The sense of experiential ownership is not about ownership of body parts or a whole body, but about whether I represent myself as the unique subject of experience. As I am typing, for example, I do not only experience tactile sensations in my fingers. I also have a sense that I am the one who is having these tactile sensations. This corresponds to Wittgenstein’s assertion: “To ask ‘are you sure it is you who have pains?’ would be nonsensical.” In this section, I will (1) illustrate that the sense of experiential ownership is different from the sense of body ownership; and (2) draw two distinctions to explicate the sense of experiential ownership. I will then (3) describe some varieties of the immunity principle (IEM); and (4) provide two counterexamples against two major forms of IEM. We will see that, pace Wittgenstein and Shoemaker, we need another way of articulating the distinction between the sense of self-as-object and the sense of self-as-subject.
Moro et al. (2004) describe two patients with somatoparaphrenia. These patients suffered not only from somatoparaphrenia but also from hemispatial neglect and tactile extinction. They denied ownership of their left hand, in which they had no sensation, and their left visual field was lost. So far, we might think that these cases involve only misrepresentation of body ownership. But there is more. When the researcher moved the patients’ left hand to the right-hand side so that they could see it, their tactile sensation was restored. But despite representing themselves as the subjects who felt the sensations, the two patients still denied the ownership of their left hands (2004, p. 440–441). This shows that it is possible to have the sense of experiential ownership without the sense of body ownership. The two types of self-experience are conceptually and empirically dissociable.
To clarify the notion of experiential ownership, let me begin with the point that every phenomenal state has a what-component and a who-component. The what-component includes the representational content and the phenomenal character of that state. The who-component ties the what-component to a unique subject. The basic assumption here is that every phenomenal state has one and only one subject. The sense of experiential ownership is exclusively about the who-component—it concerns whether one experiences oneself as the subject of a phenomenal state. I will now draw two distinctions to further clarify this point.
The first distinction is between the fact of experiential ownership and the sense of experiential ownership. When a subject experiences a phenomenal state, there exists a fact that he is the subject of that state. This fact of experiential ownership is constitutive of every conscious experience—i.e. every experience has a unique subject. For every conscious experience, we can ask “Who is the subject of that experience?” and there exists a fact of the matter. For example, right now it is me, not you, who is experiencing lower-back pain. The fact of experiential ownership is objective in that it refers to a biological fact about whether a subject undergoes a phenomenal state.
When a subject experiences herself as the unique subject of a phenomenal state, she has the sense of experiential ownership, i.e. she experiences herself as the subject of that state. This aspect is captured by the Wittgenstein Question: “Are you sure it is you who has pains?” When a subject answers this question, she relies on her sense of experiential ownership. When I have a tactile sensation, I experience what it is like for me to undergo that sensation. The what-it-is-like aspect, i.e., the phenomenal character, belongs to the what-component. The for-me aspect refers to the subjective sense that I am the one who is having the sensation.[13]
The fact of experiential ownership and the sense of experiential ownership are two different aspects of experiential ownership: the factual aspect and the subjective aspect. These are not numerically different states or events that can be detached from a phenomenal state. Rather, they are two ways of characterizing the who-component of that state. The factual aspect addresses whether a subject experiences a phenomenal state; the subjective aspect concerns whether the subject is conscious of the factual aspect. But many philosophers do not see that these two aspects are not the same. To sustain this distinction, I will later argue that the factual aspect of experiential ownership can be misrepresented, which means that sometimes the Wittgenstein Question can be perfectly intelligible. Misrepresentation, as I shall explain, happens when the subjective aspect fails to match the factual aspect of experiential ownership.
The second distinction is between the first-personal sense and third-personal sense of experiential ownership. Suppose I experience a phenomenal state—say, lower-back pain. Not only do I experience the phenomenal character of the pain but also, in the very same experience, I have the sense that it is me who is experiencing that particular pain. This sense of experiential ownership is first-personal, since it is part of the pain experience rather than resulting from a separate act of reflection. I experience a sense of experiential ownership by experiencing the pain without requiring any further attention or introspection.
Now suppose I participate in an experiment where several subjects receive tactile stimulations in a random order and everyone is simultaneously scanned with fMRI equipment.[14] Later, using the fMRI data on my somatosensory cortex, I can judge whether it was me who experienced a particular stimulation a few minutes ago. In this case, the sense of experiential ownership is considered from the third-person point of view, where the sense of experiential ownership is the content of a further judgment or reflection rather than an integral part of the respective phenomenal states.
I suggest that the sense of self-as-subject is captured by the first-personal sense of experiential ownership. Being conscious of oneself-as-subject just is to experience oneself as the subject of a phenomenal state. This implies that the sense of self-as-subject is exclusively about the who-component of a phenomenal state—no parts of the what-component belong to it. The sense of self-as-subject concerns whether I experience myself as the subject of a phenomenal state and nothing else. For the rest of this paper, I will use the term “the sense of experiential ownership” strictly in the first-personal sense.
Can one’s sense of self-as-subject go wrong? Following Wittgenstein and Shoemaker, most philosophers believe that the answer to this question is negative. According to Shoemaker, “in being aware that one feels pain one is, tautologically, aware, not simply that the attribute feel(s) pain is instantiated, but that it is instantiated in oneself” (1968, pp. 563–564; emphases in original). Hence, when I consciously feel a sensation, I cannot be wrong about whether it is me who feels it. This immunity (IEM) is widely considered to be a conceptual truth.[15]
I want to argue, however, that both Shoemaker and Wittgenstein are wrong. IEM is not a conceptual truth, and sometimes it makes perfect sense to ask the Wittgenstein Question—namely, “Are you sure it is you who is having a so-and-so experience?” Using my own terms, I will argue that the sense of experiential ownership can misrepresent the fact of experiential ownership. First, let me briefly mention some varieties of IEM. (1) Pryor (1999) distinguishes between de re misidentification and which-object misidentification.[16] De re misidentification is false identification of two particular objects. It occurs when a mental state that a is F involves an assumption that a = b, but in fact a ≠ b. For example, when looking in the mirror, I misidentify someone else as myself (Pryor 1999, p. 276). A mental state enjoys de re immunity just in case it is not possible for the state to be in error through de re misidentification. In the case of which-object misidentification, one makes an existential generalization that there is something that is F based on suitable grounds, but misidentifies which thing is F (Pryor 1999, p. 281). For example, when listening to a symphony orchestra, I can tell that one of the trumpet players is slightly out of tune, but I misidentify which one it is. A mental state enjoys which-object immunity just in case it is not possible for the state to be in error through which-object misidentification. (2) De Vignemont (2012) recently distinguished bodily immunity from mental immunity. Mental immunity concerns whether certain self-ascriptions of mental states, including thoughts, judgments, or sensations, etc., enjoy IEM. By contrast, bodily immunity is not about mental states but about bodily properties. It concerns whether certain self-ascriptions of bodily states enjoy IEM, e.g. “my legs are crossed”.[17]
My target is a form of mental immunity that I call experiential immunity. Experiential immunity concerns phenomenal experiences. It is a form of relative immunity―that is, it is relative to first-personal access to phenomenal states, such as introspection, somatosensation, proprioception, etc. Experiential immunity is then the phenomenon that, when I am aware of a phenomenal state through first-personal access, I cannot be wrong about whether it is me who feels it. Experiential immunity can be construed as de re or which-object immunity. In the following section, I present counterexamples against both versions of experiential immunity. This will show that the sense of self-as-subject can be erroneous.
Bottini et al. (2002) describe a somatoparaphrenia patient (“FB”) who has lost tactile sensation in her left hand and insists that her left hand belongs to her niece. They conducted the following tests on the patient, each involving several trials: (i) FB was blindfolded and told by the researcher that her left hand would be touched. Then the researcher actually touched the dorsal surface of her left hand. The result was that FB always reported feeling no sensation. (ii) FB was again blindfolded and was told that her niece’s hand would be touched. The result in this case was that, when the researcher touched the dorsal surface of her left hand, surprisingly, FB reported feeling the touch.[18] The relevance of this case to IEM lies in the fact that, since FB was blindfolded during these tests, she relied on internal and first-personal access (e.g., introspection, somatosensation, proprioception) to determine whether or not she felt the touch. The perplexity lies in the difference between tests (i) and (ii). For the researcher, the only difference between the two was the verbal cues given to FB before touching her hand. The remaining conditions were the same. But for FB, the difference was dramatic. Why is it that FB felt nothing when she expected that she herself would be touched, but felt the sensations when she expected that her niece would be touched? What is the best description of this strange phenomenology?
My view is that, during test (ii), FB misrepresented her tactile sensations as belonging to someone else, namely her niece. For the sake of argument, Shoemaker and I can agree on the following claims: (1) for every phenomenal state there must be a subject who experiences it; (2) every phenomenal state is in principle available to first-personal access (Shoemaker 1996); (3) every phenomenal state is experienced by the one who has first-personal access to that state. The crucial point is that (1)–(3) do not imply that (4) every phenomenal state is, from the first-person point of view, represented as experienced by the one who has first-personal access to that state. In FB’s case, (4) fails. FB fails to represent from her first-person perspective that she is the owner of the sensations. During test (ii), the factual aspect of her experiential ownership of the tactile sensations was intact when she was told that her niece would be touched, i.e., she was indeed the one who felt the tactile sensations. What went wrong was her sense of experiential ownership. Although FB felt the sensations, she misrepresented this fact as it being her niece who felt them.[19] This shows that it is empirically possible for a subject, while being aware of a phenomenal state via a first-personal method, to commit a de re error regarding who the subject of that state is. Hence, de re immunity fails. Using my own terms, the sense of experiential ownership can misrepresent the fact of experiential ownership.[20]
The second case against Shoemaker’s IEM is the “body swap illusion” (Petkova & Ehrsson 2008, figure 6). This involves agentive experience—I experience myself as someone who is doing something. In an experiment, subjects wore a head-mounted display (HMD), and stood face-to-face with the experimenter, who wore two closed-circuit television (CCTV) cameras. The images registered by the CCTV cameras were transmitted concurrently to the subjects’ HMD, such that through the HMD the subjects saw their own body facing themselves. Both the subjects and the experimenter extended their right hands, took hold, and then squeezed synchronously for two minutes. Twenty college students participated in this experiment. The authors describe their phenomenology: “after the experiment, several of the participants spontaneously remarked: ‘I was shaking hands with myself!’” (2008, p. 5)
This strange phenomenology indicates that the subjects’ agentive experience was mistaken. It was the experimenter who was shaking their hands, not the subjects themselves. Again, Shoemaker and I can agree that: (1) for every agentive experience there must be a subject who experiences it; (2) every agentive experience is in principle available to first-personal access; and (3) every agentive experience is experienced by the subject who has first-personal access to it. However, (1)–(3) together do not imply that: (4) every agentive experience is, from the first-person perspective, represented as experienced by the subject who in fact has first-personal access to it. In this case, which-object immunity fails because (4) was violated by those who experienced the strange phenomenology in the body swap illusion. They were aware that there was someone having the agentive experience of squeezing their hands, but they misrepresented themselves as the subject of that experience.[21]
As such, it is possible for the subject of a given conscious experience, while being aware of that experience via a first-personal standpoint, to be mistaken about who the subject is.[22] Thus Wittgenstein is wrong: it would make perfect sense to ask FB and the body-swap subjects: ‘Are you sure that it is you who is having a so-and-so experience?’ And Shoemaker is wrong, too: experiential immunity is violated both in FB’s case and in the body-swap illusion. One’s sense self-as-subject can be mistaken—that is, the sense of experiential ownership can misrepresent the fact of experiential ownership. Therefore, since both the sense of self-as-object and the sense of self-as-subject can involve misrepresentation, Shoemaker’s IEM fails to distinguish between them.[23]
I propose that this distinction can be made clearer by looking again at the sense of body ownership and the sense of experiential ownership. As I suggested in the last section, the sense of self-as-object can be understood in terms of a sense of self-as-physical-body which, in turn, can be understood via a sense of full-body ownership. Hence, when one experiences full-body ownership, one is conscious of oneself-as-object. In this section, I have suggested that we take an empirical approach to understanding the sense of self-as-subject. We can understand the consciousness of self-as-subject by studying the sense of experiential ownership.[24] In the next two sections, I examine some of the most relevant empirical accounts about the sense of self-as-subject. I argue that none of them are satisfactory. The reasons for this will be valuable when we consider where to go from here.