In Defence of Pre-Reflective Self-Consciousness: The Heidelberg View

In the 1960s, a school formed in Heidelberg around Dieter Henrich that criticized—with reference to J. G. Fichte—the ‘reflection model’ of self-consciousness according to which self-consciousness consists in a representational relation between two mental states or the self-representation of a mental state. I present a new “Heidelberg perspective” of pre-reflective self-consciousness. According to this new approach, self-consciousness occurs in two varieties which regularly are not sufficiently distinguished: The first variety is egological self-consciousness that exists in connection with the use of concepts. It consists in the consciousness of having an awareness of oneself as the Ego of a conscious state. The second variety is anonymous pre-reflective self-consciousness, which is awareness of consciousness but does not contain egological information. The pre-reflective self-consciousness does not exhibit a subject-object structure and thus cannot be appropriately determined as a representation, self-representation, or even acquaintance relation. Rather, anonymous self-consciousness is characterized by an indistinction of subject and object. However, pre-reflective self-consciousness is the basis of egological self-consciousness.


What We Mean by 'Consciousness'
Were it to go unnoticed, life would be devoid of meaning, namely, did it not reveal or manifest itself in a 'field of consciousness', to use a term coined by Gurwitsch. And yet, having said that, I must correct myself, for 'unnoticed' is a misleading expression: For it to occur, consciousness need not be 'noticed'. Noticing is a mode of representation in which content and vehicle diverge: In representation the object maintains a distance from the subject and noticing is, moreover, a mode of attentive representation. Attention, in turn, implies reflection and conceptualization. 1 Now, we can be aware of what we are not attending to (Block 2003, 172), a point Franz Brentano elaborated in detail (1874). 2 He went on to distinguish between experiential (more generally, non-conceptual) and conceptual consciousness, whereby the latter, being judgment-based, is susceptible to deception, which cannot be said of the former in the same way (Brentano 1974, 25 ff.;cf. Brentano 1971, 43, 140 ff.). Feeling a certain way while tasting a sour fruit does not imply thinking (judging) about it-albeit that the sense is clearly specified and not vague. And, conversely, we may entertain thoughts-as sustained cognitive attitudes (or dispositions)-without thereby being aware of it; but obviously we cannot 'experience' something unconsciously. (We cannot feel pain without it hurting.) One further widespread and misleading notion of consciousness holds that it is something which 'accompanies' parts of our mental life. 3 Here again, the tacit assumption is of an existing duality between the accompanying and the accompanied, whereas in this case we must, in fact, assume a strict unity. Yet, no sooner do we set about describing the phenomenon do we seem committed to use the two-place verb of 'representing something to someone'. There are doubtless unconscious representations (such as-leaving to one side the Freudian 'unconscious'-thoughts, beliefs, intellectual and emotional dispositions). And, according to many philosophers, we are conscious 4 of some things without them first having to be re-presented, namely, in the form of an object as counter-posited opposite to our inner eye. Thus, we feel pain or infatuation without having to counter its content with the conscious act (or state, or vehicle) which 'grasps' it. Husserl called states in which we 'feel somehow' (states of "Zumutesein" ['what-is-it-likeness']) without thereby apperceiving an object, 'experiences' ('Erlebnisse'), and went on to characterize them by pointing out that, in contrast to intentions, content and vehicle of consciousness strictly coincide (Husserl 1980 II, 352). He referred to such states as 'non-objectual' ('ungegenständlich') 5 so as to differentiate them from objectifying representations like intentions, wherein 1 Some deny this. My point is that consciousness may work inattentively. 2 Brentano 1973, 41;cf. 42;48;Brentano 1982, 23 f., 29 f., 31 ff. "Who sees a lark in the blue of the sky, has not yet noticed it, whereas noticing implies seeing" (Brentano 1982, 23 f. [my translation]). 3 For instance, David Rosenthal, according to whom being conscious "must be a relational property of being accompanied by higher-order thoughts" (Rosenthal 1991, 474;1997, 736 f.). 4 Here, and in what follows, I do not distinguish between 'consciousness' and 'awareness'. We are even aware of occurring thoughts by virtue of consciousness, not by (additional) thoughts. 5 Pothast (1987, 21 f., 32 ff.); Shoemaker (1984. 14 f.;Shoemaker 1984a, 104 f.) und Castañeda (1991, 234) also refer to a 'non-confrontational' contact with experiential content.
represented and representing dissolve and the representation may survive the absence of its object. (When leaving my desk in pain, the desk remains in my study, while my  pain may persist.) Whenever arising, consciousness presents itself as 'having already been there before,' 6 namely, prior to any act of attention, re-flection or some other kind of selfnoticing-in other words, without our having recourse to a higher-order attitude towards a supposedly primary, first-order phenomenon. I contend that this amounts to a structural feature of consciousness, which explains our oft-lamented inability to approach it "in a non-circular way" (Block 1995, 230;Williford 2006;. Now, it would appear that such circularity is subject to our way of speaking about the phenomenon, whereby the phenomenon itself is, of course, not circular. It seems to consist in a strict subject-object identity which our language is incapable of articulating. Fichte was the first to point this out (in 1797): And yet he concluded that there is self-consciousness. Whereas, in themselves, phenomena cannot be true or false, our theoretical approaches to them may well be. Should our theory prove circular we are then obliged to search for an alternative theory capable of explaining consciousness. One such theory would be that awareness must be acquainted with itself pre-reflectively (Fichte 1971, 526-8;Engl. 1994, 111-113).
By way of a first, tentative summary, I would argue that consciousness appears to be a variegated phenomenon. Whereas it may be both pre-conceptual (or experiential) and conceptual (where thought or whatever cognitive attitude comes into play), it is not by virtue of an initial act of either of the latter that consciousness emerges. Apparently, it is not revealed to itself by way of some 'inner gaze', whereby a quasi-observatory act is directed towards a primary or basic state which, in turn, is typically hetero-directed, namely, to an object or fact distinct from an act upon which it reflects. In consciousness, there is a seamless coincidence of that which is represented and that which does the representing. Hence, as an explanation of emerging of consciousness as such, the term 're-presentation' disqualifies itself from the outset. 7 However, this failed to prevent an entire philosophical tradition-from Descartes, Leibniz, Kant, Reinhold through to recent representationalists, and including 'self-representationalists' and 'higher-order theorists'-from conceiving of 'representation' as the instance (or, following Kriegel, the "core necessary condition") that triggers consciousness. Thus, to conclude, it is not "by virtue" of being self-(or hetero-) represented that consciousness initially arises (Kriegel 2009, 13 f.).

Self-Consciousness: A Unary or a Binary State?
On the other side, consciousness seems to be essentially familiar with itself and so to imply self-awareness. Clearly, saying 'I am aware of this fact, but unaware of experiencing it', would be nonsensical, which is why some philosophers refer to the "ubiquity of self-awareness" (Kapitan 1999;. Insofar it is like something for the subject to have experiences, there must be some awareness of these experiences themselves; in short, there must be self-awareness. But this originates not by way of an inner duplication into first and second order: The experience arises first, which is then followed by the awareness of it. Experiences seem to be conscious of something (different from themselves) "immediately", namely, concurrently and by the same act by which they become aware of themselves as that through which the external fact or object manifests as content in the field of consciousness. Thus, we must be cautious of conceptualizing self-awareness according to the "model of re-flection" (Henrich 1966,192 [ff.];Henrich 1982;Henrich 2003, 241-245, 246-262), which introduces an inner hiatus into the phenomenon, the original structure of which is "unary" rather than "binary" (Wehmeier 2012). Structurally, self-awareness is different from object-awareness and we are prone to be misled when conceiving the 'self' of self-awareness as the proxy of an object to which we can point independently. It would not be possible for an object to be revealed as identical to the subject had not the subject previously identified the object as itself (Shoemaker 1984a, 105). In other words, objectual self-knowledge presupposes non-objectual self-familiarity-a point to which I will return in due course.
This conclusion I hold to be of particular relevance, and find it worthwhile reiterating: The nominalized reflexive pronoun 'self' in the composite 'self-consciousness' stands precisely for the known 8 self-sameness (the identity) of the phenomenon, not for a division which splits it into two. The apparent duality of the relata is simply denied by the actual semantics of the reflexive pronoun: It indicates a unity, which nevertheless falls into consciousness. In other words, the unity, as a matter of principle, never goes 'unregistered'. (Here, I refer again to my introductory reservations with respect to such terms. I simply mean-along with Husserl or Kripke-that pain, for instance, entails or rather is awareness of pain without the event of feeling thereby being distinct from the instance of registering [Kripke 1971, 146 f.;1980, 152]. To make this clearer, Sartre, for example, places the distancing preposition "de" into brackets when speaking of "conscience (de) soi" [Sartre 1947, 382]. "In self-consciousness", he explains, "there is no subject-object distinction," as is characteristic of the reflective or intentional relation of "self-knowledge" where I make myself the object of a distinct thought [380]). Being conscious (of) oneself is different from entertaining an intentional or representational relation to oneself or to one's mental states.)

A Very Brief Digression on the Identity of Self-Consciousness
Some authors think that the unity, as manifested in self-consciousness, may also be present independently of our consciousness (Strawson 1959). They think of the identity of our body to which I refer non-ascriptively without identifying markers from a first-person perspective, whereas others refer with identifying markers from a hetero-perspective with ''you'', ''he'' or ''she'' (Tugendhat 1979, 83 f.). The apparent binarity inherent in the reflexive pronoun would not be more divisive insofar as it constitutes the repetition of a linguistic type: saying the same thing twice does not mean saying two different things. 9 Now, Chisholm has shown why one cannot detach the identity of the epistemic subject from its self-consciousness and, as it were, affirm it independently. Indeed, identity here is itself constituted by consciousness. Were self-consciousness an identifying consciousness of that entity which is myself alone, it would necessarily be the property of haecceity that individuates me (Chisholm 1981, 15 f.). But the representation of my haecceity does not assure me of the fact that I am the bearer of this unique property (or set of properties), unless, that is, I derive this knowledge from a pre-objective source. This, then, brings 'epistemic terms' into play (23 f.; cf. Parnas / Zahavi 1998, 553-5). Put differently: from the objectual knowledge of my identity there is no non-circular path that leads to self-awareness. Hence, self-awareness cannot be grounded in the identification with an ens omnimodo determinatum, but in the mechanism of its epistemic appropriation. 10 For this reason, to describe things thus would be wrong: In self-consciousness, a self-sameness presents itself under two guises, once as the object (with individuating properties) and again as the subject of representation. I would simply have to identify the two as two guises of the same referent, as in water and H 2 O. Wrong: The representing subject is not represented under the guise of the representing. What corresponds to it is a "representation sui generis". In other words, the problem is that on one side of the identity we have a representation of an unconscious object whereas on the other side we have a fully-fledged conscious experience which is non-objectually and non-representationally self-disclosed (Levine 2006, 188, 193). Either I am aware of it immediately and non-objectually or else I radically lack self-awareness.
The "mongrel" nature of consciousness: Representing outer things, including consciousness-as-object, and being non-representationally and non-identificatorily selfpresented (or self-disclosed) is, what Husserl had in mind when remarking that, "All consciousness is consciousness of something [typically distinct from consciousness itself], but all consciousness is also conscious" (Husserl 1966, 126).
Although we use the word 'consciousness' in both instances, the meaning has shifted considerably between the first and the second application. The first use is transitive, the second intransitive. Though often noted, the inner structure of this socalled intransitive consciousness has never really been elucidated. Rather, the reflec-tion model has obscured and misconstrued these explanations. The same holds for Husserl himself and, indeed, of his teacher Brentano (Cramer 1974;Frank 1991, 526-557).
I will address the regresses and circularities triggered by the reflection or higherorder model of self-consciousness in sections three and four.

A Further Differentiation of 'Self-Consciousness': Non-Egological and Egological
I have so far been considering self-consciousness as a kind of non-objectual consciousness-consciousness despite having used the term 'subject' of consciousness and referring in the above to the subject's 'identity'. It is time to remove the prevalent ambiguity. By self-consciousness, do we not commonly understand the owner (or agent), the Ego of a conscious state?
Indeed, we do. Yet the nominalized reflexive pronoun 'self' in the composite 'self-consciousness' is receptive to a narrower interpretation. It may be understood as referring to consciousness itself, conceived of impersonally as, for the most part, I have done in the previous section. This latter may, in turn, be understood as a single episode (a "momentary minimal subject" [Russell 1956;Strawson 1997;2008;]) or as diachronically unified "stream of consciousness", as we witness in Husserl's Phenomenology of Inner Time-Consciousness without a synthesizing I external to the flow of time itself.
Having now made this distinction, we may assign impersonality or anonymity to non-conceptual states or events (roughly, experiences or bodily sensations): 11 "it hurts", "it feels good", "how disgusting!"), while conceptual mental events seem typically to be I-governed. We thus encounter one of Kant's basic intuitions, namely, that it is the 'I think' which erects the conceptual framework of our phenomenal world, thereby transforming it into an 'objective' one: as he says, one to which we may refer by uttering 'true judgments'. In short, Kant held that all and only our cognitive capacities (as they are manifested in the faculty of applying 'categories') have their origin in the I think, and such that he would have never directly ascribed self-consciousness to sensory or emotional states. For him, self-consciousness is just a kind of cognitive, as opposed to immediate or pre-reflective, self-knowledge, an "I-owned" knowledge.
Unless I am mistaken, many if not all contemporary philosophers of mind incline towards a Kantian position on this. Thus, Sydney Shoemaker refers to a "self-blind", by which he means an 'I-blind' subject incapable of gaining access to the sphere of rationality. It would self-ascribe mental states in the third person, without knowing "by self-acquaintance" whether such self-ascriptions were true (Shoemaker 1996, 25 ff., 31). One may also think of Tyler Burge, who considers a subject, which is not elementarily familiar with itself, incapable of translating worldly constellations into motives for action. A self-blind subject would be motivationally insensitive (Burge 1998). Both Kant and more recent philosophers, including myself, assume an inner bond between egoity and rationality.
Similarly, several modern thinkers in the Leibnizian tradition attributed to egogoverned thinking the property of spontaneity, that is, a form of intellectual activity, as did, above all, Kant and Fichte. In his essays of 1970 and 1971, Dieter Henrich referred to an 'active principle'. This principle is, for instance, responsible for such performances as "conscious concentration on an object, the solution of a problem, the decision in favour of a plan of action, or the anxious expectation of an event. Whatever this [cognitive] self might be, it is at least an active principle of organization in the field of consciousness" 12 (Henrich 1970, 276;1971, 20;cf. 1971a, 5, 7, 10-12, 14-16). But the active ego principle also seems responsible for the fixation of the absolute zero point of the 'now', from which all conscious orientation takes its starting point (Henrich 1971a, 16;2016, 202).
Henrich proposed a 'multifactorial model' of consciousness, according to which I-thoughts cannot be derived from (impersonal) consciousness, as Sartre attempted in La transcendance de l'Ego in 1936. Fichte's reverse attempt to make non-conceptual representations intelligible from the fully developed 'I think' was also mistaken (Henrich 1971a, 6 f.). In Sect. 4 I shall present this model and adopt it-in amended form-in Sect. 5.
As mentioned, Henrich's 'active principle' governs propositional and de-se attitudes, whereby a subject not only apprehends itself as an object but apprehends itself as itself. The distinction is easily exemplified by an oft-cited example. The Viennese physicist and philosopher Ernst Mach recounts having once boarded an omnibus in a state of fatigue after a trying railway journey by night. While trudging up the bus entry stairs he caught sight of a gentleman entering from the opposite side in the same rhythm, and on catching sight of him the thought flashed through his mind: "What shabby pedagogue is that chap who has just entered," without noticing the large mirror hanging opposite him (Mach 1903, 3). Had self-consciousness exhausted itself in the conscious conception of the right object to which we ascribe properties, Mach would have given an accurate de-re description of himself. (The haeceitas problem previously discussed.) But he indeed failed to apprehend himself as himself*. 13 Mach's deception did not occur in his consciousness, and was thus a failure of the 'active', intelligent or I-principle. It is not responsible for the emergence of elementary awareness, since, of course, Mach attributes his perception of the alien master to his own field of consciousness. An image suggests itself: Thinking, insofar as it is conscious, rides on the current of already constituted consciousness. I would venture to say that this is simply what is meant by those philosophers of mind who speak of the 'phenomenality of the intentional': Consciousness of cognitive representations is not the product of an additional thought event, but is an event as occurs in phenomenal awareness ['Innesein', interiority]. Separating the two would amount to "separatism" (Horgan/Tienson 2002;Pitt 2004;Soldati 2005;Horgan/Kriegel 2007).
Let us call the two varieties of self-consciousness, following Sartre (1947) selfawareness and self-knowledge.

Impersonal Consciousness Considered as an Anonymous 'Field'
It would be worthwhile dwelling for another moment on anonymous consciousness, since it is here that the 'reflection model' suffers decisive defeat.
As mentioned, along with Aron Gurwitsch, some consider the latter's structure not only ego-less, but field-like (Henrich 1970;Searle 2004)-and many philosophers of mind employ the metaphor without being entirely clear about its implications. They think the field is a) neither conceptually manifest, nor a result from reasoning or inference, and b), that self-awareness is immanent in the experience itself (Gurwitsch 2009, 207) and does not arise by way of an additional act of reflection. Reflection, Gurwitsch says, "creates nothing new"; it simply discovers what was already there before, namely self-disclosed pre-reflectively (ibid., 127; cf. 176 and Gurwitsch 2009a, 327), 3. The anonymous field does not have an I or Ego as owner, agent or field unifier. The Ego, Gurwitsch thinks, is not constitutive of any content, let alone object, of consciousness, but is itself something constituted. It is not an agent, but a product, not active, but rather passive, something that arises or emerges in the margins of the field of consciousness without anyone's doing. Adopting Sartre's view, Gurwitsch considers it the result of a misguided, indeed, an "impure" act of reflection which, over time, illicitly "blurs" its instantaneous content ("noema") (thus rendering a short-term episode of anger a purported long-term disposition of hatred, thereby substantiating empty awareness) (Sartre 1978, 45-51, 82). 4. According to Husserl's mereological terminology (Logical Investigation III; cf. Gurwitsch 2010Gurwitsch , 28-33: 2009a, the field of consciousness is gestalt-like in structure, namely, as consisting in "parts" ("Teile"), not "pieces" ("Stücke"). Parts are "dependent" ("unselbständig"), pieces are "independent" ("selbständig"). Parts constitute segments of a "complex" and cannot be detached without destroying, or at least altering, the entire "gestalt". Pieces form a "sum" when composed and may be differently combined no less easily. Were the field of consciousness composed of pieces, it would disintegrate into as many individual occurrences of consciousness as the pieces of which it is composed. 14 Similarly, it is this holistic view of consciousness, not as a single event but as a field, that underlies Van Gulick's Higher-Order-Global-State (HOGS) view (Van Gulick 2000, 296), with which Gennaro's Wide Intrinscicality View (WIV) also works (Gennaro 2012, Chap. 4).

Higher-Order and Same-Order-Monitoring Theories of Self-Consciousness Fall Prey to the Reflection Model
According to the non-egological view, primary self-consciousness or, as it is sometimes referred to, "subjective character", is a matter of a mental event, episode, or state, which either represents itself, is represented by a distinct mental state (etc.) belonging to the same mind, or is an internally represented part of a unified representational whole (a "complex"). This type of internal reflection, not to be confused with introspection, is meant to explain how a "normal", unconscious mental state becomes a conscious mental state, or else becomes, according to the egological view, one of "my representations" [à la Kant in CPR, B 131 f.]. This type of position survives in so-called higher-order and same-order representation theories of consciousness, the majority of which are egological.
Higher-order and same-order theories (whether egological or non-egological) have come to be subdivided into theories, which view consciousness (with its "selfconsciousness" or "subjective character") as consisting in at least two numerically "distinct" mental states or mental state parts, one representing the other, whether the former ("reflecting") state be a perception (e.g., Armstrong, Lycan) or a thought (e.g., Rosenthal, Gennaro, Weisberg). 15 Higher-order-perception theories (HOP) start out from a structural analogy between perceiving an outer object and introspecting a mental act. One advantage of this model is that, unlike representation, perception as well as introspection warrant the existence of their object. This becomes all the more evident when, following Russell (1914), we speak of 'acquaintance', as does Williford (2006a;. And yet in both cases, a minimal gap opens up severing perceiving and perceived: that the perceiving act should recognize itself* in the perceived state amounts to an implausibility. HOT aggravates the problem. It is not a perception-like introspection, but a thought that is supposed to do the job of effecting second-level reflection on firstorder representation (Rosenthal 1997, 730). This creates a propositional, a de-dicto attitude towards the first-order state. However, philosophers from Castañeda to Lewis, Chisholm and Bermúdez, have sought to point out that attitudes de se cannot be reduced to attitudes de dicto: 16 not a question of epistemology, incidentally, but one of logic (Chisholm 1981, 18 f.).
The crux of the matter is expressed in the sentence, 'S knows that S is φ', wherein S is an epistemic subject and φ a mental state. How does the subject of the cognitive main clause make itself aware that the subject of the subordinate clause is itself* φ? By a higher-level thought? Surely not, for the intentional act of reflection encounters an object which Rosenthal explicitly asserts is 'numerically distinct' from the reflective act. There may even be a lapse of time between the two (Rosenthal 1991, 465)-a conceptual sloppiness of which same-order-monitoring theorists (Horgan/ Kriegel 2007, 131) are culpable, too. And both states are said to be unconscious prior to reflection. The term 'introspection' is reserved for a third-order thought intentionally directed upon a second-order mental state which subsequently becomes conscious (ib., 745; 1971, 466).
In any case, a HOT is an optional "extrinsic" additive to a FOR (which may or may not occur) and which to that extent "bear[s] a relation [not to itself, but] to something else" (Rosenthal 1997, 735). By way of an amendment, one might, along with Gennaro, argue that higher-and lower-level acts/states both belong to one and the same complex (call it mind or flow or field of consciousness) (Gennaro 2012, 21-24). In order to identify the unity of this complex with the reflexive act, the subject performing the reflection-whether conceived egologically or non-egologically-must have already possessed pre-reflective self-awareness (Parnas/ Zahavi 1998, 695;in reference to Shoemaker 1968in reference to Shoemaker , 561 [= 1984).
According to the HOP or HOT view, an extensive regress is naturally formed upwards in this manner such that an unconscious act remains at the top. Indeed, Rosenthal wryly admits this: "It may seem slightly odd that each of these hierarchies of conscious mental states has a nonconscious thought at its top. But whatever air of paradox there seems to be here is dispelled by the common-sense truism that we cannot be conscious of everything at once" (466).
Fichte believed that this pulls down the whole chain of preceding states into the unconscious (Fichte 1971 I, 526-8;Engl. 1994, 111-113). More precisely, he thought that, taken as a whole, the chain of reflections may not only lack consciousness, but self-consciousness.
Rosenthal, by contrast, sees this kind of regress as harmless given that with the third-order state introspective self-consciousness is already achieved.
In so doing, he overlooks the de-se constraint. In order to achieve self-consciousness the third-level act must refer to the second-level act not only as a 'distinct (or external) object', but to it as to itself*. And-as we have already noted-here we run into a circle in the argument (circulus in probando). If the higher-level act is to be aware of its identity with the lower-level act a de se relation must prevail. How else in the object state could the higher-level act recognize itself* and not something other or somebody else's state? A kind of brainwashing would have to have taken place to bring it to identification with a state which obviously deviates from its own state, or from itself*.
Gennaro's WIV thus seeks to redress the deficiencies of Rosenthal's theory, which he dubs External Higher-Order Theory (EHOT): The first-order representation (FOR) (Gennaro 2012, 57) must bear an intrinsic or essential relation to HOT, by 'intrinsic' is meant a property which necessarily and not merely coincidentally belongs to FOR, and which is intimately connected with a 'proper part' of HOT. Both form kind of a "complex mental state", a "whole" (in the familiar mereological or field sense) which, numerically speaking, does not prevent them from being "distinct" states. For fear that regress or circularity would otherwise "rear their ugly heads", Gennaro insists that however "intimately" HOT may be connected to FOR, they must not coincide and must maintain a minimum 'representational' distance.
The situation does not fundamentally alter when passing from HOT theory to Kriegel's self-representationalist view, irrespective of his attacks on Gennaro. Again, according to the mereological schema, a fully-fledged conscious mental state M is supposed to contain various logical components or parts, among which M*, the 'representor' of M, is the represented. Here, Kriegel assumes that M* cannot represent M ("its whole-self": Kriegel 2009, 204;215) directly or in any straightforward manner (201 ff.), but only in a "crooked" way (200, 216), namely, by way of the representation of another, distinct (mereological) part or component: M ◊ , provided, that is, that both (M* and M ◊ ) are not only 'pieces' of the same sum, but 'proper parts' of one and the same single complex-in the familiar sense.
Along with Gennaro, Kriegel accepts this laborious explanation simply for the sake of naturalising self-consciousness. As naturalism purportedly cannot work with identity it must work with a transitive, binary, hetero-relation à la causality, the "cement of our reality". And yet, why should a naturalistic explanation renounce reflexivity or the usage of the identity sign? With respect to self-awareness, we simply cannot renounce the de-se constraint proper to phenomenology since we are not merely obliged to explain the underlying causal mechanism, but to account for M*'s being aware of its being identical with M, or rather, for M*'s becoming aware of itself* as being M. Such self-conscious self-identification cannot go the way of inference without ipso facto forfeiting its immediacy and non-conceptual self-certainty. (Little wonder, then, that Gennaro locates his own position midway between Kriegel's and Rosenthal's position [Gennaro 2006].) It would seem, then, that everything points in one direction: We should abandon our recourse to relation as the basic structural feature of self-consciousness. None of the above representationalist proposals aptly captures the structure of a single mental state that presents itself to itself*.

The Heidelberg View
The first to have paid due attention to the whole range of problems was the founder of the 'Heidelberg School', Dieter Henrich, 17 who insisted on his indebtedness to Fichte (Henrich 1966;English 1982). Indeed, around 1797, Fichte claimed that in contrast to modern thinkers from Descartes through to Leibniz and Kant, we cannot explain the phenomenon of self-consciousness considered the 'first principle' of philosophy, namely, according to a model of a mental state that represents itself, or two representational states as being stacked one on top of the other. Either consciousness is aware of itself immediately ('pre-reflectively'), or else there can be no consciousness.
Henrich's critique of the 'reflection model of self-consciousness' has been translated into English soon after the publication of the original (1971 and 1982). Not only 17 As dubbed by Tugendhat (1979, 10). did it converge strikingly with the insights of other contemporaries such as Castañeda (1966) and Shoemaker (1968), but Chisholm (1981) and Nozick (1981;2011) also adapted it, especially after Henrich had rearticulated his critique in the legendary 1973 Harvard lectures to an audience of the likes of Quine, Rawls, Cavell, Putnam and Nozick (printed version: Henrich 2003). 18 Analytic philosophy has nevertheless largely ignored Henrich's original insight. 19 In his 1966 essay (English 1982), Henrich not only defended the pre-reflectivity of self-consciousness, but also discovered-along with Castañeda around the same time-the de-se constraint. Again, he attributes this discovery to Fichte who had, indeed, modified his famous formulation of 1797 from "The Ego posits itself purely and simply (schlechthin)" to "The Ego posits itself as positing itself" (Fichte 1978, 33). 20 To dismiss the reflection-model of self-consciousness oriented on the egological structure of a self-referential 'I think', Henrich introduced a new explanatory attempt in (Engl. 1971. Here, he oriented himself on the model of an 'egoless consciousness'. To this he even assigns "definite advantages" (Henrich 1970, 275;1971, 20).
Again, his point of departure pivots on the insight that "familiarity with consciousness cannot be understood as the result of any initiating activity", but must precede it. "Such familiarity must already be present whenever consciousness comes into being. No one can say that he tried to come to consciousness in the way he can try, e. g., to reflect, introspect, or observe" (Henrich 1970, 271;Engl. 1971, 16).
There is no way in which self-consciousness should be conceived as existing in a relation; at least not in the Fichtean manner of a "knowing self-reference" or a cognizing Ego, which identifies itself under the guise of an object with itself under the guise of a subject (Henrich 1970, 275;1971, 19). The crucial point here is "to not even conceive the genuine structure of the self as the result of a self-relating reference" (Henrich 1971a(Henrich /2007. And again, "One cannot, by reflecting on a state of affairs, bring it to consciousness for the first time" (Henrich 1970, 271;1971, 11). It is this context which makes Gurwitsch's operation with a self-disclosing but self-less 'field of consciousness' appear attractive (Henrich 1970, 276 f.;1971, 20).
There are some essential features of pre-reflective self-awareness: It exists exclusively in actu, never as a mere disposition. "There may be consciousness of potentiality, but no potentiality of consciousness" (Sartre 1947, 382). Within the field, it is true, data are situated in relation to each other-and necessarily so, since the differentiation of its contents is a necessary condition of consciousness (Henrich 1970, 277;1971, 22). 21 Yet, neither the epistemic form-component "nor the dimension containing it refer to themselves" (Henrich 1970, 278;1971, 22). Metaphorically speak- 18 Henrich reports of his exchange with these analytic philosophers in Henrich (2019, chap. 15). 19 There are four significant exceptions: Dan Zahavi (1999), James Hart (in: Castañeda 1999, 17 f. 25, 29), and Dieter Freundlieb (2003). See also Borner et al. (2019). 20 For reasons of brevity, I shall not further elaborate on the problems entailed by Fichte's improved formula: They give rise to an 'intensive' or Matryoshka-doll regress (Cramer 1974;Williford 2006; ing: When in the field the I raises its head, it has no concern for its consciousness. It is already guaranteed this by its being embedded in the 'field'. Ulrich Pothast, one of Henrich's former doctoral students, goes further: Consciousness results from a "wholly objective process, in the sense that no moment of a knowing self-relation enters into it" (Pothast 1971, 76)-precisely what Husserl claimed of non-conceptual experiences (see Sect. 1, in the present article).
But what is the unity of subjectivity? Henrich speaks of a 'functional' connection of a) the field, b) the active (ego) principle and c) what he calls the 'form component': the 'familiarity'. The connection is referred to as 'functional', as opposed to 'linear', whereby 'linear' is a bond of derivability between the three components that comprise the full structure of self-consciousness.

In Accord with my Goal of Advancing the Heidelberg view
I find Henrich's 'multifactorial analysis' of 1970/71 attractive, albeit that its originator has since returned to an egological, cognitivist, and consequently relational view (Henrich 2019). In letters, e-mails, and oral statements, he dismisses his earlier attempt as 'artificial', all too closely resembling an 'ad hoc solution' to a far more complex problem.
Here, I should propose some shifts in emphasis. Henrich would make it seem as if self-awareness, insofar as it is not derivable 'linearly' from self-knowledge (and vice versa), had little if any influence on the latter. As already mentioned, however, I endorse the 'non-separatist' argument (e.g., Horgan/Kriegel 2007): that which makes I-governed cognitive states conscious is precisely their embeddedness in the 'field'. Occurrent cognitive (typically propositional) attitudes should have a certain 'feel' to them; there should be something it is like to 'experience' them. Since experiential states can be characterized as 'unary' unproblematically and without circularity, their structure may also dedramatize purported circularities and regresses that threaten the reduplicative structure of attitudes de se.
One further remaining difficulty is the relationship of the non-distinction of subject and object to the relationality of attitudes de se. David Lewis argues that the latter are rooted in a relational property, namely, as being inhabitants of an ego-governed world (where propositions are defined as sets of possible worlds in which such propositions hold [Lewis 1983]). Now the knowing self-reference-the relation one bears to oneself as oneself*-is indisputably a relation, albeit one that appeals to a unity manifest in knowledge, if also one that cannot be explained in terms of the knowledge relation itself. Would it be altogether odd to recognize in this the subject-object indistinction, about which we anyway assumed that it precedes cognitive self-knowledge and sustains it in all its conscious activities?
We may, finally, make a concession to the recalcitrant fact that it would hardly be possible to articulate self-consciousness without using reflexive turns, which is, indeed, what Sartre did (Sartre 1947, 388). As inventor of the phrase 'pre-reflective self-consciousness,' he insisted that consciousness implies a minimal reflexivityand thus a minimal distance from itself-without which we could not make such phenomena as time-consciousness and self-deception (mauvaise foi) intelligible (Sartre 1947, 388 [ff.]). Sartre thinks, however, that we should picture the phenomenon as a 'reflet reflétant' (Sartre 1947, 386 [ff.];1943, 167, 221). So long as the mirror remains empty, namely, reflects only itself, it does not endanger the unity and transparency of consciousness. And world-objects are ipso facto expelled from its empty interiority.
Sartre calls the-virtual-mirror-unity of consciousness 'indistinction' and clearly distinguishes it from 'identity' proper, which is exclusive to en-soi objects (Sartre 1943, 33). Indistinction may unfold in difference (a past conscious state does not coincide with a future one), identical is "that which is what it is" (Sartre 1947, 386-8). I strongly incline towards this model, though to further unpack it would breach the confines of the present contribution/article. 22 To conclude: Recent Philosophy of Mind clearly tends towards naturalizing consciousness, and representation is a term susceptible to naturalization, while 'irrelational self-consciousness' is not. Currently, any theory which fails to show how the phenomenon it seeks to describe could be explained in physicalist terms is peremptorily shelved. Yet in the field of psychic phenomena there are those "objects," the structure of which is more susceptible to empirical explanation; and there are those the structure of which is less susceptible to such an explanation; and my descriptions of phenomena are expressly less so. This will doubtless disappoint philosophers of a neurobiological suasion. But whoever intends to spell out the structure of consciousness in terms of brain activities, must first know precisely what it is for which he seeks empirical explanation. Descriptions of consciousness, untenable simply on logicalstructural grounds, are, namely, a priori incapable of sustaining naturalistic theory. And herein one detects the lasting truth of Frege's and Husserl's anti-psychologism.
Funding Open Access funding enabled and organized by Projekt DEAL.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/ licenses/by/4.0/.