Jump to ContentJump to Main Navigation
Unlocking the BrainVolume 2: Consciousness$

Georg Northoff

Print publication date: 2013

Print ISBN-13: 9780199826995

Published to Oxford Scholarship Online: April 2014

DOI: 10.1093/acprof:oso/9780199826995.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 16 January 2019

Temporal Nestedness and “Duration Bloc”

Temporal Nestedness and “Duration Bloc”

Chapter:
(p.48) Chapter 15 Temporal Nestedness and “Duration Bloc”
Source:
Unlocking the Brain
Author(s):

Georg Northoff

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780199826995.003.0003

Abstract and Keywords

The concept of the “duration bloc” describes the integration of past, present, and future into one homogenous though threefold experience of time in consciousness. The book hypothesizes that the “global” temporal continuity of neural activity predisposes and makes possible such extension of the present into past and future on the phenomenal level of consciousness. By temporally structuring and organizing neural activity in a particular way, the global temporal continuity of the brain’s neural activity makes necessary and unavoidable the constitution of the threefold temporal structure with past, present, and future, the duration bloc, on the phenomenal level of consciousness.

Keywords:   temporal pattern of neural activity, low-frequency fluctuations, NREM sleep, resting state, slow wave activity, slow cortical potentials, temporal nestedness, difference-based coding, duration bloc, global temporal continuity

Summary

I discussed the neuronal mechanisms of the “width of present” in Chapter 14. The width of present was associated with the constitution of “local” temporal continuities of the neural activity in different regions. This raises the question how the different regions’ local temporal continuities are linked and connected to each other and ultimately integrated into a more “global” temporal continuity. I here discuss findings that show each region to display a specific and idiosyncratic temporal pattern of neural activity. This will be complemented by showing that loss of consciousness is associated with decreased linkage and integration between high- and low-frequency fluctuations; that is, temporal nestedness. This leads me to hypothesize that consciousness is directly related to the degree of temporal nestedness between high- and low-frequency fluctuations. How is temporal nestedness constituted? I propose that the temporal nestedness may directly depend on difference-based coding. The larger the temporal differences encoded into neural activity via difference-based coding, the higher the possible degrees of both temporal nestedness and global temporal continuity of neural activity. And that in turn leads to increased degrees of consciousness. How is the global temporal continuity manifested in experience and thus in consciousness on the phenomenal level? I propose that the global temporal continuity of neural activity predisposes on the phenomenal level what phenomenological philosopher E. Husserl called “duration bloc.” The concept of the duration bloc describes the integration of past, present, and future into one homogenous (though threefold) experience of time in consciousness. I hypothesize that the “global” temporal continuity of neural activity predisposes and makes possible such extension of the present into past and future on the phenomenal level of consciousness. By temporally structuring and organizing neural activity in a particular way, the global temporal continuity of the brain’s neural activity makes necessary and unavoidable the constitution of the threefold temporal structure with past, present, and future, the duration bloc, on the phenomenal level of consciousness.

Key Concepts and Topics Covered

Temporal pattern of neural activity, low- frequency fluctuations, NREM sleep, resting state, slow wave activity, slow cortical potentials, temporal nestedness, difference-based coding, duration bloc, global temporal continuity

Neuroempirical Background IA: Region-Specific Temporal Patterns and “Local Temporal Continuities”

Chapter 14 discussed slow cortical potentials and how they are related to the constitution of local temporal continuity of neural activity. Thereby, the concept of local temporal continuity referred to the integration of the different discrete time points of different stimuli in the neural activity of one particular region; that is, especially in its superficial layers 1 and 2 as manifested in slow cortical potentials (SCP) (see Chapter 14).

One may now propose the same process of temporal integration to occur in the various regions of the brain. Do the different regions’ neural activities show the same local temporal continuities in their neural activities or different ones? This will be the focus of the present section. And one wants to know how that is related to the experience of time and thus to consciousness; (p.49) that is, “inner time consciousness.” This will be discussed in the subsequent sections.

Let us start with the empirical data. Different regions receive different inputs and stimuli that differ in their respective statistical frequency distribution across different discrete points in physical time. As shown in Chapter 14, the constitution of local temporal continuity of neural activity is supposed to be based on the encoding of the stimuli’s “temporal statistics”; that is, their statistical frequency distribution across different discrete points in physical time. Due to their different inputs and stimuli, one may consequently propose that the different regions show, not only different degrees of neural activity, but also different degrees of “local” temporal continuities. Different regions may thus show a different temporal pattern in their neural activities. This seems to be indeed the case, as is proposed by Bartels and Zeki (2004, 2005).

Bartels and Zeki (2004) investigated the visual cortex during conventional and natural stimulation in fMRI. Applying a model-free data-driven, that is, independent component analysis to their fMRI data, they observed that distinct subregions in the visual cortex show distinct time courses. For instance, the primary visual cortex (V1 and V2) showed negative signal changes during the stimulus period while returning afterward back to a high resting-state activity level. In contrast, signal changes and thus the waveform in V5 were very different from the one in V1/V2. Unlike V1/2, V5 exhibited lower resting-state activity and higher stimulus-induced activity.

Most important, the waveforms and thus the activity time curves (ATCs) were specific for each area/region, thereby distinguishing them in temporal regard. Furthermore, the time activity curves for each area were consistent across subjects. This is evidenced by the fact that the different subjects’ ATCs in the same; that is, corresponding areas/regions highly correlated with each other. In contrast, the correlation of the ATCs between different areas/regions within the same subject was much lower than the correlation between the same regions across subjects. Finally, no correlation at all could be observed between different regions’ ATCs from different subjects.

Taken together, this suggests area- or region-specific time curves that hold specifically for one particular region across different subjects. Moreover, since the correlation between the different regions within the same subjects was rather low when compared to the one of the same region across different subjects, Bartels and Zeki (2004) propose functional independence between the different regions in at least temporal regard.

Neuroempirical Background IB: From “Local” to “Global” Temporal Continuities

The different regions of the brain do thus seem to have their specific temporal pattern of neuronal activity, which may distinguish them from each other. When conducted for the whole brain, this may ultimately generate what Bartels and Zeki (2004) call “chronoarchtitectonic maps.” A chronoarchitectonic map is a time-based map of the brain’s neural activity that illustrates the different temporal patterns in the different regions’ neuronal activities.

What does the assumption of such region- specific different temporal patterns of neural activity imply for the constitution of local temporal continuity? Following the assumption of difference-based coding, different temporal patterns of neural activity are supposed to reflect the encoding of different degrees of temporal differences into the neural activities in the different regions. The temporal differences and consequently the respective stimuli’s temporal statistics as the input to the different regions should therefore differ in the different regions. This implies that the resulting “local temporal continuities” for each region’s neural activity should also differ from region to region.

More specifically, the degree of temporal extension of neural activity should be different between the different regions. Regions that encode larger temporal differences between their predominant stimulus’ inputs may show a more extended “local” temporal continuity. (p.50) In contrast, local temporal continuity of neural activity may be less extended in those regions where the temporal differences between the predominant stimulus’ inputs are shorter.

How are the different local temporal continuities from the different regions linked and connected to each other and how that is related to consciousness? For that, I now turn to sleep, specifically NREM sleep, and its slow wave activity.

Neuronal Findings IA: Slow Wave Activity in Non–Rapid Eye Movement Sleep

Sleep has been much investigated and is often considered a paradigmatic example of unconsciousness. One has to differentiate between different phases of sleep, however. There are early and late sleep stages where one cannot find rapid eye movements (REM), which led to the characterization of these phases as non-REM (NREM). The NREM sleep has to be distinguished from stages (in the middle of the night) with strong rapid eye movements (REM). Both NREM and REM have also been distinguished on phenomenal and electrophysiological grounds. (I here follow the traditional and broadly known classification with the distinction between REM and NREM sleep rather than adhering the most recent re-classification of the different sleep stages, which is known more to insiders at this point in time.)

The REM sleep has traditionally been associated with dreams (see Chapters 25 and 26 for details on dreams), while the NREM sleep is usually characterized by the absence of dreams, but this theory has been questioned more recently (see Nir and Tononi 2010). I here focus on NREM sleep and how it is distinguished from REM sleep; dreams will be investigated separately in Chapter 26.

How can we characterize NREM sleep? Electrophysiologically, NREM sleep can be distinguished by two particular features from REM sleep: NREM sleep shows slow wave activity (SWA) that is characterized by slow oscillations (<1 Hz) that usually last for around a second (see also Riedner et al. 2011 for an overview). These slow oscillations are supposed to be related to synchronization of the majority of cortical neurons, which oscillate between a depolarized upstate and a hyperpolarized downstate. In addition to the slow oscillations, one can also observe spindles that peak at 13–14 Hz and can be considered the second electrophysiological hallmark feature of NREM sleep.

Neuronal Findings IB: Slow Wave Activity (SWA) Versus Slow Cortical Potentials (SCP)

How are these SWA in NREM sleep related to the slow cortical potentials (SCPs) discussed in Chapter 14? SWA are spontaneous rhythmic oscillations of the membrane potential between a hyperpolarized downstate and a hypopolarized upstate. In addition to NREM sleep, SWA can also occur during anesthesia, where they are supposed to reflect the absence of specific attentional and other cognitive functions.

Are the SWA identical to the SCPs? He et al. (2008 supplementary material, note 3; see also He and Raichle 2009) deny that and distinguish between SWA and SCP for various reasons. First, their frequency ranges differ: SCPs show a large frequency spectrum ranging from 0.3 Hz to 4 Hz, while the one of SWA is rather narrow, centering around 0.8 Hz. Second, SWA are observed only during NREM sleep but neither in REM sleep nor in the awake state. This distinguishes them from SCPs, which, as described in Chapter 14, occur during all three, REM and NREM sleep and in the awake state (see also He et al. 2008; van Someren et al. 2011; Riedner et al. 2011; and Mascetti et al. 2011).

Third, the distribution of the SWA seems to be more or less independent of the underlying functional anatomy in that they seem to occur throughout the whole brain. In contrast, the SCPs are closely related to especially the midline regions of the default-mode network as discussed in Chapter 14 (see He and Raichle 2009). Fourth, SCPs may modulate the SWA so that the latter (p.51) may be considered to be dependent upon the former.

Finally, in the preceding chapters we spoke of “fluctuations,” whereas now we use the term “oscillation.” What is the difference between fluctuations and oscillations? He and Raichle (2009) make a principal distinction between fluctuations and oscillations: oscillations describe rhythmic activity in EEG that centers on a specific frequency. This is the case in SWA that describe oscillatory activity centering on 0.8 Hz.

SCPs, in contrast, do not describe such oscillatory activity. Instead, they reflect fluctuations of neuronal activity that are not yet rhythmic but still arrhythmic. Hence, to equate SWA with SCPs would be to confuse rhythmic and arrhythmic neural activity and thus oscillations and fluctuations. Following this distinction, I here focus on SWA as oscillations, while the preceding chapter targeted fluctuations when discussing SCPs. (However, I will follow this distinction only rather loosely in this and the following chapters where I speak predominantly of fluctuations.)

Neuronal Findings IC: Slow Wave Activity and Midline Regions

Let us go back to the NREM sleep and its electrophysiological patterns. G. Tononi is a researcher from Italy. After having studied with Gerald Edelman and his theory of re-entrant connections (see Introduction I in Volume II), Tononi widened and extended that approach to information integration as a neural correlate of consciousness (see Appendix 1 for a detailed discussion of his information integration theory). In that context he is also very much interested in sleep and its loss of consciousness, especially during NREM sleep.

Tononi and his group applied 256-channel high-density electroencephalography (EEG). This use of such special EEG allows for high spatial resolution with the determination of the spatial and thus anatomical location of the signal, thereby complementing the high temporal resolution of the EEG. Concerning sleep, this makes it possible to localize the origin of SWA and to investigate their spread and distribution across the rest of the brain; that is, their traveling waves (see Tononi 2009; Massimini et al. 2009, 2012, Nir et al. 2011; Riedner et al. 2011; Mascetti et al. 2011).

The data from the group around Tononi (see also Nir et al. 2011; Riedner et al. 2011; Mascetti et al. 2011) show the predominantly local origin of SWA: large currents of SWA (around 0.8 Hz, range between 0.3 and 6 Hz) appeared predominantly in the midline structures, including the anterior cingulate cortex, the posterior cingulate cortex, and the precuneus. From there the SWA seem to propagate preferentially to medial temporal regions, including the hippocampus (see Nir et al. 2011 for details). Hence, the midline regions seem to have a special role in constituting and processing SWA, in particular, and the low-frequency fluctuations in general (see Fig. 15-1).

What do these findings tell us? They show the regions that are implicated in generating the SWA. In contrast, the findings do not reveal themselves the kind of neuronal processes involved. How are SWA generated? They are generated locally and seem to reflect predominantly synaptic strength and thus local synaptic changes as shown in a combination of electrophysiological and simulation experiments (see Tononi 2009; Nir et al. 2011; Riedner et al. 2011). Let me specify this in the following.

Taken all these findings together as obtained in simulation models, rat’s electrophysiological recordings, and human EEG let the authors suggest what they call “synaptic homeostasis hypothesis” (see Tononi 2009 as well as Tononi and Cirelli, 2003). The “synaptic homeostasis hypothesis” proposes that the SWA in NREM sleep may reflect the local synaptic strength and its decrease in sleep. Such synaptic decrease may serve the more general functional purpose of recalibrating neuronal circuits during sleep by desaturating them. This may prepare the neural circuits well for novel saturation in the subsequent awake state (see also Chapter 16 for more extensive discussion of physiological mechanisms).

(p.52)

Temporal Nestedness and “Duration Bloc”Temporal Nestedness and “Duration Bloc”

Figure 15-1 Temporal pattern of neural activity during the absence of consciousness.

(a) Example of EEG and single-unit activity during global sleep slow waves. Example of EEG and unit activities in multiple brain regions during 11.5 s of deep NREM sleep in one individual. Rows (top to bottom) depict activity in scalp EEG (Cz), right supplementary motor area (R-SMA), left entorhinal cortex (L-EC), right entorhinal cortex (R-EC), left hippocampus (L-HC), and left amygdala (L-Am). Horizontal line in dark gray, scalp EEG; horizontal line in black, depth EEG; Vertical lines, unit spikes. Rounds dots in gray show individual slow waves detected automatically in each channel separately. Gray and white vertical bars through out the whole figure mark ON and OFF periods occurring in unison across multiple brain regions. (b) Sleep slow waves propagate across typical paths. (A) Left: Average-depth EEG slow waves in different brain structures of one individual illustrate propagation from frontal cortex to MTL. All slow waves are triggered by scalp EEG negativity. Black, scalp mean waveform. Right: Distributions of time lags for individual waves in supplementary motor area (SM) and hippocampus (HC) relative to scalp. (B) Mean position in sequences of propagating waves in all 129 electrodes across 13 individuals. Each circle denotes one depth electrode according to its precise anatomical location.. (C) Quantitative analysis: mean position in propagation sequences as a function of brain region. Abbreviations: SM, supplementary motor area; PC, posterior cingulate; OF, orbitofrontal cortex; AC, anterior cingulate; ST, superior temporal gyrus; EC, entorhinal cortex; Am, amygdala; HC, hippocampus; PH, parahippocampal gyrus. (D) An example of individual slow waves propagating from frontal cortex to MTL. Rows (top to bottom) depict activity in scalp EEG (Cz), supplementary motor area (SM), entorhinal cortex (EC), hippocampus (HC), and amygdala (Am).The dots mark center of OFF periods in each brain region based on the middle of silent intervals as defined by last and first spikes across the local population. Diagonal lines are fitted to OFF period times via linear regression and illustrate propagation trend. (E) Left: The average unit activity in frontal cortex (top, n = 76) and MTL (bottom, n = 155), triggered by the same scalp slow waves reveals a robust time delay (illustrated by vertical arrow). Right: Distribution of time delays in individual frontal (top) and MTL (bottom) units reveals a time delay of 187 ms. Red vertical arrows denote mean time offset relative to scalp EEG.

Reprinted with permission of Cell Press, from Nir Y, Staba RJ, Andrillon T, Vyazovskiy VV, Cirelli C, Fried I, Tononi G. Regional slow waves and spindles in human sleep. Neuron. 2011 Apr 14;70(1):153–69.

(p.53)

Neuronal Hypothesis IA: “Temporal Nestedness” Mediates the Level/State of Consciousness

What do these findings entail for the present and absence of consciousness? First and foremost, the SWA is by itself not sufficient to induce consciousness. Otherwise, there would be no loss of consciousness in NREM sleep that is characterized by predominance of SWA. This is further supported by the occurrence of SWA in another nonconscious state; namely, anesthesia, which, like NREM sleep, is also characterized by loss of consciousness.

How is it possible that we lose consciousness in NREM sleep (and anesthesia) despite the present of SWA? Let us recount. SWA describe oscillations in the lower frequency range centered around 0.8 Hz, which predominate during NREM sleep. In contrast, other higher frequency oscillations are rather rare in NREM sleep, with the exception of the aforementioned sleep spindles (12–13 Hz) and some slow waves with multiple negative peaks in the upper delta frequency range (2–4 Hz); that is, delta waves.

How do these waves, the spindles and the multiple negative peaks, occur? They seem to be closely related to the occurrence of SWA and may be quasi-nested in them. More specifically, Tononi (2009) proposes that the delta waves are generated on the basis of asynchronous SWA with different regional origins and/or different transregional propagation (see also Nir et al. 2011; Riedner et al. 2011; van Someren et al. 2011; Mascetti et al. 2011). If so, the delta (p.54) waves simply represent an overlay or nesting of different SWAs and thus of multiple slow oscillations with distinct spatial and temporal features. In short, the delta waves may result from the temporal nesting of higher in lower frequency oscillations.

Interestingly, an almost analogous temporal nesting of different frequency waves was already described in the previous section on slow cortical potentials (SCP). Here too, based on He et al. (2010), higher frequency oscillations were proposed to be nested within lower frequency oscillations and ultimately within slow wave fluctuations like the SCP.

What then is the main difference between NREM sleep and the awake state and thus between absence and presence of consciousness? I hypothesize that the difference may consist in the degree of temporal nestedness between different frequency ranges. Higher degrees of temporal nestedness between high- and low-frequency fluctuations may go along with a higher degree in the level or state of consciousness as in the awake state. Conversely, lower degrees of temporal nestedness should lead to lower degrees in the level or state of consciousness as in NREM sleep.

Neuronal Hypothesis IB: Consciousness is like the Russian Dolls

What exactly do I mean by “temporal nestedness”? The concept of temporal nestedness describes the relationship between high- and low-frequency fluctuations in the neural activity of the brain. It is important to note that temporal nestedness goes beyond mere co-occurrence of high- and low-frequency fluctuations.

In addition to the presence of both high- and low-frequency fluctuations, they also need to be directly connected and linked, that is, integrated. The smaller time windows of the high-frequency fluctuations should be integrated and situated; that is, nested, within the longer phase durations of the low-frequency fluctuations. Such temporal nesting is supposed to be mediated, in part, by the temporal alignment of low frequency fluctuations’ phase onsets to the high frequency fluctuations, including their phase onsets and power: their cross-frequency phase-phase and phase-power coupling (see Chapter 5 as well as Chapters 19 and 20 for details).

How can we better illustrate such temporal nestedness? Let us compare the difference between temporal nestedness and temporal co-occurrence to the well-known Russian dolls. Usually, there is one large Russian doll, which, if we open its head, contains another slightly smaller Russian doll, and so forth. The Russian dolls are thus nested within each other. If one puts different Russian dolls of different sizes on the table, they can be said to merely co-occur, while they are not nested into each other.

The same is the case in the low- and high-frequency fluctuations. If the high-frequency fluctuations are integrated in the longer phases of the lower frequency fluctuations, one can speak of temporal nestedness. This corresponds to the sorting the Russian dolls according to their size and ultimately putting them all together into one big doll. If, in contrast, there is no such integration, high- and low-frequency fluctuations stand only side by side, just like the different Russian dolls of different sizes lying beside each other on the table.

Based on these considerations, I suggest the following. I hypothesize that the difference in the degree of temporal nestedness between different frequency ranges predisposes the absence or presence of consciousness: the higher the degree of temporal nestedness between different—lower and higher—frequency ranges, the more likely consciousness is to occur and thus to be present. This parallels the case when we find 20 Russian dolls within one big one (see Fig. 15-2a).

Conversely, the lower the degree of temporal nestedness between different—lower and higher frequency—ranges, the more likely consciousness will remain absent. This parallels the situation when there are only two Russian dolls within one big one, or all three lying side by side on the table. That may, for instance, be the case in NREM sleep, anesthesia, and vegetative state, where consciousness remains absent.

Metaphorically speaking, it is the number of dolls and how they are linked that ultimately (p.55)

Temporal Nestedness and “Duration Bloc”

Figure 15-2a and b Temporal nestedness and consciousness. The figure displays the relationship between the fluctuations of neural activity and the degree of consciousness. (a) The figure shows the dependence of the degree of the state or level of consciousness on the degree of temporal nestedness between different frequency ranges in the fluctuations of the neural activity. The better the high and thus shorter frequency fluctuations are linked and connected and thus nested into the longer phases of the low-frequency fluctuations, the higher the degree of the level or state of consciousness that can possibly be constituted. Hence, a lower degree of such temporal nestedness will then go along with low degrees in the level or state of consciousness, as in vegetative state, anesthesia, and NREM sleep. (b) The figure shows the dependence of the degree of the state or level of consciousness on the degree of temporal differences as they are encoded between the different frequency ranges in the fluctuations of the neural activity. The higher the degree or range of the temporal differences between the different frequency ranges in the neural activity fluctuations that can possibly be encoded into neural activity, the higher the possible degree of the level or state of consciousness that can be constituted. Hence, lower degrees or ranges of encoded temporal differences should go along with lower degrees of the level or state of consciousness as in vegetative state, anesthesia, and NREM sleep.

predisposes whether consciousness will be present or remain absent. Accordingly, the Russian dolls are not just traditional symbols of Russian culture, but also a wonderful (metaphorical) symbol of consciousness itself and its particular temporal (and spatial) structure.

Neuronal Hypothesis IC: Difference-Versus Stimulus-Based Coding

Now let us go back to NREM sleep and consciousness. As described earlier, NREM sleep can be characterized by the predominance of (p.56) local and asynchronous out-of-phase SWA and the concurrent absence of higher frequency waves in its resting-state activity; that is, in the absence of specific extrinsic stimuli (see also Nir at al. 2011). In contrast, higher frequency waves are well present and very abundant in the awake state during its resting-state activity. Accordingly, the resting-state activity’s degree of temporal nestedness between different frequency ranges may be much higher in the awake state when compared to NREM sleep.

Due to its different resting-state activity pattern, neural activity during NREM sleep may no longer be able to properly process extrinsic stimuli. A 2011 fMRI-EEG investigation by Dang-Vu et al. (2011) demonstrated less consistent neuronal responses in auditory cortex and thalamus during auditory stimulation in NREM sleep. The neural processing of the auditory stimulus was severely hampered by the slow wave oscillations and the spindles typically occurring in the NREM sleep.

Accordingly, the ongoing rather slow resting-state activity in NREM seems to prevent the auditory stimulus from being properly processed in the brain. This may make it impossible for the stimulus to become linked and integrated into the brain’s ongoing intrinsic activity; that in turn, could be crucial for the association of the resulting stimulus-induced activity with consciousness (see Chapters 11 and 29 for the detailed neuronal and neurophenomenal mechanisms of such rest–stimulus interaction).

What does this mean exactly? The auditory stimulus may not be properly processed in the brain anymore because it cannot be linked and connected to the brain’s intrinsic activity. Why? The brain and its intrinsic activity are busy with other things; the SWA, and, metaphorically speaking, “have no time to take care of the extrinsic stimulus.” More specifically, the auditory stimulus cannot be encoded into neural activity relative to the resting-state activity level. The degree of difference-based coding of the auditory stimulus will consequently be rather low which, as we have seen in Volume I (see Chapter 1), goes along with a high degree of stimulus-based coding. Even if the auditory stimulus induces some activity changes in the brain, these may only be related to the stimulus itself while remaining independent of the brain’s intrinsic activity.

Why does the resulting stimulus-induced activity remain more or less independent of the brain’s intrinsic activity? The stimulus is no longer encoded into neural activity in relative difference to the intrinsic activity but is rather independent of it in a more stimulus-based way. Accordingly, the neural processing of the stimulus may then be characterized by a low degree of difference-based coding and a rather high degree of stimulus-based coding.

Neuronal Hypothesis ID: Difference-Based Coding Mediates the Level/State of Consciousness

This leads me to the following hypothesis. I propose the hypothesized relationship between temporal nestedness and consciousness to be mediated by difference-based coding. The encoding of temporal differences between high- and low-frequency fluctuations allows for integration of the different frequency ranges and thus for their temporal nesting. Higher degrees of difference-based coding should thus go along with higher degrees of temporal nestedness between high- and low-frequency fluctuations and consequently with higher degrees in the level or state of consciousness.

In contrast, lower degrees of temporal nestedness (and consequently higher degrees of mere temporal co-occurrence) may signify a high degree of stimulus-based coding. The balance between difference- and stimulus-based coding is here shifted toward the latter, which in turn decreases the likelihood that the resulting neural activity is associated with a high level or state of consciousness. That may, for instance, be the case in NREM sleep, anesthesia, and vegetative state, with all three supposedly showing a high degree of stimulus-based coding (when compared to the degree of difference-based coding) and rather low, if not absent, level or state of consciousness (see Chapters 28 and 29 for details).

This amounts to the following relationship between difference-based coding and the state or level of consciousness: the higher the degree (p.57) of difference-based coding (and the lower the degree of stimulus-based coding) in the neural coding between high- and low-frequency fluctuations, the higher the degree of the level or state of consciousness. Accordingly, I propose the degree of consciousness to be directly related to the degree of the encoding of temporal differences into neural activity; that is, difference-based coding.

Conversely, low degrees of difference-based coding and consequently high degrees of stimulus-based coding should decrease the likelihood of high degrees in the level or state of consciousness. In the extreme case of abnormally high degrees of stimulus-based coding, one would expect loss of consciousness, which is exactly what one observes in NREM sleep, anesthesia, and vegetative state (see Fig. 15-2b; and see Chapters 28 and 29 for details).

Neuronal Hypothesis IIA: Difference-Based Coding and Temporal Nestedness

Why do I propose that the degree of temporal nestedness between the different frequency ranges predisposes consciousness? Let me first detail what temporal nestedness implies in neuronal terms. During the resting-state activity, temporal nestedness between different frequency ranges may be traced back to the neural overlay of extrinsic stimulus-triggered high-frequency fluctuations onto the intrinsic low-frequency fluctuations (see Chapter 5 in Volume I for details as well as Part VI in Volume II).

This means that the resulting higher frequency wave must really be considered the product of a temporal difference: the temporal difference between the phase onset and duration of the low-frequency fluctuations on the one hand and the discrete point in physical time associated with the stimulus on the other. The encoding of the temporal difference between the intrinsic activity’s phase onset and the extrinsic stimulus’ discrete time point is possible, however, only on the basis of difference-based coding, whereas it remains impossible in the case of stimulus-based coding.

I consequently hypothesize that the temporal nestedness of different frequency waves may directly depend on the degree of temporal differences encoded into neural activity via difference-based coding: The more fine-grained the temporal differences are encoded into neural activity via difference-based coding, the higher degrees of temporal nestedness between different ranges of frequency fluctuations can be constituted in the resulting neural activity.

Consider again the analogous example of the Russian dolls. The smaller the differences in size between the different dolls, the more dolls that can be fitted within the largest one, resulting in a higher degree of nestedness. The same now applies to the degree of temporal differences between the different ranges of frequency fluctuations as they are encoded into neural activity. As in the case of the Russian dolls, the encoding of more fine-grained and thus smaller temporal differences results in higher degrees of temporal nestedness.

Neuronal Hypothesis IIB: Encoding of Temporal Differences by the Intrinsic Activity’s Low-Frequency Fluctuations

In addition to the different degrees of temporal differences between different ranges of frequency fluctuations, the brain is also confronted with different degrees of temporal differences as encoded in the different regions’ intrinsic activities. We recall the findings from Bartels and Zeki, who showed that each region has its specific temporal pattern of neural activity.

This indicates that the degree of temporal differences encoded into the regions’ neural activities must differ between the different regions. Some regions may predominantly encode larger temporal difference, while other regions, based on the temporal statistics of their predominant stimulus input, may encode smaller and more fine-grained temporal differences (see earlier for details).

How are the different regions’ different temporal activity patterns and thus their different local temporal continuities linked and connected to each other? For that, the fluctuations in neural activity may be central. The fluctuations in neural activity, especially the low-frequency (p.58) fluctuations, operate across different regions and their respective local temporal continuities.

If now the fluctuations show a broad frequency range with many intermediate frequency ranges, they will be well able to link and connect a higher multitude of local regional temporal differences and thus different local temporal continuities. I consequently propose a broad frequency range and a high variability in the frequency range to be central for the integration of neural activities in different regions (see also Garrett et al. 2011; as well as McDonnell and Ward 2011, for the relevance of variability).

This leads me to the following hypothesis. I propose that the degree to which the different regions’ temporal differences can be linked and connected to each other depends very much on the degree of the range in the fluctuations’ frequencies. A higher range of the fluctuations’ frequencies implies a large difference between the highest and lowest frequencies with many intermediate frequency ranges.

That makes it more likely that the different regions’ temporal differences can be matched and thus be connected to each other. The diverse “local” temporal continuities can consequently be integrated and nested into one “global” temporal continuity. The concept of global temporal continuity describes the linkage, integration, and ultimately synchronization between the local temporal activities of the different regions.

Neuronal Hypothesis IIC: “Global” Temporal Continuity of Neural Activity Mediates the Level/State of Consciousness

“Global” temporal continuity in this sense comes close to what is described as the “global neuronal workspace” in the current neuroscience literature on consciousness (see Baars 2005; Dehaene and Changeux 2011), though specified in temporal regard (see Chapters 18 and 19 and also Appendix 1 for a detailed discussion of the concept of the global neuronal workspace in relation to my neurophenomenal account).

Why is the global temporal continuity important? The global temporal continuity may predispose the association of the resulting neural activity with consciousness, as is the case in the awake state and, to some degree, also during dreams in REM sleep. The converse case is the one when the range of the fluctuations’ frequencies is rather low. This means that the difference between the highest and lowest frequencies is rather small and/or that not many intermediate frequency ranges are present. Such a lower frequency range is less likely to be able to link the different regions’ different temporal differences to each other. (see Fig. 15-3a).

The different regions’ local temporal continuities may therefore no longer be well integrated and thus nested into each other so that the degree of the resulting global temporal continuity is rather low. That in turn decreases the likelihood of associating the respective neural activity with a high level or state of consciousness, as can indeed be observed in NREM sleep, anesthesia, and vegetative state (see also Chapter 16 as well as Chapters 28 and 29 for more detail). This is well in accordance with the observation of decreased spatial and temporal spread and propagation of externally induced neural activity changes these three states, as will be described in further detail in Chapter 16 (for NREM sleep and anesthesia) and Chapters 28 and 29 (for vegetative state) (see Fig. 15-3b).

In sum, I propose the number and the degree of temporal differences that are encoded into neural activity via difference-based coding to predict the degree to which the different regions’ “local” temporal continuities are extended into a more “global” temporal continuity. That, in turn, may predispose the possible degree of the state or level of consciousness: higher degrees of global continuity of neural activity make more likely the association of a higher degree of the level or state of consciousness. Accordingly, larger degrees of temporal differences during the encoding of neural activity predispose a higher degree of global temporal continuity and consequently a higher level or state of consciousness.

Phenomenological Excursion IA: “Inner Time Consciousness” and “Duration Bloc”

The question now is how such “global” temporal continuity of the brain’s neural activity is manifested in our experience and thus in consciousness. Recall from the previous chapter that the (p.59)

Temporal Nestedness and “Duration Bloc”

Figure 15-3a and b Temporal nestedness and “global” temporal continuity. The figure depicts the relationship between temporal nestedness of different frequency ranges and the degree of “global” temporal continuity of the brain’s neural activity. (a) The figure shows the dependence of the degree of temporal extension of local temporal continuities into a more “global” temporal continuity on the degree of frequency ranges in the fluctuations of the brain’s neural activity. The higher the degree of frequency ranges (i.e., differences between highest and lowest) and the number of intermediate frequency ranges, the more the “local” temporal continuities (from the different regions/networks of the brain) can possibly be integrated and extended into a more “global” temporal continuity that spans across the whole brain and all its regions’ neural activities. I propose the range of frequency fluctuations to be rather low in vegetative state, anesthesia, and NREM sleep (see Chapters 28 and 29 for empirical support), which consequently go along with lower degrees of temporal extension from “local” temporal continuities to a more “global” temporal continuity in the brain’s neural activity. The extension from “local” to “global” temporal continuity of neural activity is supposed to be mediated by difference-based coding; for example, by encoding the temporal differences between the regions’ different inputs (bars in uppermost line on the time arrow and second-highest line as well as right upper part). (b) The figure demonstrates how various “local” temporal continuities in particular regions (upper and middle upper part) are integrated by connecting high- and low-frequency fluctuations to each other, resulting in temporal nestedness with “global” temporal continuity of neural activity and, ultimately, consciousness (middle and lower part). This is supposed to be mediated by difference-based coding; for example, by encoding the temporal differences between different stimuli (right upper part) and the different frequency ranges into the resulting neural activity (right lower part).

(p.60) local temporal continuity was associated phenomenally with what phenomenological philosopher E. Husserl described as the “width of the present” (see Chapter 14). The question now is how what I described as the “global” temporal continuity is manifested in “inner time consciousness.” For the answer, I again turn to Husserl.

Husserl argues that the width of the present can be stretched and extended deeply into both past and future (we remember, for instance, the imaging findings on “prospection” and retrospection as described in Chapter 13). Such extension and stretching of the width of the present into the future and past may result in what Husserl described as “duration bloc.” The duration bloc comprises and interconnects previous, present, and next moments, reflecting the three temporal modes of past, present, and future (Husserl 1991, 23, 113–114). Husserl calls these three temporal modes of the duration bloc “primal presentation,” “protention,” and “retention” which shall be described briefly in the following explanation.

Let us start with the “primal presentation.” The duration bloc includes the right-now moment, the moment the object appears or when the tone is actually played—this may be called “primal presentation” (Zahavi 2005, 56). The here and now of the primal presentation is, however, not abstracted and isolated from the previous and next here-and-now moments, which are built into and thus enclosed in the current right-now moment.

How is such integration possible? In addition to primal presentation, there is a second element in the duration bloc: namely, retention. Retention is the component that provides us with conscious access to the just-elapsed phase of the preceding object, the previous moment that occurred just before the current right-now moment. The preceding object is retained and can therefore be carried over to the current object and thus be enclosed in the experience of the current right-now moment object. This comes close to what we, relying on William James, described as “sensible continuity” in Chapter 13.

Due to retention (and “sensible continuity”), we hear the current tone in relation to the previous one, with this temporal connection between past and present tones enabling us to decipher both previous and current tones as part of a melody. The previous tone is carried over to the current one. Without such retention both tones could not be connected in consciousness, which in turn would make experience of the tones as a melody impossible.1

In addition to retention of previous tones, we also anticipate the next tone, which enables us to complement the melody, even if not all tones are actually played. This leads to the third element of the duration bloc: protention (Husserl 1991). Listening to a melody, we often anticipate; we expect a particular tone and will be surprised if the anticipated tone does not match with the actually occurring next tone. Similar to retention, “protention” is connected to the current object while at the same time extending it to the beyond the actually occurring right-now moment to the next not-yet occurred moment.

Phenomenological Excursion IB: Threefold Temporal Structure and Mutual Modulation

Let me now shed a more detailed light on how the three elements, primal presentation, retention, and protention, are related to each other. We listen to the actually occurring tone within the context of the potentially occurring (anticipated) next tone. This enables us to decipher the present tone as part of a continuous melody extending from the past, over the present, to the future. If, for example, we anticipate the tone E after C, we listen to C in relation to the (potential) tone E.

If we anticipate another C rather than E, we would listen to the present C in a completely different way than when we were anticipating E or F. The primal presentation is thus strongly impacted, modulated, and changed by protention. The anticipated tone seems to have some feedback effect (as one may call it) upon the present tone in our perception. And the same holds for retention. There is thus not only connection but mutual modulation between primal presentation, retention, and protention. The concept of mutual modulation describes that all three, presentation, retention, and protention are (p.61) interdependent on each other with the constitution of the one being intrinsically related to the respective others and vice versa.

Let us describe such mutual modulation in more detail. Go one step further and imagine you anticipated E and now finally the real tone kicks in. If it is indeed the tone E, you are happy and continue singing the melody.

If, in contrast, it is not the tone E but another one, there are two options. Either the next tone is not E but F, which may continue the melody, but in an unexpected way. You are then surprised and are probably unable to sing the melody completely, but your attention is nevertheless caught by the unexpected turn of the melody. Alternatively, the next tone may be C, which does not continue the melody at all, but rather disrupts and terminates it completely. You may be disappointed and turn your attention away from the tones and stop listening altogether.

Taken together, this results in what Husserl described as the threefold structure of our experience of objects in time, including primal presentation, retention, and protention:

In this way, it becomes evident that concrete perception as original consciousness (original givenness) of a temporally extended object is structured internally as itself a streaming system of momentary perceptions (so-called primal impressions). But each such momentary perception is the nuclear phase of a continuity, a continuity of momentary graded retentions on the one side, and a horizon of what is coming on the other side: a horizon of “protention,” which is disclosed to be characterized as a constantly gradated coming. (Husserl 1977, 202)

Neurophenomenal Hypothesis IA: “Global” Temporal Continuity of Neural Activity Mediates the “Duration Bloc” in Consciousness

How, then, is what Husserl describes as “duration bloc” on the phenomenal level of consciousness related to the neuronal processes in the brain? First, I propose that the duration bloc corresponds well to what I described earlier as “global” temporal continuity of neural activity. The duration bloc describes the continuity between past, present, and future on the phenomenal level of consciousness.

How does that relate to the global temporal continuity of neural activity? The concept of the global temporal continuity concerns the integration of different local temporal continuities into one more general and thus “global” temporal continuity. As such, the “global” temporal continuity is supposed to span the different regions’ local temporal continuities in their neural activities, including their respective different temporal differences.

By integrating different “local” temporal continuities, the “global” temporal continuity links and connects different temporal differences. The more diverse temporal differences are linked and connected, the more the resulting “global” temporal continuity can extend across different discrete points in physical time from the present into both past and future discrete points in physical time. In short, higher degrees of “global” temporal continuity lead to higher degrees of temporal extension of neural activity.

How is that related to the “threefold temporal structure” and thus the “duration bloc” on the phenomenal level of consciousness? The higher the degree of temporal extension of the “global” neural activity from present into past and future, the more and better the threefold structure of time, with present, past, and future, can be constituted and therefore comes close to what Husserl described as a “duration bloc” (see Fig. 15-4a).

Based on these considerations, I propose that the degree of “global” temporal continuity of neural activity predisposes the degree of the “duration bloc” on the phenomenal level of consciousness: the higher degree of “global” temporal continuity in the brain’s neural activity, the higher the possible degree of the “duration bloc” on the phenomenal level of consciousness. If, in contrast, the degree of “global” temporal continuity is rather low, the degree of temporal extension of the “duration bloc” will abnormally shrink, with a more limited time range between past and future (see Fig. 15-4b).

We have seen earlier that the “global” temporal continuity of the brain’s neural activity ultimately depends on the degree of the (p.62)

Temporal Nestedness and “Duration Bloc”Temporal Nestedness and “Duration Bloc”

Figure 15-4a-d Neural predispositions of the “duration bloc.” The figure displays how different neuronal mechanisms (a, b, c, d) predispose the degree of the temporal extension of the duration bloc from the present into the past and future. (a) The figure shows the different stages from the extrinsic stimuli’s occurrence (upper level) via their encoding into the brain’s intrinsic activity in terms of temporal differences (second from upper level) and the “width of present” (second from lower level) to the constitution of the “duration bloc” (lower level). The most important step is here the encoding of the different stimuli’s discrete points in physical time in terms of temporal differences by the brain’s intrinsic activity and its low-frequency fluctuations. That makes it possible to extend or “stretch” the single discrete point in time beyond itself, as indicated in the “width of present,” which corresponds to the regional activity in the brain. The overlay of the different regional activities and their respectively associated “width of present” leads to the “duration bloc”; the concept of “duration” describes temporally homogenous stretches of neural activity where it does not change, which corresponds to phase durations that are not interrupted either by other frequencies or stimuli (see horizontal lines in the lower part). (b) The figure shows the dependence of the degree of the temporal extension of the duration bloc from the present into the past and future on the degree of the “global” temporal continuity of the brain’s neural activity. The higher the degree and the larger the extension of the “global” temporal continuity of the brain’s neural activity, the higher the number of past and future discrete time points covered by the neural activity, and the larger the possible extension of the duration bloc in “inner time consciousness.” (c) The figure shows the dependence of the degree of the temporal extension of the duration bloc into past and future on the degree of the temporal differences between the different frequency ranges in the fluctuations as they are encoded into neural activity. The higher the degree (and number) of temporal differences encoded into neural activity, the more the actual discrete time points in the present can be extended into future and past ones, which predisposes a larger extension of the duration bloc in inner time consciousness. (d) The figure shows the dependence of the degree of the temporal extension of the duration bloc into past and future on the degree of temporal nestedness between the different frequency ranges in the fluctuations of neural activity; that is, phase-phase or phase-power coupling. The higher the degree of temporal nestedness, the more the actual discrete time points in the present can be extended into future and past ones, which predisposes a larger extension of the duration bloc in inner time consciousness.

(p.63) temporal differences encoded into neural activity and the range of the different frequency fluctuations. This implies that the degree of the “duration bloc” on the phenomenal level of consciousness is ultimately predisposed by the degree of temporal differences encoded into neural activity via difference-based coding and the range of different frequency fluctuations and their degree of temporal nestedness (see Fig. 15-4 c, d).

Neurophenomenal Hypothesis IB: Predictive Coding Versus “Global” Temporal Continuity

One may now be surprised to see the parallels of (especially) the protention in the threefold temporal structure to the assumption of a predicted input as in predictive coding. We recall from Chapters 7 through 9 in Volume I where we discussed predictive coding. Predictive coding means that the brain generates a predicted input, an anticipation of the forthcoming or expected stimulus, that is then compared and matched with the actual input. The result is described as the “prediction error,” which is supposed to determine the degree of stimulus-induced activity.

How is such predictive coding, and especially the predicted input, related to the protention in the threefold temporal structure of the “duration bloc”? First and foremost, the concept of predictive coding is a functional concept that is applied to the brain and its neural activity. This distinguishes the concept of predictive coding from those of “duration bloc” and “protention,” which are phenomenal rather than functional concepts. As such, both need to be distinguished from my (p.64) concept of “global” temporal continuity, which is a purely neuronal concept.

How does my concept of “global” temporal continuity stand in relation to predictive coding? As detailed in Chapters 7 through 9 in Volume I, predictive coding presupposes the processing of contents and thus stimulus-induced activity, while largely neglecting the relevance of the brain’s intrinsic activity independent of any stimulus processing (whether real or anticipated). This is different in my concept of “global” temporal continuity, which is supposed to operate across the boundaries of resting-state and stimulus-induced activity. Even stronger, “global” temporal continuity is supposed to be already at work in the resting-state activity of the brain itself and therefore characterizes the temporal structure of the brain’s intrinsic activity.

That has important implications. The extrinsic stimulus does by itself not generate the “global” temporal continuity (or a global neuronal workspace), as seems to be often presupposed in predictive coding and also by the proponents of the global workspace theory of consciousness. Instead, the extrinsic stimulus encounters an already existing “global” temporal continuity when it interacts with the brain’s intrinsic activity. This means that the stimulus must be linked and integrated into the already existing “virtual” temporal structure of the brain’s intrinsic activity.

Such linkage and integration is accounted neuronally for by what I described as “rest–stimulus interaction” in Volume 1 (see Chapter 11). I now propose that such rest–stimulus interaction is central for associating the newly resulting stimulus-induced activity with consciousness and thus the “duration bloc”.

How is such association of the purely neuronal stimulus-induced activity with a phenomenal state that is consciousness possible? The degree of integration between extrinsic stimulus and intrinsic activity predisposes the degree to which the intrinsic activity’s temporal structure is transferred to the extrinsic stimulus and its stimulus-induced activity. The degree of the “duration bloc” in consciousness may thus ultimately depend on the degree of rest–stimulus interaction and more specifically its degree of GABA-ergic mediated nonlinearity (see Chapter 29 for neurophenomenal details).

How now is my neurophenomenal account related to predictive coding? I assume that the here-suggested neurophenomenal mechanisms precede and are thus more basic than the cognitive processing of contents as focused upon in predictive coding and its generation of the prediction error (see Chapter 9 for an extensive discussion).

Neurophenomenal Hypothesis IIA: Neurocognitive Versus Neurophenomenal Approaches to Mental “Time Travel”

How does my neurophenomenal (rather than neurocognitive account) of the “duration bloc” stand in relation to the results of mental time travel as discussed in Chapter 13? Let us recall the imaging experiments by the Belgian scientist d’Argembeau from Chapter 13, where subjects had to actively “prospect” future events or “retrospect” past events. He showed that the neural activity in the midline regions was central in the temporal extension to past and future during mental imagery.

What exactly happens during such mental time travel? The proponents of predictive coding and the global workspace theory would probably suggest that the strong neural activity changes in the midline regions are due to the mental imagination of particular stimuli and their respective contents; that is, the events or objects the subjects imagined. The temporal signature of the mentally imagined events or objects and thus the respective stimuli themselves, including their discrete points in physical time, are then supposed to cause the neural activity in the midline regions. The contents themselves and their processing are thus supposed to cause the neural activity changes in the midline regions which in turn makes possible the mental time travel with prospection and retrospection. Contents are thus processed first while their temporal signaturing comes second.

Moreover, the observed neural activity may probably be assumed to reflect a predicted input as described in predictive coding (see (p.65) Chapters 7–9 for details). Psychologically, this may correspond to the anticipation or expectation of a particular event in response to a particular cue. The anticipation of the event and its particular content is supposed to cause the extension into the future time. Hence the temporal extension and associated “inner time consciousness” follow the imagination and prospection of the contents. Accordingly, time follows the processing of contents in predictive coding and any other neurocognitive account of mental time travel.

How does such neurocognitive account stand in relation to my neurophenomenal approach? I do not deny that subjects imagine the event and that there is anticipation of particular contents. But, and this is important, the event and the anticipation of the respective contents do not cause by themselves the temporal extension as suggested in the neurocognitive account. Instead, the events and thus the contents follow the degree of neurotemporal extension that is predisposed in the midline network, the “dynamic temporal network” as Lloyd called it (see Chapter 13).

Accordingly, the neurophenomenal account claims that the processing of contents follows the prior and more basic constitution of time. This is clearly different from the neurocognitive approach where the contents are supposed to be processed first while their temporal signaturing occurs only after that in a second step. The neurophenomenal approach thus reverses the neurocognitive account: instead of temporal extension following the anticipation of content, the anticipation of the content follows the temporal extension of the midline regions’ intrinsic activity and their degree of global temporal continuity. To put it more strongly still, the neurophenomenal approach considers the temporal extension provided by the intrinsic activity’s degree of global temporal continuity to be a necessary condition of the possible anticipation of contents.

If there were no such underlying global temporal continuity in the brain’s intrinsic activity, the subjects could probably still imagine the event. But, and this is important, they would no longer be able to anticipate the event and thus to shift its mental occurrence into the future. Why? There would be no longer a temporal matrix that allows the subjects to link their present discrete point in physical time with the ones in the future as presupposed in the anticipation of the event. Due to the lack of such a linkage, anticipation of the event would remain impossible. This is what the neurophenomenal account postulates.

Neurophenomenal Hypothesis IIB: Empirical Plausibility of the Neurophenomenal Approach to Mental Time Travel

How can we decide between neurocognitive and neurophenomenal approaches to mental time travel? The data themselves shall decide. The neurophenomenal approach claims that temporal extension is related to the brain’s intrinsic activity and provides the very basis for the subsequent anticipation of future events in mental time travel. One would consequently expect neural overlap between mental time travel and intrinsic activity, especially in the midline regions.

If, in contrast, one favors the neurocognitive approach, one would expect the temporal extension to be based on the stimulus-induced or task-related activity associated with the anticipated event itself, rather than the brain’s intrinsic activity. There should thus be no neural overlap between intrinsic activity and mental time travel which may then be considered as two distinct dissociable neural processes. These are clear hypotheses that can be tested and have indeed been addressed in the study by Oestby et al. (2012; see also Chapter 13).

We recall from Chapter 13 that Oestby et al. (2012) observed strong neural overlap between the midline activity during mental time travel and the same regions’ high activity during the resting state (see Chapter 13 for details). This means that the resting-state activity itself must already contain some information about the temporal extension into past and future as it is applied to specific contents during mental time travel. Otherwise there would be no such neural overlap between mental time travel and intrinsic activity.

How is such a neural overlap between mental time travel and intrinsic activity possible? I suggest that this can be explained only in the (p.66) neurophenomenal rather than the neurocognitive model. More specifically, we need to postulate a particular temporal structure in the neural activity of the resting-state activity itself. This temporal structure, as detailed earlier, is supposed to be manifested in the “local” and “global” temporal continuity of the neural activity in the resting state.

The subject’s instruction to perform mental time travel by imagining certain extrinsic events or objects therefore only modulates the preexisting temporal structure of the brain’s intrinsic activity (rather than causing it as presupposed in the neurocognitive model). By modulating the resting state’s temporal structure, the event and thus the content becomes integrated into the already existing temporal structure of the brain’s resting state; this in turn makes possible to extend the imagined event in time and to shift it from the present to the future.

Most importantly, the shift of the content from the present into the future allow us to anticipate or prospect the respective content. The neurocognitive function of anticipation is thus directly dependent upon the more basic neurophenomenal function of the constitution of time. Put conversely, the constitution of time—namely, the extension of the present time point into future ones— provides the basis here for the subsequent cognitive function, the anticipation or prospection.

Neurophenomenal Hypothesis IIC: “Cognition Follows Phenomenology” Rather than “Phenomenology Follows Cognition”

Let us summarize. Constitution of time precedes anticipation of events in time. Since the constitution of time is associated with “inner time consciousness,” neurophenomenal functions precede neurocognitive functions like anticipation or prospection. Cognition follows phenomenology, rather than the reverse, phenomenology following cognition (as it is tacitly presupposed in the neurocognitive approach).

I postulate that it is necessary and unavoidable that cognition follows phenomenology. Why? Our brain and its intrinsic activity operate in such a way that it is necessary and unavoidable. Due to the way the brain encodes its neural activity, including its own intrinsic activity, the constitution of global temporal continuity and consequently of “threefold temporal structure” and the “duration bloc” occur by default.

Since any stimulus or event, even imagined ones, cannot avoid interacting with the brain’s intrinsic activity and its temporal structure, the cognition of the event, as in anticipation or prospection, has to follow the phenomenology: the consciousness of that same event. There is thus priority of time and phenomenology rather than priority of contents and cognition. This implies what I will describe as the “priority hypothesis” in Chapter 17 when discussing the relationship between cognition and the loss of consciousness in anesthesia.

Put in a nutshell, the “priority hypothesis” basically postulates that we need to switch our allegiances and follow the brain itself (and its intrinsic activity) rather than our cognition (of contents and their events and their related extrinsic or stimulus-induced activity in the brain). I provided empirical evidence for the priority of time and phenomenology over the cognition of contents in time. How about phenomenal evidence? If the brain’s intrinsic activity does indeed provide temporal extension by means of its global temporal continuity, one would expect that, even in the resting state, we should be prone to continuously shifting and extending our present point in physical time to future ones.

This is indeed the case, as is well described by Blaise Pascal in the following quote where he distinguishes between “physical present” and “subjective present,” with the latter obviously coming close to what I described as the “threefold temporal structure” of “inner time consciousness”: “We never keep to the present. We anticipate the future as we find it too slow in coming and we are trying to hurry it up, or we recall the past as if to stay its too rapid flight” (Pascal 1966, 47).

Open Questions

We here propose temporal nestedness between high- and low-frequency fluctuations to be central in constituting “global” temporal continuity of neural activity and ultimately the “duration bloc” on the phenomenal level of consciousness. (p.67) However, we were not able to provide direct empirical evidence to support our neurophenomenal hypothesis.

One of the problems here is that the respective neuronal and phenomenal variables have not yet been operationalized. We need to develop a measure, an index of the degree of temporal nestedness between different frequency waves.

One would also need to relate the index of temporal nestedness to the degree of the duration bloc. One possible measure of the duration bloc could be the degree of temporal extension into both future and past during the experience of, for instance, mental time travel. Once these variables are operationalized and quantified, they may be tested in different states of consciousness, in awake state and in REM and NREM sleep, as well as in disorders of consciousness like vegetative state.

Another interesting question is the one for the degrees of “global” temporal continuity and the duration bloc in species other than humans. Other species may, for instance, show a lower degree of temporal extension into past and future of their “global” temporal continuity. If so, one would expect a lower degree of temporal nestedness, a lower number and less fine-grained temporal differences that can possibly be encoded in neural activity, and a lower number in the ranges of the fluctuations’ frequencies in the brain’s neural activity of these species. These are testable hypotheses and may be related to the degree of how deeply and extended animals can reach in their behavior into past and future (see also Chapter 31 for the discussion of consciousness in animals).

Finally, one may want to know how my neurophenomenal hypothesis stands in relation to other hypotheses about time and neural processing postulated by other neuroscientists. F. J. Varela, for instance, developed a neurophenomenological hypothesis of the neural mechanisms underlying Husserl’s concept of the duration bloc. S. Gallagher also oriented himself strongly on the phenomenological model of time, as have others like J. Fuster and, in part, also E. Poeppel, M. Wittmann, and A. C. Craig. For the discussion of their hypotheses and how they compare to the one put forward here, I devote a separate appendix to them (see Appendix 2).

Note

(1) . Dainton (2008) contrasted Husserl’s retentional concept of temporality with an extensional one. The extensional model claims that there needs to be only an overlap of both past and future with the present in order to establish temporal continuity, while the retentional model argues for complete integration of past and future into the retentional (and “protential”) temporal structure. I here follow the Husserlian model of retention since it seems to be more in accordance with the complete integration of low- and high-frequency waves yielding temporal nestedness and “global” temporal continuity. This, however, does not necessarily exclude the model by Dainton of only partial overlap. Neuronally, both complete and partial integration can coexist in terms of different degrees of the same neuronal mechanisms (like temporal nestedness), even if on a conceptual level they seem to be contradictory.

Notes:

(1) . Dainton (2008) contrasted Husserl’s retentional concept of temporality with an extensional one. The extensional model claims that there needs to be only an overlap of both past and future with the present in order to establish temporal continuity, while the retentional model argues for complete integration of past and future into the retentional (and “protential”) temporal structure. I here follow the Husserlian model of retention since it seems to be more in accordance with the complete integration of low- and high-frequency waves yielding temporal nestedness and “global” temporal continuity. This, however, does not necessarily exclude the model by Dainton of only partial overlap. Neuronally, both complete and partial integration can coexist in terms of different degrees of the same neuronal mechanisms (like temporal nestedness), even if on a conceptual level they seem to be contradictory.