(p.563) Appendix 3: Brain and Unity
(p.563) Appendix 3: Brain and Unity
The discussion of unity in the context of consciousness led us deeply into philosophical territory, as in the discussion of the concepts of unity and subjectivity. There is another point of convergence with philosophy, more specifically with the framework of German philosopher Immanuel Kant, whose transcendental approach I believe can be linked to the brain and neuroscience by advocating what I describe as a neurotranscendental approach (see also Northoff 2011, 2012a and c, 2013, for the linkage between Kant and neuroscience; as well as Churchland 2012, 1–5, 19). Interestingly, a connection of Kant’s philosophy to neuroscience has also been observed by one of the main neuroscientists of visual consciousness, Semir Zeki. His consideration of Kant shall be discussed here and will be put into the current framework.
One concept centrally figuring in Kant’s philosophy is that of transcendental unity, which he suggested is necessary for making consciousness possible. I here specify Kant’s concept of transcendental unity by what I described earlier as the environment–brain unity that I suppose to occur prior to any subsequent unity; that is, prephenomenal unity and phenomenal unity. I also enrich Kant’s concept of synthesis by postulating particular neuronal mechanisms that are supposedly involved in constituting the environment–brain unity as transcendental unity. I conclude the section with the charge of a possible category error; that is, the confusion between natural and logical levels of investigation.
Neuroempirical Remark IA: Zeki’s Theory of “Micro-Consciousness”
Semir Zeki (2003, 2008), based in London, has made major contributions to the understanding of the visual system. His neuroscientific (besides his aesthetic) work focuses mainly on the visual cortex and how it relates to visual consciousness, which he takes as a paradigmatic example of consciousness in general. Let us start with the visual cortex. The visual cortex contains neuroanatomically different systems for visual motion (V5) and color (V4) that have distinct anatomical inputs and are functionally segregated from each other. This is further supported by lesion studies. Patients with lesions in V5 show color blindness (achromoatopsia), while they remain able to see and therefore conscious of visual motion. In contrast, lesions in V4 lead to motion blindness (akinetopsia), whereas the perception and thus consciousness of color is preserved. Since perception here is taken to be identical to consciousness, one cannot deny that these patients show consciousness, albeit limited to either visual motion or color with deficits in the respective other. Zeki speaks here of what he calls “micro-consciousness,” which is “micro” because it is limited to certain contents like color or visual motion.
How is such micro-consciousness generated? Zeki (2003, 2008) conducted a series of imaging experiments where he presented either two identical or non-identical visual stimuli at the same time, for example, same or opposite faces and (p.564) same or opposite houses. The same or identical faces/houses induced high activity in the visual cortex, for example, the face and house areas, and most important, were consciously perceived to 100%. In contrast, neither of the non-identical faces/houses were consciously perceived at all (0%) and went along with lower activity in the respective face and house regions. Interestingly, both regions, face and house regions, were also active when their respective stimulus type remained absent albeit to a much a lower degree.
From these results Zeki concludes that the difference between consciousness and non-consciousness does not lie in the presence or absence of neural activity in particular regions (as, for instance, the involvement of higher order regions like the prefrontal cortex). Instead of the involvement of a particular region, he proposes the degree of neuronal activity in the region processing a particular content (like faces or houses) to be central for inducing consciousness. The higher the activity in the region processing a particular content, the more likely it is that the content will become conscious. In contrast, lower activity levels in the same region will make consciousness of the particular content less likely, or even impossible.
Accordingly, the region’s activity levels predict whether the respective contents will become conscious, entailing what Zeki calls “micro-consciousness.” In addition to their spatial differences, that is, the regions associated with different conscious contents like color or visual motion, micro-consciousness must also be characterized in temporal terms. For instance, color is temporally perceived prior to visual motion, while locations are perceived earlier than color, which in turn precedes the perception of orientation. Different forms of micro-consciousness and their respective contents are thus not only distributed across space, that is, regions, but also across time. There is thus a certain temporal sequence in the occurrence of the different contents and their respective micro-consciousness.
Such intraregional temporal characterizations must be distinguished from interregional temporal synchronization, that is, binding, which must be assumed to occur later following the activation of a particular region at one particular point in time. Since it binds together different features or attributes of a stimulus into a whole, interregional binding and synchronization may be characterized by what Zeki (2003, 2008) calls “macro-consciousness” (and “unified consciousness”; see later), which must be assumed to temporally follow micro-consciousness.
Neuroempirical Remark IB: Zeki’s Theory of “Micro-Consciousness” and Kant’s Concept of “Transcendental Consciousness”
Zeki proposes a clear temporal hierarchy with micro-consciousness occurring early and first, followed by macro-consciousness, and ultimately the overall and final “unified consciousness” (see below) as he calls it. What does Zeki mean by the concept of “unified consciousness”? “Unified consciousness” describes the final and ultimate stage that allows us to perceive ourselves as the perceiving person; it is my self (and no other person’s self) that perceives the visual motion and the color, including their linkage in my perception. It is at this point where Zeki sees the similarity (or correspondence) to Kant, who, according to him, established the connection of micro- and macro-consciousness to the unified consciousness.
What Zeki describes as micro- and macro- consciousness corresponds to what Kant called “empirical consciousness,” while Zeki considers his concept of “unified consciousness” as analogous to Kant’s concept of “transcendental consciousness.” Zeki (2008, 16) cites from Kant the following passage (without giving the exact location in Critique of Pure Reason):
All representations have a necessary reference to possible empirical consciousness. For if they did not have this reference, and becoming conscious of them were entirely impossible, then this would be tantamount to saying that they do not exist at all. But all empirical consciousness has a necessary reference to a transcendental consciousness (a consciousness that precedes all particular experience), viz., the consciousness of myself as original apperception.
What does Zeki think about what Kant described as “transcendental consciousness” and (p.565) its relationship to empirical consciousness? Kant argues that any empirical consciousness can only occur on the basis of a prior transcendental consciousness, implying that the former is necessary related to the later. In contrast, Zeki disagrees with Kant in that any empirical consciousness, that is, micro- and macro-consciousness, must have a necessary relation to transcendental consciousness. This is so because there are cases where micro- and macro-consciousness can easily occur even without the consciousness of myself as the perceiving person that is, unified consciousness. In other words, empirical consciousness, i.e., micro- and macro-consciousness, can occur without and thus disassociated from unified consciousness. Therefore, these instances shed some empirical doubt on Kant’s assumption of the necessary relation of micro- and macro-consciousness to transcendental consciousness.
In addition, Zeki also doubts Kant’s assumption that any transcendental consciousness is prior to any experience, meaning that it precedes the occurrence of micro- and macro-consciousness. He, concedes however, that there must be special cortical programs in, for instance, the visual cortex that must indeed be present before any experience and thus visual consciousness can be acquired so that “all experience must therefore be read into them” (Zeki 2008, 16). The exact nature of the a priori cortical programs remains unclear, however. What is clear though, following Zeki, is that these a priori cortical programs must concern micro-consciousness and thus empirical consciousness, rather than unified consciousness, that is, transcendental consciousness:
The cortical programs to construct visual attributes must also be present before any experience is acquired and all experience must therefore be read into them. It seems more likely that, ontogenetically, the micro-consciousness precedes the unified consciousness and that the programs for them are also present at birth. Hence, even though in adult life the unified consciousness is at the apex of the hierarchy of consciousness, ontogenetically, it is the micro-consciousness that occupies this position. (Zeki 2008, 16)
Neurophilosophical Remark IA: Nonlinear Interaction and Consciousness
There are two points where Zeki (2008) himself admits that he does not know the exact neuronal mechanisms. First, he admits that the neuronal mechanisms underlying the different levels of neural activity that are predictive and decide upon the presence or absence of consciousness remain unclear. Second, he does not give any indication of the exact neuronal nature of the a priori cortical programs that are necessary a priori for subsequent micro-consciousness to occur.
Let us start with the first point, the neuronal mechanisms that allow for the distinction between high and low levels of cortical activity and subsequently between conscious and unconscious perception. Zeki observes that even in the absence of a particular stimulus, for example, house or face, the respective region, that is, face or house region, still shows some degree of neuronal activity. Since the stimulus remains absent here, this neuronal activity must be characterized as what I have described as resting-state activity.
Most important, the level of activity in these regions during the resting state is apparently not sufficiently high enough to induce consciousness of, for instance, houses or faces during the resting state itself. For that, as I propose, a stimulus must interact with that region’s resting-state activity, entailing (usually except in extreme cases of rest–rest interaction as in dreams; see Chapter 26) rest–stimulus interaction to change activity to a sufficient degree (see Chapter 29). But, as Zeki observes, only certain stimuli, that is, similar face or house stimuli, increase the respective regions’ resting-state activity to such levels such that conscious perception of the stimulus becomes possible. In contrast, other stimuli, for example, different face or house stimuli, do increase the respective regions’ resting-state activity, but not to a sufficiently high level as to induce conscious perception.
Why do only certain stimuli, the identical ones, induce consciousness? Neuronally, both cases, the conscious and the unconscious one, show rest–stimulus interaction. This suggests (p.566) that, in both instances, the level of neural activity increases. However, the same neuronal mechanism, rest–stimulus interaction, leads to different levels of neural activity, which then seem to account for the difference between presence and absence of consciousness. There must therefore be some additional neuronal mechanisms at work during rest–stimulus interaction that accounts for the phenomenal difference.
I propose this additional neuronal mechanism to consist of the occurrence of non-linearity during rest-stimulus interaction which makes it possible to associate a phenomenal state, that is, consciousness, with the otherwise purely neuronal stimulus-induced activity (see Chapter 29). How does that stand to the earlier-described results by Zeki obtained during the presentation of similar or different house/face stimuli? I hypothesize that in Zeki’s case of similar faces or houses, nonlinear interaction and consequently higher levels of neuronal activity are more likely to occur than in the case of different faces or houses. In the case of different stimuli, in contrast, rest–stimulus interaction may remain only linear, which makes the induction of sufficiently high changes in the levels of neuronal activity to associate consciousness with the stimuli impossible.
Taken all together, this amounts to the following neurophenomenal hypotheses. I hypothesize that the degree of nonlinear interaction during rest–stimulus interaction (see Volume I, Chapters 11 and 12) is directly related to the degree of neuronal activity in the respective regions and consequently to the degree of consciousness: The higher the degree of non-linearity during rest–stimulus interaction, the higher the degree of stimulus-induced activity, and the higher the likelihood that the resulting stimulus-induced activity will be associated with consciousness. This remains to be explicitly demonstrated in the future, however (see also Chapters 28 and 29).
We are now able to provide an answer to Zeki’s first question, the one about the neuronal mechanisms that predict the high levels of neural activity and their association with consciousness. Based on the considerations discussed in this section, my answer to Zeki’s first point is that the degree of nonlinearity during rest–stimulus interaction accounts for sufficiently high neural activity levels to subsequently associate consciousness with the stimulus and its purely neuronal stimulus-induced activity.
Neurophilosophical Remark IB: “Cortical Programs” and Difference-Based Coding
The argument in the preceding section leads me to the second point. There must be some information encoded in the resting state that programs it to enable nonlinear rather than merely linear rest–stimulus interaction. I argue that this information is encoded in the resting state in the gestalt of the kind of neural code the brain’s resting state applies to the processing of all changes in its activity levels (during either the resting state itself or during stimulus-induced activity).
Moreover, I propose this kind of neural coding to be difference-based coding, as detailed in Volume I, which I also suggest to predispose the occurrence of consciousness as discussed here in Volume II. I postulate that difference-based coding is central in allowing for nonlinear interaction, such as that between identical stimuli, as in Zeki’s experiment with same or different houses and faces (see Chapter 29). Such nonlinear interaction in turn may allow the association of a phenomenal state, that is, consciousness, with the purely neuronal stimulus-induced activity as, for instance, related to the identical stimuli in Zeki’s experiment.
Zeki’s results of the neuronal and phenomenal difference between same and different face/house stimuli are nicely compatible with the assumption of difference-based coding and its nonlinear character in the presence of the “right,” for example, identical stimuli. In both cases, same and different face/house stimuli, the stimulus material remains the same. Despite the presentation of the same stimulus material, there are differences in both regards—neuronally, i.e., in the degree or level of neural activity, and phenomenally, i.e., in the presence or absence of consciousness.
How can we account for these neuronal and phenomenal differences in the presence (p.567) of the same stimulus material? If the stimuli were encoded by themselves, i.e., in isolation and thus independently of their respective combinations or constellations, there should be neither a neuronal nor a phenomenal difference. In other words, if stimulus-based coding were at work, there should be neither a neuronal nor a phenomenal difference between same and different face/house stimuli. This suggests that the constellation or combination between the stimuli as either the same or different is central for determining the neuronal and phenomenal differences. In other words, this suggests that difference-based coding is at work that allows the brain to encode the spatial and temporal differences between stimuli rather than the stimuli themselves into neural activity. Accordingly, Zeki’s results speak in favor of difference-based coding rather than stimulus-based coding.
How does such difference-based coding relate to what Zeki described as the “cortical programs”? I propose that what Zeki calls “cortical programs” describe the programming of the brain’s resting-state activity to apply a specific kind of encoding to its own activity changes; namely, difference-based coding rather than stimulus-based coding. In the same way the computer is programmed to apply the 0-1 code to its own processing of any incoming information, the brain and its resting-state activity apply difference- rather than stimulus-based coding to its own processing of any activity changes as induced either by the resting state itself or extrinsic stimuli. In short, I propose what Zeki calls a “cortical program” to consist of a particular neural code, difference-based coding.
Neurophilosophical Remark IIA: “Cortical Programs” and “Content- Based Concepts of Consciousness”
How does all that relate to Kant? As discussed earlier, Zeki contests Kant’s assumption of transcendental consciousness’s being a priori and necessarily related to empirical consciousness. Why? Because, following him, micro- and macro-consciousness occur temporally prior to unified consciousness. He proposes this because the perception and thus consciousness of attributes and features, e.g., micro- and macro-consciousness precede the consciousness of the self who perceives these attributes and features, e.g., unified consciousness.
However, Zeki does not contest Kant’s assumption that there must be something occurring prior and thus a priori to micro- and macro-consciousness when he proposes specific cortical programs that make consciousness possible. Accordingly, Zeki does not deny the necessity that there must be something prior to empirical consciousness, e.g., micro- and macro-consciousness, in order for it to be possible. In contrast to Kant, Zeki does not associate this “something prior” with the concept of transcendental consciousness, but rather with what he describes as “cortical programs.” This implies that the a priori cortical programs cannot be associated with the perception and consciousness of the perceiving self, the unified or transcendental consciousness. How can we clarify this conceptual puzzle? I argue that we need to distinguish different concepts of consciousness, which shall be detailed in the following.
Zeki presupposes a concept of consciousness that is based on contents. His concepts of micro- and macro-consciousness and unified consciousness are all based on different contents, features/attributes in micro-consciousness, objects and events in macro-consciousness, perceiving self in unified consciousness. This follows Kant’s characterization of empirical consciousness and its determination by contents that are by their very nature empirical. I therefore speak of a “content-based concept of consciousness” (see Fig. A3-1).
Neurophilosophical Remark IIB: Kant’s Transcendental Consciousness as a “Mode-Based Concept of Consciousness”
How does Kant’s concept of transcendental consciousness fit in? This is where the trouble starts. Kant seems to characterize consciousness as content based when he associates it with the “consciousness of myself as original apperception” (see above quote from Zeki). This corresponds well to Zeki’s characterization of (p.568)
However, such a content-based determination of transcendental consciousness is incompatible with necessity of its occurring prior to empirical consciousness; that is, micro- and macro-consciousness. If transcendental consciousness is defined by the self that links, integrates, and thus unifies all preceding contents, including the ones associated with micro- and macro-consciousness, transcendental consciousness must follow rather than precede empirical consciousness. In short, a content-based determination of transcendental consciousness makes its characterization as a priori impossible. If, however, transcendental consciousness is no longer characterized as a priori, empirical consciousness can no longer show necessary reference to it.
This is well observed by Zeki (2008, 16), who, as described earlier, denies Kant’s assumption of necessity; that is, necessary reference. I propose the problem here to stem from an ambiguity in Kant himself in his determination of consciousness. When claiming for transcendental consciousness to be a priori and being the necessary reference for empirical consciousness, Kant does not presuppose a content-based determination anymore but rather a “mode-based concept of consciousness.” (p.569) What do I mean by “mode-based concept of consciousness”?
The concept of mode refers here to the principally conscious mode as distinguished from the principally non-conscious mode thus referring to what, relying on Searle, I described in the introduction as “principal consciousness” as distinguished from “principal non-consciousness.” Such mode-based concept of consciousness, that is, “principal consciousness,” can well be characterized a priori, thereby signifying a specific “cortical program,” as Zeki says, and a particular way of neural coding, difference-based coding as suggested earlier.
How can we characterize the concept of cortical programs in further detail on the basis of a mode-based concept of consciousness? I would propose that a particular coding strategy, that is, difference-based coding, to account for what Zeki calls “cortical programs” and what Kant describes as transcendental consciousness. Therefore, I consider that difference-based coding takes on the role of what may be called a transcendental (or better, neuro-transcendental) condition or, in my own terms, a neural predisposition, that is, necessary, non-sufficient condition, of possible consciousness, that is, mode-based consciousness (or Kant’s transcendental consciousness, as determined in a mode-based way). This has important implications for the determination of Kant’s concept of transcendental consciousness. Instead of implying a neural correlate and thus actual consciousness, that is content-based consciousness or, as Kant would say, empirical consciousness, transcendental consciousness must be characterized as mode-based (rather than content-based) and requires the search for neural predisposition (rather than neural correlate).
Neurophilosophical Remark IIC: Mode-Based Concept of Consciousness Requires a Neurotranscendental Approach
I postulate that that difference-based coding is a transcendental (or neuro-transcendental) condition (or neural predisposition) of consciousness in a mode-based (rather than content-based) way. This however is not compatible with Kant’s concept of transcendental consciousness when he presupposes it in a content-based way (as in the context of the “consciousness of myself as original apperception”). That means to confuse the mode of consciousness with its contents since being conscious of myself pertains to a content (one’s own self) rather than a mode.
Such a content-based determination of transcendental consciousness in Kant, however, is to be distinguished from his assumption of a necessary and a priori role of transcendental consciousness for possible empirical consciousness. This pertains to what I said above: that there must be “something additional” besides empirical consciousness itself for it to be possible. This “something additional” is apparently what Kant refers to when he signifies transcendental consciousness as a priori and necessary for the possibility of empirical consciousness.
How can we characterize Kant’s concept of transcendental consciousness and thus the “something additional” in further conceptual and empirical detail? I argue that what I described as mode-based consciousness signifies Kant’s concept of transcendental consciousness, including its a priori and necessary character. Conceptually, the a priori and necessary character of Kant’s concept of transcendental consciousness resurfaces in what I described in the second Introduction as “principal consciousness,”.
What exactly does the concept of “principal consciousness” refer to? The concept of “principal consciousness refers to the principal possibility of the occurrence of a phenomenal state, that is, consciousness, independently of whether it is actually realized or not. Empirically, as stated earlier, the concept of transcendental consciousness and its conceptual analogue in my framework, “principal consciousness,” are supposed to be related to a particular kind of neural coding, difference-based coding, that the brain and its resting-state activity apply to all changes in the brain’s own activity.
Finally, another point of convergence with Kant shall be mentioned. Kant considered the mind as an active organ that provides an input that structures and organizes the stimuli from the environment such as that we can cognize them. This active input refers to the transcendental (p.570) level and that is where he “located” the mode of consciousness and thus transcendental consciousness.
The same is true in the case of the brain. I characterize the brain as an active organ that provides an input, that is, its spatiotemporal structure of the resting state and its specific neural coding, that is, difference-based coding, that predispose the brain to process the stimuli from the environment in a certain way. This is what I here described as neural predisposition, which, taken from a Kantian perspective, may well be described as neurotranscendental (see also Northoff 2011, chapters 1 and 2 herein; Northoff 2012, 2013).
Kant’s mode-based concept of consciousness, that is, transcendental consciousness, may consequently well be associated with the active input to the brain to its neural processing of stimuli from body and environment. This implies a neurotranscendental approach, which is here conceptualized as the search for the neural predispositions of consciousness. More specifically, this brain’s active input may consists in its resting state’s spatiotemporal structure and its specific way of neural coding, difference-based coding, which predisposes the brain to associate a phenomenal state, that is, consciousness, with its purely neuronal activity changes during either rest–rest or rest–stimulus interaction.
Philosophical Remark IA: Concepts of Transcendental and Empirical Unity
So far, I have used the concept of the transcendental without going into further detail. I therefore will discuss the concept of the transcendental now in the subsequent sections.
The notion of the transcendental was introduced by Immanuel Kant. Roughly, Kant introduced the term “transcendental” to characterize all knowledge that focuses more on the form of our cognition and knowledge of ourselves and the world than on the content of our knowledge.1 Since the form of our knowledge must also be “cognized” and known by us, Kant speaks here of a certain form of cognition or knowledge, a priori cognition, which remains independent of the specific contents: “I call all cognition transcendental that is occupied not so much with objects but rather with our mode of cognition of objects insofar as this is to be possible a priori” (Kant 1998, A11–A12).
The concept of the “transcendental” must be distinguished from that of the “transcendent” that “goes beyond” or transcends any possible knowledge of humans into a world that lies beyond the world we inhabit. Let me rephrase this important distinction. The concept of “transcendental” concerns the possible knowledge of the objects within the world that we can possibly cognize. In contrast, the notion of “transcendent” goes beyond the objects we can possibly cognize by postulating some objects in a world that lies beyond our possible cognition and knowledge, for example, a transcendent world.
In sum, the notion of the “transcendental” concerns the mode in which we cognize objects in the world, whereas “transcendent” refers to objects in a non-natural world we cannot cognize at all (see also footnote 6 on p. 717 in the Introduction by P. Guyer and A. Wood in Kant 1998). In other words, the concept of “transcendence” has ontological-metaphysical implications, thus belonging to the ontological-metaphysical domain, while the concept of the transcendental remains (supposedly) purely epistemic.2
Philosophical Remark IB: Transcendental Unity as Form or Structure
What exactly is meant by the “form” (or mode) of our knowledge? Kant refers here to a specific structure and organization that is inherent in our knowledge. One such central form that structures and organizes our knowledge is unity. Unity provides the most basic form or structure and organization of our cognition and knowledge of ourselves and the world.
In other words, unity is the basic form or structure and organization of consciousness:
Every necessity (i.e., the necessity of connection) has a transcendental condition as its ground. A transcendental ground must therefore be found for the unity of consciousness in the synthesis of the manifold of all our intuitions, hence also the concepts of objects in general, consequently also of all objects of experience without which it would be impossible to think (p.571) of any objects for our intuitions; for the latter is nothing more than the something for which the concept expresses such a necessity of synthesis. Now this original and transcendental condition is nothing other than the transcendental apperception. (Kant 1998, A106–A107; see below for the determination of the terms “apperception”3 and “synthesis”)
The unity of consciousness reflects the most basic form or structure and organization, which as such must be distinguished from a more empirical unity, the unity we encounter in the contents of our perception, or outer sense as Kant would have said, and the contents in introspection, or inner sense in Kant’s terms. Kant considered the more empirical unity of inner and outer sense, that is, perception and introspection, to be dependent upon the unity as basic form or structure and organization, i.e., a transcendental unity, as one might say. Without the unity as basic form or structure and organization, i.e., the transcendental unity, no unity, i.e., empirical unity, in either perception or introspection would be possible at all.4
Taking all this into consideration, the unity as basic form or structure and organization must be characterized as transcendental and thereby be distinguished from the unity in perception and introspection that is then empirical rather than transcendental. One may consequently distinguish between transcendental unity and empirical unity with the former providing the ground or necessary condition for the possibility of the latter.5
Philosophical Remark IC: Synthesis of Transcendental Unity
How is the transcendental unity of consciousness generated? Kant considers the transcendental unity to be the very basis of the empirical unity and, even more radical, of any other form or structure and organization in consciousness. He therefore proposes that the transcendental unity is a priori given.6 Where, however, does the transcendental unity come from? Kant suggests some kind of process that generates the transcendental unity, and this process is described by the term “synthesis”:
Only the spontaneity of our thought requires that this manifold first be gone through, taken up, and combined in a certain way in order for a cognition to be made out of it. I call this action synthesis. By synthesis in the most general sense, however, I understand the action of putting different representations together with each other and comprehending their manifold in one cognition. (Kant 1998, A77/B102–B103)
The concept of synthesis refers to a “putting together,” “combination,” “composition,” and “nexus” (see footnote a in Kant 1998, A77/B103) of what Kant called the “manifold” resulting in unity: “But in addition to the concept of the manifold and of is synthesis, the concept of combination also carries with it the concept of the unity of the manifold. Combination is the representation of the synthetic unity of the manifold” (Kant 1998, B130–B131). This7 entails that the transcendental unity is also a synthetic unity that (unlike an “analytical unity”) underlies certain processes yielding its generation.8 That is well reflected in the following quote where Kant speaks of a “synthetic unity of apperception”:
This synthetic unity (of apperception), however, presupposes a synthesis, or includes it, and if the former is to be necessary a priori then the latter must also be a synthesis a priori. Thus the transcendental unity of apperception is related to the pure synthesis of the imagination....Now we call the synthesis of the manifold in imagination transcendental if, without distinction of the intuitions, it concerns nothing but the connection of the manifold a priori, and the unity of this synthesis is called transcendental if it is represented as necessary a priori in relation to the original unity of apperception. (Kant 1998, A118; see also B135, where Kant speaks of the transcendental synthesis as the “faculty of combining a priori”)
Neurophilosophical Remark IIIA: Environment–Brain Unity as Neurotranscendental Unity
How does Kant’s concept of transcendental unity relate to the here-suggested “environment–brain unity”? Analogous to Kant’s transcendental unity, the environment–brain unity is the most basic form or structure and organization upon which any kind of subsequent (p.572) neuronal processing and ultimately consciousness depends and is built (see Chapters 20 and 21). The environment–brain unity is supposed to be based upon a statistically based spatiotemporal continuity between the environmental stimuli and the brain’s resting-state activity. Such a statistically based spatiotemporal continuity leads, in an ideal case, to the constitution of a virtual spatiotemporal unity between environment and brain (see Chapter 21). This virtual spatiotemporal unity between environment and brain, the environment–brain unity, is supposed to bias and predispose the subsequent constitution of the phenomenal unity and thus consciousness during rest–stimulus interaction (see Chapters 18 and 29).
How is such biasing and predisposition of consciousness by the environment–brain unity possible? For instance, a rhythmic presentation of environmental stimuli may lead to a higher degree of a statistically and spatiotemporally based environment–brain unity than a non-rhythmic presentation of the same stimuli (see Chapter 20 for details). And the higher the degree of the environment–brain unity, the more likely it is that a phenomenal state, that is, consciousness, can be associated with the resulting change in the resting state’s neural activity. This suggests that the environment–brain unity does indeed provide the basic form or structure and organization, that is, a particular temporal and spatial template, for the subsequent phenomenal unity as a hallmark of consciousness.
The dependence of the phenomenal unity on the preceding environment–brain unity is also reflected in the relationship between the different underlying neuronal mechanisms. More specifically, the phase of the resting state’s low-frequency oscillation is adjusted in a specific way to the statistical structure of the environmental stimuli; as such, it biases and predisposes the phases and amplitudes of the more stimulus-related high-frequency oscillations, including their degree of entrainment by the low-frequency oscillations of the resting state.
How, then, is the relationship between low- and high-frequency oscillations related to the association of consciousness to the stimulus-induced activity? The degree of the high- by low-frequency oscillation entrainment biases and predisposes how the actual stimulus, and its specific temporal (and spatial) discrete point in time (and space), will be processed during subsequent rest–stimulus interaction: The better the stimulus’ discrete position in time and space corresponds to and matches with the phase durations of the ongoing high-by-low-frequency entrainment in the resting state, the higher the likelihood that consciousness will be associated with the stimulus and its respective stimulus-induced activity.
Taken together, this demonstrates that the phenomenal unity of consciousness can indeed be ultimately traced back to the virtual and statistically and spatiotemporally based environment–brain unity.. This means the environment–brain unity must be considered a necessary condition of possible consciousness. Moreover, the environment–brain unity must occur prior to the actual stimulus that is to be associated with consciousness. The environment–brain unity can consequently indeed be characterized as a transcendental unity in very much the same way Kant used this concept when presupposing it in a mode- rather than content-based way. Since it is based, at least in part, on the brain and its neuronal states, i.e., its intrinsic activity, one may want to characterize the environment–brain unity as neurotranscendental unity rather than merely as transcendental unity.
Neurophilosophical Remark IIIB: Empirical Evidence for the Neurotranscendental Role of the Environment–Brain Unity
Let me describe this striking analogy to Kant in slightly different terms. Both unities, Kant’s transcendental unity and my environment–brain unity, are supposed to provide the base for any subsequent unity, be it the empirical unity of consciousness in Kant, or what I (and others) describe as phenomenal unity of consciousness (which for Kant would be subsumed under what he describes as empirical unity). I consequently propose that what Kant called transcendental unity (in a mode- rather than content-based way) may correspond more or less to the concept of environment–brain unity as posited here (see (p.573) later for more detailed discussion of the suspicion of what philosophers call “category error”; as well as Northoff 2011, 2012, 2013; see also Fig. A3-2).
One may go even one step further. Kant associates the empirical unity with inner and outer sense, that is, perception and introspection. This is strikingly similar to what I here describe as phenomenal unity that can occur in either perception of the outer environment, that is, outer sense, or the perception of one’s own self, that is, introspection or inner sense. Hence, I propose that what I here describe as phenomenal unity may more or less correspond to what Kant called empirical unity.
Most important, Kant claimed the empirical unity to be prevalent during perception and introspection, that is, inner and outer sense, and to depend on the preceding transcendental unity. I demonstrated here empirical support for the perception and thus outer sense to depend on the preceding environment–brain unity. This was, for instance, shown in the case of schizophrenia, where an abnormally altered environment–brain unity leads to bizarre perception of the environment and one’s own self, thus including both inner and outer sense (see Chapters 22 and 27).
However, I did not show any data supporting that the perception of one’s own self in introspection, that is, inner sense, is also dependent upon the prior environment–brain unity. I only showed that the environment–brain unity has an indirect impact via the resting state’s self-specific organization upon the degree of self-specificity assigned to subsequent stimuli (see Chapters 23 and 24). In contrast, empirical support for the environment–brain unity’s impact on the self (and self-consciousness) was only gathered indirectly, via the alteration of the self in schizophrenia (see Chapter 27). Hence, future research is warranted to demonstrate the dependence of our sense of self, i.e., self-consciousness, on the degree of the spatiotemporally and statistically based environment–brain unity.
In contrast to the here-presupposed characterization of the environment–brain unity as transcendental, Kitcher (1992) proposes higher-order cognitive functions like working memory and attention to be crucially involved in the generation of the transcendental unity. This, however, is challenged here by showing that what Kant called “transcendental unity” is very much synthesized by and based on a specific method of neural coding of stimuli at the interface between brain and environment. Hence, rather than going up to the highest logical functions as Kant did, or the highest cognitive functions as Kitcher does, I claim that we need to go down to the lowest
Neurophilosophical Remark IIIC: Kant’s Concept of “Synthesis” and the Constitution of the Nvironment–Brain Unity
So far, I have applied the conceptual framework of Kant to shed more light on the environment–brain unity using his concept of the transcendental unity to enrich and detail my own concept. However, the transfer may also go the reverse way, from my neuroscientifically based concepts to the more philosophical= ones of Kant. The neuroscientific data and findings may help to detail and further specify some of Kant’s concepts like his concept of synthesis, as described earlier. Kant characterizes synthesis by “putting together,” “combination,” “composition,” and “nexus” (see earlier). Though Kant distinguishes between distinct kinds of synthesis (mostly with regard to different material or content that is synthesized), the details of such “combination,” “putting together,” “composition,” and “nexus” remain unclear (in either case of the different concepts of synthesis).
This is the point where I propose that the here suggested neuroscientifically based concepts and particularly their underlying neuronal mechanisms may contribute to fill the gap Kant left open in his concept of synthesis. Let me be more specific and detail the concept of synthesis by discussing each of its different features like “putting together,” “combination,” “composition,” and “nexus” with regard to the environment–brain unity (as transcendental unity). What exactly is “put together” in the synthesis of the environment–brain unity? Our empirical data provide a clear answer, as described in detail in Part VI). The environment and more specifically the occurrence of its stimuli across time (and space) are put together, integrated, and linked with the low-frequency oscillations in the brain. The phases of the resting state’s low-frequency fluctuations are aligned to the onset of the stimuli from the environment as reflected in their statistical frequency distribution, i.e., their natural statistics.
This, in turn, makes possible the constitution of a statistically based virtual temporal (and spatial) continuity between the environment’s stimuli and the brain’s resting-state activity. Both brain and environment are thus directly linked together in a statistically based virtual temporal (and spatial) continuum, resulting in what I called “environment–brain unity.” What Kant described as synthesis and “putting together” may thus be empirically specified by the resting state’s neural alignment to the statistical frequency distribution of the environment stimuli, i.e., their natural statistics.
How are both stimuli and low-frequency oscillations “combined” in the synthesis of the environment–brain unity? They are combined by means of their statistical structures. More specifically, the statistical structure of the environmental stimuli’s occurrence across time is “combined” with the phase of the low-frequency oscillations, that is, their cycling across time—hence, the statistical-based rather than physical-based nature of the environment–brain unity (see Chapter 20). Kant’s concept of “combination” as hallmark feature of synthesis can consequently be empirically specified here by the matching processes between two different statistical frequency distributions, the one from the environmental stimuli and the one from the brain and its resting state’s stimuli.
What kind of “composition” is going on in the synthesis of the environment–brain unity? The environment–brain unity is synthesized and thus composed by spatial and temporal differences between the different stimuli and their respective discrete points in time and space that (p.575) are encoded into the brain’s neural activity via difference-based coding. What is composed by the synthesis of the environment–brain unity is thus spatial and temporal differences, which is possible on the basis of difference-based coding. Hence, Kant’s concept of “composition” can be specified by the difference-based (rather than stimulus-based) nature of the environment–brain unity.
Finally, based on its continuity-, statistical-, and difference-based nature, the environment–brain unity may well be described as “nexus” between environment and brain. A nexus is where two distinct set of properties or features overlap at a particular point in space and time while diverging in others. This is exactly what happens with regard to the resting state’s low-frequency fluctuations, including their phases on one hand, and the rhythmic structure of the environmental stimuli on the other.
Let me be more specific. The phase onsets of the resting state’s low-frequency fluctuations may align themselves to the onsets of the environmental stimuli and their rhythmic structures, thus accounting for the overlap between environment and brain. In contrast, it may be impossible for the resting state’s low-frequency fluctuations to align their phase onsets to some other stimuli in the environment (simply because the former’s phase onsets do not correspond to the latter and their rhythmic or non-rhythmic structure). Taken together, this means that the resting state’s neuronal mechanisms of phase shift and neural alignment may empirically specify Kant’s more metaphorical description of synthesis by the term “nexus.”
Neurophilosophical Conclusion IA: Kant and the Brain—Nothing but a Category Error?
Leaving aside and neglecting most of the difficulties and controversies in interpreting Kant (this is left to the philosophers and especially the Kant specialists), I shall nevertheless briefly mention one central argument against my neurotranscendental interpretation (see Northoff 2011, 2012, 2013, for a more detailed neurotranscendental account of Kant). The traditional philosopher and especially the Kant specialist may be very much puzzled by the fact that I propose correspondence between Kant’s transcendental unity and the environment–brain unity.
Why may the traditional philosopher be puzzled? He may diagnose what is called a “category error” in philosophical circles. My assumption of a correspondence between Kant’s transcendental unity and my concept of “environment–brain unity” is faulty in that it confuses logical and empirical categories: Kant used the term transcendental unity in a predominantly logical (and epistemic, I would claim) context, which is by definition devoid of any reference to empirical reality, including space and time.
Such a predominantly logical (and epistemic) domain excludes any empirical characterization. This means that the characterization by space and time, which are deemed to be empirical (or metaphysical-ontological), are also excluded: any concept that directly refers to time and space or at least indirectly presupposes them can only be characterized as empirical, not as transcendental. What does this imply for the characterization of my concept of the environment–brain unity? I characterized the concept of the “environment–brain unity” in strongly spatial and temporal terms; namely, by the spatial and temporal continuity between the environmental stimuli’s statistical frequency distribution and that of the resting state’s spatial and temporal neuronal measures. This even led me to describe the environment–brain unity as a spatiotemporal unity.
This, however, following the Kantian philosophers, designates the environment-brain unity as empirical rather than as transcendental. When characterizing the “environment–brain unity” as transcendental, or better, neurotranscendental, I thus confuse Kant’s notion of the transcendental with the concept of the empirical: due to its reference to space and time, the concept of the “environment–brain unity” can at best be characterized as empirical but not as transcendental., I consequently commit what the philosophers call a “category error” that consists in the confusion between transcendental and empirical levels (and ultimately between logical and natural contexts).
(p.576) Neurophilosophical Conclusion IB: Kant and the Brain—Immunity of the Neural Predispositions of Consciousness Against the Category Error
Is the charge of a “category error” justified? Presupposing Kant’s predominantly logical context, the charge of a category error is certainly justified. This, however, changes once one interprets Kant no longer in an exclusively logical context but rather in the natural context of embodiment and embeddedness (see, for instance, Svare 2006). The transcendental unity is then no longer supposed to be generated by “reason” and “understanding,” as Kant himself proposed, but rather by the body and its integration within the environment. Most important, such a shift from a purely logical of reason to the more natural context of the body entails the reference to and inclusion of space and time. Space and time are then an integral part of the natural reality of the body, and, to put it even more strongly, they may constitute that natural reality by providing some kind of template or grid.
What does the shift from the logical to the natural context entail for the alleged “category error”? The terms “transcendental” and “empirical” do then need to be redefined in their relation to space and time. What in Kant’s purely logical context would be deemed to be empirical, that is, the environment–brain unity, could then be designated as transcendental within a natural context, as I suggested earlier. This needs to be detailed.
My focus here is one our natural world rather than on the purely logical world as Kant presupposed it. This shift in context that is from the logical to the natural world entails a redesignation of the role of time and space as part of that very same natural world: If time and space necessarily precede consciousness in an a priori way, and do henceforth predispose possible consciousness (as distinguished from actual consciousness), space and time need to be characterized as transcendental. If, in contrast, time and space enable the manifestation of actual consciousness and thus correlates rather than predispositions, they must be characterized as merely empirical rather than transcendental.
I now claim that the spatiotemporal continuity that characterizes the environment–brain unity takes on exactly such transcendental, or better, neurotranscendental, role with regard to consciousness: The environment–brain unity precedes the occurrence of consciousness and is as such a necessary condition of its possibility, that is a predisposition (rather than a correlate). In other words, I propose the environment–brain unity that allows to constitute time and space to predispose possible consciousness and thus be a neural predisposition of consciousness (NPC). This clearly fulfills the criteria for a transcendental, or better, neurotranscendental, role of the environment–brain unity.
Does the environment-brain unity has a special and thus transcendental rather than merely empirical role? As NPC, the environment–brain unity must be distinguished from the neural conditions that are sufficient to induce actual consciousness, like nonlinear rest–stimulus interaction, which I consider the neural correlate of consciousness (NCC) (see Part VIII for details). In contrast to the NPC and their transcendental or neurotranscendental role, the NCC take on an empirical role. This means that the charge of a “category error” can well be directed toward the NCC, whereas it does not apply in the case of the NPC and particularly the environment–brain unity.
Neurophilosophical Conclusion IC: Kant and the Brain—Natural Versus Logical Worlds
The traditional philosopher may now claim that all that sounds plausible. Yes, presupposing the framework of the natural context rather than the one of the logical context leads indeed to a re-definition of the transcendental and the empirical. And that in turn may indeed rule out the diagnosis of a category error in the case of the NPC and thus the environment–brain unity.
Kant, however, was not interested at all in the natural reality itself. Instead, he (and many other past and current philosophers) focus on the logical conditions, the transcendental conditions, that are necessarily presupposed by the natural, i.e., the empirical world. This, however, implies that the concept of the transcendental cannot (p.577) be associated at all with the natural and thus the empirical world, but only with the logical domain. Since time and space are excluded by the logical domain, they cannot be associated at all with the concept of the transcendental and consequently with the environment–brain unity. That in turn means that the environment–brain unity can at best be characterized as empirical but not as transcendental. In other words, the diagnosis of a “category error” needs to be maintained as long as one follows Kant’s original framework with its dichotomy between logical and natural domains.
How can we continue from here on? Now it all comes down to whether one accepts Kant’s original framework with its dichotomy between natural and logical domains. Or whether, alternatively, one shifts Kant’s logical (and epistemic) domain, including its transcendental conditions, into the natural world, thus presupposing what is described as “naturalization” in philosophy. This raises the question for what philosophers describe as ‘naturalization’. I do not want to go into the philosophical debate here over whether one can “naturalize” Kant or not. I leave that to the philosophers and to the search for conceptually and logically plausible answers.
The only point I want to make here is a neurophilosophical one. From a neurophilosophical perspective, the question of the naturalization of Kant comes down to the question of empirical and more specifically neuronal plausibility. If such a naturalization of Kant and his concept of the transcendental is empirically plausible, i.e., in accordance with the empirical data of the brain, I can avoid the charge of a category error at least for claims that are limited to the natural world. My characterization of the environment–brain unity as transcendental or neurotranscendental is therefore at least valid in the natural world.
This however changes once one presupposes the logical world as the traditional philosophers do. My characterization of the environment–brain unity as transcendental is indeed a category error and may therefore not be valid in the purely logical world, which the philosophical traditionalists claim to hold in Kant. This has important implications.
Do I need to be concerned about the charge of a category error in the domain of the logical world? No, because as a neurophilosopher (and neuroscientist) I am primarily interested in the natural world, so that my work is done once I show that the naturalization of Kant’s philosophy is empirically plausible. If so, I do not need to care much about the charge of a category error that applies only to the logical world because that world is simply not my primary concern as neurophilosopher. My aim is to explain how the brain and consciousness are related to each other in the natural world we live in, rather than in some merely logically possible world we do not actually live in.
(1) . Obviously, I will not be able to recount the details and the difficulties in interpreting Kant’s stance here. I leave this to the philosophers to discuss.
(2) . This also makes it clear that the here-supposed concept of the “transcendental” does not refer to the possible cognition of objects outside the limits of our cognition and a priori cognition (see also McDowell 1994 for a more refined post-Kantian concept of the concept of the transcendental). Hence, there is nothing mysterious about the transcendental view of the mind’s input to our cognition of objects and events of the world, while the search for transcendent objects is mysterious in its search for objects beyond and thus outside our (cognition of the) world.
(3) . It should, however, be noted that in his attempts at a deduction in between the A- and B-version (Prolegomena, Kant 1977); and after the B-version, Kant did not rely on apperception as a primary tool for the deduction. He does not even mention the term “apperception” in either the Prolegomena to Any Future Metaphysics or Metaphysical Foundations of Natural Science.
(4) . Within Kant’s mainly logical and epistemic (and higher-order cognitive) framework (as I propose it to be), this requires the distinction of the transcendental unity of consciousness from the unity as one category within his list of categories, so that one may designate the former unity as extra categorical as distinguished from the categorical use of the term “unity” (see Caygill 1995, 407–409). This is also reflected in the following quote from Kant:
This unity, which precedes all concepts of combination a priori, is not the former category of unity (p.578) (§10); for all categories are grounded on logical functions in judgements, but in these combination, thus the unity of given concepts, is already thought. The category therefore already presupposes combination. We must therefore seek this unity (as qualitative, §12) someplace higher, namely in that which itself contains the ground of the unity of different concepts in judgements, and hence of the possibility of the understanding, even in its logical use. (Kant 1998, B131).
(5) . Kant seems to describe the cognition of both transcendental and empirical unity by the term “apperception” (I deliberately say “seems” because the definition of the term “apperception” is highly controversial), where he correspondingly distinguishes between transcendental and empirical apperception. This is well reflected in the following quote that concerns the introspection, that is, the inner sense, of one’s own self:
Now this original and transcendental condition is nothing other than the transcendental apperception. The consciousness of oneself in accordance with the determinations of our state in internal perception is merely empirical forever variable; it can provide no standing or abiding self in this stream of inner appearances, and is customarily called inner sense or empirical apperception. (Kant 1998, A107; see also B132)
However, Allison (1983, 273–274) points out that the identification of empirical apperception and inner sense is problematic, because Kant thereby undermines the distinction between both kinds of apperception. Rather than treating both types of apperception as distinct activities or faculties, which is suggested by the identification of empirical apperception with inner sense, both shall rather be conceived as distinct ways in which apperception can be conceived. Allison suggests that empirical apperception may be regarded as the consciousness of its activity during cognition of objects (o), while transcendental apperception may be described as a thought about the same activity. This, however, is problematic, because then transcendental apperception must be characterized as an analytical unity, which, according to Kant, it is not; it is rather a synthetic unity (see later). Allison’s characterization of transcendental apperception of the thought of synthesizing and cognizing activity, as distinguished from the consciousness during actual synthesis and cognition, opens the door for a purely logical account of transcendental apperception (see later). In this case, however, transcendental apperception could no longer be characterized as an epistemic function, as I propose was Kant’s intention, but rather by a purely logical role.
(6) . Ì am aware that the notion of the “given” is by itself problematic and could include different meanings; this though is left for the philosophers to discuss.
(7) . I am well aware that this carries plenty of room for interpretation, which I leave to the Kant specialists.
(8) . Kant seems to speak here also of a “pure synthesis” that is pure because it neither concerns specific contents nor space and time, thus being beyond or “outside of space and time”; see later for further discussion of this point.
(9) . This is almost literally reflected in the following quote when Kant characterizes the transcendental unity as the highest point: “And thus the synthetic unity of apperception is the highest point to which one must affix all use of the understanding, even the whole of logic and, after it, transcendental philosophy; indeed this faculty is the understanding itself” (Kant 1998, B134*).