|M.C. Escher, Self-Portrait in a Spherical Mirror|
The easiest way to understand prereflective self-consciousness is by contrast with reflective self-consciousness, which is self-consciousness in the mode of ordinary object-consciousness. In reflective self-consciousness, the subject is aware of itself in much the same way it is aware of other objects in the world. The claim that object-consciousness suffices to explain self-consciousness is known as the “reflection model of self-consciousness”: it basically sees self-consciousness as resulting from a turning around (re-flection) of object-consciousness away from external objects and unto the subject itself. Despite its prominence in Western philosophy, notably in early modern philosophers from Descartes to Kant, the reflection model has come under increasing attack in philosophy since Kant. It has become increasingly clear that the reflection model suffers from a number of paradoxes, infinite regresses and vicious circles. To explain self-consciousness, then, reflection does not suffice: we must postulate a sui generis form of self-consciousness, different in kind from reflective object-consciousness. The adjective “prereflective” indicates this special type of self-consciousness.
Below I will discuss the paradoxes of the reflection model in more detail. For now, a few examples will suffice. To repeat: the reflection model explains self-consciousness as the redirection of object-consciousness away from external objects and unto the subject itself. But how does the subject know that its new object, of which it thus becomes aware, is indeed itself and not another external object? The difficulty is nicely illustrated by the example of Ernst Mach who, sitting in a Vienna bus, noticed “a shabby-looking school teacher” (“ein herabgekommener Schulmeister”) sitting across from him… until he realized he was looking in a mirror (Mach 1922: 3, n.1). The lesson is that mere object-consciousness, if it is accidentally turned towards the subject, does not intimate that the object one is aware of is indeed oneself – to achieve that self-awareness, a further act of the mind is required, a mental act irreducible to object-consciousness. Thus, the reflection model fails to explain self-consciousness.
The model can evade this difficulty only by claiming that the turning around of object-consciousness towards the subject happens by no means accidentally but with the intention to get the subject in view: the subject intends to know itself and therefore turns its object-consciousness towards itself. This solves the problem of failed self-recognition (as in the Mach example), since the object is intended as oneself from the start, but only at the price of circularity. For how can the subject intend to know itself by means of object-consciousness if it isn’t already aware of itself to some extent? If the subject were completely oblivious of itself, it could not even intend to know itself. As the analytical philosopher Sydney Shoemaker notes:
“[I]f one were aware of oneself as an object in such cases (as one is in fact aware of oneself as an object when one sees oneself in a mirror), this would not help to explain one’s self-knowledge. For awareness that the presented object was φ, would not tell one that one was oneself φ, unless one had identified the object as oneself; and one could not do this unless one already had some self-knowledge, namely the knowledge that one is the unique possessor of whatever set of properties of the presented object one took to show it to be oneself.” (Shoemaker 1984: 105)
The reflection model, then, can explain self-consciousness only by presupposing self-consciousness. Thus, the model either fails or is guilty of circularity. Of course, it is not to be denied that reflective self-consciousness is in fact possible: I can, and occasionally do, observe and think about myself as one object among the other objects that populate the world. The point is, however, that this reflective self-consciousness is facilitated by a pre-existing – and therefore pre-reflective – self-consciousness, in a mode different from object-consciousness. As Dan Zahavi notes: “[W]hen one does in fact succeed in taking oneself as an object, one is dealing with a self-objectification which in its turn presupposes a prior nonobjectifying self-awareness as its condition of possibility.” (Zahavi 1999: 6-7)
The self-registration view of consciousness
The primary motivation behind the notion of prereflective self-consciousness may be the correct understanding of self-consciousness as such, but it certainly is not the only motivation. The notion of prereflective self-consciousness is central to philosophy of mind in general because self-consciousness is taken to be crucial for consciousness as such. That is, even conscious states such as thinking about and perceiving an external object, say, a tree, although they are ostensibly not about the thinking and perceiving subject, nevertheless seem to presuppose self-consciousness. This claim, that all consciousness presupposes self-consciousness, and thus that self-consciousness is ubiquitous in all conscious states, is known as the Ubiquity Thesis (the term was coined by Kapitan 1999; following common usage, I will refer to this thesis as “Ubiquity”). If Ubiquity is correct, and if reflective self-consciousness presupposes prereflective self-consciousness, then the latter must be central to our understanding of consciousness in general. A closer look at Ubiquity supplies us with further evidence of the paradoxes ailing the reflection model and hence the need for a notion of prereflective self-consciousness.
Ubiquity is motivated by a particular view of consciousness which has been and still is fairly dominant in Western philosophy and cognitive science. We can call it the “self-registration view”. On this view, which has been elaborated in many different ways, consciousness is due to a special “internal monitoring” (Lycan 1997) or “self-registration mechanism” (Frank 2015) enabling the mind to register its own processes. On this view, then, a perception of an external object is conscious because not just the object is registered by the mind but also the perception itself. Likewise, a thought is conscious because not just the propositional content of the thought is registered but also the thought itself. Mental process that are not thus registered by the mind remain unconscious. Since self-registration of mental processes by the mind amounts to a form of self-consciousness, we can summarize this view by saying that self-consciousness underlies consciousness (= Ubiquity). In other words: a mental process becomes conscious because the mind is self-conscious with respect to that process, i.e. it is conscious of its own mental process, which thereby is a conscious process. As said, this view of consciousness has been and still is fairly dominant in philosophy and cognitive science. It can be traced back to Aristotle, who argues in different places of his work that mental processes are conscious because they have, besides their external objects, also themselves as objects (De Anima III, 2,425b, 12; Metaphysics Δ, 9). As Kenneth Williford notes: “Its distinguished history, prominence in careful descriptions of consciousness, and visible if disputed place in the philosophy of mind, AI, and neuroscience lend the claim substantial prima facie credibility.” (Williford 2006: 111)
Problems for the higher-order theory of consciousness
So how does the self-registration view of consciousness provide further evidence for the paradoxes of the reflection model and the subsequent need for a notion of prereflective self-consciousness? The point is that the self-registration view remains problematic as long as we operate within the reflection model of self-consciousness. On the reflection model, the mind’s awareness of a mental state, which lifts the latter into consciousness, is conceptualised as an additional mental state, separate from the first. Mental states are primarily aimed at external objects, and as such they are unconscious. They become conscious only insofar as the mind turns its attention away from those external objects and unto those mental states themselves. A mental state, then, is lifted into consciousness by an additional mental act of reflection. On closer inspection, however, this leads to several problems.
It leads, first of all, to the same problem of self-recognition that we first encountered in the Mach example: if one becomes aware of oneself as an object, how does one know that this object is oneself? One can recognize the object as oneself only if one already has self-awareness to some extent. Or, in terms of the self-registration view, how does the mind know that the mental state, of which it becomes aware through an additional reflection, is indeed its own mental state? Clearly, this already presupposes at least some minimal form of self-awareness, which must therefore be prereflective and in a mode different from object-consciousness. Secondly, the reflection model can be seen to lead to a vicious regress in the context of the self-registration view of consciousness. If a mental state becomes conscious only by becoming the object a further mental state, what then ensures that this second state is also conscious? On the reflection model, a third act would be required to lift the second act into consciousness, and a fourth act to lift the third into consciousness, and so on. It seems, then, that the self-registration view, when married to the reflection model of self-consciousness, can ‘explain’ consciousness only by accepting an infinite regress of higher-order mental states – which means, of course, that it cannot explain consciousness at all.
As said, however, most philosophers find this solution to the regress problem questionable. It seems paradoxical to explain consciousness in terms of unconscious mental states. One objection that is often raised against the HOR explanation of consciousness in terms of unconscious mental states is that it violates Ubiquity. This thesis, after all, states that consciousness presupposes self-consciousness. But how can the unconscious registration of a mental state by a higher-order state be classified as self-consciousness? True, it is a form of self-registration, insofar as the mind registers its own mental states by means of higher-order states. But insofar as this self-registration remains unconscious, it is questionable whether it amounts to self-consciousness. The phrase “unconscious self-consciousness” is, after all, a clear contradiction in terms. Insofar as HOR theories aim to explicate Ubiquity, then, they seem to fail. As Williford writes: “Classic higher-order representation (HOR) theories do not really do justice to the phenomenology behind ubiquity… Such theories arguably push the self-representational aspect of consciousness into the unconscious and thus betray the likely original experiential motivation for their theories.” (Williford 2006: 111)
Is consciousness grounded in the unconscious?
One might come to the rescue of HOR theory by making a distinction between strong Ubiquity and weak Ubiquity. Whereas strong Ubiquity states that full-blown self-consciousness is necessary for consciousness, weak Ubiquity states that mere self-registration of mental states by the mind is required, where this self-registration can remain unconscious. There is something to be said for weak Ubiquity, and thus for HOR theory. Weak Ubiquity still conforms to the basic intuition behind the self-registration view of consciousness. Moreover, HOR theorists ask, what is the alternative? The only way to avoid both the regress of higher-order states and the grounding of consciousness in the unconscious is to accept the existence of mental states that are not just aware of other mental states (thereby lifting the latter into consciousness) but also of themselves. Only such mental states, that are aware of themselves, can do without higher-order states, as they lift themselves into consciousness by being self-conscious. But HOR theorists generally find this a paradoxical solution, and thus prefer their own solution of grounding consciousness in unconscious higher-order states, which they find – if not totally unparadoxical – at least less paradoxical. As David Armstrong puts it: “[I]t is impossible that the introspecting and the thing introspected should be one and the same mental state. A mental state cannot be aware of itself, any more than a man can eat himself up.” (Armstrong 1968: 324) I will say more about this issue below.
In the final analysis, however, HOR theory remains unsatisfactory, for two reasons at least. First of all, we do not just want to explain consciousness, we also want to explain self-consciousness. Even if HOR theory succeeds in explaining consciousness in terms of the mind’s self-registration of mental states by higher-order mental states, the fact remains that this self-registration occurs unconsciously and therefore falls short of self-consciousness, since – as noted earlier – “unconscious self-consciousness” is clearly paradoxical. Self-consciousness, then, seems definitely out of the range of HOR theory. It is, moreover, questionable whether HOR theory can even explain consciousness, given the Hard Problem of Consciousness (HPC). The HPC seems to show that reductionism vis-à-vis consciousness is a dead end: it suggests that consciousness cannot be explained in terms of something else, i.e. something without consciousness, e.g. the brain as a purely physical object. But such reduction of consciousness to something else is precisely what HOR theory amounts to, as it explains conscious states in terms of unconscious higher-order states. This should come as no surprise, since HOR theories are often explicitly designed to facilitate a naturalist (i.e. materialist, physicalist) explanation of consciousness (hence the title of Armstrong’s 1968 classic, A Materialist Theory of Mind).
|David Chalmers coined the term|
"Hard Problem of Consciousness"
The unavoidability of prereflective self-consciousness
Let’s take stock. The self-registration view of consciousness explains the latter in terms of self-consciousness: a mental state aimed at an external object is conscious because the mind is not just aware of the external object but also of the state itself. We saw, however, that the reflection model of self-consciousness fails: the reflective turning around of object-consciousness towards the subject cannot lead to self-knowledge, unless this reflection is guided by a prior self-consciousness, which is therefore prereflective and in a mode different from object-consciousness. Prereflective self-consciousness, then, is what we need to explain consciousness as such, in line with the self-registration view.
This also became apparent from the failure of HOR theory, where the reflection model returns in the idea that mental states are lifted into consciousness by additional reflections, i.e. higher-order states. We saw that HOR theory faces the problem of self-recognition: how does the mind know that the mental state, of which it is aware through a higher-order state, is its own mental state? Doesn’t this already presuppose self-awareness? We also saw that HOR theory faces a dilemma: either accept an infinite regress of higher-order states or accept that consciousness is grounded in unconscious higher-order states. Both horns of the dilemma are undesirable. An actual infinity of higher-order states not only violates the phenomenology of consciousness, it is also mysterious how a finite object such as the human brain can contain such infinite complexity. As for the second horn, the grounding of consciousness in the unconsciousness, we noted that this ignores the HPC.
So, to avoid both the regress and the grounding of consciousness in the unconscious, we have to accept the existence of mental states that are not just aware of other mental states (thereby lifting the latter into consciousness) but also of themselves. Only such mental states, that are aware of themselves, can do without higher-order states, as they lift themselves into consciousness, by being self-conscious. This is therefore what prereflective self-consciousness amounts to: a state of consciousness that is immediately aware of itself, unmediated by reflections.
|David Armstrong: "A mental state |
cannot be aware of itself, anymore
than a man can eat himself up."
But how, then, should we respond to the objection, raised by HOR theory, that the notion of a mental state being aware of itself is incoherent? To repeat the earlier quote from David Armstrong: “[I]t is impossible that the introspecting and the thing introspected should be one and the same mental state. A mental state cannot be aware of itself, any more than a man can eat himself up.” (Armstrong 1968: 324) Note, first of all, that this is just a dogmatic assertion, without real argumentation. Also, the comparison of a self-aware mental state with a man eating himself up goes limp. A man who would – per impossibile – eat himself up entirely would not only kill himself; he would disappear altogether. In that sense, eating oneself up is a form of total self-negation. But a self-aware mental state is not self-negating – on the contrary, it is rather self-affirming or even self-producing.
To be conscious of an object, after all, is judgmental in nature, in that (a) one is conscious of the object as existing, such that existence is – at least implicitly – affirmed of the object, and (b) one is aware of the object as having one or more properties, which are therefore also affirmed of the object. For example, when I take a walk in the countryside and I (veridically) see a tree, I see the tree as existing and as green, as leafy, as beautiful, etc. Likewise, then, when a mental state is self-aware, it is aware of itself as existing and as having certain properties (e.g. awareness of itself).
A self-aware mental state, then, is self-affirmative, i.e. the complete opposite of the self-negation inherent in eating oneself up. The latter is clearly paradoxical, but where is the paradox in self-affirmation? Whereas “I don’t exist” is obviously contradictory, “I exist” is a truism. Thus, I see no paradox in speaking of a self-aware mental state… unless, perhaps, one interprets the self-affirmation inherent in self-awareness in a strong ontological fashion as self-production, as Fichte notoriously did. But it is clear that the Fichtean concept of “self-positing” is not per se needed to understand the self-affirmation inherent in self-awareness. I will return to this issue at the end of this post.
Subject-object difference vs. subject-object identity
Is, then, Armstrong’s criticism of the notion of a self-aware mental state completely unfounded? No, but whatever plausibility it has at the same time makes clear why the reflection model of self-consciousness is inherently wrong. Let me explain. The intuitive plausibility of Armstrong’s criticism derives from the common idea that some kind of subject-object difference is intrinsic to all consciousness, such that the conscious subject is always different from the object of which it is conscious. Hence Armstrong’s bald statement that “it is impossible that the introspecting and the thing introspected should be one and the same mental state”. But – and this is what Armstrong overlooks – it is precisely this idea that underlies the inadequacy of the reflection model of self-consciousness. In fact, we can use the idea of subject-object difference to clarify what object-consciousness really is – a concept we haven’t properly defined yet. Object-consciousness, we can say, is intentional consciousness and is as such inherently wedded to subject-object difference. In intentional consciousness, the subject is invariably aware of an object as different from itself.
Self-consciousness, however, is essentially characterized by subject-object identity. In self-consciousness, the subject is its own object; thus, subject and object coincide, they are numerically identical. Hence the inadequacy of the reflection model. Object-consciousness and self-consciousness pull in different directions: the first pulls towards subject-object difference, the second towards subject-object identity. The reflection model has to bring about an identity by means of conceptual tools that imply difference – an obvious impossibility. Hence the many paradoxes ailing the model. It constantly has to undo or supress the difference which its concepts equally constantly generate. Already on this abstract level, then, we see that the reflection model is in principle incapable of explaining self-consciousness: the aspect of subject-object identity keeps eluding the difference engendering conceptuality of reflection – like the tail eluding the self-chasing dog.
The non-intentional nature of prereflective self-consciousness
That the subject-object distinction is indeed the root of all trouble for the reflection model becomes clearer when we take a closer look at the phenomenological concept of intentionality. Intrinsic to that concept is the idea that intentional consciousness is inherently “thetic” or “positional”, such that consciousness essentially purports to be about an independent object, i.e. an object existing independently from the consciousness aimed at it. This, of course, harks back to what I said earlier about the affirmative nature of consciousness, albeit that the phenomenological view of the positional nature of consciousness is stronger. On the phenomenological view, consciousness not just affirms the existence its object, it affirms that existence as independent from itself. Thus, intentionality is seen to imply a strong subject-object distinction. Phenomenologists put this by saying that the object is intended by consciousness as transcending consciousness. Husserl referred to this positing of objects as transcending our consciousness of them as “the natural attitude”. As Sartre (1972: xxvii) put it: “All consciousness is positional in that it transcends itself in order to reach an object.” It should be noted that such a concept of consciousness as ‘intending beyond itself’ is by no means unique to phenomenologists; many analytic philosophers held similar ideas, notably (and influentially) Moore with his notion of the diaphanous nature of consciousness as an argument for realism.
The point is that the failure of the reflection model becomes all the more obvious if we understand object-consciousness in this strong sense as intending its object as existing independently. If self-consciousness were “self-transcending” in that sense, it would have to posit its object, a mysterious entity called “the self”, as existing independently. But then, immediately, a new regress would arise. For since self-consciousness is obviously a property of this self, self-consciousness would have to posit the self as independently being self-conscious. That is: self-consciousness would then have to presuppose a prior self-consciousness on the part of its object, the self. And this prior self-consciousness, since it too would posit its object as existing independently, would also have to presuppose an already self-conscious self as its object, and so on indefinitely (cf. Sartre 1972: xxvi-xxix; Frank 1991: 226). Again, then, we see that the reflection model leads to a regress. Hence the conclusion, explicitly drawn by Sartre in particular, that prereflective self-consciousness is non-intentional, i.e. not committed to a strong subject-object distinction. Rather, in prereflective self-consciousness, the subject is aware of itself as strictly identical with itself. Or in terms of mental states, prereflective self-consciousness is a mental state that is aware of itself as itself, not as something different.
Final considerations: Prereflective self-consciousness and Idealist Monism
Earlier we noted that Fichte interprets the self-positing inherent in self-consciousness in a strong ontological fashion as self-creation. We now begin to see the motivation behind that idea. If we cannot see prereflective self-consciousness as aimed at the self as an independently existing object, then the self becomes a function of prereflective self-consciousness, i.e. the self only exists as the object of this self-awareness. In other words: a self is that particular self only because it is aware of itself as that particular person: Socrates, for example, is Socrates only because he takes himself to be Socrates. As such, the self-creating aspect of prereflective self-consciousness underlies the radical autonomy of the self, as Fichte stressed. Hence his claims to the effect that the self is the prereflective self-consciousness it has of itself and is as such self-creating. As Fichte put it: “What was I, then, before I came to self-consciousness? The natural answer to this question is: I did not exist at all, for I was not an I. The I exists only insofar as it is conscious of itself.” (Fichte 1991: 98) This bootstrapping of the self through self-consciousness Fichte called “self-positing” (“Selbstsetzung”), saying things like: “the self begins by an absolute positing of its own existence” (Fichte 1991: 99). Note, by the way, that Fichte was not the first to draw attention to the self-creating power of self-consciousness. Similar ideas can already be found in Plotinus: see the previous post on this blog.
|Baron von Münchhausen pulling him-|
self from the swamp by his own hair.
Can self-consciousnss do the same?
-Armstrong, D.M. (1969), A Materialist Theory of Mind. Humanities Press: London & New York.
-Brentano, F. (1991), “Vom inneren Bewusstsein”, in: Frank 1991, 131-168.
-Fichte, J.G. (1991), Science of Knowledge with the First and Second Introductions. Edited and translated by Peter Heath and John Lachs. Cambridge University Press: Cambridge.
-Fichte, J.G. (1994), “An Attempt at a New Presentation of the Wissenschaftslehre (1797/98)”, in: Fichte, J.G. (1994), Introductions to the Wissenschaftslehre, translated and edited by Daniel Breazeale. Hackett Publishing Company: Indianapolis/Cambridge.
-Frank, M. (ed.) (1991), Selbstbewußtseinstheorien von Fichte bis Sartre. Suhrkamp: Frankfurt am Main.
-Frank, M. (2015), Präreflexives Selbstbewusstsein: Vier Vorlesungen. Reclam: Stuttgart.
-Kapitan, T. (1999), “The Ubiquity of Self-Awareness”, in: Grazer Philosophische Studien 57, pp. 17-44.
-Kriegel, U. & Williford, K. (2006), Self-Representational Approaches to Consciousness. MIT Press: Cambridge Mass. & London.
-Lycan, W.G. (1997), “Consciousness as Internal Monitoring”, in: Block, N. & Flanagan, O. & Güzeldere, G. (eds.), The Nature of Consciousness: Philosophical Debates, pp. 755-71. MIT Press: Cambridge, Mass.
-Mach, E. (1922), Die Analyse der Empfindungen und das Verhältnis des Physischen zum Psychischen. Gustav Fischer: Jena.
-Rosenthal, D. M. (2005), Consciousness and Mind. Oxford University Press: Oxford.
-Sartre, J.-P (1972), Being and Nothingness: An essay on phenomenological ontology (transl. H.E. Barnes). Methuen & Co: London.
-Shoemaker, S. (1984), “Personal Identity: A Materialist’s Account”, in: Shoemaker, S. & Swinburne, R. (1984), Personal Identity. Blackwell: Oxford.
-Williford, K. (2006), “The Self-Representational Structure of Consciousness”, in: Kriegel & Williford (2006), pp. 111-42.
-Zahavi, D. (1999), Self-Awareness and Alterity: A Phenomenological Investigation. Northwestern University Press: Evanston, Illinois.