1 Introduction

The problem of self-consciousness (Selbstbewusstsein) has been a fundamental problem of philosophy since the advent of modernity, particularly within Continental philosophy since the second half of the twentieth century. It is central to the dispute between the Heidelberg School (the main representatives of which are D. Henrich, K. Cramer, U. Pothast, and M. Frank), and E. Tugendhat, which lasted nearly thirty years.Footnote 1 Examining this dispute reveals various aspects of the traditional theory of self-consciousness and allows us to meditate productively upon the so-called “circular dilemma” and, possibly, find a way out.

Much earlier than either the Heidelberg School or Tugendhat, however, Husserl also paid particular attention to the basic theme of the traditional theory of self-consciousness, and the various contemporary disputes can be clarified through Husserl’s phenomenological theory. I am particularly concerned with the most fundamental of these disputes, namely, the question of whether or not the reflection-model of self-consciousness is circular. First, I will elucidate the basic approach of the traditional theory of self-consciousness (the “reflection-model”) and its dilemma (the “circular dilemma”). Based on this elucidation, then, I will inquire into what kind of theory of self-consciousness Husserl’s phenomenology offers, and, in response to Tugendhat’s critique, I will explore whether such a phenomenological theory of self-consciousness is able to avoid the circular dilemma of the traditional theory.

2 The Circular Dilemma of the “Model of Reflection” and Its Way Out

The revival of contemporary interest in the theory of self-consciousness begins with Henrich’s well-regarded article, “Fichte’s Original Insight.”Footnote 2 The entire Heidelberg School, Tugendhat, and other researchers concerned with the theory of self-consciousness are deeply indebted to this study. Henrich briefly traces the development of the theory of self-consciousness and points out that, despite various differences between Kant and his intellectual forebears in this regard, almost all share a basic model that he called “the reflection theory” of ego-consciousness or self-consciousness.Footnote 3 A clear formulation of this reflection-theory can be found in Kant:

That I am conscious of myself is a thought that already contains a twofold self, the self as subject and the self as object. How it should be possible that I, who think, can be an object (of intuition) to myself, and thus distinguish myself from myself, is absolutely impossible to explain, although it is an undoubted fact…We are not, however, referring thereby to a dual personality; only the self that thinks and intuits is the person, whereas the self of the object that is intuited by me is, like other objects outside me, the thing.Footnote 4

In brief, the reflection-model of self-consciousness holds that the ego detaches its attention from the external object, turns back to the internal field, and “reflects” upon itself. Similar to the presentation (Vorstellung) of the external object, the ego as subject (or the action of consciousness) presents itself as an object. This theory of reflection is the oldest and most evident model of self-consciousness. Frank (1986, p. 30; 1991, p. 419) traces its source back to the metaphor of the mind’s eye (the “optical metaphor”) found in Parmenides and Plato. Tugendhat (1979, pp. 17, 34) considers the entire European philosophical tradition from Parmenides to Husserl to have fallen victim to the tendency toward an epistemology of sight, or rather, to the “the metaphor of seeing.” Later we will explore whether the core of the problem is the object which is being seen or the act of looking itself.

However, there is an ineradicable dilemma in this theory of reflection. According to Henrich’s investigation, Fichte is the first thinker both to point out the dilemma and try to resolve it. It is formulated as the “circle” of the reflection-model of self-consciousness, and, following Fichte, Henrich summarizes the dilemma as having two basic difficulties. The first problem is that the theory of reflection attempts to talk about self-consciousness as the presentation (Vorstellung) or knowing of the self as object by the self as subject. The “ego-subject” is presupposed here, but how can such an ego-subject turn back upon itself if it is not conscious of itself in advance? A vicious circle appears here: the problem to be resolved (i.e. self-consciousness) has actually been presupposed from the start. Second, the theory of reflection presupposes that the ego’s cognition of itself is possible only through a turning back upon itself; it must therefore presuppose that the “ego-subject” that knows and turns back is identical with the known object-ego. In other words, the theory presupposes the identity of the act of knowing and what is known. Only in this way can one speak of self-consciousness. However, how could such an identity be given? Would this not simply be self-consciousness itself? Thus, there is a circle again (see Henrich 1966, pp. 193ff; 1982, pp. 62ff).Footnote 5

In a later article, “Self-Consciousness: A Critical Introduction to a Theory” (Henrich 1970), Henrich repeats the two basic difficulties, but there are some changes. First, he is no longer optimistic about Fichte’s attempt but holds that he is not able to resolve the circularity in the reflection theory.Footnote 6 And the more important change is that the second difficulty, only briefly discussed in the earlier article, is widely developed in this new one. In Henrich’s view, the second problem is not only a difficulty for the reflection theory but for any traditional theory of self-consciousness. Self-consciousness here certainly means that what the “I” gets acquainted with is itself. Although it is not necessary to describe it conceptually, asserting that the “(subject) I = (object) I”— that the subject-I and the object-I are identical—is crucial. However, in order to come to know the self-sameness of the subject and object, the “I” must know in advance how to attribute what this “I” encounters to itself. Thus, self-consciousness as self-attribution is not the result of reflection but is, rather, the presupposition of reflection. In effect, the reflection-model in the traditional theory of self-consciousness—and, really, any theory of self-consciousness—has presupposed such self-attribution at the outset and thus falls into the circular dilemma, which Henrich also calls “the circular subject-object relation of self-knowledge” (1970, pp. 266ff). In order to avoid the circular dilemma, the Heidelberg School, represented by Henrich, speaks of a “pre-reflective,” “non-objectifying” “self-acquaintance” (Selbstvertrautheit).

Tugendhat inquired into the traditional theory of self-consciousness, and the efforts of the Heidelberg School as well, using the framework of the two basic difficulties noted by Henrich. Like Henrich, Tugendhat deemed that the traditional theory had become muddled. Furthermore, he thought that the Heidelberg School “marks a discernible end point in the traditional theory of self-consciousness” (1979, p. 54). With this assertion, Tugendhat acknowledges the Heidelberg School’s attempt to make a radical break with the traditional theory of self-consciousness; but he also tries to place the Heidelberg School within that tradition and distance himself from it.

Tugendhat summarizes three different strands within the philosophical tradition that provide the basis for the traditional theory of self-consciousness (1979, pp. 33f). The first model is the so-called “ontological model of a substance and its states” that has determined the whole philosophical tradition since Aristotle and is deeply rooted in the basic structure of our speech – that is, in the “subject-predicate structure.” The source of such a model can be found in Aristotle’s Categories and Metaphysics. As the foundation of a theory of self-consciousness, this model takes “self” as a substance or entity, and “consciousness” is seen as a state of the substance. In this sense, then, self-consciousness would mean the grasp of a substance and its states. We find a trace of this theory when Descartes deduces the ego-substance as mind from the ego cogito. In Hume’s work, Descartes’ ideas encounter a kind of resistance, which was later echoed in William James and in Husserl’s efforts of 1901. Tugendhat evidently felt the most sympathy for this model: “The first model has been most subject to question, although in reality it is the most harmless. […] It appears very disputable, however, whether this model really must be abandoned for epistemic self-consciousness” (1979, p. 34/1989, p. 25). Tugendhat’s sympathy here derives from the following two aspects of his thought: on the one hand, his insistence on an empirical, substantial ego-individual, and, on the other, his method of analyzing language, which focuses on its subject-predicate structure.

The second model is the idea of a “subject-object relation.” According to this model, consciousness means “having something before oneself” and is the “presentation” of the object. Tugendhat dealt with the conception of “having something before oneself” in his earlier work, Der Wahrheitsbegriff bei Husserl und Heidegger (1967, pp. 300, 311, 339), showing the persistence of this model in Heidegger’s notion of object as what stands over and against the subject. In this model, to talk about “self-consciousness” means to talk about a relationship between the subject and himself qua object: one has oneself before oneself. It is the foundation of the reflection-model in the traditional theory of self-consciousness. In Tugendhat’s view, reliance upon the subject-object model is a characteristic feature of all traditional theories of self-consciousness, including those of the Heidelberg School (1979, p. 51).

The third model turns on an “epistemological orientation toward seeing” in which it is held that all immediate empirical knowledge must rest on perception. This entails that the knowledge of oneself is a kind of inner having-in-perception (Gewahrhaben). If our sight is normally directed outward in perceiving external things and objects, when we attempt to grasp ourselves we must turn our sight inward in an “inner perceiving.” The metaphorical concept of an “eye of the mind” that can look “inward” may be traced back to ancient Greek philosophers such as Plato, and it finds expression in the theory of self-consciousness that lies at the base of both modern Anglophone and Germanic philosophy.

According to Tugendhat, then, all traditional theories of self-consciousness were developed in the framework of these three models. He holds that the first model cannot contribute much toward clarifying the phenomenon of self-consciousness, though he seems to have the greatest sympathy for it. The second model, in his view, is fundamentally misguided, and leads to a circular dilemma. Since the third model is based on some kind of metaphor, even if it is a “deep-rooted and firmly planted” one, it is highly questionable. As a result, Tugendhat holds that all extant theories of self-consciousness are hopelessly muddled, and that we require a new path if we are to speak reasonably about the phenomenon of self-consciousness.

In order to avoid the circular dilemma of the traditional theory, Tugendhat gives up the model of “mental vision” in which the method of grasping oneself is through an inner perception, and with it he also gives up the model of subject-object as the form of self-consciousness in the sense of epistemology (to be discussed below). Instead, he chooses the method of linguistic analysis, in which we analyze and examine how we use the expression “to know oneself.” According to Tugendhat, only such a method can avoid the impasses of the Heidelberg School. Tugendhat summarizes the expression that describes the phenomenon of self-consciousness as “I know that I φ.” Thus, self-knowledge does not have the form of a self-relation – a relation between the subject and itself – nor does it mean that the knower and what is known are identical. Rather, the statement “to know of oneself” means that the subject knows a state of affairs, a conscious state or experience (i.e. a particular state that “I” have), which is embodied in a proposition (“I φ”) (Tugendhat 1969, pp. 56f). The key point here is that the first letter of the indexical “ich” should not be capitalized and nominalized; it cannot take the form of an object. The object of knowing is always the proposition (“I φ”). For Tugendhat, the semantic explanation “I know that I φ” is the only way to avoid the circular dilemma of “(the subject) I ‘know’ (the object) I” in the theory of self-consciousness.

At the same time, Tugendhat criticizes the Heidelberg School when it attempts to avoid the dilemma of the traditional theory by fundamentally writing off the general phenomenon of self-consciousness. Tugendhat argues that the “self-acquaintance” developed by the Heidelberg School inappropriately restricts the problem of self-consciousness to the level of pre-reflection. For the Heidelberg School, the concept of self-consciousness is so narrow that it effectively abandons the whole “objectified” dimension in talk of “Selbst” or “Ich” in the traditional theory. In contrast, Tugendhat places both the objectified and the propositional dimension under “self-knowing” (Selbsterkenntnis). Nevertheless, we must ask: Is the reflection-model that appeals to objectification entirely without merit? And does Tugendhat’s semantic approach really avoid the circular dilemma of the traditional reflection model? This will lead us to Husserl’s phenomenology.

3 Is Husserl’s “Inner Consciousness” a “Presentation” (Vorstellung)?

The Heidelberg School’s critique of the traditional theory of self-consciousness aims mainly at the model of reflection, while Tugendhat’s main problem with the traditional theory lies in its model of “presentation” (Vorstellen). Tugendhat’s understanding of the term “presentation” is broad: on the one hand, presentation can be understood as an intuition, perception, or a way of “seeing,” and this ties the model to the visual metaphor. On the other hand, the word “presentation” itself means to “put (stellen) before (vor-)”, and thus involves a subject-object relation. Thus Tugendhat’s critique of the presentation model entails a critique of these two previously discussed models.

Tugendhat summarizes the traditional theory of self-consciousness with the basic schema: “z represents x [z stellt x vor].” Additionally, in his view all traditional theories of self-consciousness can be distinguished according to “(1) whether or not x was regarded as identical with z, and (2) what x and z were regarded as designating” (1979, p. 51).

In the way the traditional theory of self-consciousness approaches question (2), x is regarded either as “the ego,”, or as the “φ” state (a state of consciousness), and so is z. Now let’s focus on the first question(1): whether x is to be regarded as identical with z in the schema. At first, we assume that x and z are not regarded as identical, that is, x ≠ z. According to what was mentioned above, x is either “the ego” or the “φ” state, and z is either “the ego” or the “φ” state. And since the “ego” can never be “represented” without positing it, it is absolutely impossible that the “φ” state as a subject (z) “represents” an “ego” (x) if we assume that x ≠ z. Thus, there are only two possibilities: 1) x is the “φ” state and z is “the ego,” and therefore “ego represents the state of consciousness”; or 2) x and z are regarded as two different kinds of “φ” state, that is, “one conscious state represents another.” Tugendhat found that, historically, Kant holds the former view, while Husserl holds the latter. Since Tugendhat’s criticism of Husserl is our focus here, we will return to this point later.

If, alternatively, we assume that x is identical to z, i.e. x = z, there are also two possibilities. Both x and z are regarded either as “the ego” or as the “φ” state, that is; (1) “the ego represents the ego” and (2) “the conscious state represents the conscious state.” Tugendhat took Fichte as the most important representative of the former possibility and Brentano as the representative of the latter. The Heidelberg School, represented by Henrich, also rejected the model of “x ≠ z” at the very outset, and worked out a theory based on the presupposition that x = z.

We cannot further develop Tugendhat’s exposition here, but we can briefly articulate the tactics of argumentation that he employs. First of all, he agrees with Henrich and the Heidelberg School in their rejection of the model of x ≠ z. Next, he concentrates his firepower on Henrich’s theory which presupposes x = z. It turns out that the various theories of self-consciousness mentioned above inevitably involve the circular dilemma, if the alternative x = z follows the schema “z represents x.” It is pointless to ask whether or not x is identical with z, as Henrich and the Heidelberg School did, so long as x and z are cut into “I” and “φ” from the whole “I φ” on the one hand, and the predicate connecting x and z as “presentation,” on the other. Tugendhat thus approaches the issue first by reuniting the cut “I” and “φ” into “I φ” again, and then radically rejecting the idea that “presentation” can be conceived as a predicate. His alternative schema is expressed as “I know that I φ”.

Let us now turn to Tugendhat’s critique of Husserl. He thought that Husserl’s phenomenological theory of self-consciousness (mainly in the first edition of the Logical Investigations) belongs to the second situation of the model of x ≠ z, that is, “one conscious state represents another.” In Tugendhat’s view, Husserl elaborated a “doctrine of inner perception […] in the extreme, uncontaminated form.” Both x and z are experiences, but different ones. Every experience can be represented only by a second “simultaneous” experience, an “act of inner perception.” Nevertheless, he argues, “the weakness of this conception is readily evident. According to it, an experience a represents another experience b; but in what sense can this experience be identified as self-consciousness?” (Tugendhat 1979, pp. 52f/1989, pp. 41–42).

The key to the problem lies in two questions: Is “inner perception” in Husserl a kind of “presentation” in the sense of a subject-object relation? And does Husserl presuppose that x ≠ z, i.e., that there are two different experiences here?

Tugendhat’s interpretation relies on the second of the three concepts of consciousness discussed in the Fifth Logical Investigation, where consciousness “is that ‘inner perception’ thought to accompany actually present experiences, whether in general, or in certain classes of cases, and to relate to them as its objects” (Hua XIX/1, A 333/B1 354; 2001, p. 86). Because of the relative expression of the Husserlian concept of consciousness, his definition of this concept is often misunderstood. Tugendhat is no exception.

The difficulty here lies in two expressions: one is to “accompany actually present experiences,” and the other is “to relate to them as its objects.” According to Husserl, there are two types of “beings that are experienced” in the stream of consciousness (or the stream of experiences) after the phenomenological reduction: “the abstract aspects of experience” and “the experience” itself. Husserl writes: “with the experiences in their total and concrete fullness, their component parts and abstract aspects are also experienced: they are real contents of consciousness” (Hua XIX/1, A 326/B1 348; 2001, vol. 2, p. 82). For example, in perception we perceive the computer in front of us. According to Husserl, the act of perceiving the computer is experienced by us (is an “experience”) so long as the act of perceiving the computer itself happens in our consciousness. The mode of givenness of perceiving the computer is fundamentally different from the mode of givenness of the computer itself as an object: the former is “experienced” and the latter is “perceived.” Meanwhile, the way of being disclosed in the act of perception is the “being-experienced”: the “component parts and abstract aspects” that compose the act of perception are experienced together, such as, in this example, the material shape and the sense data I receive from the computer. A new aspect appears here – “the abstract and manifold sense data” that contribute to the act of perception – and these are “experienced,” too. Thus Husserl’s previously cited definition of consciousness in the second sense means the accompanying grasp of all the actual present experiences in the stream of experiences, including both the grasp of “the abstract aspects of experiences” (for example, the sense-data coming from the computer) and the grasp of “the experiences” themselves (for example, the perception of the computer), and this implies “an inner perception” that “takes these experiences as its objects.” On the basis of these statements, then, we can both: (1) regard our perception of the computer as an object and “internally perceive” this as an act of perception, and, (2) simultaneously take the sense-data of the computer as another object and “internally perceive” such sense data.

Aspect (1) unquestionably belongs to Husserl’s phenomenology: the objectified “inner perception” just means “reflection,” in which our sight turns from being simply directed to the computer to the perceiving of it. It is also a reflection or “inner perception” upon the perception’s content: “It is evident that this possibility belongs to the essence of the perception. But reflection is, nevertheless, perception that is related to the perception and its content” (Hua XXIV, p. 244; 2008, p. 241). That is to say, we can transform our sight from a simple perception into a “new reflective consciousness” at will, but this new reflective consciousness as “inner perception” is also an experience in the stream of experiences, which can be internally perceived as an object again, and so on. One may worry about a possible infinite regress, and we will later consider whether or not this is really troubling.

Such an account of “inner perception” as reflection, however, seems to bring a significant difficulty in regard to (2), where “the abstract aspects of experiences” are involved, since the latter cannot be the sort of reflection we have just described. At the same time, the difficulty reveals the possibility of another way of understanding (1), which would be just the second concept of consciousness that Husserl intended.

In fact, Husserl made a dichotomy on two levels in the second chapter of the Fifth Logical Investigation. The first is the dichotomy of “intentional experience” and “non-intentional experience.” The word “intentional” is used here in a wide sense, meaning that an experience is “intentional,” i.e., “has objects,” as long as it is an act, while “non-intentional” here designates the factors or contents (e.g., sense data) of an act. Secondly, “intentional experiences” can be divided into a narrower sense (the objectifying acts) and a wider sense (all the acts of consciousness including the non-objectifying acts).Footnote 7

Returning to the question at hand, if “the non-intentional experiences” or “the abstract aspects of experiences” designate the factors or reell contents of an act, they can only “be experienced” but never have “the quality of something perceived,” and thus never “have the quality of givenness of an object of perception.” They are “the being of ‘mere’ or pre-phenomenal experience. All of naive perception and all components in it (like the material of sensation, the attending, the apperceiving) are experienced, i.e., merely experienced” (Hua XXIV, p. 244; 2008, p. 242). Apparently, these abstract aspects of experiences are only “pre-phenomenal” experiences. They are simply experienced, “but ‘experiencing’ does not then mean having-an-objectivity and ‘referring’ to the objectivity in one way or another, and taking a position toward it in one way or another, etc., but it means the unity of all phenomenological findings and possibly what has been found in connection with phenomenological time” (Hua XXIV, p. 247; 2008, p. 244). Therefore, pre-objective and non-objective being are both a part of this experience. The being of the pre-phenomenal experience is, itself, “experienced,” and it is a “mere,” “non-objective” “being-experienced” accompanying the experience itself.

In the Logical Investigations (1901), Husserl employed the easily misunderstood expression “inner perception,” but he also employed a less easily misunderstood expression, “inner consciousness” (inneres Bewusstsein), which designates this being-experienced. In the Introduction to Logic and Theory of Knowledge (1906/07) and The Phenomenology of Inner Time Consciousness it is more appropriately called “primal consciousness” (Urbewusstsein).Footnote 8 Thus, Husserl’s second concept of consciousness essentially means a primal consciousness that accompanies all experiences as such, as a non-objectifying grasp of them. It is phenomenologically one and the same as the accompanying, pre-reflective, non-objective self-acquaintance described by Henrich and the Heidelberg School.

This primal consciousness is therefore not only the way of being-given of the abstract aspects of experiences that cannot be objectified in themselves, but also the way of being-given of those experiences (e.g., perception) themselves. This means that sense data are “experienced” (one is “primally conscious” of them), while perception can be both “experienced” (in “primal consciousness-of”) and also objectified in reflection.

Thus, what Husserl means in the Logical Investigations by concepts such as inner perception, object, and so on, is relatively clear. In the flowing of the stream of consciousness, we can accompanyingly, non-objectifyingly and pre-reflectively be primally conscious of or internally conscious of the proceeding of consciousness or experiences themselves. Therefore, such a primal consciousness or inner consciousness is equivalent to what we have been calling “self-consciousness” in a rigorous sense (in contrast to “self-knowledge”). Here “self” means consciousness or experience itself, and “self-consciousness” is the consciousness and experience which is non-objectifyingly and accompanyingly conscious of it in its streaming. Such a primal consciousness or inner consciousness is fundamentally not an objective perception or “presentation,” so the first point in Tugendhat’s critique of Husserl is not tenable. Husserl’s “inner perception” is not itself the sort of “presentation” which Tugendhat attempted to replace by a “non-presentational” linguistic analysis.

For Husserl there is not first a primary act of consciousness (experience b) and then a new, secondary act of consciousness (experience a) objectively presenting b. There is only one stream of consciousness, in the process of which the primal consciousness or inner consciousness of its streaming is an indivisible element of the stream itself. And it is this accompanying and indivisible element that make a subsequent objective and reflective “experience b” possible. Following Iso Kern, who borrowed the metaphor from Yogācāra Buddhism, the relationship of the process of consciousness and its accompanying primal consciousness is not like the knife (a knife can cut everything but itself), but like the lamp (a lamp illuminates everything including itself) (Kern 1988, pp. 51ff). Hence Husserl does not adopt the model of x ≠ z, as Tugendhat maintained, but he does indeed discuss self-consciousness in the model of x = z, as do Henrich and the Heidelberg School. In contrast to Tugendhat’s stress on thematic and propositional self-knowledge, what Husserl has in view is “non-egological” pre-reflective and non-thematic self-consciousness or “primal consciousness.” Through this analysis, then, Husserl does approach Henrich’s concept of self-acquaintance.

4 The Model of Reflection in Husserl’s Theory of Self-Consciousness

However, unlike Henrich and the Heidelberg School, Husserl did not confine himself to the dimension of pre-reflection; in fact, the dimension of reflection is the most important thing in his phenomenology. Thus Tugendhat’s critique of the Heidelberg School does not apply to Husserl, since Husserl did not abandon reflection, nor did he dismiss or restrict the theme of the traditional theory of self-consciousness. In 1913 Husserl again took up the question of the “pure ego.” In doing so, he not only further developed the meaning of the “personal ego” or “habitual ego,” but also ascribed both to an “abstract ego,” after which he introduced the transcendental and concrete ego as a monad, eventually designating his phenomenology as a “transcendental egology.” How then did Husserl understand the givenness of the “I”? Our major concern here will be the role of the so-called “abstract I.”

Husserl’s concept of reflection appears to be the one that, according to Tugendhat and the Heidelberg School, leads to the circular dilemma. We agree that Husserl developed the model of reflection in his phenomenological theory of self-consciousness: both pure ego and personal ego are the objects of a reflective self-apperception (see Hua IV, pp. 247f.). Husserl writes, “It pertains in general to the essence of every cogito that a new cogito of the kind called by us ‘Ego-reflection’ is in principle possible, one that grasps, on the basis of the earlier cogito (which itself is thereby phenomenologically altered), the pure subject of that earlier cogito” (Hua IV, p. 101/1989, p. 107). Clearly, what Husserl here calls “Ego-reflection” is not a self-othering cogito in the streaming of consciousness but something new, subsequent and secondary. It is in this Ego-reflection that the pure ego is originally given as object. Neither in the patent conscious experience in which the ego actually lives, nor in the background consciousness in which the ego non-actually lives, is the ego there as an object; it is known only in reflection, which alone makes the field of pure consciousness into the object of phenomenological study.

The new type of cogito here not only means “self-perception” but also includes “the possibility of corresponding modifications of self-grasping: i.e., self-remembrance, self-phantasy, etc.” (Hua IV, p. 101/1989, p. 107). Such a secondary cogito or Ego-reflection is an objectified consciousness in which the pure ego itself is given. Naturally, the personal ego is grasped in this sort of reflection as well.

Now, the question is what the phenomenologically essential structure of the objectified grasping of such Ego-reflection is. Will it still be stuck in the circular dilemma of the traditional reflection-theory of self-consciousness? The question can be answered from two sides: first, the relationship of the objectifying Ego-reflection to primal consciousness, and second, the structure of Ego-reflection itself.

As mentioned above, the model of reflection in the traditional theory of self-consciousness holds that “a subject-I ‘reflects’ an object-I”, and the crux of the circular dilemma is that one needs to know how the reflecting “subject-I” grasps herself in order to avoid an infinite regress. According to Husserl’s phenomenological theory of self-consciousness, just like that of Henrich and the Heidelberg School, a pre-reflective and non-objectifying primal consciousness accompanies the flow of the stream of consciousness. It is this primal consciousness or self-consciousness that makes possible an “ego-polarizing of the stream of consciousness,” since the ego here primarily means a “dative I” (mir),Footnote 9 an “ego pole” in the ego cogito that always “is,” an attribution that can refer not only to a formal and empty ego pole but also to an individual and personal habitual pole. It is in primal consciousness that “the living consciousness” is non-objectifyingly acquainted with itself; primal consciousness forms the basis on which all kinds of reflection, or “the variants of self-grasping,” are subsequently possible. “Self-perception is a reflection (self-reflection of the pure Ego) and presupposes according to its essence a non-reflected consciousness” (Hua IV, p. 248/1989, p. 250). In other words, for Husserl, primal consciousness lays a foundation for the subsequent and secondary Ego-reflection, and so the dilemma of the model of reflection in the traditional theory of self-consciousness is avoided. Reflection is not originary and does not characterize all self-consciousness. Thus Husserl’s phenomenology of the essential structure of Ego-reflection provides the basis for inquiring into the nature of the ego, thereby avoiding Tugendhat’s criticism of the Heiedelberg School for writing off this important dimension found in the traditional theory of self-consciousness.

But what is the phenomenological structure of this ego-reflection? What is the object of the reflection? Is it a lone “I,” a lone “φ,” or a whole “I φ”? In the lecture First Philosophy from the early 1920s, Husserl gave a definite answer: through the phenomenological reduction, “I get my ‘I perceive’ as the experience just as it really and purely is or was in itself. I get it as a piece of my transcendental subjectivity. As transcendental subjectivity, it is no more than ego and ego-life as it is in and for itself” (Hua VIII, p. 87). Here, Husserl explains that with the aid of a phenomenological epoché—that is, by bracketing the world’s existence in order to focus on the experiences that this ego has of the world in its own life—what we achieve is the experience of “I perceive,” in which the world is contained as a phenomenon.

We can illustrate Husserl’s meaning with the help of a concrete example. Let’s suppose that when I approach the liberal arts building on campus, I simply and directly observe or perceive it. Here, I perceive in a manner of “self-lostness” (Selbstverlorenheit) or being “self-forgotten” (Selbstvergessen); that is, I devote myself totally to watching and observing the building, while the ego is not highlighted but is lost. “Lost” here means that the ego is not present thematically, but in fact in the intentional act of perception the ego is always actually living. In our example, the essential structure of the act of consciousness becomes a three-sided structure: ego-cogito-cogitatum.Footnote 10 It consists of an ego (I) that is always “actually” living there but is self-lost, a concretely proceeding cogito (perception), and a cogitatum related to the cogito (the liberal arts building). In this regard, the phenomenologically essential structure of a present act of consciousness is captured by the “subject-object-predicate” in grammar (See Hua VIII, S. 88).

Now, the act of perception that focuses on the liberal arts building flows away in the mode of being self-lost, while I begin to try to “turn back,” to reflect on the act I have just accomplished. What is the object of this act of reflection? It is not the building (the object of cogito), and not only the act of perception (cogito), let alone the self-lost I; rather, it is the three-sided structure of ego-cogito-cogitatum just passing that becomes the object of the act of reflection. Here we find that Husserl’s phenomenological approach is remarkably consistent with Tugendhat’s semantic approach. It would seem that the three-sided structure here—the “subject-object-predicate” in grammar—is equivalent to the total and indivisible “I φ” stressed by Tugendhat. If this is so, then (contrary to what Tugendhat argues) not every philosophy that approaches the problem of self-consciousness through consciousness, and not every “presentation” model of self-consciousness, will entail the division of “subject I” and “object I” in the analysis of “I φ.” Husserl’s phenomenological philosophy of consciousness insists on grasping the complete object of reflection. In this sense, we can say that the object of phenomenological reflection is “I φ,” on the condition that it is primarily a pre-linguistic total “I φ”.Footnote 11

In phenomenological reflection, I as a reflecting ego execute a reflective “grabbing-back” (zurückgreifen) upon the simple act of perception that has already passed, and the ego, which was self-forgotten, is “found” again. That is to say, in this reflection, in this “post-finding” (Nachgewahren), the reflecting I “[has] my appearance once again in the act of ‘I perceive.’ In that appearance, I turn the self-forgotten ego and previously non-perceived ‘I-perceive-the-house’ into the perceived content which is directed toward through my grasping” (Hua VIII, S. 88). It is the reflection that makes the previous conscious act as such (my perception of the liberal arts building) into an object. The three-sided structure of the conscious act is not itself perceived by the accompanying pre-reflective and non-objectifying primal consciousness-of the experience of the conscious act. But reflection does not simply take the object of the conscious act (the building) as its object. It is by this reflection that the self-lost ego in the previous conscious act can appear once again in the three-sided structure; it is due to reflection that the self-lost ego is activated and becomes something that can be grasped. So-called “self-knowledge” or “ego-reflection” gains knowledge of the ego only in the reflective grasp of the total three-sided structure. How could there ever be a reflection or knowledge of a lone “(object) I”?

It is clear that the model of reflection propounded by a phenomenological theory of self-consciousness grasps and knows the “I” by a reflective “grabbing-back” of the pre-linguistic whole “I φ”. The model of reflection is linguistically expressed as “I reflect that I φ.” If, in Tugendhat, the theory of the “self-consciousness of knowledge” stresses, on the one hand, the knowledge of the state of affairs as different from the knowledge of objects; and, on the other hand, in Husserl, the expression of the state of affairs is different from the presentation of objects, then the reflection-model of the phenomenological theory of self-consciousness also stresses the knowledge of the state of affairs and the reflection of the state of affairs, but not the knowledge of objects and presentation of objects.

However, in sharp contrast to Tugendhat’s semantic approach through an analysis of how the expression “I know that I φ” is used, the phenomenological approach appeals to intuition, evidence and self-givenness. Perhaps we can apply a sentence that Husserl once uttered to stress the “incomparable function” of phenomenology: discussion of the phenomenon of self-consciousness can only be concretely accomplished by “a kind of research that draws intuitively upon what is given phenomenologically but not by thinking of the kind that plays out value concepts, a game played with constructions far removed from intuition.”Footnote 12

5 Is the Reflection-Model of Self-Consciousness Circular?

In general, although both the Heidelberg School and Tugendhat discussed the issue under the name of “self-consciousness,” they understand the two parts of this concept (“self” and “consciousness”) differently. The only commonality is their point of departure, that is, the rejection of the understanding of “self” as “ego.” The subsequent developments of both sides rarely intersect. In Henrich, “self” means “the living conscious life” itself, and “consciousness” means the pre-reflective and pre-thematic “acquaintance” with this self; in Tugendhat, however, “self” means “I φ” (that is, my such and such state that can be propositionalized), and “consciousness” means a propositionalized cognition or knowledge.

The Heidelberg School and Tugendhat provide different models for avoiding the circular dilemma of the traditional theory of self-consciousness, but in fact their models are not as opposed as they might at first seem; rather, they are complementary. In the exploration of self-consciousness (in the broad sense), we not only need to escape from—or resolve—the circular dilemma; we also need to stress the dimension of reflection, as Tugendhat does in his semantic approach. Tugendhat firmly rejects the method of intuition and insists on the method of linguistic analysis. For him, only a semantic explanation can avoid the circle. But, a representative of the Heidelberg School might ask, how can propositionalized self-knowledge (“I know that I φ”) avoid the circular dilemma if it rejects a priori the possibility of a pre-reflective self-consciousness? Further, what is the relationship of an intentional consciousness to a propositional expression when the semantic schema fundamentally presupposes that all intentional consciousness is propositional?Footnote 13 We shall concentrate on the first of these issues.

As we have seen, Husserl not only supplies a set of phenomenological descriptions of the self and the stream of consciousness but also supplies a phenomenological theory of the ego. The phenomenological theory of self-consciousness in the broad sense comprises, on one hand, both the dimensions of the pre-reflective and the non-objectified self-consciousness (in the narrow sense), or primal consciousness, and, on the other hand, the dimension of objectified and reflective self-knowledge. Thus Husserl’s phenomenology is closer to the philosophical approach of the Heidelberg School than it is to Tugendhat’s semantic approach.

Further, Husserl stressed that the secondary, reflective dimension must always be founded on the primary and pre-reflective dimension and that pre-reflective primal consciousness is the foundation for all types of objectified reflection. It is on the basis of the division of the pre-reflective from the reflective dimension, and the grasp of the founding relation between the two, that the reflection-model in Husserl’s phenomenology of self-consciousness can avoid the circular dilemma. However, the model of reflection in Husserl is not a traditional one but lies closer to the theory found in Tugendhat’s semantic model. Through his phenomenological study, Husserl clearly reveals the essential structure of reflection and thereby resolves the basic difficulty of the traditional theory’s reflection-model.

However, if we define the reflection-model of the phenomenological theory of self-consciousness as “I reflect that I φ,” we must still confront the query posed to Tugendhat’s “I know that I φ” by Manfred Frank. This is one question that Tugendhat could not answer through his semantic approach and, in fact, he did not even think he needed to answer: in the expression “I know that I φ” (and “I reflect that I φ”), how is it possible to warrant and even to demonstrate the identity of the indexical “I” that appears in both the main clause “I know” (or “I reflect”) and in the subordinate clause “I φ”?Footnote 14

Let us turn back to the example mentioned above. I devote myself in a self-lost way to the perception of the liberal arts building and “then” I reflect or “grab-back” and “post-find” the three-sided structure comprised of the self-lost I, my act of perception, and its object. In such a formulation, one posits a separation in time between an “I” that has been previously absorbed in perception and an “I” that reflects “then.” How could we say that these two “I’s,” separated by time, are one and the same I?

First, since the separation is not necessary, there is no reason to question I’s identity on the basis of a separation in time. Thus we can say that in order to argue for this identity, one does not need to resolve the difficulty of separation. For example, if I constantly observe the building (that is, if I always alternate between reflection and direct observation), then, in the living present

I possess the dualized ego and dualized ego-act in a co-existent way; that is to say, I possess an ego that constantly observes the building now and the ego executing following act: ‘I am conscious that I constantly observe the building,’ and this act can be formulated in following way: I observe the building (Hua VIII, p. 89).

In other words, we can grasp a “dualized ego” without separation. However, the question concerning identity is now simply re-directed toward this very dualization. Even without the separation, it seems, one must still confront Frank’s question.

Second, in reflection the previously self-lost ego becomes salient, but meanwhile there is also the three-sided structure of consciousness in the present reflection, in which the reflecting I is again self-lost when I concentrate my attention on the reflected object. How do we grasp the “I”? Husserl argued that we need a secondary reflection or one on a higher level. As to the three-sided structure, the object of reflection, we can describe it in the following manner: that I am conscious of “I perceive the house.” Here, cogito becomes “be conscious of,” cogitatum is “I perceive the house,” and ego is still the “I” as the subject. And the description of the secondary reflection will become: I know that I am conscious of “I perceive the house.” And so here too the question about the grasp of the I as the subject in the main clause emerges. Husserl affirmed the possibility of a continual, and even infinite, reflection. Does this revive the problem of infinite regress that beset the reflection-model of the traditional theory of self-consciousness? Will this problem lead the phenomenological theory of self-consciousness to ruin?

Husserl might say that in trying to answer these questions we arrive at a new and different branch in the road. For the infinite iteration of reflection, which Husserl affirmed, is only a possibility. Here phenomenology keeps to its nature as essential science: there always exists the possibility of infinitely iterating a reflective “grabbing-back”. Nevertheless, such an infinite regress is not vicious; it is not necessary but only a possibility attested in free variation in fantasy.Footnote 15 But it is in this very same free fantasy variation that we grasp the essence as what is constant and we intuit what belongs essentially to Ego-reflection. Therefore, the possibility of an infinitely iterated chain of “I”’s provides no reason for claiming the impossibility of the identity of I.

How do these considerations help us to address Frank’s question? Based on phenomenological insights, we see that the stream of our consciousness is always flowing, but all along it remains one and the same. In the dimension of inner time consciousness, the direct perception, the subsequent reflection, and even the infinite reflection in possibility have, indeed, a succession in the sense of inner time, but the ego is not there, the ego does not really lie in the stream of consciousness and, of course, does not live in the succession. The so-called “separation” of ego in time is basically one in the dimension of objective time and has nothing to do with the things themselves on the level of inner time consciousness. At the same time, the ego is always there, and is numerically identical. The experience is variable, while the ego is always one. This is because the ego is primarily a “dative I”: whether in perception or in reflection, even the possible infinitely continuing reflection is given to me or accessed by me in a special way. There is not a multitude of I’s; the I is always one and the same ego in ego-cogito-cogitatum. The problem of identity is essentially a problem of polarization.

I see that I myself can establish myself as an ego reviewing in reflection on a higher level, and that I can realize the identity in multiplicity of all the act-poles and their being-modes decided by conditions in evident synthetic identification. Thus I say, here I am identical everywhere, I am identical as the reflector which is taken as un-reflected in afterwards grasping, it as the perceiver of itself regards me as the perceiver of the house, and so on (Hua VIII, p. 91; my emphasis).

The “I” follows me as a shadow.

This is the surprising fact implied in consciousness. And what the reflection- model of phenomenological self-consciousness attempts to do is just to exhibit this peculiar vision of consciousness. There can be no room here for the circular dilemma.

6 A Brief Conclusion

Henrich and the Heidelberg School, as well as Tugendhat, judged the traditional theory of self-consciousness to be circular and offered different approaches to escape from it.Footnote 16 Each accused the other, however, of failing to avoid it. Undoubtedly, since Fichte the word “circular” is tied closely with the theory of self-consciousness. But is circularity really the inextricable fate of all theories of self-consciousness?

Klaus Düsing, another important voice in the revival of the modern theory of self-consciousness, opened his own study of the theory of self-consciousness with the question, “Is there a circle of self-consciousness?”Footnote 17 On the one hand, Düsing attempted to clarify the misunderstandings of the circle problem by historical analysis of the theory of self-consciousness in philosophers such as Kant, Fichte, Hegel, Husserl, and Heidegger. On the other hand, he tried to show systematically that the circular dilemma is not a weapon by which the problem of self-consciousness is forced into a hopeless situation. He described his own position as a theory of “the models of self-consciousness” (Selbstbewusstseinsmodelle)Footnote 18 and attempted to show that there are so many kinds or models of self-consciousness that we cannot treat or deal with them in one simple way. Two main principles of the theory of models are: (1) that every model of self-consciousness will be shown to turn on method, and (2) that there is a genetic-dynamic construction of self-consciousness in content.

In these terms, it turns out that the different models supplied by the Heidelberg School and by Tugendhat are not completely opposed to each other but are complementary. In the exploration of self-consciousness in the broad sense, neither can avoid the pre-reflective dimension or the reflective dimension. The self-acquaintance concept of the Heidelberg School is concerned with the former dimension, while the semantic approach of Tugendhat pays more attention to the latter.

However, in stressing the latter dimension, Tugendhat firmly rejects any approach through a philosophy of consciousness. According to him, Husserl’s phenomenology cannot avoid the circular dilemma so long as it still employs the intuitive method. Thus Tugendhat argued that Husserl’s phenomenology has been surpassed by analytic philosophy.

But as we have seen, although Tugendhat’s work on the language-analytic explanation of self-consciousness was published in 1979, its discussion and critique of Husserl was based entirely on Husserl’s Logical Investigations of 1901. First Philosophy (1959) and The Phenomenology of Inner Time Consciousness (1966) were completely ignored. It was in Husserl’s later works that the phenomenological model of self-consciousness was developed in a more systematic way. Husserl’s phenomenology shows us how the “I” is pre-reflectively aware of “me,” how through a reflexive sight, I can grasp “me,” as well as what the “reflecting I” means, how it gains the unity of itself, and so on. Doesn’t Husserl’s phenomenological approach already demonstrate a combination of the two positions maintained by the Heidelberg School and by Tugendhat in its combination of “primal consciousness” with “I reflect that I φ”? It is precisely in determining the founding relationship between primal consciousness and reflection that the circle of the traditional theory is avoided.

The contribution of Tugendhat’s semantic approach to the modern theory of self-consciousness is remarkable, but it is one-sided thanks to his failure to recognize the philosophical novelty of Husserl’s systematically phenomenological approach as it began to appear in the 1960s. In a certain sense, then, the work of this article can be regarded both as a phenomenological response to Tugendhat’s semantic approach of self-consciousness and as a language-analytic proposal that draws on Husserl’s more complete position.