1 It is a well known and often deplored fact that the treatment of almost all important questions by philosophers leads to a wide range of normally not very illuminating answers which usually turn out to be incompatible with or even obviously contradictory to one another. This fact is confirmed, it seems, particularly well when it comes to the spectrum of philosophical answers to the question as to the nature and function of the self or the I. On the one side there are those who would go along with something like the following: What is meant by terms like ‘the (human) subject’, ‘the (individual) person’, ‘the (self-conscious) I’or ‘the (conscious) self’ everyone somehow knows but nobody can tell exactly (to use Chalmers’ characterization of consciousness [1]). There is some agreement that the phenomena designated by these terms are characterized by the ability to be conscious of oneself, but this characterization does not help much to make them more clear because again nobody can tell what exactly is meant by self-consciousness. Self-consciousness, we are quite often told, is taken to be by many an ‘elusive’ [2] phenomenon that is in an obscure way related to or even identical with equally elusive items like self-awareness, self-perception, self-feeling, self-familiarity, etc. How these phenomena can be analyzed adequately, whether they depend on one another or how they are related, nobody seems to be sure of either. Thus there is reason to suspect that the I or self-consciousness might turn out to be an illusion or at least a mystery (again borrowing from Chalmers [3]) and it might be that it will stay that way. On the other side of the spectrum there are those or at least there have been those who claim to have very definite insights into the nature and the function of self-consciousness or the I. This can be inferred from their most important philosophical messages according to which the I or self-consciousness can be shown to lie at the very basis of all experience and all knowledge (Kant, Fichte), or can be shown to be the center or even the creator of our world (Schelling, Hegel). These philosophers maintain that in order to make sense at all of whatever aspect of our relation to the world, one first must acknowledge that this cannot be done without attributing supreme reality to self-consciousness or the I. It is the ‘highest point’ on which all experience depends (Kant [4]) or it is the first unconditioned principle of all knowledge (Schelling  [5]).

2 Against the background of this puzzling situation, which is the result of centuries of discussions, the best recommendation for a philosopher might be just to give up on the topic of self-consciousness or the I. However, because it is not merely this topic which leads to such perplexing situations in philosophy, but rather, this seems to be the normal fate of almost all important topics in this field, opting for abandonment would mean running away from philosophy altogether. Instead, I suggest, one should choose a different path: one should try to reconcile the results of some of the main positions put forward so far with respect to the I or self-consciousness by developing a certain degree of sensitivity towards the context in which they are advanced and to pay attention to the specific needs they are supposed to answer. In approaching the topic that way one will find, I hope to show, that one can discover quite a lot about the nature and the function of self-consciousness or the I which might help to determine its status and its significance within the wider framework of human attitudes and accomplishments. In particular I want to point out that elusiveness and supreme reality of the I or self-consciousness are not that much opposed as one might believe at first, as can be learned especially well from Kant and the German idealistic tradition.

3 In dealing with topics like ‘I’ or ‘self-consciousness’ it seems reasonable to start with what belongs to the least controversial. And here many – at least in the Continental tradition – agree that one should approach the I or self-consciousness within the wider context of the phenomena of conscious experiences. Thus it seems reasonable that in order to find out what can be said about the I or self-consciousness, one should look at consciousness first.

4 Within the philosophical tradition it has been proven to be a temptation to think of the self-conscious I as the subject of all conscious experience and for that reason to assign to it a fundamental status. This temptation should be resisted, at least if one understands by the self-conscious I an item that has a conscious relation to itself, because this understanding leads to a misleading conception of conscious experience and the I as well, and makes it somewhat difficult to see where and in what sense the self-conscious I is indeed fundamental. In order to substantiate this assertion, let me begin with an observation that has been emphasized and exploited heavily by the phenomenological tradition of the 20^th century (Husserl, Sartre, Merleau-Ponty) and which can be stated like this: There are a lot of conscious experiences a human subject can have which do not presuppose that the subject is conscious of itself, i.e. that it is a self-conscious I. This statement is not very helpful as long as we are not told what it is supposed to mean. First of all one has to settle the meaning of the terms used in the statement. In the phenomenological context, I take this observation to rely on a quite plausible and common-sensical reading of what is meant by ‘conscious experience’ or ‘consciousness, ’ and a less perspicuous, ambiguous understanding of the terms ‘being conscious of itself’ or ‘self-conscious I’. According to our normal comprehension, which is shared by the phenomenologists, the first pair of terms (‘conscious experience’, ‘consciousness’) is meant to characterize all states a subject is in as long as and insofar it is not unconscious, i.e. as long as and insofar it has some degree or other of intentional or non-intentional awareness of its (inner and outer) environment. The second pair of terms (‘being conscious of itself’, ‘self-conscious I’) is ambiguous – at least when used in philosophical contexts – in that it can refer to the fact that the subject is conscious of itself either as an object of experience, or as the subject of experience. [6] In order to avoid confusion I will use the terms ‘self-presentation’ or ‘being present to itself’ to designate the situation in which a subject is conscious of itself as an object, whereas the terms ‘the self-conscious I’, ‘self-consciousness’ or ‘being conscious of itself’ are used to refer to the non-objective consciousness someone has of himself as a subject which never can become an object. This non-objective consciousness I wish not to be identified with what is called ‘self-feeling’ or (bodily) ‘self-awareness’ with which I do not deal here as special phenomena at all (actually, I take both these phenomena to be cases of non-intentional awareness). Because of this ambiguity of the second pair of terms the observation of the phenomenologists can be interpreted in two ways. The first amounts to the claim that a human subject can be aware of a lot of things without at the same time being present to itself (as an object of experience). The second maintains that in order to have conscious experiences a subject does not have to have non-objective consciousness of itself (as the subject of these experiences). I take this latter reading of the observation to capture what the phenomenologists want to claim and hence will refer to it as ‘the phenomenological thesis’. Whereas the first interpretation, in my eyes, is not very informative because it boils down to the trivial platitude that my being aware of whatsoever is not necessarily connected with my being present to myself at the same time, it is the second reading, i.e. the phenomenological thesis, which is interesting and thus has to be elaborated a little bit.

5 But before elaborating the phenomenological thesis, I need to consider a possible objection to my attributing it to the phenomenologists mentioned above (Husserl, Sartre, Merleau-Ponty). I have to admit that it is by no means obvious that they defend any such thesis. This is so because their views as to the relation between consciousness and self-consciousness on the one hand and as to the distinction between non-objective, i.e. radically subjective, self-consciousness and self-consciousness as an object on the other hand are far from clear.

6 As far as Husserl is concerned, he holds in his early period of the Logical Investigations that one has to give primacy to consciousness over self-consciousness or the I because he takes the I to be nothing but the unified entirety of experiential content (einheitliche Inhaltsgesamtheit der Erlebnisse). For him, the I is nothing special (Eigenartiges) but something that floats (schwebt) on top the manifold conscious experiences. [7] Thus it seems that Husserl in fact endorses the phenomenological thesis. However, at the same time he criticizes in connection with a discussion of Natorp’s notion of a pure I the idea of a self-conscious I that is radically subjective in the sense that it can never become an object of experience. [8] And even later when he acknowledges what he then calls a transcendental ego he thinks of this transcendental ego as something that can become an object of experience. [9] This gives rise to the impression that Husserl could not endorse the phenomenological thesis because he denies the distinction between the self-conscious I as subject and as object for the reason that for him there is no way to assert the existence of non-objective consciousness of oneself. This impression would be justified if one took the phenomenological thesis to imply that there indeed is something like non-objective self-consciousness. Though this interpretation of the thesis is by no means obvious, one could nevertheless take Husserl to be committed to the phenomenological thesis even under this reading at least by implication, by pointing out that for Husserl there has to be an I which is not an object (because otherwise there would be nothing which could become an I as object) and that consciousness does not depend on this I. It is on the basis of this somewhat ‘creative’ conjecture that I take him to be a supporter of the phenomenological thesis.

7 Things are even more complicated in the case of Sartre. Whether he can be said to advocate the phenomenological thesis depends on how one interprets his conception of a non-positional or non-thetic self-consciousness as presented in The Transcendence of the Ego and his notion of a pre-reflective cogito as introduced in Being and Nothingness. At first glance, Sartre seems emphatically to endorse the primacy of consciousness. Thus he writes in The Transcendence of the Ego: “The phenomenological conception of consciousness renders the unifying and individualizing role of the I totally useless. It is consciousness, on the contrary, which makes possible the unity and the personality of my I.” [10] However, whether he really endorses the primacy of consciousness becomes a bit dubious as soon as one pays attention to Sartre’s explicit characterization of consciousness: it looks as if for him an irreducibly subjective consciousness of oneself is an essential characteristic of consciousness. Thus he claims: “Indeed, the existence of consciousness is an absolute because consciousness is consciousness of itself. This is to say that the type of existence of consciousness is to be consciousness of itself. And consciousness is aware of itself in so far as it is consciousness of a transcendent object.”  [11] This awareness of itself in being aware of an object he describes as non-positional self-consciousness. From this characterization one could get the impression that for Sartre, in the end self-consciousness is constitutive of consciousness and as such the basis of all conscious phenomena, thus repudiating the phenomenological thesis. This repudiation appears to be confirmed by his claiming in Being and Nothingness that “every positional consciousness of an object is at the same time a non-positional consciousness of itself”. [12] And yet if one pays attention to the broader context in which statements like the ones quoted are situated, it shows that in fact they rather strongly support the phenomenological thesis. An outline of this broader picture can be sketched thus: In all conscious states a subject has to have some basic form of self-awareness which is not to be identified with consciousness of oneself as a subject. This basic self-awareness or, in his terminology, this non-positional consciousness of oneself is a necessary element of what Sartre calls “immediate consciousness, ” where there is no difference between me, the conscious subject, and the object of consciousness. This immediate consciousness does not require my being conscious of myself as a subject because it “does not permit me either to judge or to will or to be ashamed” – whatever the object of my immediate consciousness may be. It is when approaching these objects with ‘reflective’ activities (like judging, willing, or being ashamed) that a self-consciousness comes into play which can never become an object because these activities presuppose a radically subjective consciousness of oneself (that is different from the self-awareness of immediate consciousness) in order to be experienced as my activities. This radically subjective self-consciousness Sartre names ‘pre-reflective cogito’: it is a necessary and sufficient condition for all my ‘reflective’ activities, though not involved in immediate consciousness. Thus immediate consciousness to which necessarily self-awareness or non-positional consciousness of oneself belongs has to be distinguished from and is prior to “reflecting consciousness” and its pre-reflective cogito, i.e. its radically subjective self-consciousness. This sketch of a broader picture might still look a little bit fuzzy, but if one goes along with this outline of Sartre’s refined distinctions and their place in his general conception of consciousness, then one has to attribute to him the phenomenological thesis as well.  [13]

8 In the case of Merleau-Ponty one also has to rely on indirect evidence to think of him as a supporter of the phenomenological thesis. This is so because he – at least in the third part of his Phenomenology of Perception, Being-for-Itself – is not really interested in giving a detailed account of the relation between conscious and self-conscious states. He rather wants to explore (in the chapter: The Cogito) what can be said about the role of the cogito in the process of perception. I take him to argue in the same direction as Sartre does when he insists that at the very basis of all our particular thoughts or of all our states in which self-consciousness is comprised, there is what he calls ‘the silent cogito’ (§ 12) or ‘the primordial I’ (§ 13), an ‘original subjectivity’ which in turn is grounded in our condition as experiencing subjects in a non-thetic way. This must mean that there is an experiencing subject before there is an attitude to the world that goes together with a ‘primordial I’. Here again it might be controversial whether this reading does justice to the details of Merleau-Ponty’s views concerning consciousness and self-consciousness, nevertheless his overall attitude seems to suggest to me that he too would be in favor of the phenomenological thesis.

9 So much for the historical background of the “phenomenological thesis.” I now want to consider its legitimacy: is it really the case that I can be in the state of being aware of something or other without being conscious of myself as the subject of this awareness? The most sensible answer seems to me to be: yes. This is so because there are numerous examples of states of awareness that do not seem to presuppose a distinct self-consciousness or a peculiar consciousness that they are my states or that it is me who is in them. Here some cases which I take to be convincing: when I walk around aimlessly through a city, maybe admiring a building, I am certainly aware of my walking in the sense that I somehow coordinate the movements of my body in order to adjust to my present environment; but in doing so I do not have to be conscious of the fact that it is me who is walking – I am just aware of my walking, and if while walking I am conscious of myself as the subject of a state I am in, then this self-consciousness might relate to my admiring the building but not to my walking. The same holds in another case I take to be obvious: when I sit at my desk, look out the window into the garden and try to remember what was said to me the other day, I definitely do perceive my surroundings without paying attention to anything in particular. This state of non-focused perceiving can be said to be a state in which I am aware of a lot of things without having a consciousness of myself as the subject of this awareness.

10 One could object that this is so in both of the cases I just described only because of the proviso that it is not just the walking or the perceiving I am engaged in, but in addition I also consciously do something else: I admire, and I remember. One could argue that my not self-consciously being aware of my walking and perceiving is merely due to the fact that at the same time I am conscious of my being the subject of my admiring or my remembering – if I were just walking or perceiving and not involved in something else also (admiring and remembering), either I would be unaware of my walking or perceiving, or I would be conscious of being the subject of these states. This objection amounts to the claim that though it might be possible that there exists the phenomenon of indeterminate awareness, this phenomenon cannot occur independently of my being conscious of myself as the subject of some other state. And if this claim is correct, the objection goes on, then self-consciousness still turns out to be a necessary condition of all conscious experiences. In other words: though it might be possible that I can be in the state of being aware of something or other without being conscious of myself as the subject of this awareness, the self-conscious I is still fundamental in that it can be shown to be an enabling condition of awareness. This leads to the suspicion that the phenomenological thesis is unpersuasive in the end.

11 This objection makes it advisable to distinguish between two senses in which the phenomenological thesis can be understood: a weak and a strong one. In the weak sense, the claim is that there are conscious experiences without self-consciousness (at least in the form of states of awareness) though it might be that these experiences or states are parasitical on self-conscious states. According to this weak reading, the phenomenological thesis does not exclude the fundamental role of self-consciousness even for consciousness in the form of awareness, because it does not say anything about the enabling conditions of states of awareness. All it claims is that there are those states whatever their conditions might be. Against this weak reading, the objection mentioned has no force because it does not deny what the phenomenological thesis asserts, and the phenomenological thesis does not deny what the objection states. However, this defense of the phenomenological thesis by giving it a weak reading is not convincing, because it amounts to accepting the possibility that all conscious experiences are based either directly or indirectly on self-consciousness. But it is exactly this possibility the phenomenologists want to rule out with the phenomenological thesis. They hold this thesis to imply: consciousness comes first, self-consciousness is just a second-rank phenomenon. Thus they are committed to a strong reading of this thesis, a reading according to which it claims that there are conscious experiences which are neither directly nor indirectly dependent on self-consciousness as an enabling condition. And against this reading the objection indeed sticks if one cannot come up with cases which confirm the phenomenological thesis in its strong interpretation. Now, these cases are not difficult to find. Again one has to ponder cases of awareness. When I am dreaming, I am definitely in a state of awareness, whatever it might be I am dreaming, without being conscious of myself as the subject of my dreaming. (This at least seems to be a good example if one assumes that in dreaming it is possible that I play no role at all, or if I play a role, then in an undetermined manner.) Or when I am engaged in some physical exercise or other (like jogging) I surely have a conscious experience (an awareness) of my doings but not necessarily a non-objective consciousness of myself. In those cases, awareness is not even indirectly linked to self-consciousness. Thus the objection fails and the phenomenological thesis can be defended even in its strong version.

12 This result obviously implies that self-consciousness cannot be taken to be more basic than consciousness in the sense that without self-consciousness there would be no consciousness. [14] All the states and activities mentioned in the examples (walking, perceiving, dreaming, jogging) are sure enough mine but there is no need for my being conscious of them as mine in order to be in these states. These states and activities of mine are what they are and have their specific content (walking, perceiving, etc.) without my having to be conscious of them as their subject. All this amounts to denying – to allude to a Kantian way of speech – that the self-conscious I is something that accompanies all my conscious experiences or that the self-conscious I is the subject of all my conscious states and activities.

13 If one agrees that the phenomenological thesis has some credibility and that there are indeed conscious states (i.e. states of awareness) a subject can be in without being conscious of itself as the subject of these states, then the second question is what to do with the idea of such a subject, i.e. the self-conscious I. Is it merely a fiction made up in order to satisfy a strange and suspicious metaphysical instinct according to which we cannot help but making up substance-like sub-strata for every state of affairs (e.g. Nietzsche)?  [15] Is it just a metaphorical expression intended to do justice to the omnipresent conviction that in order to accept conscious states as my states one has to think of oneself as the self-conscious subject of these states (e.g. Campbell)? [16] Such a conclusion does not seem to be right. And it is warranted neither by the acceptance of the phenomenological thesis nor – more importantly – by the facts. The phenomenological thesis as presented here does not amount to a sweeping denial of self-consciousness, it just claims the primacy of consciousness over self-consciousness. Whatever one is going to claim with respect to the role and the function of the self-conscious I as the subject of conscious states, it is perfectly compatible with the phenomenological thesis as long as one does not deny the reality of conscious states without a self-conscious subject. And as to the facts, it seems to be obvious that there are conscious states in which I am conscious of my being the subject of these states. Especially with respect to those states which are characterized by a conscious mental attitude towards a propositionally structured content (whatever can be said about the constitution of this content considered as independent of me), the self-conscious I seems to be a necessary element: these states are necessarily not merely states of a conscious subject (though trivially they are these too) but states whose possibility depends on the existence of a self-conscious I. I will call the states which are characterized by such a conscious mental attitude towards a propositionally structured content ‘conscious propositional states’. Examples are states in which attitudes like hoping, believing, (focused) perceiving, judging that something is so and so play a role, or states like wishing something or other (if they are conscious at all). This characterization of a conscious propositional state is rather vague. Another way to describe these states a bit more comprehensively is the following: A conscious propositional state as conceived here is a state which comes about whenever a person performs a propositional act, i.e. whenever someone performs an operation which results in his or her entertaining (either grasping or forming) a proposition to which he or she relates in a specific mode. Thus when I am in the conscious propositional state of believing that something is so and so I am actively engaged in two things: (1) I grasp or form a proposition and (2) I connect with this proposition in the specific way or mode of believing. Or again when I am in the conscious propositional state of hoping that something is (or will be) so and so (1) I grasp a proposition and (2) I connect with this proposition in the specific way or mode of hoping.

14 What is important to realize is that these states are such that they presuppose not just a subject but that this subject is in a special, fundamentally ‘subjective’ way conscious of being the subject of these states. This at least is suggested by what one implicitly takes for granted when dealing with verbal expressions of these states. If someone says ‘I hope the weather will be fine’ it goes without saying that I am in this special ‘subjective’ way conscious that it is me who is in that state of hope. It would sound rather odd to say ‘I hope the weather will be fine, but I am not conscious of me as hoping’ or ‘I hope the weather will be fine, but I am not sure whether it is me who is hoping’. The same is true of the other states whose content is propositionally structured: they all require a subject that is a self-conscious I, i.e. an I which cannot become an object. Thus it seems that one can and has to distinguish between a conscious subject and a self-conscious I by specifying the area where they play a necessary role: Whereas all conscious states must have a subject, only some of them, e.g. propositional states, are such that in order to account for them one is in need of a self-conscious I. If this is correct then the acceptance of the phenomenological thesis has no damaging consequences for the belief in the reality and significance of a self-conscious I.

15 It might be worth mentioning here explicitly that the claim that a self-conscious I is a necessary condition for being in propositional states is meant to imply that the significance and reality of the self-conscious I is restricted to propositional-state contexts. This may seem to be much too strong a claim and, in addition, a rather implausible one: many people will take it to be an evident fact that there are non-propositional conscious states that also presuppose a radically subjective self-conscious I. As obvious examples they are likely to draw attention to states in which emotions play a role. When I feel very bad or when I feel happy without having the thought ‘I feel bad’ or ‘I am happy’, or when I am in the mood of being afraid though of nothing in particular I can point to, I am certainly neither related to myself as to an object of introspection, nor am I in a propositional state. Nevertheless, in such situations I seem to be intensely and immediately conscious of myself as a subject in a radically subjective way – so they say. Though I am not convinced this analysis is correct, even if it were it would not pose a problem for my claim that self-consciousness is restricted to the context of propositional states. All one needs here is a distinction between consciousness of myself as the subject of my conscious propositional states and what could be called ‘self-feeling’. If one accepts this distinction – which I do not (for reasons that are independent of any considerations involving the relation of self-consciousness to propositional content but which I cannot elaborate here) – one could argue as follows: The subjective feature allegedly present in those cases were emotions have a part, is not this consciousness of myself as a subject claimed by me to be necessary for my being in a conscious propositional state, but rather what one could call a ‘selfish’ element in basic conscious awareness itself. Consciousness (understood in the sense of non-propositional conscious awareness) might well include in some cases like those of feeling bad or happy a component of basic self-feeling which manifests itself in my immediate feeling as the subjective center of these conscious states. And yet, this basic self-feeling is not to be identified with my being conscious of myself as the subject of a conscious propositional state though both of them are irreducibly subjective items. This is so because self-feeling is lacking what will turn out in what follows to be distinguishing marks of a self-consciousness that can never become an object, i.e. distance and activity. Thus the claim mentioned above is not that strong in the end: the restriction of a radically subjective self-consciousness to propositional states would leave room for a radically subjective self-feeling in other (non-propositionally) conscious states – if there were such a feeling.

16 But these considerations do not answer satisfactorily the second question asked above as to the role and function of self-consciousness because they give no indication concerning the details of the relation between self-consciousness and conscious propositional states. Why is it that without being conscious of myself as the subject I cannot be in or entertain states whose content is propositional? The reasons or at least two of the reasons seem to be the following: (1) In order to be in such a state one has to be conscious not only of a content, i.e. of what the proposition is about, but also of a specific mode in which I am conscious of that content. There are as many modes as there are different mental attitudes towards a propositional content. I might hope or fear or believe or desire or judge that something or other is the case (or is not the case). Now, these modes each of which can be considered to be a particular case of what is called ‘thinking’ are definitely not part of the content I am conscious of when conscious of a proposition. Nevertheless, they are without doubt an integral part of a conscious propositional state. Thus in order to be conscious of a propositional content, i.e. a content that is about something, not just a subject of this consciousness is presupposed but a subject who is conscious of itself as thinking in one of the customary modes (expecting, believing, desiring, judging, etc.) this propositional content because otherwise I cannot account for the very modes in which, in the very act of thinking the proposition, I relate to its content. Consequently my being in a state whose content is propositionally structured means that I am conscious of myself as thinking (expecting, desiring, judging, etc.) that something or other is or is not so and so. Or to emphasize it again: in order to actually be in such a state I have to be conscious of myself not as an object, i.e. as someone who is conscious of a proposition, but as the subject of this very propositional state if this state is supposed to be mine. (2) The proposition that is the content of a conscious propositional state consists of a combination of different elements which can have different functions within a proposition. In order to form or even to grasp a proposition these elements have to be synthesized in such a way that they yield a proposition. Now, this synthesizing requires a subject who does it. This subject has to be me if it is me who entertains (thinks) a proposition because in entertaining a proposition I have either established or comprehended a specific connection between its elements. And I have to be conscious of myself as being this subject because otherwise I could not account for my being conscious of these very connections.

17 Things might get a bit clearer by way of an example highlighting a judgmental state. While walking absentmindedly to the station I have all these visual, acoustical, olfactory and tactile impressions I happen to have at the same time. I am aware of all of them without noticing anyone of them in particular or, in the terminology used above, I am just consciously aware of them. Suddenly I realize that on the left side of the street in which I am walking, there is a house with a blue door. It is as though out of nowhere, the as yet amorphous content of my conscious awareness becomes organized in such a way that I can make both a distinction between different items and characteristics of my conscious environment and a connection between them in accordance to propositional practices. What has happened is that in focusing for reasons normally beyond my grasp on a particular collection of my impressions, I end up being in a propositionally structured state (i.e. that this house has a blue door). This state I am compelled to acknowledge as my state is the outcome of an activity of mine, of my bringing together into a specific relation certain elements I have not been aware of in a determined manner until then. Some of these elements I synthesize in such a way that they give rise to something that fits my concept of a house or a door, and I attach to them specific characteristics like color predicates. It might be debatable whether this activity depends on my being conscious of me as the subject of this activity. What is less doubtful is that the activity of arranging these conceptually organized elements (which come into being out of my initial awareness situation) into a propositional state of affairs which can be expressed by the judgment ‘this house has a blue door’ or ‘this blue door belongs to that house’ cannot occur without my being conscious of me as its subject. The consciousness of that activity of mine presupposes my being conscious of me as the subject of that state, because without that self-consciousness there would be no conscious subject of that state at all, as follows from what was said above. Thus the self-conscious I understood as the self-conscious subject is a necessary condition for being in a conscious state whose content is a proposition. What this means is among other things that the propositional structure my environment (my world) displays if it is taken as an object of my propositional attitudes is a result of an activity of mine that requires me to be conscious of me as its subject or that demands a self-conscious I. Whether the self-conscious I is the creator of this propositional structure in that it imposes it on the conscious environment or whether this I just makes accessible to consciousness an already underlying structure of the environment is a different and very complex question with which I will not deal here.

18 I take it that it is this intimate connection between the consciousness of myself as the subject and the consciousness of a proposition Kant wants to emphasize when in the second edition of his Critique of Pure Reason he famously claims that what he calls the ‘I think’ must be able to accompany all my representations (§ 16 of the Transcendental Deduction) if they are to become recognizable elements of my self-conscious life. Relying on Kant as a witness in this context is in need of some clarification because it might seem by no means obvious that this connection has a firm basis in a Kantian way of thinking about self-consciousness. Thus one could object that Kant’s ‘I think’ dictum does not restrict what he calls ‘pure’ or ‘original apperception’ (CpR, B 132) to propositional contexts but that he wants to think of this apperception as fundamental to all representations, whether propositional or not. It is hard to deny that this view appears to be supported by many of Kant’s formulations (especially those which are to be found in the very same § 16 that starts with the ‘I think’ dictum). However, I take it that Kant’s considered view has to be that in the end there are no non-propositional representations for us if we make it a condition that only those items are representations for us which qualify as potential elements of thoughts. [17] The main reason for ascribing this view to Kant is that according to him the unity of apperception (or in the terminology used above: the radically subjective self-conscious I) is required by as well as depends on synthetic activities with respect to some manifold or other (cf. CpR, B 135 f.). This in turn is supposed to imply that for something to be my representation it must be the product of my bringing together a manifold into the unity of an object. Thus if one thinks of representations as elements of thoughts and regards thoughts to be states of a self-conscious subject, one has to take representations to be organized in such a way that they can fit into a propositionally structured content if they are to be mine at all.

19 Kant points in the same direction when, at the beginning of the chapter on the Paralogisms of Pure Reason in the first Critique, he proclaims that the I as subject is a representation that is in itself without any content (A 345) or that the I (as subject) has to be regarded as a mere consciousness that accompanies all concepts (A 346), or when – also in the chapter on the Paralogisms – he maintains that the I (as subject) signifies “nothing further… than a transcendental subject of thoughts = x, which is recognized only through the thoughts that are its predicates, and about which, in isolation [abgesondert], we can never have even the least concept” (A 346). [18] Though Kant is not interested in the conditions of propositional states in general but mainly, if not exclusively, in the conditions of those that involve knowledge-claims, his message is quite clear: as soon as one is in the state of entertaining a proposition (a ‘thought’ in his terminology) a consciousness of oneself as the subject of this state, i.e. a self-conscious I, shows up, because for a subject to entertain a proposition or, in more Kantian terms, to make a judgment, it is necessary for that subject to be conscious of itself as the thinker of the proposition, i.e. the maker of the judgment. In an even more radical and explicit way, Fichte insists on the close relation between the self-conscious I and conscious propositional states when he claims, e.g. in § 1, 7 of the Doctrine of Science from 1794 that the very existence of conscious propositional states depends on my being conscious of myself as the subject of these states: “One cannot think at all without subjoining in thought one’s I, as conscious of itself; from one’s self-consciousness one can never abstract.”  [19]

20 Thus the result reached so far is that though there are indeed conscious states that do not presuppose a self-conscious I as their subject, some conscious states, i.e. propositional states, must have a self-conscious I as their subject. This result has some interesting and somewhat surprising implications. The most important is that the self-conscious I cannot be identified without qualifications with the subject of all conscious states or with me as the subject of all my conscious experiences. This means inter alia that those who are inclined to identify the subject of all conscious experiences with my body [20], or to claim that the very idea of a radically subjective self-conscious I “is intimately bound up with awareness of the subject… as a physical object in a world of physical objects” (i.e. as a body) [21], are not well advised to think of this bodily subject in terms of the self-conscious I if they agree, as I take them to do, that not all of my conscious experiences are propositional states. It might be necessary or at least unavoidable to think of the body as the ultimate subject of all conscious states of a person in the sense that without the body a person would have no conscious states at all, but the particular subject required in order to account for conscious propositional states, i.e. the self-conscious I, is much too special to be identified with the body (though it might have to be embodied). Another implication is that the self-conscious I can indeed never be an item of which I can be conscious as an object. For if it is the case that the only role the self-conscious I can play with respect to conscious propositional states is to function as their subject, then my making it an object of consciousness would mean being able to transform it into a propositional structure that could do without a self-conscious I as its subject. This seems to lead to a paradox: I would have to think of the self-conscious I in terms of propositions like ‘the thinker of this thought is me’ or ‘I am the subject of this thought’ in order to comply with the demand to be conscious of myself as an object. And I would have to add in each case ‘but this I (or me), because it is an object of my thought, cannot be the thinker of this thought’ in order to fulfill the requirement to be conscious of myself as a subject. Thus there seems to be no way for me to represent myself as the self-conscious I or the subject of my conscious propositional states. [22] I take it that this obstacle is what lies at the basis of the distressing assessment that self-consciousness is an elusive phenomenon, which indeed it is under the perspective of its discursive accessibility.

21 But even if one is prepared to follow these suggestions as to the role of the self-conscious I in our conscious lives, i.e. that it comes about whenever a propositional state occurs, quite a lot of deeply puzzling questions remain. I want to address two of them, albeit too briefly: (a) what are the features of the self-conscious I and how does it realize them and (b) how does this self-conscious I come about and what is its ontological status? Because of the importance I ascribe to conscious propositional states in this context, I begin with a recapitulation of what I take to be a conscious propositional state and its peculiarities. A conscious propositional state consists in a mental attitude of a subject towards a propositionally structured content. The familiar candidates for these attitudes are all forms of mental activities in which thinking is involved like hoping, believing, wishing, knowing, judging, etc. The propositionally structured content can be described in its most general form as an ordered relation between a couple of items like concepts or judgments. A specific case of such a structured content is exemplified by a ‘normal’ affirmative categorical judgment like the one mentioned before ‘this house has a blue door’. Thus to be in a conscious propositional state means for a subject to be conscious in one or the other of the appropriate modes (believing, wishing, etc.) that such and such (e.g. this house) is so and so (e.g. has a blue door). Now what is special about these states is (1) that in order for them to obtain, a subject must have access to or must be able to employ conceptual resources because propositions comprise conceptual elements. This seems to be true independently of any ontological claims regarding propositions: whether one thinks of them as ontologically independent abstract objects out there in the world that are grasped or taken hold of by a subject, or whether one thinks of them as achievements of the constructive efforts of the subject – in both cases one has to acknowledge that propositions are conceptual entities. The other thing special about them is (2) that their conceptual elements have to be related in an ordered way. Not every combination of these elements yields a proposition. Thus when I am in a situation where I perceive a blue door over there (and here I think of perceiving as a propositional attitude) the proposition I should entertain could be expressed by ‘the door over there is blue’ or maybe even by ‘something blue over there is a door’ but it should not be by ‘the blue is dooring’ or by some other nonsensical grammatical construction like this one. Whether the acceptable ways in which these elements must be ordered if they are to form a decent proposition depend on factors which are due to the world or whether they are products of the subject is again another question which has no impact on the fact that there has to be a particular order. Granted that these two peculiarities are characteristic of conscious propositional states, then two observations seem difficult to avoid: (1) that to be conscious of a propositional content already requires conceptual and classifying activities of the subject who is conscious of that content or who is in or who has a conscious propositional state; and (2) that this subject has to be conscious of itself as the subject of that state because otherwise it could not be its conscious state. This summary in place one can tackle the questions mentioned above.

22 (a) One might object to these observations that the activities claimed so far as essential for bringing about a propositional content and thus making possible the occurrence of a propositional state at all do not have to be regarded as achievements of the self-conscious I, even if one agrees that being a self-conscious I is a necessary condition for being in conscious propositional states. One could argue that these activities are physiological reactions of a subject or that a super-human mind is engaged in executing them. I do not know yet how to respond to this objection in a more substantial way than before. But fortunately enough, there are some other features that can be shown to be necessary if a subject is to be in or to have a conscious propositional state, and which can only be accounted for by having access to the conception of the self-conscious I. One of these features has to do with the phenomenon of the indispensable distance between me as the subject of a conscious propositional state and the proposition or the thought I am entertaining. In thinking (believing, judging, etc.) the thought that something is so and so, i.e. in being conscious of a propositionally structured content, I have to be able to be conscious at the same time of a difference between me who is thinking that thought and the thought (the structured content) itself. In thinking, I am not identical with my thought, i.e. I am not my thought but the subject of my thought. After all, I am the subject who is thinking that thought or who has that thought, and because it is to be my thought, I better be able to be conscious of myself as its subject. Thus it seems to be a necessary condition for being conscious of a propositionally structured content or for having a thought that I be able to distinguish between me and my thought or to put a distance between me, as the subject having that thought, and the thought itself, and that I be conscious of myself as its subject.

23 In order to get a better sense of what is meant by talk of distance and difference in this context, one has to appeal once again to a phenomenological observation. According to the picture presented so far, to get into a conscious propositional state is an accomplishment of a conscious subject which is based on the ability to actively abandon the original or primary state of pure conscious awareness, where there is no difference between the conscious subject and its environment but just undifferentiated unity. The transition from the state of undifferentiated conscious awareness to the conscious propositional state is done by the subject by paying attention to, or by focusing on, specific elements of its (inner and/or outer) environment, by singling out particular situations and by giving these elements and situations a propositional structure. It is along these lines of reasoning that in the context of the foundational discussions of early German idealism, people like Hölderlin found it important to distinguish between what they called ‘being (Sein)’ and ‘judgment (Urteil)’. The term ‘being’ was meant to characterize the original undifferentiated unity of conscious awareness, whereas ‘judgment’ (Urteil, interpreted as ‘ursprüngliche Teilung’, original division, by relying on a rather dubious etymological observation) referred to the mode of presence of ‘being’ as propositionally structured object for a subject. But in this transition from the state of amorphous conscious awareness to the conscious propositional state, it’s not only a propositionally structured object that shows up, but also a subject for which this propositional object is present, and which is conscious of itself as this subject by being conscious of its distance to the object, i.e. to the proposition it is entertaining. This is where – to put it in the language of the German idealist period – the subject-object split occurs and the self-conscious I is constituted as a radical non-object, as something whose whole function consists in serving as the self-conscious subject for a propositional object by opening up a distance between itself and the object. If one is prepared to give some credit to such a phenomenological description, one can understand why and in what sense some of the German idealists, especially Fichte, came to declare that the self-conscious I is a condition for objectivity (in a weak sense of sheer otherness): Fichte’s bizarre-sounding claim that the I posits the Non-I can be read as drawing attention to the fact that the very idea of otherness in the sense of objectivity is due to some activity of the self-conscious subject which is directly related to the process of its constitution. [23]

24 (b) However, what still has to be addressed in the context at hand is the question of the origin and the ontological status of the self-conscious I. According to the suggestion elaborated here, the self-conscious I is not a self-standing entity identical with the subject of all my conscious states (presumably my body) but a necessary active correlatum (again a term I borrow from Kant, cf. CpR, A 123) of conscious thoughts, i.e. of a propositionally structured content. This means that this self-conscious I comes into being in virtue of there being a propositionally structured item which can function as its object, and that at the same time the propositionally structured item can occur as an object only if there is a self-conscious I. Now, if one accepts that susceptibility to propositional structure is a result of an activity, an achievement, of the subject, and if one agrees that there is the intimate mutual relation of ‘Gleichursprünglichkeit’ (reciprocal basicality) between propositional structure and self-consciousness, then it is likely the most natural move to attribute the formation of the self-conscious I to the very same activity which is responsible for our responsiveness to propositional structure. This is so because to introduce other factors as responsible for the constitution or evolution of the self-conscious I would seem to be – at least in the context of the approach delineated here – both ad hoc and unwarranted.

25 As to the ontological status of self-consciousness, the proposal presented here suggests that the self-conscious I is a fleeting phenomenon, something that is embedded in the fabric of our conscious lives as a necessary element in the spontaneous process of generating conscious propositional states. Thus self-consciousness is quite well conceived, as Kant did, as something which is produced by an “act of spontaneity (Akt der Spontaneität)” (CpR, B 132) or as something “which exists in the act (was in der Tat existiert)”  [24] (B 423 fn.) of thinking, i.e. of entertaining propositional states. One could even be tempted to follow Fichte’s radicalized version of Kant’s insight and simply assert: “The I posits originally and absolutely its own being (Das Ich setzt ursprünglich schlechthin sein eignes Sein)” (Grundlage der gesamten Wissenschaftslehre 1794/95, § 1, 10).

26 Though I am pretty sure that for many of you what I have outlined so far might ultimately sound a bit too metaphysical, it at least succeeds to bring together the two apparently conflicting philosophical intuitions with regard to self-consciousness mentioned at the beginning. The first of these intuitions was that the self-conscious I is elusive and thus a mystery, perhaps even an illusion, the second was that the self-conscious I has to be given supreme reality. If one follows the line of thought sketched here one can realize that both these intuitions are warranted though not without restrictions. The first one is right in this: the self-conscious I is elusive. Elusiveness is one of its characteristics because this I can never become an object. But this does not mean that it is a mystery or an illusion. The second intuition is right concerning the supreme reality of the I as subject. But this intuition has to be qualified by the observation that it has reality only within the special domain of conscious propositional states. [25]