Hallucination

In this paper, I argue that the indirect realists’ recourse to mental representations does not allow them to account for the possibility of hallucination, nor for the presentational character of visual experience. To account for the presentational character, I suggest a kind of intentionalism that is based on the interdependency between the perceived object and the embodied perceiver. This approach provides a positive account to the effect that genuine perception and hallucination are different kinds of states. Finally, I offer a tentative suggestion as to how a hallucinatory experience may still be mistaken for a genuine perceptual experience.

The first part of the argument from hallucination is intended to show that hallucinations involve the awareness of mental particulars, and the second, generalizing part concludes that the same kind of particulars are involved in veridical perception. All parties agree that what necessitates the adoption of mental particulars in the case of hallucination is the acceptance of an assumption that Robinson (1994, 32) called “Phenomenal Principle”:¹

If there sensibly appears to a subject to be something which possesses a particular sensible quality then there is something of which the subject is aware which does possess that sensible quality.

Thus when a subject hallucinates a blue square, there is something of which the subject is aware that is actually blue and square; and since there is no blue square in the external environment, the subject is aware of a mental particular that is blue and square.

Intentionalists attempt to resist this conclusion by pointing to the phenomenology that characterizes both hallucinatory and genuine experience: “In turning one’s mind inward to attend to the experience [of a blue square], one seems to end up concentrating on what is outside again, on external features or properties. And this remains so, even if there really is no blue square in front of one” (Tye, 1995, 30). Hallucinatory experience, like genuine experience, is transparent; introspection of it does not reveal any mental particulars and their qualities, but only objects and their qualities appearing in the external environment. Thus the acceptance of the Transparency Thesis, prima facie, seems to undermine the idea that when one hallucinates, one is aware of mental particulars and their qualities.²

But an indirect realist could respond that though what we are aware of appears to be an external object with its qualities, it is actually a mental particular with its qualities. In other words, we are aware of the mental particulars not as mental particulars but as external objects. This means that the Phenomenal Principle can be compatible with the Transparency Thesis.

In this paper, after setting the stage in the first section, I argue in the second section that the acceptance of this compatibility does not allow the indirect realist to account for the possibility of hallucination. This calls for a close examination of the postulated mental particulars and of the way they are taken to be involved in perception and hallucination. Here I critically examine one dominant way mental particulars are conceived by indirect realists, namely, as mental representations of a particular kind, and I argue that this conception does not account for the possibility of hallucination. In the third section, I argue that the way indirect realists could use mental representations to account for the presentational character of visual experience does not do justice to its phenomenological difference from visual imagination or visual recall. In the fourth section, I suggest a kind of intentionalism that is based on the interdependency between the perceived object and the embodied perceiver in order to account for the presentational character of perceptual experience. This approach provides a positive account to the effect that genuine perception and hallucination are different kinds of states. Finally, in the fifth section I offer a tentative suggestion as to how a hallucinatory experience may still be mistaken for a genuine perceptual experience.

Indirect realists seem, prima facie, better placed to account for the distinctive phenomenology of perceptual-like experiences. By postulating an act of direct awareness of actually present mental particulars as the core of perception, they make these particulars constituents of perceptual content. Thus the presentational-sensory character of perception supposedly stems from the very involvement of actually present particulars in the content of perception, and in that sense, perceptual content literally presents them. The same account holds for the presentational-sensory character of hallucinatory experiences: it is supposedly explained by what hallucination shares with veridical perception, namely, the direct awareness of these actually present mental particulars. Moreover, since no similar kind of awareness is involved in thought, we have readily an account of why thought lacks presentational-sensory character. It is this account of the presentational-sensory character of experience that explains the indirect realist’s motivation for accepting the Phenomenal Principle in the case of hallucination: thus, if there sensibly appears to a subject to be something that possesses the sensible qualities of being red and cubical, then there is something of which the subject is aware, namely, a mental particular, that does possess these sensible qualities. In other words, the indirect realist accounts for the presentational-sensory character of perceptual-like experiences by reducing it to the awareness of the instantiated properties of a mental particular.

But how does this move allow the indirect realist to deal with perceptual error? Let us take the case of hallucinating a red cube. How could this be possible within the indirect realist’s framework? The obvious answer would be: by being directly aware of a mental particular that is red and cubical. It is the ability to grasp these instantiated properties of the mental particular that accounts for the possibility of perceptual error. But since, according to the indirect realist approach, the same holds in the case of veridically perceiving a red cube, this cannot be the whole story. What ultimately differentiates veridical from nonveridical perception is external to the act of direct grasping of a red and cubical mental particular, and it has to do with the causal chain that brought about the mental particular. Thus there is nothing within the experience itself that could allow the subject to differentiate a veridical from a hallucinatory experience. And this opens up the metaphysical possibility that for every veridical experience there could possibly be a phenomenologically indistinguishable hallucination—an assumption that is crucial for the success of the generalizing step of the argument from hallucination.

This solution raises the well-known skeptical problem of how I can know that things are as my experience presents them as being. But the formulation of this epistemological problem already presupposes that our experience presents things as being a certain way, for example, as being of a red cube. Thus indirect realism faces a more fundamental problem: “How can my experience so much as be intelligibly of an external world? . . . What sort of unity must experience possess in order to be able to present an appearance about which the question could arise ‘Shall I endorse it?’” (Conant, 2004, 100).⁴ The indirect realist seems to take for granted the intentionality of experience, and, a fortiori, the possibility of having a nonveridical experience (say, of a red cube), and just infers from this a thesis about its nature, namely, that a nonveridical experience involves the awareness of instantiated properties of mental particulars. The indirect realist does not explain how the mere awareness of a red quality and of a cubical quality of a mental particular suffices to give rise to a nonveridical experience of a red cube.

Thus the indirect realist does not seem to address at the right level the problem of the possibility of perceptual error: he seems unable to bridge the gap between, on the one hand, the mere sensory consciousness that the purported awareness of mental particulars provides and, on the other, the perceptual-like consciousness that we appear to entertain—as it is presupposed in the Transparency Thesis. How could the indirect realist bridge this gap?

A first response would be to insist that there is no gap to be bridged: the experience of the red cube should be intended as nonconceptual. Thus the perceiver does not have to recognize that what he sees is a red cube to have the particular experience. But even so, the very phenomenology of perceptual-like experience is quite unlike any awareness of instantiated properties of mental particulars: the experienced entities appear as situated in the external world and as having a perceptual unity. In other words, we do not experience disparate sensible qualities of an entity floating around in some “inner” space; we experience distinct entities possessing spatiotemporally organized sensible qualities and situated in the external environment along with other entities of this kind. In other words, we experience the entities as objects in the external world. This is the way we perceptually entertain the world before we attempt to introspect our perceptual experience. The Transparency Thesis presupposes this phenomenological description of perceptual experience and adds that it does not change if we attempt to introspect our experience.

The indirect realist cannot deny as a purely phenomenological claim that the experienced entities appear as situated in the external world and as having a perceptual unity. In fact, he seems to presuppose this phenomenological description in the first part of the argument from hallucination, where he refers to how things look in a hallucinatory experience. Actually, this description is involved as an antecedent in the Phenomenal Principle (“If there sensibly appears to a subject to be something that possesses a particular sensible quality [say, a red cube] . . .”). The indirect realist also presupposes it in the generalizing step of the argument from hallucination in the way he describes the phenomenologically indistinguishable hallucinatory and veridical experiences. That is, the indirect realist explicitly appeals to the phenomenology of perceptual-like experience, and when he does so, he describes it in a commonsensical way—a description also presupposed by the Transparency Thesis.

So the problem that the indirect realist faces is twofold. On the one hand, he has to give an account of how the awareness of instantiated properties of mental particulars brings about the awareness of external entities with perceptual unity. On the other hand, he has to respect the restriction set at the phenomenological level that from the subject’s perspective there appears only one kind of object—that is, an external entity with perceptual unity—and this remains so even if we attempt to introspect to our perceptual experience.

One idea for dealing with this problem is to suggest that some kind of systematic error is involved here: perceivers systematically mistake the awareness of instantiated properties of mental particulars for the presentation of external objects with a perceptual unity.⁵ This solution allows the indirect realist to retain the Transparency Thesis as a purely phenomenological thesis, namely, as a thesis that describes the phenomenology of perceptual-like experience. On this account, the indirect realist—in accordance with the Phenomenal Principle—must reject any further claim that the Transparency Thesis does justice to the nature of perceptual-like experience. However, for this suggestion about the systematic mistake not to be just a naming of the problem, the indirect realist ought to give a positive account of what makes the systematic error possible. In other words, he has to answer the question of in what way the mere awareness of certain instantiated properties of a mental particular is systematically mistaken for an experience—veridical or not—of an entity that appears (i) to be situated in the external environment and (ii) to have a perceptual unity. Let us call this “the bridging problem.”

Another option is to equip the mental particular itself in such a way that the mere awareness of it would be sufficient to give rise to the experience of an external entity with a perceptual unity. This could be achieved by taking the mental particulars to be mental representations, namely, entities that stand for something else. In what follows, I examine this second option and argue against it.

Indirect realists are forced to appeal, in particular, to pictorial mental representations, namely, to mental pictures, to preserve the sensory-presentational character of the experienced entities. The reason is that only the vehicle of a pictorial representation instantiates (some of) the sensory properties of the represented object. For example, an external picture of a red cube has a vehicle that instantiates redness and cubicalness. Thus if mental particulars are mental pictures, the perceiver is aware of instantiated properties that account, according to the indirect realist, for the presentational-sensory character of perceptual-like experience, namely, for their distinctive phenomenology.

But how do mental pictures account for the bridging problem? This is supposed to be achieved specifically by the picturing function of mental particulars. In short, the idea is that just as an external picture with the help of its sensible properties depicts certain other entities, so a mental picture depicts external entities with perceptual unity.

But one should be cautious not to overemphasize this analogy between the awareness of the supposed mental pictures and the perception of an external picture. When we perceive an external picture, we are able to differentiate the pictorial vehicle from the pictorial content. This is exemplified in our ability to focus our attention either on the vehicle or on the content. For example, when I perceive the picture of a tree, I can focus my attention either on the vehicle of the picture (e.g., on the quality of the paper on which it is printed) or on the content of the picture, namely, on the tree. The depicted object is taken as having no causal influence on the objects of our current environment: the tree is not taken as something that can actually provide me shelter. If one does not distinguish the pictorial vehicle from the pictorial content, one has the illusion that what one sees is actually present and not just a depiction—in other words, one does not take what one sees as a picture at all.

Figure 16.1

René Magritte, La condition humaine I.

For example, Magritte’s painting La condition humaine I depicts a second painting within it. However, when one sees Magritte’s painting for the first time, one hardly distinguishes this second painting. The reason for this is that Magritte has eliminated most of the features that depict the vehicle of the second painting. As a result, the spectator manages to distinguish or infer some features that depict the vehicle of the second painting, and thus to see it as a painting, only after attentive examination. Such features are the easel on which paintings are usually set, the edge of the left curtain, and finally the right side of the canvas.

Thus this analysis shows that the awareness of a putative mental picture should be more like the perception of an external picture that we do not recognize as such than like the perception of one that we do recognize as a picture. Only in this way can the indirect realist remain faithful to the constraint that, though what we are actually aware of are supposedly the instantiated properties of a mental particular, we erroneously take them as properties of an external object. This shows that the putative mental pictures function more like what we ordinarily would call “substitute” than like what we ordinarily call “picture”—in the sense of a species of representation proper. But a substitute is not a picture. For example, saccharin is usually used as a substitute for sugar, but this does not make the former a picture of the latter. The reason is that a substitute of A—if successful—should not be distinguishable from A, whereas a picture of A should be distinguishable from A to function as a picture.

Thus the suggestion that mental particulars could be taken as mental pictures cannot be what the indirect realist wants, since the analogy with perception of external pictures breaks down at a crucial point for his argument; namely, perception of pictures requires a twofold awareness of the depicted object and of the pictorial vehicle. But the putative awareness of mental particulars should not provide this double object awareness (of a mental vehicle and an external object), because that would falsify the phenomenology of perceptual-like experience as it is described by the Transparency Thesis.⁷

Thus the indirect realist is left only with the notion of mental substitute to deal with the bridging problem. But how is it possible for the awareness of a mental substitute to give rise to an experience of an external entity with a perceptual unity? Indirect realists have traditionally appealed to resemblance and to causation: mental substitutes resemble (some of) their causes.⁸ Of course, indirect realists themselves did not find it plausible to apply the idea of resemblance to secondary qualities. Moreover, the application of this idea to primary qualities, such as shape and size, would require presupposing some kind of inner space much like external space. Some theorists suggest that this inner space should be two-dimensional; others speak of a three-dimensional space.⁹ Either way, the idea of resemblance forces indirect realists to attribute to this inner space, and to the mental particulars that populate it, a significant part of the perceivable properties of the external world. Thus whereas the bridging problem asks for an account of how the mere awareness of instantiated properties of mental particulars brings about the experience of entities that have a perceptual unity and appear to be situated in an external space, the acceptance of resemblance on the part of indirect realists forces them to already endow mental particulars with a perceptual-like unity and a kind of spatiality. In that sense, they seem to already presuppose much of what they need to show.

In addition to this fundamental problem, resemblance and causality introduce indeterminacy to perceptual content, since they are both very “cheap,” so to speak. Everything resembles everything in some respect or other.¹⁰ On the other hand, the causal chain that produces the putative mental image has a great number of causes both before and after the distal cause that is supposedly represented; so how is this particular cause selected instead of any other?¹¹ Moreover, given that there are infinite different causes that could produce the same mental image, how is it that the content of the mental image is not as vague as an infinite disjunction of causes?¹²

The problem of reducing the content of mental representations to some kind of causal relation is a hotly debated issue in contemporary representational theory of mind. But this new approach constitutes a radical reconceptualization of the problem because it naturalizes mental representations. In particular, it takes them as having a physical vehicle that is in principle inaccessible to the subject. In that sense, mental representations are conceived as subpersonal and as situated in the brain. Thus subjects are supposed to have access only to the content of (some of) the mental representations, but not to their vehicle.

The accordance of this new approach with the Transparency Thesis, but, mainly, the prospect of providing a naturalistic account of the intentionality of the mental, led many contemporary supporters of intentionalism to embrace the representational theory of mind. But the assumption that the intentionality of perceptual states can be accounted for by reducing it to the possession of subpersonal mental representations is not constitutive of intentionalism, since one could hold that perceptual states are intentionally directed toward an object—or represent, in this weak sense, an object—without being committed to the existence of subpersonal mental representations.¹³ Such an intentionalist uses the expression “perceptual experience represents a red cube” as a synonym for the expression “perceptual experience is intentionally directed toward a red cube,” while he rejects the further claim that the intentionality of perceptual experience is due to the existence in the brain of a particular that functions as a subpersonal mental representation. In that sense, an intentionalist is not necessarily a representationalist.

Thus far, I have argued against the idea that the postulation of mental pictures available to awareness suffices to account for what I called “the bridging problem,” which is actually one way to formulate the problem of intentionality for indirect realism. On the other hand, we saw that intentionalists who embrace subpersonal mental representations (i.e., representationalists) attempt to account for the intentionality of perception by recourse to some kind of naturalistic relation holding between the subpersonal mental representation and the represented entity or feature.¹⁴ However, I am not going to examine these attempts, since their possible success establishes, at most, the claim that there are physical states in the brain that possess intentionality and are inaccessible to awareness; and this claim does not support the indirect realist’s claim that perceivers are aware of mental pictures, namely, of nonphysical vehicles that are systematically mistaken for external entities with perceptual unity. Thus we can conclude that the postulation of mental pictures does not provide the indirect realist with an account of the bridging problem or, a fortiori, of the possibility of hallucination.

But how could we account for the presentational character of visual experience if not in terms of the awareness of mental pictures? One option is to embrace a strong version of naive realism and to attempt to account for the presentational character in terms of the direct awareness of certain instantiated properties of physical particulars. This approach makes physical particulars and their instantiated qualities the constituents of perceptual content; in that sense, the approach is structurally analogous to that of indirect realism, but it has no implication of attributing presentational character to visual imagination or visual recall, since only perceptual content has physical particulars as constituents. Moreover, it seems to assimilate presentational character with sensory character in the sense that it can attribute the presentational character to the presence of the instantiated sensible properties of physical particulars.

However, the postulation of an act of direct awareness of certain instantiated properties of physical particulars makes this strong version of naive realism vulnerable to a version of the bridging problem: in what way is the mere awareness of certain instantiated properties of a physical particular systematically mistaken for an experience of an entity that appears (i) to be situated in the external environment and (ii) to have a perceptual unity?

For this reason, I focus on a weaker version of naive realism that seems not to be vulnerable to this general formulation of the bridging problem. This weak version holds that the constituents of perceptual content are object-dependent senses, namely, a special kind of intentional content.¹⁶ McDowell (1994/1996) is one prominent supporter of this view.

Can this view account for the presentational character of visual experience? I think not. The problem is that object-dependent senses may constitute not only perceptual content but also the content of object-dependent thoughts. More particularly, McDowell attempts to differentiate perceptual experience from perceptual judgment by recourse merely to the attitude that they involve, while he holds that experiential content may be shared by perceptual judgments: “A judgement of experience does not introduce a new kind of content, but simply endorses the conceptual content, or some of it, that is already possessed by the experience on which it is grounded” (McDowell, 1994/1996, 48–49; italics mine).¹⁷ Thus the presentational character of perceptual experience cannot be accounted for by recourse to the object-dependent senses, because this would commit one to attributing presentational character even to object-dependent thoughts.

But perhaps we could use the way McDowell draws the distinction between perceptual experience and thought to account, in a different way, for the presentational character of perceptual experience. McDowell’s suggestion is that experience, unlike belief, involves a kind of awareness without endorsement: “We need an idea of perception as something in which there is no attitude of acceptance or endorsement at all, but only, as I put it, an invitation to adopt such an attitude, which, in the best cases, consists in a fact’s making itself manifest to one” (McDowell, 2002, 279). McDowell uses this idea of awareness without endorsement to capture the subject’s transparent experiential openness to the world. In this sense, he minimizes any effect on the part of the subject on the object-dependent perceptual content. Thus we end up with a nonanalyzable act of awareness and an object-dependent perceptual content that, as I have argued, is not sufficient to account for the presentational character of perceptual experience.

McDowell’s image for experience as “an invitation—a petition . . . to accept a proposition about the objective world” (2002, 278; italics mine) does not capture the force with which perceptual experience is imposed on us, a force that puts an end to the need for further justification. This force stems from what is distinctive about experience: its presentational character. I would like to suggest that to account for the presentational character, we should fully exploit what is relational about perception. I will call this approach to perceptual experience “relational intentionalism.”

Visual content, unlike the content of object-dependent thoughts, does not depend just on the existence of the object; rather, it depends on the current availability of the object itself within our visual field. This is why visual content is not “portable”: though I can think (or visually recall) now of an object I saw somewhere yesterday, it is impossible to see it now. Thus perceptual content is relational not only because it is object dependent but also because it is egocentric, dependent on the embodied perceiver. We do not just perceive external objects; we perceive them from here, to our left or to our right, far away or close to us. This is why what we see does not depend only on the object; it depends also on our position and our movement or rest. If we do not integrate this element in our approach to perceptual experience, the mere object dependency confines the subject to a view from nowhere, transforming the perceiver into a disembodied passive spectator of the outer show that unfolds completely independently of him. Such a perceiver would not feel to be in the environment of which he is aware.¹⁸ And that would result in a kind of experience without presentational character.

An approximation of this experience is the experience we have when we see a movie or a photograph. As Currie (1995, 66) notes, photographs, unlike ordinary seeing, do not convey egocentric information:

With ordinary seeing, we get information about the spatial and temporal relations between the object seen and ourselves. . . . Photographs on the other hand do not convey egocentric information; seeing a photograph does not tell me anything much about where the object photographed is in relation to me.

This claim should, however, be moderated, since even when we see a photograph, we get egocentric information about its representational vehicle (e.g., we are aware of its position relative to us); this is necessary, as I argued in section 2, for seeing something as a kind of depiction. Thus seeing a photograph or a movie is just an approximation of an experience where the perceiver does not feel to be in the environment that he is aware: though the perceiver of a photograph does not feel to be in the depicted environment, he feels to be in the environment that the representational vehicle belongs.

Thus, to account for the presentational character of perceptual experience, we have to analyze the interdependency between the perceived object and the perceiving subject. I suggest that we can account for the presentational character of visual experience in terms of the relation between the awareness of the object and the implicit bodily awareness of the perceiver. To take a simple example, when we look at a static object and move our eyes, we do not experience the object as moving; we experience it as remaining still, due to the implicit bodily awareness of our movement. Thus there is a close coordination, or rather coupling, between the awareness of the object and the implicit bodily awareness of the perceiver. When this coordination breaks down, what we are aware of loses its presentational character. This is why the appearance of an afterimage in our visual field is not experienced as perceiving. The afterimage does not afford us any appropriate way to explore it, since it depends completely on the movement of our eyes.

To summarize, the presentational character of visual experience does not rest on the way things sensorily look or in how the way they sensorily look changes. Rather, it rests on how the way things sensorily look changes relative to our position and our bodily movement or rest. In other words, the presentational character is manifested amodally and does not amount to any kind of sensory presence. Thus even if it were possible to have a visual-like experience where the changes in how things sensorily look were indistinguishable from such changes in a genuine visual experience, the subject could still be aware of a difference that would be revealed as a sense of passivity that would characterize the former case of awareness. In other words, hallucinatory content, by not being dependent on any objects in the subject’s visual field, is independent from the subject’s exploratory eye movements, and this explains the sense of passivity that may accompany a hallucinatory experience.

Thus I have suggested that to account for the presentational character of visual perception as an intrinsic feature of it, we should analyze the act of perceptual awareness. I proposed that it involves a double awareness, which is, however, different from the one that some indirect realists may tend to introduce—namely, an awareness of two objects. My suggestion is that besides the awareness of the perceived object, visual perception also involves a kind of bodily self-awareness that is implicit and intransitive or pre-reflective.¹⁹ This means that it does not inform us of our experiencing as an object; it is not an intentional mode of awareness.

Thus when I see a rat, I am not only aware of the rat; I am also implicitly and pre-reflectively aware of my orientation, my distance, and my movement compared to the rat. But this presupposes that I am pre-reflectively aware of my seeing the rat. Minimally, I cannot be in a state of perceiving without being pre-reflectively aware that this perceiving is mine.²⁰ My awareness of the sensible qualities of the particular and of the way it changes relative to my position and my bodily movement or rest makes me pre-reflectively aware of the kind of experience I entertain, namely, of whether, for example, I see, hear, or visually imagine something. Of course, I can be wrong about the kind of awareness, but the point is that there must always be a pre-reflective awareness of this sort when I am intentionally directed toward something, since perceiving something, as opposed to, say, imagining it, has a different cognitive significance for the subject. When I visually imagine a lion, as opposed to seeing it, I do not have, ordinarily, the tendency to run away. The awareness that I am perceiving is not inferred at a later stage from the content of perception but is an integral part of conscious perception.

Does this suggestion undermine the Transparency Thesis? It depends on how one understands this thesis. If it amounts to the claim that if one attends to one’s experience, the only objects and properties that one is aware of are the objects and properties of the world that one’s experience presents, then the foregoing suggestion does not question this formulation of the Transparency Thesis. But if the thesis implies also the clause that no other kind of awareness is involved, the suggestion questions the Transparency Thesis. Thus relational intentionalism requires a modification of the Transparency Thesis. Following Thompson (2008, 402), we can state the modified thesis as follows:

Moderate Transparency of Attention: We can (with effort) attend to (intrinsic mental features of) our experience, but not by turning our attention away from what that experience is of (i.e., what is presented by that experience).

This point can be reinforced by giving a positive account of how it is possible to mistake a hallucinatory experience for a veridical one. Very briefly, I would like to tentatively put forward a hypothesis about a possible mechanism of hallucination production, namely, a mechanism of production of mental states that are intended as visual experiences. The suggested hypothesis is that a visual hallucination can be produced by a loss of the sense of agency in visual imagination or visual recall.

Let us focus on visual imagination. Concerning the intentional structure of visual imagination, I will follow Martin’s Dependency Thesis that “to imagine sensorily a φ is to imagine experiencing a φ.”²² Before applying this in relation to the production of visual hallucinations, I would like to add a small but important modification to the Dependency Thesis suggested by Thompson in line with his moderate transparency of attention thesis. Thompson (2008, 408), based on Husserl’s work, suggests that “the intentional object [of the imagining] is the visualized object[;] . . . the correlative and co-imagined visual experience is experienced only intransitively and pre-reflectively.” Thus when I visually imagine a pink rat, though I am imagining that I am visually experiencing a pink rat, I am transitively aware only of the pink rat, while my awareness of the experiencing act is intransitive.

Now, if a loss of the sense of agency occurs while I am visually imagining a pink rat, that loss would result in a peculiar mental state that would have as its content a visually appearing pink rat and the intransitive awareness of my visually experiencing it. In this case, we could hypothesize that the loss of the sense of agency results not in an alien imagination but in a visual-like experience of a pink rat because of the information—involved in the content of the imagining—that the experiencing of the pink rat is mine. This further explains the sense of passivity that characterizes hallucination since, although the patient takes the hallucination to be a perceptual experience, he cannot affect its content by his bodily movements unlike what happens in the case of perceptual experience proper. If this suggestion proves tenable, the problem posed by hallucinatory content will be reduced to the problem posed by the content of visual imagination (or visual recall). However, these issues need further examination, which I hope to undertake on a future occasion.

This work was supported by the Research Committee of the University of Patras, Greece, Program Karatheodori, project no. C597.

Conant, J. (2004). Varieties of skepticism. In D. McManus (Ed.), Wittgenstein and skepticism. London: Routledge.

Crane, T. (2006). Is there a perceptual relation? In T. Gendler & J. Hawthorne (Eds.), Perceptual experience. Oxford: Oxford University Press.

Currie, G. (1995). Image and mind: Film, philosophy, and cognitive science. Cambridge: Cambridge University Press.

Fodor, J. (1990). A theory of content and other essays. Cambridge, MA: MIT Press.

Goodman, N. (1972). Seven strictures on similarity. In N. Goodman, Problems and projects. Indianapolis: Hackett.

Harman, G. (1990). The intrinsic quality of experience. In J. Tomberlin (Ed.), Philosophical perspectives (Vol. 4). Atascadero, CA: Ridgeview.

Haugeland, J. (1998). Having thought. Cambridge, MA: Harvard University Press.

Husserl, E. [1900] (1970). Logical investigations (Findlay, J. N., Trans.). London: Routledge & Kegan Paul.

Jacob, P. (1997). What minds can do: Intentionality in a non-intentional world. Cambridge: Cambridge University Press.

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

McDowell, J. (1994/1996). Mind and world. Cambridge, MA: Harvard University Press.

McDowell, J. (2002). Responses. In N. Smith (Ed.), Reading McDowell: On mind and world. London: Routledge.

McDowell, J. (2008). Avoiding the myth of the given. In J. Lindgaard (Ed.), McDowell: Experience, norm, and nature. Oxford: Wiley Blackwell.

Millikan, G. R. (1984). Language, thought, and other biological categories: New foundations for realism. Cambridge, MA: MIT Press.

Putnam, H. (1994). Sense, nonsense, and the senses: An inquiry into the powers of human mind. Journal of Philosophy, 91(9), 445–517.

Sellars, W. (1989). The metaphysics of epistemology. Atascadero, CA: Ridgeview.

Siegel, S. (2006). Direct realism and perceptual consciousness. Philosophy and Phenomenological Research, 73(2), 378–410.

Smith, A. D. (2002). The problem of perception. Cambridge, MA: Harvard University Press.

Sterelny, K. (1990). The representational theory of mind: An introduction. Oxford: Blackwell.

Strawson, P. F. (1966). The bounds of sense: An essay on Kant’s Critique of Pure Reason. London: Routledge.

Thompson, E. (2008). Representationalism and the phenomenology of mental imagery. Synthese, 160, 397–415.

Zahavi, D. (2005). Subjectivity and selfhood: Investigating the first-person perspective. Cambridge, MA: MIT Press.

1. Robinson introduces this principle in the context of his analysis of the assumptions involved in the argument from illusion, but the same principle must be used in the argument from hallucination in order, from the commonsensical idea that in hallucination we appear to see something that does not exist, to infer that in hallucination we are aware of a mental particular. See also Smith (2002, 194) and Crane (2006, 135).

3. There are, of course, intentionalist accounts of the phenomenal character of perceptual-like experience that focus on other phenomenological features of perceptual content, such as the fineness of grain or the analog character. These accounts attempt to reduce the particular features to the possession of some kind of nonconceptual intentional content. But these are features that perceptual content shares, to some extent, with the content of quasi-perceptual states such as visual imagination and visual recall, and in that sense, they do not constitute what is distinctive of the phenomenology of perceptual-like experience.

5. See, e.g., Prichard (1950, 68): “What we call seeing or feeling a body consists in genuinely mistaking certain sensa for a body.” See also Sellars (1989, 36–44) for a critical discussion of this approach.

7. Moreover, perceptual-like experience presents external objects as actually present, unlike the depicted objects of a putative experience of a mental picture. For more on this issue, see sections 3 and 4.

10. See, e.g., Goodman (1972) and Goodman (1968) for further arguments against the idea that one can account for representation in terms of resemblance.

13. See, e.g., Husserl ([1900] 1970), Haugeland (1998), McDowell (1994/1996), and Putnam (1994). For example, an intentionalist could take intentionality to be an irreducible feature of perceiving and attempt to illuminate the concept of intentionality by showing its connections with other concepts, such as those expressing the possession of certain capacities, like discriminatory, recognitional, linguistic, and inferential capacities.

14. Most notable among these projects are the asymmetric dependence theory (Fodor, 1990) and the teleosemantics (Millikan, 1984).

15. As I noted at the beginning of section 2, there is a different path (which I will not explore in this paper) that an indirect realist could follow for dealing with the bridging problem and, more particularly, with the problem of the presentational character of perceptual-like experience, and this is to appeal to some kind of cognitive intervention. A suggestion along these lines is to account for the presentational character in terms of a belief in the “physical reality” of sensations. Russell (1921) examines and rejects this suggestion. An obvious problem that the particular suggestion faces is how the addition of a belief can account for the perceptual presence of an object, especially in view of the fact that perceptual content is much less fragile than belief.

16. In that sense, McDowell’s well-known disjunctivism is compatible with a particular kind of intentionalism.

17. McDowell’s view on this issue has recently changed (McDowell, 2008). He now holds that perceptual content is not propositional but still conceptual. In my paper, I will not examine this new approach.

18. McDowell (1994/1996, 102–104) raises a similar objection against Kant’s approach. He attempts, based on his notion of second nature, to overcome this objection by “accommodating the fact that a thinking and intending subject is a living animal” (104). But he does not spell this suggestion out in more detail.

19. This notion has been elaborated in the phenomenological tradition. For a recent discussion, see Zahavi (2005) and Thompson (2008).

20. For a defense of the thesis that awareness of objects is interdependent with self-awareness, see the reading in Strawson (1966, 72–117) of Kant’s Transcendental Deduction.

21. See Crane (2001, 133). If we were infallible about our mental acts, then it would not be possible to experience a hallucination.