1 Introduction

In our daily life we are conscious of the objects around us, but we are also conscious of ourselves. We are conscious that we perceive those objects, that we see their colors or smell their aromas; conscious of our bodies and that we are different from the objects we perceive; and also of our plans, hopes, and fears. Influenced some way or other by Kant, many philosophers have claimed that self-consciousness is constitutive of our conscious experience. For example, William James (1895) suggested this when he claimed that “whatever I may be thinking of, I am always at the same time more or less aware of myself, of my personal existence” (p. 42). The plausibility of this claim will depend on the way in which such self-consciousness is unpacked; for it is not entirely obvious in what sense one is aware of one’s personal existence when one is, say, fully immersed in watching a movie or concentrated while playing chess.

Although we are self-conscious when we consider our plans or the kind of person we are, this form of self-consciousness is definitely not present in all our experiences. Gallagher (2000) distinguishes the narrative self, corresponding roughly to the idea that we and others have of ourselves – dependent on episodic and autobiographical memory, as well as on imagination related to the planned and expected future – from the minimal self. Those who defend the claim that self-consciousness is constitutive of our conscious life do not think of self-consciousness as consciousness of the narrative self but rather as involving a minimal sense of self-consciousness:

\(SC_{MIN}\)

Consciousness entails a minimal form of self-consciousness.¹

In the phenomenological tradition, conscious episodes are taken to be present to the subject in an immediate first-personal way. This is unpacked in terms of a minimal form of self-consciousness which is manifest in our experience prior to reflection and introspection – pre-reflective self-consciousness. In line with these ideas, some philosophers in the analytic tradition understand the subjective aspect of our experiences in terms of self-consciousness. They have distinguished two aspects in which the phenomenal character can, at least conceptually, be decomposed: the qualitative and the subjective character (Kriegel, 2009; Levine, 2001). The qualitative character is what distinguishes different conscious experiences: what it is like for one to see red is different from what it is like for one to see green or to smell coffee. Moreover, conscious experiences are subjective. Despite the differences, there is something that all those episodes share; namely, their subjective aspect – there is something it is like for oneself. The idea is that conscious states present the world from a first-person perspective; they do not merely inform us about the way the world is but also about our relationship to it. For example, in having a visual experience one is not merely aware of one’s environment, but also that one oneself is bearing a certain relationship to it.² The subjective character is unpacked as a minimal form of self-consciousness and intended to be a feature constitutive of all and only conscious states.

This short introduction is, of course, insufficient to show that \(SC_{MIN}\) is true, but this is not the aim of the paper. Instead, I will more modestly argue that \(SC_{MIN}\) cannot be ruled out by available empirical evidence, because alleged “counter-examples” to \(SC_{MIN}\) are in fact compatible with the claim consciousness entails a minimal form of self-consciousness. Many authors have considered that a version of \(SC_{MIN}\) is true if we restrict its scope to our ordinary experiences.

\(SC_{MIN-ORD}\)

Ordinary Consciousness entails a minimal form of self-consciousness.

But some have claimed that the unrestricted version (\(SC_{MIN}\)) is false for empirical reasons.

In this paper I want to dispute this claim. I will assume that a minimal form of self-consciousness accompanies all our ordinary experiences (\(SC_{MIN-ORD}\)) and evaluate whether the evidence from empirical research on altered states offer reasons to think that this self-consciousness is not constitutive of consciousness (\(SC_{MIN}\)).³

This requires first a proper understanding of the alleged minimal form of self-consciousness, one that makes it plausible that \(SC_{MIN-ORD}\) is true. Section 2 is devoted to this task. In particular, I will argue that minimal self-consciousness should be understood in terms of perspectival first-person (core de se) awareness:

PFP-Awareness

A non-conceptual identification-free self-attribution that characterizes the first-person perspective that consciousness offers us.

I will characterize PFP-Awareness in terms of its semantic and epistemological features (section 2.1); sharply distinguish it from other forms of self-consciousness (section 2.2), as well as from other notions in the vicinity which often accompany our waking experiences but not constitutively so; and explain how self-consciousness can be disrupted to different degrees while PFP-Awareness remains intact (section 2.3). With this tool in hand, in section 3, I deal with some altered states of consciousness. In particular, I analyze the evidence from empirical studies of drug-induced states, dreamless sleep and meditation, as they present the clearest cases in favor of a radical dissolution of self-consciousness.

2 Perspectival first-personal (PFP) awareness

When we theorize about the nature and mechanisms underlying our conscious experiences, a fundamental element to be considered is their phenomenal character, what is phenomenologically manifest. This can be characterized by what the experience conveys to its subject or what the subject is aware/conscious of during the conscious episode.

We can think of what is conveyed, or the content of a state, in terms of accuracy or correctness conditions. Imagine that you read in the newspaper that there has been a celebration in Mexico City. This news tells you that the world is a certain way, one in which there has been a celebration in Mexico City; and it is accurate if the world is such a way and inaccurate if it is not. Likewise, we can think of what the experience conveys to its subject as the condition under which the experience is adequate (Siegel, 2010). Consider an ordinary perceptual experience like the one I have when looking at my black mug of coffee. Such visual experience (non-conceptually)⁴ conveys to me that there is a black mug. The experience is adequate if the world is such that there is a black mug. The mug and its properties constitute, in the previous example, what we can call “the primary object of experience”. It doesn’t seem controversial to claim that visual experience conveys the presence of its primary object.⁵ However, the primary object does not exhaust our conscious awareness, and few would endorse this radical form of transparency. The object of experience and the relationship our experience represents (or makes us aware of) ourselves as being in change across experiences for example, when seeing a black mug, tasting a wine or thinking that two plus two is four.⁶ But all these conscious episodes remain subjective. There is a first-person perspective that remains constant along, at least, all our ordinary experiences. Those who think that \(SC_{MIN}\) is true hold that, in each of these cases, one is aware that one oneself is in a certain state, which is, in turn, responsible for the subjective aspect of experience. In this sense consciousness is claimed to entail a minimal form of self-awareness. In such cases, my conscious experience does not merely convey to me the way the environment is, but also that I relate in a certain way with such an environment; namely, that I am in a certain relational state – a visual one in this case – to the mug. A proper understanding of this claim in the presented framework requires us to spell out in what sense the experience is supposed to concern or carry information regarding its subject; that is, we need to spell out the role that the subject plays in the adequacy conditions of the state. The distinction between awareness of oneself as-object and as-subject is a classical starting point for this task.⁷

Consider the following thoughts I (Miguel Sebastián, MS) might form after looking at myself wearing a sweater in a mirror:

1. that there is a sweater.

2. that MS is wearing a sweater.

3. that I see a sweater.

4. that I see me.

5. that I wear a sweater.

With the exception of 1, all other thoughts involve awareness of oneself. Nonetheless, we would deny that 2 involves self-awareness. The content of self-awareness is expressed by means of a first-person pronoun in direct speech. Now, there are two ways in which one can be self-aware: as-subject and as-object. Example 3 illustrates the first one. In this case, I am aware both of myself and of a sweater the primary object. However, there is a difference between the ways I am aware of them. There is a sort of intimacy with the subject of experience that is missing with the sweater, as revealed by the fact that I can be misguided with regard to whether it is a sweater what I am seeing, but not with regard to whether I am the one seeing it. This intimacy is also missing in self-awareness as-object (in example 5, 4 exemplifies both kinds). Indeed, the kind of awareness involved in self-awareness as-object is similar to the awareness I have of the mug in the visual example. As we will see in some detail, the difference is that in self-awareness as-object I identify myself with the entity I am aware of, and hence there is room for misidentification.

Many would doubt that the experience I have when seeing a black mug or, even more clearly, when deeply immersed in watching a film takes me as an object – and that there is any entity that all my experiences take as object, as the visual experience of the example takes the mug. If these doubts are justified and, at least, \(SC_{MIN-ORD}\) is to be plausible, then minimal self-awareness ought not to be characterized in terms of self-awareness as-object.

This is of course insufficient to evaluate further disputes with regard to disruption of self-awareness in altered states of consciousness. My aim in the rest of the section is to address this and characterize the minimal form of self-awareness that is suitable to make \(SC_{MIN-ORD}\) plausible. I will argue (section 2.1) that we should think of minimal self-awareness as perspectival first-personal, characterized by a semantic and an epistemological feature: i) indexicality corresponding to the fact that the content has to be expressed in direct speech with a sentence that deploys the first-person pronoun – recall that this does not demand that the subject has to possess the concepts required to express the content – and ii) Immunity to Error through Misidentification (IEM) corresponding to the sort of intimacy above mentioned. Perspectival awareness might seem puzzling to some. In order to mitigate perplexity, I will present a different but well acknowledged form of perspectival awareness – the spatial-centric (section 2.1.3). I then explain how perspectival first-person awareness can accommodate straightforward objections against \(SC_{MIN-ORD}\) (section 2.1.4). Once the intended notion of minimal self-consciousness is clarified, I discuss how PFP-Awareness relates to other notions that have been used to characterize minimal self-awareness (section 2.2), and finally how to make sense of the idea that one can be more or less self-conscious (section 2.3).

2.1 Perspectivalness: Indexicality and IEM as individuative features

2.1.1 Indexicality

The kind of experience I have when looking at the black mug conveys that I am visually related to the black mug. This experience is silent with regard to identity of the entity I happen to be identical with: it does not convey that MS is visually related to the black mug, despite the fact that I am MS. And it is silent with regard to any other property that might individuate the subject: it does not convey that property F is visually related to the mug either, even if I happen to be the only one that has the property. In line with these ideas Williford nicely stresses that “it is a serious confusion to identify subjective character with one’s individuality or particularity” (2015, p. 1).

Moreover, we think that different individuals can have experiences with the same phenomenal character. This is an intuitive folk psychological assumption that a theoretical model should better not rule out. We want to make room for the possibility that (at least in some cases) when two subjects, \(S_{1}\) and \(S_{2}\), look at the mug they have the same kind of experience. If the experience of \(S_{1}\) depends on representing a certain entity beyond the mug (\(S_{1}\)), then \(S_{2}\) cannot have this experience. \(S_{2}\) will rather represent another entity (\(S_{2}\)) and this will entail a difference in their awareness. Nor can it be a matter of representing a generic entity (a subject). \(S_{1}\)’s experience does not convey that a subject is seeing the mug: her experience would not be correct if another subject, say \(S_{2}\), were seeing the mug.

The plausible claim is that all one’s experiences commonly convey something that one would express deploying an essentially indexical expression, the first-person pronoun: “I am in such-and-such a state”. If the hypothesis that conscious experiences constitutively make us aware of ourselves is to leave the ground, then such an awareness is better understood as being first-personal or de se (Castañeda, 1966; Chisholm, 1957; Lewis, 1979; Perry, 1979).⁸

My awareness that I am in such-and-such a state and that MS is in such-and-such a state are different even if I am MS. This is illustrated by the fact that I can have one without the other, if, for example, I ignore the fact that I am MS. We can capture the difference between them in terms of content – their adequacy conditions – with the help of a semantic theory.

Contentful mental states make partitions in the space of possibilities (Stalnaker, 1999). In the case of non-indexical representations such partitions are taken to be among ways the world might be. And we typically make those partitions by attributing properties to objects. My awareness that MS is in such-and-such a state distinguishes worlds in which MS is in such-and-such a state and worlds in which he is not. It attributes the property of being in such-and-such a state to MS, and it is adequate depending on whether MS has this property in the actual world. However, possible worlds seem insufficient to capture the adequacy conditions of de se representations. Lewis argues that indexical representations do not determine partitions of possible worlds, but rather of centered worlds: they are adequate depending not only on the way the world is but also on an individual. Inspired by Lewis (1979), we can think of first-person representations as involving a self-attribution of a property – that of being in a such-and-such a state. And this kind of representation is correct if the subject has the property she self-attributes.

If self-awareness is understood in this indexical way, the tension between \(SC_{MIN-ORD}\) and the idea that \(S_{1}\) and \(S_{2}\) can have experiences of the same phenomenological kind disappears. Their experiences can convey alike and we can capture that in terms of the self-attribution of the same property – that of being in such-and-such a state – to their respective experiences.⁹

2.1.2 Immunity to error through misidentification (IEM)

Unfortunately, the notion of self-attribution is insufficient to characterize minimal self-awareness, for it does not allow us to distinguish between a first-person awareness as-subject, and a first-person awareness as-object. In the previous section I appealed to a sort of intimacy between the subject of the experience and what they are aware of. Such an intimacy can be unpacked in terms of the way in which the subject cannot be misguided.

In general, when one predicates F-ness of certain object a – as when one is aware that a is F – there are two ways in which one can be wrong: one might be wrong about whether a is really F, and one can be right about an object being F, but misidentify such object as a (Coliva, 2006). Call these two kinds of errors mispredication and misidentification respectively.

We can think of the self-attribution in self-awareness as-object as involving two components: i) the attribution of a certain property to a certain object O, and ii) the identification of myself with O. When I come to believe that I wear a sweater after looking at the mirror, I can be wrong because the person reflected in the mirror does not wear a pullover or because, despite the fact that person wears a pullover, I am not this person.

In the sort of self-awareness involved in experience there seems to be no room for misidentification (Shoemaker, 1968). We can be wrong with regard to the state we attribute to ourselves in experiences (mispredication): one might, for example, be hallucinating rather than seeing a sweater in the case described above (cf. Weisberg, 2011). But it cannot be the case that we self-attribute such state because we have misidentified ourselves with someone who is in that state. The kind of self-attribution in first-person representation as-subject does not involve an identification, it does not require to take any entity as an object of awareness and identify ourselves with such an entity.¹⁰
Summarizing, I have argued that if \(SC_{MIN-ORD}\) is to be made plausible, then our conscious experience involves an awareness that i) does not require the possession of the corresponding concepts by the subject; ii) has to be evaluated not only with regard to the way the world is, but also with regard to the very same individual who is doing the representation; and iii) cannot be evaluated with respect to any other entity. We can call this form of non-conceptual de se awareness as subject “core de se” or “Perspectival First-Person (PFP) Awareness.”¹¹

Due to the lack of familiarity with some semantic tools used to spell out the correctness conditions of indexical de se or first-person representation and the phenomenon of IEM, some readers might find PFP-Awareness a bit mysterious. I will deal with this problem next (section 2.1.3), and then explain how perspectival first-person awareness can accommodate straightforward objections against \(SC_{MIN-ORD}\) (section 2.1.4).

2.1.3 Perspectival spatial-centric awareness

I want to deflate any suspicion that perspectival awareness might raise by presenting another form of representation that the reader might find more familiar, which is also perspectival – indexical and IEM: spatial-centric representation.

We can be aware of locations. We can think of the middle point of the Plaza del Zocalo in Mexico City and wish we were there. If we visit the place, we can see that at this point there is a huge flag and we can attend to this location independently of the flag. In these cases a certain location, the center of the plaza, is an object of awareness. Alternatively, we can walk to the center of the plaza and look at the environment. We can see, for example, that the presidential palace is to the right of the cathedral as seen from there. As in the previous cases, a certain location is also part of what is conveyed; it is part of the correctness conditions of the state. However, it doesn’t seem adequate to think of such location as an object of awareness – at least not in the same sense as the presidential palace is in the previous example. An adequate sentence that expresses the content of our visual awareness will deploy the essential indexical expression “here”. In the case of the experience I am currently having, this would be something like – setting the subject aside for the sake of simplicity – that a black mug is located to the right of the computer from here. It does not merely convey the way the world is, but the way it is from, or relative to, a certain location. We can call this sort of awareness de hinc to contrast it with de se awareness. In the Lewisian framework presented above the content is a set of centered worlds where the center is a location rather than an individual.

If you were located in a very different location instead – say Madrid rather than Mexico – but in front of an identical visual set up, we would probably have the same kind of experience, individuated by its phenomenal character. And you would express what is conveyed by your experience using a similar expression. Indeed our experiences do not reveal any further feature about the location from which the awareness is to be evaluated for correctness. This representation does not depend on identifying a particular object-location, and it is then immune to the second sort of error (IEM). If my visual experience represents that there is black mug to the left of a laptop from here, one might wonder whether there is really a mug or whether it is located where one takes it to be. What cannot happen is that there is another location from which the black mug is to the left of the laptop and I misidentify it as the pole of representation. We can call this form of de hinc representation “perspectival spatial-centric”.

Although it is uncontroversial that most, if not all, of our visual experiences involve a perspectival spatial-centric representation, this is hardly a constitutive feature of all conscious experience since nothing spatial seems to be involved in many ordinary experiences, such as the experience of happiness or the experience one has when one considers whether Madrid is the capital of Spain. However, in all those cases it seems that it is conveyed that I am in a certain state. \(SC_{MIN-ORD}\) requires PFP-Awareness.

2.1.4 PFP-Awareness in ordinary experiences

Some have rejected that \(SC_{MIN-ORD}\) is true – and \(SC_{MIN}\) with it. One possible reason for that is to think that self-consciousness commits us to the existence of a metaphysically unacceptable entity, the self – as is the position of Buddhism for example. However, the rejection of an entity like the self is no good reason for denying \(SC_{MIN}\). First, consistently with the existence of self-consciousness, many have argued that this is derived from hallucinatory representation (where there is no such a thing as the self) or illusory representations (where the self does not have the properties it appears to have).¹² And secondly, and more importantly, PFP-Awareness does not give us any reason to postulate a mysterious entity, because it does not reveal any detail with regard to the individuality or particularity of the subject. A theory of mental content that spells out what is required from a system to entertain this sort of representational state will then show the metaphysical commitment of the theory. In its absence there is no reason to postulate any sort of mysterious entity, in the same respect as spatial-centric awareness does not lead us to postulate the existence of any mysterious location-entity (cf. Fine, (2005).

Once we understand perspectival awareness it is simple to explain the mistake that those who fail to appreciate self-consciousness in our ordinary experiences tend to make. For example, some (e.g. Prinz, 2012) have read Hume’s famous inability to find himself in introspection as claiming that we are not self-aware even in ordinary experiences.

For my part, when I enter most intimately into what I call myself, I always stumble on some particular perception or other, of heat or cold, light or shade, love or hatred, pain or pleasure. I never can catch myself at any time without a perception, and never can observe any thing but the perception. (Hume, 1739, p. 252)

However, there is no entity to be found in introspection because PFP-Awareness is not a matter of being aware of an entity, as we are aware of the primary object of experience – the mug in our visual example.¹³ Consider again de hinc awareness – a non-controversial feature of our visual experience, which tells us how the world surrounding us is relative to a certain location. There is no such location to be found in introspection, there is no “hereness” that we can find when we introspect the content of our visual experience. Analogously, in PFP-Awareness there is no self to be found in introspection, but this should not lead us to conclude that there is no self-consciousness. When we look for the subject in introspection we are unable to find it, in the very same sense as when we look in introspection for the here we do not find a privileged location: they are not objects of awareness in the sense that the black mug in our example is. A semantic model of perspectival representation, like the one presented above, can offer a characterization of the differential role that the subject (or the location) plays. It also explains how this is fully compatible with the claim that visual experiences convey how the objects are located from here or that all experiences convey that one oneself is in a certain state.

Up to this point, I have offered a characterization of PFP-Awareness in semantic and epistemological terms and dispelled some worries that the notion might bring. In the next subsection, I discuss how PFP-Awareness relates to other forms of self-consciousness that have been claimed to be minimal, and hence suitable to make \(SC_{MIN-ORD}\) true.

2.2 PFP-Awareness as minimal self-consciousness

Gallagher (2000) has associated the minimal form of self-consciousness with the sense of agency and ownership. The sense of agency is a form of self-awareness present in states that convey that I am the one who is causing or generating an action – the sense that, for example, I am the one who is causing the door to close or generating thoughts in my stream of consciousness. It typically accompanies our conscious life, but not constitutively so, as there seems to be no sense of agency involved in the experience one has when one contemplates the environment, and it is lost in such cases as thought insertion (ibid.). The sense of ownership is presented as the sense that I am the one who is undergoing the experience and in this sense closely related to PFP-Awareness. However, the use of the term “ownership” is confusing, for it suggests a reading corresponding to a state that conveys that a certain state is mine, rather than the intended sense that what is conveyed is that I am in a certain state. This has led philosophers such as Metzinger (2003) to claim that the sense of ownership is lost in thought insertion as reflected by the fact that subjects claim that the thought is not theirs. However, their experience still conveys that they are in a certain state, that of thinking (Campbell, 1999; Zahavi & Kriegel, 2016).

Some authors have also thought of some sense of embodiment or bodily self-awareness as being a basic form of self-awareness. For example, Thompson (2015, p. 4) considers “the embodied self-experience of being alive in the present moment, or the experience of being sentient”, and Wider (2006, p. 82) thinks of “a conscious, bodily orientation toward the world” as the most basic form of self-awareness (see also Blanke & Metzinger, 2009). Metzinger (2013) maintains that being conscious of one’s location in egocentric space is the most basic form of self-awareness. Although some form of bodily self-awareness and spatial-centric location does typically accompany our daily-life experience, there doesn’t seem to be any spatial or bodily component in the phenomenal character of the experience one has when thinking, for example, that Madrid is the capital of Spain. Such an experience is still subjective, and plausibly conveys that one is in a certain state.

One possible reaction to the characterization of minimal self-consciousness as PFP-Awareness is that this should not count as full-blown self-consciousness. It concerns oneself – something about the subject is conveyed, and the subject is part of the adequacy conditions of the state –, but oneself is not an object of awareness. One might consider that full-blown self-consciousness requires that oneself is an object of representation, as the primary object is, and not what we can call “a mere pole” or “perspective” – that is, a condition beyond the state of the world for the adequacy of the state. Accordingly, if I am aware that I see the black mug, or that I hear the radio, I am self-conscious only in a weak sense; whereas if I am aware that I see myself, or when I identify myself with a certain entity, as happens for example when I am aware that I am an entity with certain boundaries or identical with certain body, I am self-conscious in the strong sense. Often in the debate the strong or full-blown notion of self-consciousness seems to be assumed. For example, Blanke and Metzinger (2009, p. 8) think that the simplest form of self-consciousness requires “a globalized form of identification with the body as a whole”;¹⁴ Millière (2019, p. 5) suggests that “states of self-consciousness may broadly speaking represent the same intentional object – the self”; Letheby and Gerrans (2017, p. 1) assess in this respect that we represent “the self as an entity which sustains interoception, affection, cognition, and perception […]”. I do not intend to contend for the most adequate use of the term “self-consciousness”. The point that I want to make, following my reasoning in the paper up to this point, is that \(SC_{MIN-ORD}\) is plausible only in so far as we understand self-consciousness in the weaker sense. Moreover, this weak understanding of self-consciousness as PFP-Awareness is indeed the way to capture the idea behind such a claim. As Zahavi and Kriegel stress:

Our view is not that in addition to the objects in one’s experiential field—the books, computer screen, half empty cup of coffee, and so on—there is also a self-object. Rather the point is that each of these objects, when experienced, is given to one in a distinctly first-personal way. (2016, p. 38, my emphasis)

I have argued that if \(SC_{MIN-ORD}\) is to be plausible, then minimal self-consciousness has to be understood in terms of PFP-Awareness. However, PFP-Awareness does not seem to admit degrees. So, before moving into the discussion of the empirical evidence from altered states, we need to make sense of the idea that self-consciousness can be partially impaired. This is the purpose of the next subsection.

2.3 Degrees of self-consciousness

If we are to assess the possibility that self-consciousness can be disrupted but not fully so, we need to understand to what extent it makes sense to say that one can be more or less self-conscious. Self-consciousness in the weak sense does not admit degrees. As I have characterized it, awareness is either PFP or it is not. This is in accordance, if \(SC_{MIN}\) is to be true, with the view that consciousness is not gradable (Bayne, Hohwy, & Owen, 2016).

It is not clear that the strong notion admits of degrees either. Millière (2019) argues that the most plausible way in which we can make sense of degrees of (strong) self-consciousness is in terms of a scaffolding model, such that some ways in which we are aware of ourselves are more fundamental than others. He considers bodily self-awareness and spatial self-awareness as the most plausible potential candidates to occupy the fundamental level, and presents a detailed collection of experiences in which both are missing to conclude that we should abandon a scaffolding model of self-consciousness, in favor of a multidimensional one. I partially agree with this result. I think that the multidimensional model is adequate for strong self-consciousness. This is consistent with the claim that there is a fundamental form of self-awareness: PFP-Awareness. In PFP-Awareness we self-attribute a property in the absence of an identification, and this form of self-awareness grounds – if Shoemaker is right (see footnote 10) – any stronger sense in which we are aware of ourselves as an entity that satisfies certain features. By attributing other features to ourselves, we end up with an image of the self, an image that corresponds to a singular entity that has all of one’s experiences across time, that possesses a body and occupies a certain location in space and time. And probably other features like being immaterial and irreducible to neural processes, having free-will, initiating our conscious actions, having a history recorded in one’s biographical memory, or an essential personality consistent across times and contexts, etc.¹⁵ When we have an experience that conveys something that conflicts with the model we have of ourselves – i.e., with the properties that we attribute to ourselves – then a report of disruption of the self is likely to be produced as a result of the feeling that some feature of oneself is missing. This is important to make sense of the subject report as we are about to see.

3 Altered states

If \(SC_{MIN}\) is not a priori justified, but based on the phenomenological observation that there is a subjective aspect, a first-person perspective that accompanies all our ordinary experiences, then it is open to empirical falsification (even if we accept \(SC_{MIN-ORD}\)): is consciousness possible in the absence of this minimal form of self-consciousness? Although many of us might not be able to make sense of this idea, it is intellectually enriching to keep the mind open. Altered states, conscious episodes that differ some way or other from our ordinary waking experience, might illustrate what it might be like to have those experiences. And reports of these experiences have been interpreted as showing that \(SC_{MIN}\) is false.

I think that this conclusion is not justified for two reasons. First, there is a misunderstanding of the minimal form of self-awareness involved: although there are good reasons to think that forms of self-awareness that typically accompany our ordinary experiences are compromised in altered states, there are reasons to doubt that PFP-Awareness is missing. Secondly, there are indeed good reasons to mistrust the subject’s report due to the confirmation bias derived from their expectations and their cosmological and metaphysical views. I will start by presenting the case of dreamless sleep, which will allow me to present the minimal form of self-attribution. Then I will illustrate these two reasons in discussing the evidence from studies of meditative and drug induced states.

3.1 Dreams

According to some eastern traditions – the Advaita Vedanta and Yoga – consciousness persists during dreamless sleep. To motivate this view, Thompson (2015) brings back a classical philosophical debate between the Advaitins and the Nyayayikas.

When we wake up in the morning we are not surprised by the arrival of consciousness and we have the feeling of having slept peacefully. Advaitins argue that this cannot be the result of an inference and that it is rather the result of a state of consciousness lacking any particular thoughts or images. If we accept the argument we have to acknowledge a kind of experience that is radically different from waking experiences.¹⁶ But does this jeopardize \(SC_{MIN}\)? A positive answer seems to be suggested by Thompson, who remarks that when we sleep deeply we “do not cognize anything – there is no object to be known and no awareness of oneself as knower” (p. 5), and by Windt (2015), who insists that this kind of experience is characterized by an absence of “the subject of experience, or the ‘I’.” (p. 15).

However, the argument above offers no good reason to think that dreamless sleep experiences have to be characterized by an absence self-consciousness. Windt et al. (2016) clarify that what these experiences lack is the sense of agency or the sense of being a potential possessor of knowledge (Metzinger, (2013), and any sense of self-other distinction. But as we have already seen none of these features are constitutive of PFP-Awareness. Thompson, on the other hand, proposes that the state of dreamless sleep should be characterized by a minimal form of sentience consisting of the feeling of being alive. I honestly fail to appreciate this feeling of being alive during sleep. This might well be a failure on my side and this sort of phenomenological disputes is difficult to adjudicate. However, as McClelland (2015) stresses, one should not attribute to the opponent a kind of mistake that she is not disposed to make in contexts other than the assessment of the phenomenological claim under dispute. And I do appreciate the feeling of being alive upon awakening (cf. section 2.1.4).

I think that alleged dreamless sleep experiences can be better characterized by the feeling of emptiness or nothingness, fully consistent with one’s report of peacefulness and absence of knowledge – and consistent with a feeling of duration (Windt, 2015). There is an ambiguity in the characterization of this state crucial to the debate at hand. This can be understood as a state that conveys that there is emptiness or nothing (jeopardizing the truth of \(SC_{MIN}\)); or as a state that conveys that I am in no state or at least in a contentless state, a state that says nothing (fully compatible with the truth of \(SC_{MIN}\)). The first reading is closer to Vedanta tradition which involves the absence of any form of self-awareness. One problem with this proposal is that, as there is no feeling of ignorance belonging to an “I”, an explanation of the first-person report upon awakening is still required. Advaitins reply that upon awakening the ego sense is again operative and “appropriates the lingering impression or retention of not-knowing and associates this retention with itself, thereby generating the retrospective thought” (Thompson, 2015, p. 10). However, a straightforward explanation of the formation of an episodic memory that explains one’s report that “I slept well and I did not know anything” is available in the second case.¹⁷ This absolute minimal form of self-ascription, where I self-attribute the property of being in a contentless state, is then consistent both with the existence of PFP-Awareness and what can be characterized as undergoing an experience of nothingness or emptiness.

Independently of the strength of this abductive reasoning, the important thing to note for current purposes is that there is no reason for thinking of dreamless sleep experiences as experiences in which there is no self-awareness – once this is properly understood as PFP-Awareness – because the alternative explanation that I have offered is available. Moreover, if the proposed absolute minimal form of self-ascription can adequately characterize the experience of dreamless sleep, then it might also characterize some alleged selfless states of meditation. Indeed, Sankara, probably the most famous of all Vedanta teachers, stresses the similarities between dreamless sleep and samadhi – the highest state in meditation –, when he remarks that there is a “natural eradication of difference in deep sleep and in samadhi” (Sankara, 1980, p. 365; quoted in Comans, 1993).

3.2 Meditation

It has been claimed that one of the mechanisms of action in meditation is related to an altered sense of self (Hoelzel et al., 2011). Indeed, there is evidence showing that the activity of brain networks linked to the narrative self and self-related activity is influenced – for a review of evidence see Dor-Ziderman, Berkivich-Ohana, Glicksohn, & Goldstein (2013). This has made of mindfulness studies an interesting field to study the possibility of conscious experience in the absence of self-consciouness. And recent results might be read as pointing in this direction.

Dor-Ziderman et al. (2013) recorded, using magnetoencephalography (MEG), the brain activity of 16 meditators and then asked the participants to – freely and in their own words – describe their experience. The experimenters were interested in three conditions associated with the narrative self, the minimal self, and a selfless experience, and gave participants instructions to guide their meditation accordingly: “Try to think what characterizes you”, “Try to experience what is happening to you at the present moment”, and “Try to experience what is happening at the present moment, when you are not in the center.” They categorize the reports of meditators in the last condition into three groups. The most interesting for current purposes is the one that they labeled “lack of ownership”, where subjects “reported experiencing what was happening, only with the sense of agency/ownership absent” (Dor-Ziderman et al., 2013, p. 6). Four participants, the most advanced ones, offered this kind of report, which is difficult to interpret. As we have seen, PFP-Awareness is orthogonal to the sense of agency and, depending how it is understood, ownership. Although the reports suggest a strong deflation of self-consciousness, there is no clear absence of PFP-Awareness. In what I take to be the most significant report in the study, the subject said: “There was an experience but it had no address, it was not attached to a center or subject.” This suggests a loss of PFP-Awareness. But then she clarifies: “It was not 100%, but there was no sense of an object there running the show. Emptiness is the best word”, suggesting rather a disappearance of the self-as object, and fully consistent with the minimal self-attribution presented in the case of dreamless sleep.

In Eastern tradition samadhi, the most advanced state in meditation, is described as a feeling of union with the universe or the divine. According to Hinduism, in samadhi, the atman (self, soul) frees itself of habits, prejudices, desires, impulses (ahamkara) and reveals itself as being part of, or participating in, the universe or the divine (Brahman). It is in this sense that the fundamental distinction subject-object or self-other is claimed to be obliterated (Comans, 1993). The feeling of union with the universe might entail a disruption of the distinction between self and other, but this distinction is not something revealed in any sense in PFP-Awareness, and hence is not a challenge to \(SC_{MIN}\). The experience I have when I look at the mug (other), or when I look at my chest or at myself in a mirror (self) involves in the very same respect PFP-Awareness. It is true that the latter might additionally be accompanied, at least typically, with some additional feeling that it is me who I am looking at. However this is not the kind of minimal self-awareness appropriate to make \(SC_{MIN}\) true.¹⁸

There is one further reason to doubt that current empirical evidence from meditation studies should lead us to think that \(SC_{MIN}\) is false. Meditation studies have focused on Buddhist practices (Lutz, Dunne, & Davidson, 2007). In Buddhist traditions there is no such a thing as the self, which is a mere illusion of our minds. Buddhism rejects the foundational premise of the Vedas and Upanishads that “Atman exists”, accepted by all major orthodox schools of Hinduism – Mimamsa, Nyaya, Samkhya, Vaisesika, Vedanta, and Yoga. The root of human suffering derives from our identification with such an illusory entity and meditation is a way to learn this by reaching a selfless mode of experience where the identification with the self is abandoned (e.g., Dalai Lama, 1991). Samadhi in Buddhism is often described as a selfless experience. This is a problem, because we all are prone to a confirmation bias, a well established tendency to store, recall, and interpret information in a way that confirms one’s preexisting beliefs or hypothesis. Are the most advanced meditators in Buddhism and Hinduism having different kinds of experiences or are they offering different descriptions of an otherwise similar experience? I am not calling into question that advanced meditators undergo experiences whose phenomenology is indeed different from ordinary waking experiences, nor that their experiences differ from less experienced ones – as their reports and neuroimages suggest. The claim is rather that there are good reasons to think that the information stored and recalled from their experiences as well as the interpretation of these memories as reflected in their reports is likely to be influenced by their beliefs. A control for cognitive bias in the experiments is required.

Considering a distinctive form of meditation might contribute to advance the discussion. The subject-object distinction is claimed to be disrupted in a more profound sense in very advanced states in non-dualistic forms of meditation, such as open-presence (Josipovic, 2010, 2019; Millière, Carhart-Harris, Roseman, Trautwein, & Berkovich-Ohana, 2018). The state of open-presence is the main meditative state that practitioners of Chag-zôg style attempt to cultivate (Lutz et al., 2007). The aim of the practice is to understand the nature of the conscious experience, by focusing on the invariant feature of all experiences, hence essential to consciousness. According to Chag-zôg theorists, what is constitutive is a reflexive form of awareness: “as an object is being presented to an experiencing subject, reflexive awareness also presents the process or occurrence of that experience”¹⁹ (ibid., p. 512). Following a series of steps the object and the subject are de-emphasized, making it possible to reach the most advanced state in which there is no object nor subject of awareness, just reflexive awareness.

But there are reasons to doubt that such a state is really possible. First, as Lutz and colleagues stress (p. 511), the practice itself presupposes what is at stake in this debate, namely that the invariant elements of experience are selfless. Only very few practitioners in each generation recognize having truly reached this level, and the most advanced ones are likely to characterize the differential phenomenology of their experience in the terms prescribed by the tradition. Second, the aim of the practice is to let the experience continue even though “one de-emphasizes the particularity of the object and subject” (p. 515). In looking for invariants, the practice aims at abstracting away from the differences with regard to what is accidental about the object – the difference in the experience one has when looking at a cup and at a computer – but also what is accidental about the subject “for example, its temporal location in the narrative of personal identity or the particular emotional state that is occurring within the subjectivity” (p. 512). However, those elements, that indeed change from experience to experience, might contribute to the self-as-object, but not to PFP-Awareness – which is, according to those who maintain that \(SC_{MIN}\) is true, an invariant element of experience. Indeed, it is not clear that this element is missing, for a sustained state of open-presence is one in which one may “continue to experience phenomena without objectifying them and, ideally, without having a sense of an agentive or narrative subjectivity” (p.515). This all is consistent with there being a first-personal view that is characterized in terms of PFP-Awareness.

3.3 Drugs

Psychedelic experiences are sometimes characterized by distortions of self-consciousness (Huxley, 1954; Leary, Metzner, & Alpert, 1964; Lebedev et al., 2015), and several studies using classical psychedelic drugs (\(5\textnormal{-HT}_{2A}\) receptor partial agonist), such as psilocybin, mescaline, LSD (lysergic diethylamide) or DMT (dimethyltryptamine), as well as other psychoactive substances, such as ketamine, have observed a reduction of self-awareness in comparison with ordinary experiences – see Millière (2017; Millière et al., 2018) for a recent review. On this basis, it has been claimed that the study of psychedelic experiences offers evidence against \(SC_{MIN}\) (Letheby & Gerrans, 2017; Nour, Evans, Nutt, & Carhart-Harris, 2016; and especially Millière, 2017). I think that this conclusion is unjustified for the very same reason as in the case of meditation. First, there are good reasons to think that trip reports are subject to a similar cognitive bias as the one discussed in the case of meditation; secondly, a sharp distinction between PFP-Awareness and other forms of self-awareness, together with a proper understanding of the former, offer the tools to make sense of the subjects’ reports – in a way that is also consistent with the formation of autobiographical memories on the basis of these allegedly selfless episodes – without any need to give up on \(SC_{MIN}\).

Early research on mescaline before the popularization of psychedelics focused on perceptual hallucinatory experiences (Kluever, 1928), with few mentions of absence of self-consciousness and rather mentioning bizarre body experiences, detachment from one’s body and mental states or the feeling of merging with one’s surroundings (Mayer-Gross & Stein, 1926). The term “psychedelics” was coined by Humphrey Osmond in correspondence with Aldous Huxley to refer to a drug induced altered state of consciousness, in which “the mind” (psyche) is “revealed” (delos). Huxley is indeed the most important player in the popularization of these drugs, not only through his literary work but through his intense social life promoting its use among intellectuals and artists (Stevens, 1998). Huxley was greatly influenced by eastern religions (Huxley, 1945), and the resemblance between his descriptions of trips to samadhi should be of no surprise. Although in The doors of perception, Huxley considers the feeling of unity with the environment as the “final state of egolessness”, and this experience doesn’t seem to entail the absence of minimal self-awareness as PFP-Awareness, later development in the dissemination of psychedelics might help to explain the view that psychedelic experiences refute \(SC_{MIN}\). Huxley introduced the Tibetan Book of the Dead to Timothy Leary, which influenced the well known manual for LSD trips (Leary et al., 1964). Leary and colleagues considered the Tibetan Book of the Dead to be a key for “those who are seeking the spiritual path of liberation” (p. 11) that Huxley thought psychedelics were aimed to offer. They construed the effect of LSD finding parallels between the stages of death and rebirth in it and the stages that Leary had identified during his research (Gould, 2007), and defined ego-dissolution as “[...] complete transcendence – beyond words, beyond space – time, beyond self. There are no visions, no sense of self, no thoughts. There are only pure awareness and ecstatic freedom […].” Despite the criticism in religious studies of the idea that a common self-less experience is a basis of Buddhism (McMahan, 2008, ch. 1), and the connection between ego-dissolution and psychedelics with Tibetan Buddhism (Reynolds, 1989), such a connection remains popular in the public image. Hence, reports of ego-dissolution by drug users might be the result of their expectations rather than a pure reflection of their phenomenology.

This alternative hypothesis finds support in anthopological studies – as the difference between the reports of Hindu and Buddhist meditators evidences a possible cognitive bias in meditation studies. Psychedelic drugs have been traditionally consumed by communities with different cosmological views outside western culture. For example, psilocybin mushrooms have been traditionally used in Mesoamerica (Wasson & Wasson, 2011), as well as the mescaline containing peyote (La Barre, 2011); ayahuasca has been used in Amazonia (Chaumeil, (1998) and anadenanthera seeds in South America (Torres & Repke, 2006), both containing dimethyltryptamine. Anthropological studies show that drug experiences in these communities are all about meeting spirits and getting power and knowledge from them, without reference, to the best of my knowledge, to ego-dissolution or mystical union with the universe.²⁰

The influence of Huxley’s work is also present in scientific studies. One of the tools most widely used to measure the features of altered states is the 5D-ASC (3D-OAV+2D) questionnaire (Dittrich, Lamparter, & Maurer, 2010), consisting of 94 descriptions in which the participants have to rate how adequately they describe their experience. The descriptions are intended to measure five dimensions of the experience, two of which, oceanic boundlessness and anxious ego-dissolution, are analogous respectively to the “Heaven” and “Hell” aspects of Huxley’s mescaline account (Dittrich, 1998). Importantly, although some of the descriptions entail a loss of self-consciousness, they do not indicate a loss of PFP-Awareness. Indeed, Studerus and colleagues have shown that the descriptions in the 5D-ASC are better clustered in eleven dimensions (Studerus, Gamma, & Vollenweider, 2010, fig. 1), of which the two dimensions related to loss of self-awareness are associated with a sense of unity²¹ – which as we have seen Huxley associated with ego-dissolution, and loss of the sense of embodiment.²² None of them jeopardizes \(SC_{MIN}\).

An interesting complementary set of questions to 5D-ASC, intended to track ego-dissolutions in particular, has been recently validated by Nour et al. (2016). Unfortunately, it is also subject to significant deficits considering current purposes. The “Ego-Dissolution Inventory” they present includes descriptions that seem to be irrelevant to evaluate the absence of PFP-Awareness (“I felt a sense of union with others”, “I experienced a decrease in my sense of self-importance”, “I felt far less absorbed by my own issues and concerns”, Nour et al., 2016, p. 3, table 1). Their statistical validation considers a unique phenomenon and cannot distinguish the possibility of ego dissolution in two different dimensions (as-subject and as-object). In fact, there is no significant difference with regard to how effectively they track the studied condition of ego dissolution – as measured by Spearman’s rank correlation coefficient (r)²³ – between descriptions that concern the self-as-object (“I felt at once with the universe”, r=0.830; “I experienced a disintegration of my ‘self’ or ego”, r=0.897) and descriptions that are candidates to track the self-as-subject (“I lost all sense of ego”, r= 0.883).

Moreover, it is far from clear that such descriptions have to be interpreted as showing an absence of PFP-Awareness. The description “I experienced a disintegration of my ‘self’ or ego” (ibid.) might correspond to an experience that conveys that me, the entity I am, disintegrates. But PFP-Awareness cannot be a matter of being aware of a particular entity as we have seen. Likewise, the description “I lost all sense of ego” (ibid.), might correspond to an experience that conveys to its subject that they themselves are in a certain state, namely one in which there is no self-awareness in the strong sense described in section 2.2. So, the report is consistent with an experience where self-awareness is disrupted but PFP-Awareness remains unimpaired. Furthermore, this alternative interpretation has the advantage that it can not only accommodate such reports but also the willingness to self-attribute, in a seemingly non-inferential way, the experience.

Millière (2017) – see also Millière et al. (2018) – argues that the possibility of total disruption of self-consciousness is suggested by open descriptions of psychedelic episodes such as those offered in Erowid.org. He remarks that in these narrative reports, some drug users are reluctant to use the first-person pronoun when describing their experience, suggesting that they lack a “first-personal aspect altogether” (Millière, 2017, p. 14). As a paradigmatic example, he presents a description of an experience induced by the ingestion of psilocybin mushroom where the subject reports that in their experience “There existed no one, not even me, just One!” (ibid.). This seems to rather describe a typical case in psychedelic experience involving ego-dissolution in which there is a feeling of unity with one’s surroundings that is related to the disruption of ego-boundaries and hence irrelevant to PFP-Awareness. It is from this disruption that the subject concludes that “Yet thought continued, so would it be proper to still speak of ‘I’ even as the notion of ‘I’ seemed palpably illusory?” (ibid.). Millière interprets this as a total absence of first-person perspective, whereas I am suggesting that that the report reflects the perplexity of experiencing the dissolution of the subject-object duality, as a consequence of the feeling that the entity we identify ourselves with merges with the environment into “One.” This description is indeed very close to the Hindu’s description of samadhi, and as we have seen it is fully compatible with the truth of \(SC_{MIN}\) once the minimal form of self-consciousness is understood as PFP-Awareness. Further empirical data are required to decide which interpretation is the right one. Moreover, in the absence of the corresponding control, we cannot decide if these reports are the result of a cognitive bias rather than a straightforward report of the phenomenology.²⁴

4 Conclusions

Many authors are sympathetic to the idea that a minimal form of self-consciousness is constitutive of consciousness. Altered states of consciousness have been presented as an empirical challenge to this view.

In this paper, I have argued that such a minimal form of self-consciousness – one suitable to make it plausible that, at least, all our ordinary experiences entail self-consciousness – has to be understood as a non-conceptual identification-free de se awareness (a core de se representation that characterizes PFP-Awareness). Furthermore, I have argued that the conclusion that consciousness does not entail a minimal form of self-consciousness cannot be derived from current empirical research on altered states for two reasons.

First, most of the reports are consistent with experiences where PFP-Awareness is not compromised, and more evidence is required to decide how the reports should be interpreted. Secondly, there are good reasons to suspect that these reports are the result of a cognitive bias derived from the Buddhist belief that there is no self in meditation, and the influence of these ideas in the public imagination of psychedelic experiences.

However, I do think that altered states of consciousness open an interesting door to test the claim at stake. This requires addressing the flaws of current evidence by:

1. Validating new questionnaires that are sensitive to the difference between PFP-Awareness and other forms of self-awareness.

   One of the main challenges for empirical research on ego-dissolution, as we have seen, is that it is hard to determine what subjects are really reporting. One possibility might be to explain to the subjects what PFP-Awareness is and ask them explicitly about the presence of this feature in their experience. If we accept that PFP-Awareness is present in our ordinary experiences and the aim is to test whether it is a constitutive feature appealing to altered states, ordinary experiences can be used to validate the subject’s understanding or the adequacy of any description suitable to track the absence of PFP-Awareness, such as “I lost all sense of ego.” For example, some subjects might tend to think that the offered description is satisfied by an immersive experience such as the one one might have while watching an engaging film. This would show that this is not an adequate description to capture PFP-Awareness. Experiences where strong forms of self-awareness tend to vanish, such as those involving immersion, and high demands of working memory can be used in validation

   Moreover, PFP-Awareness does not admit degrees unlike other forms of self-consciousness. This can also be used to validate the understanding of the subjects or the suitability of the description used. We can control for the confidence of subjects in their reports. Those who have a high degree of confidence should rank the adequacy of the description only with extreme values – a three point scale might be preferred to a seven point Likert scale for this purpose.

   Alternatively, other methodologies such as open interviews with subjects who understand the difference between PFP-Awareness and other forms of self-consciousness after drug use or meditation are of use.

   More details have to be considered to refine the adequate experimental set-up – some of them beyond my expertise. However, I see no reason to think that it is not possible to determine experimentally the form of self-consciousness whose absence the subject is reporting.

2. Include controls regarding preexisting cultural beliefs and expectations to avoid cognitive bias. For this purpose, experiments with non-Buddhist meditators in meditation studies, and with non-western influenced subjects in drug studies could be performed. Anthropological studies and open descriptions of psychedelic episodes by non-western users are particularly interesting.

There are many things to be learned from the empirical study of altered states with regard to the nature of our conscious experience. It seems that, at least for the moment, the possibility of a radical dissolution of self-consciousness is not one of them.

Acknowledgments

For discussion and comments I am very grateful to Kirareset Barrera, José Luis Díaz, Martin Fortier, Rubén Sebastián, to my students at the SCC-PhiLab and the participants at the 5th International Workshop in Cognitive Sciences (UAEMor). I am specially grateful to two anonymous referees and the editors of this special issue, Thomas Metzinger and Raphaël Millière, for their thoughtful comments and suggestions.

Financial support for this research was provided by DGAPA projects IA400218 and IG400219, and the Spanish Ministry of Science, Innovation and Universities via research Grant PGC2018-095909-B-100.

References