M. (1995). Ten problems of consciousness. Cambridge, MA: MIT Press.

Williamson, T. (1990). Identity and discrimination. Oxford: Blackwell.

 

1. Schellenberg (2010) also puts forward a mixed position. But her conception of the intentional element of perception is different from the one suggested here. While she promotes a Fregean picture, my proposal is more Husserlian in spirit.

2. I use the term “perceptions” exclusively to denote veridical perceptual experiences. The expression “perceptual experiences” is intended to cover both perceptions and perception-like (or “perfect”) hallucinations—but not hallucinations that are, from the inside, discriminable from perceptions. The class of “experiences,” finally, is meant to include not only perceptual experiences but also other sensory episodes, such as episodic memories, imaginings, and subjectively discriminable hallucinations.

3. This strengthening of the epistemic conception of perception-like hallucinations has been defended in Martin (2004, 2006) and Fish (2009) and criticized in Sturgeon (2000) and Siegel (2008), among others.

4. Just like other proponents of the epistemic conception, I do not have much to say in this chapter about hallucinations that are subjectively discriminable from perceptions. I address the issue of what they have in common with perception-like hallucinations in Dorsch (2010c).

5. There is one notable difference between the two readings. If we cannot discriminate the character of a given experience first-personally, we cannot discriminate it at all. Even reductionist physicalism has to assume that our canonical access to what is to be reduced is first-personal. Otherwise it would be impossible to decide which brain states are to be identified with pain, say, and which with pleasure. So if the character of a given hallucination is distinct from that of perceptions but cannot subjectively be known to be distinct from perceptions, we cannot say anything positive about it at all. By contrast, the structure of experiences is open to third-personal investigation. Indeed, identifying the subjectively accessible aspects of the structure may be impossible without relying on such a third-personal access (cf., for instance, the external determination of content). Hence, even if we cannot tell from the inside that a given hallucination differs structurally from perceptions, the sciences or metaphysics can still reveal their structure (Dorsch, 2010c).

6. See the writings of Martin, especially Martin (2004, 2006). In §§13–14 and 22, I characterize in more detail both this version of disjunctivism—which is sometimes also called “naive realist disjunctivism” (see Martin, 2002; and Nudds, this volume) or “phenomenal disjunctivism” (Macpherson & Haddock, 2008)—and its understanding of subjective indiscriminability in terms of introspection. The term “disjunctivism” and its counterpart “conjunctivism” are, if taken literally, perhaps not particularly apt, but I follow the tradition in using them for the positions at issue.

7. Martin (2004, 2006) and Nudds (2010) are notable exceptions, though they limit their discussion to introspection.

8. The following considerations draw heavily on Williamson (1990) and, to some extent, also on Martin (2006).

9. This should be obvious for cases in which both entities are accessible to us at the same time and in the same way (e.g., when we can simultaneously see them). In other cases, however, it is less clear whether indiscriminability is symmetric (e.g., when we see one entity while merely remembering the visual appearance of the other). Consider the example of my coming across a certain twin earlier today and being unable to tell which of the two he is, although I was able to recognize the identity of the other (or the same) twin when meeting him yesterday. This case would not pose any problem for the symmetry claim if it were true that, if I would have met today the other twin instead, I would not have been able to tell him apart from his twin as well; or if it were true that a change in the identity of the twin met earlier today would have led to a significant change in the epistemic situation (e.g., if I would have noticed a distinctive feature of the second twin—such as a mole—when seeing him, which I did not notice when remembering his visual appearance; or if I would not have remembered a distinctive feature of the first twin, which I did see). However, it is unclear whether the truth of the idea that the actually seen twin is indiscriminable for me at the time of seeing him from the merely remembered twin requires that also at least one of these counterfactuals is true.

10. The relevant comparison class of perceptions might perhaps be demarcated more precisely by reference to a certain shared content of some sort—assuming that this would also take into account parameters like spatial point of view, lighting conditions, and so on. One difficulty with this approach is that it might not be able to capture all factors that influence whether perceptions are mutually indiscriminable from the inside or not. Another problem is that it would not be compatible with views on perception that deny their having a content of that kind (Travis, 2004).

11. Note that the thesis labeled “(S)” in Martin (1997a) is a completely different claim from the one discussed here.

12. In what follows, I ignore two possible mixed views. The first claims that conjunctivism is true of some cases, and disjunctivism of others. My objections to the general form of disjunctivism have the same force against the disjunctivist part of this view. The second mixed theory adopts a middle way between conjunctivism and disjunctivism about character by maintaining that the hallucinations share some, but not all, character determinables with the corresponding perceptions. Although I think that this position merits more detailed discussion, I would surmise that it, too, faces problems similar to those for disjunctivism about character in its pure form. Besides, it is not clear what could plausibly motivate us to endorse it, rather than one of its two more radical rivals.

13. It is perhaps more precise to say that disjunctivism about character takes the satisfaction of (S) to be the distinctive mark of perceptual experiences but then understands (S) in terms of introspective indiscriminability. In §44, I return to this issue and argue that a conjunctivist about character should reject this focus on introspection and replace it with a focus on experience.

14. See Martin (2004, 2006). Again, I ignore the possibility of weaker variants of experiential disjunctivism, such as the view that only certain, but not all, character determinables of perceptions are relational properties—which would be compatible with, but not necessitated by, the view that hallucinations may share some of their character determinables with perceptions.

15. See section 42 for a sketch of an intentionalist explanation of how hallucinations render certain beliefs reasonable in situations in which we are unaware of their hallucinatory status, even if they do not enable us to satisfy truth- or knowledge-related norms. If one—like McDowell (1998)—prefers to count judgments and beliefs as reasonable only if they conform to all epistemic norms (in addition to being formed in the light of the evidence available), my subsequent considerations should be read as being exclusively concerned with the partial reasonableness coming with the blameless formation of a judgment in response to a perceptual experience, the perceptual status of which is not under doubt. For the sake of simplicity, I do not mention further in what follows this potential partiality of the reasonableness of our reliance on perception-like hallucinations.

16. See Martin (2004, 66) for a similar proposal. Again, this line of thought is compatible with the idea that perception-like hallucinations cannot figure as grounds of knowledge (McDowell, 1998). Rendering the formation of a certain perceptual or introspective judgment reasonable relative to the subjective take on reasons is distinct from putting the subject into the position to acquire the related piece of perceptual or introspective knowledge.

17. A similar general thought is central to Siegel’s observation that some instance of unknowability (namely, the subjective indistinguishability of a hallucination from a perception of, say, a sausage) does not suffice to ground some instance of knowability (namely, that we can come to know, from the inside, that the hallucination in question is not a perception of something else, say, a pyramid) (Siegel, 2008, 218).

18. See Martin (2002) and Nudds (2010). See also McDowell (1998) for a very similar epistemic claim in the context of a slightly different form of disjunctivism about perceptual experiences.

19. One question here is whether conscious presentation can be such as not to allow for error, as proponents of the idea of relational awareness are claiming. In particular, it is unclear how a presentation can get it “right” (e.g., count as “veridical” or a “good case”; see Martin, 2006) if it could not go “wrong” in any way (see Dretske, 1986).

20. This challenge does not arise for views that maintain that we are aware only of external objects, but not of how we are aware of them, or of any other aspect of our experiences (cf., e.g., Tye, 1995; Dretske, 1995). But these views are perhaps to be rejected on independent grounds.

21. One issue here is how our recognition of the relationality of perceptions fits together with the observation that attending to our experiences just leads to attending to their external objects (Evans, 1982; Martin, 2000b). I propose a solution to this problem, which is compatible with experiential intentionalism, in Dorsch (2010c). The central thought is that perceptual experiences are given to us as relational by means of our awareness of the external objects involved as the determinants of our experiences.

22. Again, I ignore any middle position according to which hallucinations share such a character determinable with the corresponding perceptions, without thereby sharing the whole character of the perceptions.

23. The second view, which assumes that we introspectively judge that it seems as if we are perceiving, naturally combines with the idea of a preceding experiential or intellectual seeming (e.g., an intuition) that we are perceiving. However, this latter idea introduces the error involved in taking perception-like hallucinations to be perceptions already at a pre-introspective level and therefore cannot be adopted by someone trying to capture this error in purely inferential terms. It is therefore more plausible to prefer talk of introspective judgments about having a perceptual experience over talk of introspective judgments about it seeming that one has a perception.

24. See Dorsch (2009) for a discussion of our experience of mental images and thoughts as active or passive. The distinction between the two ways in which things may be given to consciousness may also be cashed out in terms of “experiencing something” versus “experiencing something as being a certain way”—whereby “experiencing” is equivalent to the German erleben, and to be distinguished from the more narrow notion of “experiencing something perceptually or sensorily.”

25. See Dorsch and Soldati (2010) for a detailed discussion of the motivation and nature of the resulting experiential intentionalism about perceptual experiences and their subjective character.

26. Of course, there is more to intentionality—notably subjectivity. Intentional presentation is always presentation to a subject, or a subjective perspective, and therefore to a waking or dreaming conscious mind. The normative element then distinguishes intentional from nonintentional consciousness, that is, something being presented to consciousness as being a certain way from something merely being present in consciousness.

27. Perhaps not all attitudinal and presentational differences among intentional states can be captured in terms of normative differences. But this approach may promise to go a long way. For example, it may be plausible to say that desires should occur only if something is valuable, instances of visualizing only if something is visible, and imaginative thoughts only if something is possible or, perhaps more generally, thinkable.

28. I discuss this feature of perceptual experiences in more detail in Dorsch (2010b).

29. Again, these considerations stay neutral on the issue of whether hallucinations may provide us with reasons or merely seem to do so. Correspondingly, they stay neutral on whether perceptual reasons consist in the relevant aspects of the world and our relation to them or, alternatively, in our fallible awareness of those aspects and our relation to them.

30. In Dorsch (2010b), I discuss another sense in which perceptions enjoy priority over perception-like hallucinations: such hallucinations possess a specifically perceptual character, that is, a character that is characteristic of perceptions, and not vice versa.

31. See Siegel (2008), who takes this to be another problem for experiential disjunctivism. But see also Martin’s reply to this objection in Martin (2006). Martin nonetheless accepts the falsity of higher-order accounts of consciousness.

32. See also Dorsch (2010b) for the related discussion of how experiential intentionalism can provide an account of the transparency and nonneutrality of episodes of visualizing, which is very similar to that put forward by experiential disjunctivism.

33. See Dorsch (2010a) for a similar conclusion in the case of our experiences of color similarities.