Neural Correlates of Temporality: Default Mode Variability and State-Dependent Temporal Awareness

Presenter: Dan Lloyd, Trinity College

Commentator: Geoffry Lee, University of Berkeley

Advertisements

15 Comments

  1. Hello! The last slide of references is too blurry to read, so I’ve pasted them here. Also, I’d like to add my thanks to my research assistant, Brian Castelluccio, for his help in this study.
    REFERENCES
    Buckner, R. L., Andrews-Hanna, J. R., & Schacter, D. L. (2008). The brain’s default network: anatomy, function, and relevance to disease. Ann N Y Acad Sci, 1124, 1-38.
    Calhoun, V. D., Adali, T., Pearlson, G. D., van Zijl, P. C., & Pekar, J. J. (2002). Independent component analysis of fMRI data in the complex domain. Magn Reson Med, 48(1), 180-192.
    Chapelle, O., & Vapnik, V. (2000). Model selection for support vector machines. In S. A. Solla, T. K. Leen & K.-R. Muller (Eds.), Advances in Neural Information Processing Systems (Vol. 12). Cambridge, MA: MIT Press.
    Cox, D. D., & Savoy, R. L. (2003). Functional magnetic resonance imaging (fMRI) “brain reading”: detecting and classifying distributed patterns of fMRI activity in human visual cortex. Neuroimage, 19(2 Pt 1), 261-270.
    Fransson, P. (2006). How default is the default mode of brain function? Further evidence from intrinsic BOLD signal fluctuations. Neuropsychologia, 44(14), 2836-2845.
    Grady, C. L., Springer, M. V., Hongwanishkul, D., McIntosh, A. R., & Winocur, G. (2006). Age-related changes in brain activity across the adult lifespan. J Cogn Neurosci, 18(2), 227-241.
    Greicius, M. D., & Menon, V. (2004). Default-mode activity during a passive sensory task: uncoupled from deactivation but impacting activation. J Cogn Neurosci, 16(9), 1484-1492.
    Gusnard, D. A., & Raichle, M. E. (2001). Searching for a baseline: functional imaging and the resting human brain. Nat Rev Neurosci, 2(10), 685-694.
    Hahn, B., Ross, T. J., & Stein, E. A. (2007). Cingulate activation increases dynamically with response speed under stimulus unpredictability. Cereb Cortex, 17(7), 1664-1671.
    Haynes, J. D., & Rees, G. (2006). Decoding mental states from brain activity in humans. Nat Rev Neurosci, 7(7), 523-534.
    Husserl, E. (1966 (1928)). Zur Phänomenologie des inneren Zeitbewusstseins (Phenomenology of Inner Time Consciousness) (Vol. 10). The Hague: Martinus Nijhoff.
    James, W. (1890). Principles of Psychology. New York, NY: Henry Holt & Co.
    LaConte, S., Strother, S., Cherkassky, V., Anderson, J., & Hu, X. (2005). Support vector machines for temporal classification of block design fMRI data. Neuroimage, 26(2), 317-329.
    Lloyd, D. (2002). Functional MRI and the Study of Human Consciousness. Journal of Cognitive Neuroscience, 14(6), 818-831.
    Ma, J., Zhao, Y., & Ahalt, S. (2002). OSU-SVM Toolbox for Matlab.
    Mason, M. F., Norton, M. I., Van Horn, J. D., Wegner, D. M., Grafton, S. T., & Macrae, C. N. (2007). Wandering minds: the default network and stimulus-independent thought. Science, 315(5810), 393-395.
    McKiernan, K. A., Kaufman, J. N., Kucera-Thompson, J., & Binder, J. R. (2003). A parametric manipulation of factors affecting task-induced deactivation in functional neuroimaging. J Cogn Neurosci, 15(3), 394-408.
    Norman, K. A., Polyn, S. M., Detre, G. J., & Haxby, J. V. (2006). Beyond mind-reading: multi-voxel pattern analysis of fMRI data. Trends Cogn Sci, 10(9), 424-430.
    Opper, M., & Winther, O. (2000). Gaussian processes and SVM: Mean field results and leave-one-out. Advances in large margin classifiers (pp. 43-65). Cambridge, MA: MIT Press.
    Raichle, M. E., MacLeod, A. M., Snyder, A. Z., Powers, W. J., Gusnard, D. A., & Shulman, G. L. (2001). A default mode of brain function. Proc Natl Acad Sci U S A, 98(2), 676-682.
    Schreiber, T., & Schmitz, A. (1996). Improved Surrogate Data for Nonlinearity Tests. Phys Rev Lett, 77(4), 635-638.
    Schreiber, T., & Schmitz, A. (2000). Surrogate time series. Physica D: Nonlinear Phenomena, 142(3-4), 346-382.
    Svoboda, E., McKinnon, M. C., & Levine, B. (2006). The functional neuroanatomy of autobiographical memory: a meta-analysis. Neuropsychologia, 44(12), 2189-2208.

  2. Hello Dan and Geoffry,

    Thank you very much for the interesting paper and comments.

    I share Geoffry’s questions about how the data connects up with the psychological description of the experience of temporality. So, I am eager to hear back from Dan on those points. However, to those questions I would like to add my own, and perhaps either of you could address it.

    1) What about the offered data favors the Husserlian model as opposed to some other model of temporal experience?

    2) Given these talks, I’m not sure what the experience of temporality is supposed to be.

    On the one hand, there is the experience of the passing of time. This kind of experience is *not* the default mode–at least for me. As I’m sure we can all testify, time sometimes “flies by” and we do not notice it. We need to attend to our experience in a particular way to end up having an experience of the passage of time, or so it seems to me.

    On the other hand, there is the experience of continuity between our experiences. When I have an experience of a car moving in front of me I do not experience a series of snapshots of the car, but a flow of one experience into the other. This kind of experience seems to be ubiquitous, and a better candidate for the Default Mode Dan worries about.

    If this is just a terminological point, and what you guys call the experience of temporality is what I call the experience of continuity, then I have another worry.

    3) I gathered from some of what you both said that representing other temporal features of experience, like duration or ordering, isn’t problematic in the way that representing the experience of continuity is. I’m not sure why. One way of explaining how representing continuity is possible is to say that a single experience represents a series of perceptions as connected. Am I missing something?

  3. My thanks to Richard Brown for getting this fascinating enterprise up and running, and to Geoff Lee for this very useful response. I think Geoff raises interesting questions which I’d like to keep open, because, first, the experience of time is fundamental to consciousness and fascinating in itself, and second, the possible uses of cognitive neuroscience data go far beyond their typical explanatory and confirming roles. So many voxels, so little time.

    I’ll begin with the experience of time, as explored in the classic _On the Phenomenology of Inner Time Consciousness_ of Husserl. (This will take about four paragraphs.) Husserl (“H”) makes several claims about perception in general and time perception in particular that will help, I hope, with Geoff’s questions. Overall, H’s phenomenology is expansive, offering an intricate (and not always consistent) picture of the contents and dynamics of consciousness. Perception, for H, is not limited to sensory properties of the occurrent stimulus. Sensuous “givens” are always supplemented by non-sensory “apprehensions,” and these together comprise appearances. The non-sensory contents of consciousness are explicit, and may or may not be attended. H’s regular example (in another book, _Thing and Space_) is a solid cube, which presents some of its faces but hides others; but the hidden side is nonetheless in our awareness (even if it is indefinite) with the same presence as the visible faces. Moreover, for H the hidden faces are not just in awareness, but perceived. Other examples include the familiar cases of seeing-as, where identity of sensory features between two objects is compatible with different interpretations. The objects appear differently, despite their visible indiscernibility.

    The ubiquitous exemplar of non-sensory perception is indeed temporality, which H elaborates in two principal ways, speaking of the experience of flow (or flux) and of the experience of temporal structure (the temporal context of events). All conscious experiences involve _both_ flow and structure. Experiences, whether perceptual or purely endogenous, do not simply occur and then disappear. Rather, they occur and leave their traces in “retention,” like a comet’s tail. As Paul McCartney lands on “Jude,” the “Hey” is retained though no longer sensed. Likewise, as “Jude” sounds, we anticipate something to follow (“Don’t make it bad,” if one knows the song, or something less definite) — this is protention. The immediate sensation, then, is called the “primal impression.” This is the tripartite structure of consciousness at every moment, and at every moment this entire structure is sinking (flowing) into the past. To perceive a song requires that its notes be held in awareness in a temporal structure; no occurrent sensation of a particular tone is the percept of the song. But this is no less true for a static percept. The immobility of a boulder is also perceived temporally. It’s there now, and still there a second later, and _still_ there many seconds after that. So H’s expansive concept of perception extends into the past and future, as these are realized in retention and protention.

    One other big point: For H, all of the above is necessary. It is presupposed in the very concept of human experience. We can only see objects as objects by simultaneously supposing they have hidden sides; we can only experience situations as either static or changing by also maintaining a simultaneous awareness of their immediately preceding conditions. (Temporality is equally implicated in other psychological states – recollection, fantasy, etc.) For H, then, _all_ intentional objects are temporal objects, experienced in the framework of protention/primal impression/retention, a frame that is constantly sinking toward the past. Thus, time perception is not merely a species of perception to be found among others, like color perception, tone perception, etc. Rather, it is an essential aspect of all consciousness, giving it its characteristic thickness.

    I think H is right about all of this, including his Kantian argument for temporality (and non-sensory perception) as a precondition of experience. Temporality is pervasive and basic, and for that reason is taken for granted as we theorize about consciousness. Temporality is so interwoven in perception and action, so tuned to the world, that it rarely breaks, and when it does (as, perhaps, in schizophrenia), we can hardly characterize what is wrong.

    What, then, is the experience of temporality? I reach into my backpack and produce a red cube, which I place on the table. Then I do the same to find a green cube, which I place next to the red one. And a third, yellow, cube. The green cube is part of a temporal order of appearing cubes; it has been on the table for a smaller duration that the red cube, but longer than the yellow. H claims that I will perceive a non-sensory, temporal difference between the cubes. I can pick up the second cube without having to infer their order, because I saw it placed second; `its temporal position is part of the percept. Meanwhile, all three cubes are clocking more time on the table. Their temporal order of arrival is fixed, but their steady presence is lengthening in duration. They were interesting at first, but now they’re a bit boring. And now even more boring still. That too is an aspect of the perception. To be a stodgy, boring cube is exactly to be a cube that stays put for a longish time. I don’t need to infer this aspect of my cubic awareness – it creeps up on me, until I spontaneously blurt out, “enough with the cubes already!” (I might confuse the original order or be wrong about the real duration, but both order and duration will still be presupposed.) Look around you: your old jacket hangs on an even older coathook; today’s mail is on top of that pile of papers; but that form from two weeks ago is buried in the pile. And who left their mug on the table? That wasn’t here yesterday! Among the many perceptual features of these objects are temporal properties, optionally available to attentive scrutiny but present in every case (as one can see by our sensitivity to violations of expected timing).

    H’s account, by the way, bears a superficial resemblance to William James’, who responded to similar phenomenological worries that the present instant was too thin to sustain the many temporal distinctions that suffuse experience. James offered the “specious present,” a moving window of nowness reaching a few seconds into the future and past. Within the window, everything was part of the now. The idea is not totally clear in James. Sometimes he suggests that within the window everything is part of the now and temporal order is indeterminate. In that case, any specious present beyond a few dozen milliseconds will be too broad. H’s account, unlike James’, does not make future, present, and past into a single entity, but rather coordinates them in a structure, while preserving their radical phenomenological differences. Nor does H limit himself to the short span of the specious present. At longer intervals, retention fades in distinctness, but the temporal horizon is variable according to the salience of the retained experiences.

    As this overview sinks into retention, I can suggest some Husserlian answers to Geoff’s questions about the nature of the experience of temporality. I agree that temporality is ubiquitous; it is at least a background content in all conscious experience, available to attention like many other stimulus features. My characterization of it as implicit was misleading, sorry, insofar as it suggested that temporality might not be a feature of some experiences. I should have said that it is always explicit, but non-sensory. Geoff asks if it’s the experience of flow, or a representation of a frame of reference – it’s both. But with respect to the frame of reference, it’s not so much a map or window in which experiences occur, but rather a changing structure attached to every experience. It’s as if every object of perception has a timeline attached, however rudimentary that timeline might be.

    So now we have twenty five subjects in a scanner, alternating tasks and “rest” in 24 second blocks. What’s their phenomenology? Geoff is right that the target of data analysis here isn’t the “feel of flow,” the “sense of a moving now,” as he puts it. Rather, among the various temporal properties that might be in and out of consciousness is the awareness of duration. This can be variously characterized as the awareness that a task block has ended or as the awareness that another is approaching. But the landmark events need not be the block starting and ending points. In the Husserlian structure, subjects’ awareness of their environment includes a protention of events to come and a retention of events just past, both constantly sliding toward to past. The whole structure moves at once, so the entire contents of consciousness are subject to the same continuous change as they slide into the past. At each moment, the timeline of salient events, whatever it may be for each subject, is the landscape against which the now appears. The Husserlian hypothesis suggests that each moment will have a different content than moments before or after, but that these contents nonetheless represent the current temporal position in a sequence of experiences. So, numbering the time points during default mode from 1 to n is arbitrary. The Husserlian entailment is that consciousness at each time point encodes temporal structure that specifies its position in the temporal landscape. So, the brain must encode that content as well. The default mode blocks in this experiment were attractive because the experiment has nothing to do with time. Probably subjects did attend to passing time, but less so than they might in an experiment explicitly directed at timing. Husserlian ubiquitous temporality will make its most convincing showing in a context where timing is not explicitly at issue.

    So much for the explanandum! This will be a two part reply (or I’ll never get to post it), with comments on the explanans to follow. (The difference between being @ a conference versus being at a conference is that real life doesn’t get out of the way of virtual conference sessions. That, and there’s no minibar.)

  4. Hi Michal, thanks for your comment.
    I like the idea of the experience of continuity, which is sort of a middle ground between flow (which could just be change, without continuity) and structure — I guess continuity is the reflective judgment that flow is smooth. In Husserlian terms, this reflective judgement would be based on structural temporality, the whole protention/impression/retention sandwich. Like flow, continuity could be experienced some times and not others. It could, accordingly, be experimentally manipulated (say, in watching movies), and from there any brain imaging technique might be employed — that would be nice!
    I agree that the default mode drift is not always occupied with noting the passage of time. But the phenomenological argument (in the reply to Geoff) suggests that temporality is built into perception (and consciousness overall), whether passing time is attended or not.
    Regarding your last suggestion, if continuity does represent a series of perceptions and their connection, then it does begin to resemble the Husserlian temporal structure. And that wouldn’t be problematic, if there is reason to suppose it’s there pretty steadily during default mode. So, if continuity is a contingent judgment, you couldn’t count on it during DM and my analysis wouldn’t get at it (but another experiment might). But if continuity is an everpresent pulse, then you could look for it with something like the analysis in the paper (and it would be harder to manipulate experimentally).
    Actually characterizing the phenomenology of time is challenging…
    dan

  5. Thanks, Dan, for a really good presentation. The original use of the fMRI data is a real contribution all by itself. Your extension of the NCC idea to temporal brain dynamics as is also fascinating. And I share your belief that the default or resting mode of brain activity deserves much more attention.

    First, a question about the data: As you noted at the outset:

    “‘Raw’ fMRI images reveal continuous metabolic activity nearly everywhere in the brain, overlaid with a great deal of noise. Unlike other medical and scientific imagery, single images reveal little or nothing. Interpretation thus often rests on contrasts (to detect small changes between all-over activation patterns) and averages (to smooth away some of the noise). Both are inimical to detecting temporality.” (p.2)

    But as I understood your ensuing analysis (please help me understand it better), it does indeed rely on contrasts to detect small changes between overall activation patterns. So the first question is a straightforward one. Why isn’t the information summarized in figures 2 & 3 subject to that problem?

    The second question concerns whether the NCC idea is really fit to handle the temporal dynamics of consciousness. Intuitively, the thought makes perfect sense. There will be patterns or structures in neural activity that ‘mirror’ or reflect the structures of experience. And it would seem that temporally extended processes would be prime candidates for correlation, since the persistence of high-order invariance across first-order flux can be extracted from the imaging data and modeled mathematically.

    But on closer inspection of the NCC idea, there is more to it than correspondence between neural & experiential parameters. I think the criticisms voiced by Geoffry belie that the NCC idea is at heart a classically computational approach to consciousness. The NCC will be the neural representation of the content of consciousness, in the sense that the content of the subsystem N will specify the content of the experience. From this point of view, some of Geoffry’s concerns become fairly pressing: What is the evidence that the neural activity here revealed actually specifies some identifiable conscious content? In particular, why should we think that the retention/protention structure of temporality is here encoded?

    Perhaps you are suggesting that this neural activity is part of the NCC for creature consciousness, rather than content consciousness? Part of the force of the Husserlian analysis (and of the whole Kantian tradition) is that consciousness per se has a meaningful structure, regardless of what it is consciousness of. Should we expect to find a ‘representation’ of this structure in the system?

    Another way of approaching this issue is in regard to figure 1 (which is really impressive, btw). As I understand it you have found a way to extract information from the imaging data that independently specifies temporal sequence. This is really neat. But it seems a similar technique could be applied to any set of sequential data whatever. We could train the system to discriminate earlier/later cases in the tides, the age of faces, the sunlight on the trees. There is information in all kinds of patterns that specifies temporal order. But in these other systems, there is no suspicion that this information represents the passage of time. It just, as it were, expresses or reflects the passage of time. Why is this brain data different?

    Sorry to be so long-winded, but your stuff is really engaging. At this point I’ll take a break to look for that confounded minibar . . .
    Thanks, -joe

  6. Hi Dan,
    Thank you for your wonderful reply both to me and to Geoff. I am grateful we can have this exchange, and have already benefitted from it. I look forward to your further posts.

    First, I agree with your characterization of continuity as a Husserlian sandwich. That’s what I had in mind. There are some problems with this view, but I don’t think they cannot be overcome. The prospect of manipulating continuity in movies is especially interesting–I wonder what kind of experiment you have in mind.

    But, I am skeptical about the Kantian-Husserlian view that the temporality of experience is a condition for the possibility of any experience whatsover. It seems to me that the temporality of the neural processes is necessary for experience, but not the experience of temporality itself. One reason I think this is motion-blindness. Another is that we are often wrong about what our experiences are like, and I imagine it is entirely possible that experiential temporality is a grand illusion. Hence, I wouldn’t put too many eggs in that basket.

    Consequently, it seems to me that ‘flow’ can be subsumed under continuity, as one of the ways we represent our mentality in experience. If that were so, then we would expect activation in the PFC to correlate with the experience of continuity (I think Hakwan Lau’s work is especially illuminating in this regard), but we would not find a ubiquitous hum all over the brain.

    I think that maybe DM and your analysis are pointing to something deeper than the NCC of the experience of flow or continuity. Maybe the everpresent hum is the neural continuity that is necessary to have mental states at all. Just a thought to throw out there.

  7. Hello Dan,
    I just thought up what is hopefully a nice experiment for continuity. I hope you will be back to to comment on it.

    A rapid display of images can, as movies do, give the illusion of continuity between the images. There is definitive threshold for how rapid that display has to be to achieve the effect. I think it’s 6 images per second.

    Presumably, a paradigm could be set up where the subject is watching a sequence of images while inside a scanner, at a rate lower than 6 images per second. At this time, the subject would probably report no experience of continuity. After a few seconds, the rate of display can be increased above the threshold. At that point, we could (I hope) expect the change to be reflected in the scanner. I hope this can work!

  8. One of the nice things about fMRI is that you only have to think of it, and it’s already been done:
    http://www.psychol.ucl.ac.uk/attention.lab/reprints/Carmel-curr-bio2006.pdf
    I haven’t had a chance to read through the paper, but it’s clearly relevant to your ideas of continuity. It does look like the areas they found are not the same as what I called the Dynamic Temporal Network, but I’m not surprised as I think these are quite different phenomenological aspects. I think in your proposed experiment you’d tap two different forms of continuity. To recognize flicker also involves continuity to experience the alternation of frames, just as you’d need it to experience smooth flow. The paper may spark additional thoughts. It’s fun to try to think up experiments, and especially nice to _not_ have to do them. Continuing thanks,
    dan

  9. Hi Joe! (I’m just back from a “real” conference; sorry for my slow answer.) Thanks for the questions — very provocative about the whole approach I’ve taken… I’ll comment in two parts.
    1. Simplify, simplify — In general I regard the contrastive methods of many neuroimaging studies as inappropriate for capturing consciousness, especially since consciousness is present on both sides of the contrast (C-C = 0). But temporality is something that does change from moment to moment; the problem then becomes capturing a moving target. Any method that throws images into a bin and averages them will flatten the dynamic changes over time. Figure 2 shows each time point separately, no contrasts, no binning. The simplification there is that it shows just the activity which enabled the machine learning algorithm to correctly identify the approximate current time point. Other activity surely undergirds other aspects of consciousness at that time, so in that sense this is a simplification. Figure 3A abstracts from Figure 2 to show the areas that are changing in Figure 2. So 3A does not depict a steady state or state of activation, but rather the “first derivative”, the regions where things are changing, and where the changes are indexed to passing time. A large part of the success of the machine learning algorithm hinges on these areas, but this network is only an aspect of the total conscious experience of any moment in the sequence. So there is indeed aspects of conscious brain activity that are left out. One other aspect of simplification is a conflation of images from the same time position in each of the default mode blocks. That is, during training the pattern recognizer looked at time points 1, 48, 96, and 144 because these are in the first position in each of the first four default mode blocks. And then it tried to find the first image from each of the remaining default mode blocks. Those training and test images are first relative to the start of the block, but in the absolute flow of time they are separated by 48 seconds from each to the next; the subjects would not have gotten them mixed up — they were different experiences, and the difference was a temporal one (and nothing else — constant sensory environment). So the large real flow of time is obscured by exactly the binning I decry; speaking strictly, the Dynamic Temporal Network should be called the Relative Short-term Dynamic Temporal Network. Some big Husserlian structure might be averaged out. On my behalf, however, one could note that temporal experience is not simply or exclusively the flow of life time from birth to death. Always we are tracking our position in demarcated chunks, the time from one marker to another; there were plenty of these short term markers for the subjects to experience as temporal objects, and the time scale is close to what Husserl imagined for protention and retention.
    But now it’s _time_ to sign off for a while. Tomorrow, NCC and other issues!
    thanks,
    dan

  10. This picks up the thread of Joe’s comments from a few days ago…
    Joe writes:
    “What is the evidence that the neural activity here revealed actually specifies some identifiable conscious content? In particular, why should we think that the retention/protention structure of temporality is here encoded?” — this in the context of a general question about the NCC of temporality. In general, an NCC picks up supporting evidence when we have some criterion for supposing that a subject has X as part of the contents of consciousness, and we find that X is correlated reliably with some brain state B. Temporality is a stimulus feature like spatiality, which we assume to be part of consciousness whether or not it is attended and whether or not it is reported. Why do we assume this? Because every other perceptual achievement depends on the orderly perception of these underlying structures (and because violations of that orderliness immediately capture attention). Spatiality is quite general, but not every content is spatialized. But temporality is completely general, a sort of constantly updated index of order of events, whether internal or external. My analysis exploits both generality (ubiquitous time) and orderliness (sequential, serial events). The phenomenological observations in the comments above (deriving from Husserl) seek to establish the X of a conscious content and its properties; the machine learning analysis distills the B, the brain areas which produce dynamic states that track the temporal X. The analysis works — it might not have — but the success of the analysis supports the hypothesis that specific B areas encode temporal facets of perceptual or reflective states that may differ in every other respect. Accordingly, I propose that my NCC rests on the same sorts of evidence as other NCCs.

    Joe further states:
    “Perhaps you are suggesting that this neural activity is part of the NCC for creature consciousness, rather than content consciousness? Part of the force of the Husserlian analysis (and of the whole Kantian tradition) is that consciousness per se has a meaningful structure, regardless of what it is consciousness of. Should we expect to find a ‘representation’ of this structure in the system?”
    Interesting! If temporality is indeed very general, then you’re exactly right that it’s a structure that unfolds (in continuous flow) regardless of contents. (The only qualification is that all contents are inflected by temporality: a rose at T1 is not the same as a rose at T2; that is, time is not a frame in which static contents are deployed, but rather the spine that organizes them as the contents that they are.) Because time is always systematically changing, it’s not creature consciousness per se, but it would be very useful as a necessary condition for creature consciousness, at least for humans.
    And then Joe writes:
    “As I understand it you have found a way to extract information from the imaging data that independently specifies temporal sequence. This is really neat. But it seems a similar technique could be applied to any set of sequential data whatever. We could train the system to discriminate earlier/later cases in the tides, the age of faces, the sunlight on the trees. There is information in all kinds of patterns that specifies temporal order. But in these other systems, there is no suspicion that this information represents the passage of time. It just, as it were, expresses or reflects the passage of time. Why is this brain data different?”
    Another interesting question, and you’re right that the machine learning technique would work on any monotonic process that duplicated a temporal progression on the schedule that this experiment does. But why would the brain be doing this? We have systematic changes on a 24 sec cycle. Brains may be rhythmic on shorter scales (breathing and heartbeat, EEG cycles) and perhaps on longer scales, but it would seem unlikely that there would be a spontaneous rhythm on just this time frame. Unless the dynamic was entrained to the time of the experiment — but that’s just the correlate with temporality again. Overall, then, my argument is that the analysis identifies a continuous change in brain dynamics that matches a conspicuous aspect of the experience of the experiment. The brain observations are of the same type as in all brain imaging experiments, and the temporal experience is real and necessarily has dynamic features that match the brain dynamics. Other candidate interpretations for either side of the correlation are less plausible. Or so I hope…
    thanks again,
    dan

  11. Dan,

    Many thanks for the paper reference. I look forward to reading it. I agree with you that there are probably two sense of continuity.

    I think _not_ doing experiments is all the rage in philosophy.

    Anyway, thanks again.
    Michal

  12. Dan,
    Thanks for your detailed responses! They are very helpful, indeed.

    One remaining worry concerns your response to the question: “Why should we think that the retention/protention structure of temporality is here encoded?” Your answer, roughly, is that temporality is a “stimulus feature” and so it can be correlated with neural activity just as well as any other stimulus. This appears to set aside the distinction, drawn in the paper, between temporality and timing. Timing is clearly a controllable aspect of stimulus presentation, and hence it is something that can be integrated into an experimental design. Temporality, in contrast, is a background structure that conditions experience itself. In the argument, then, the phenomenological analysis of temporality is substituted for the experimental manipulation of a stimulus variable. I am not necessarily opposed to this idea – it is really interesting. But the resulting argument is not, it seems, “resting on the same sorts of evidence as other NCCs.” Thanks again.

  13. Joe,
    I agree — the problem with a structural feature of consciousness is that it can’t be manipulated in an experiment. This is a big problem for a science of consciousness, as omitting the explanation of structural features is omitting a necessary element in any convincing account. Nonetheless, if the phenomenology of time is correct, it is a changing parameter, and that’s been my target here. Not a standard approach to an NCC, but not, I hope, a crazy one. The proof is in the pudding. (By the way, the manipulable variables are not so rock-solid either. The experimenter can say “do this” but the subject can respond with more, less, or different conscious acts.)
    I hope we can continue this conversation in other venues — it’s been very interesting. Thanks.

  14. Thanks Dan for your responses, and Joe and Michal for discussion. A really helpful interchange! I’ve got few remarks on all this, sorry for being very very late in the game posting….

    Dan’s clarification of the Husserlian picture of temporal perception was really vivid and helpful, thanks for that. I think its plausible that we have an “implicit” awareness of the temporal context of our current percept, analogous to the “implicit” perception of the backs of objects. I am intrigued by Dan’s suggestion that an implicit awareness of currently unperceived temporal parts of enduring 3D objects is essential to our awareness of them as enduring 3D objects, in the way that a sense of the presence of unperceived spatial parts of an object may be essential to perceiving it as an ordinary material object. If that is right, it would indeed show that “implicit” temporal awareness is essential to normal human consciousness of the world.

    There’s at least a suggestion in Dan’s discussion that all awareness of change in the world has this “implicit” character. But some temporal perception is just as phenomenally vivid as visual perception of the spatial features of facing surfaces can be – for example, if I perceive a flickering array, my perception of the flicker, and the rate of the flicker need not be “implicit” in this sense. Similarly, the perception of an object as moving or stationary need not having anything “implicit” about it – motion (or lack of motion) can be part of the phenomenal content of a percept in just the same way as color or shape. It seems to me that the most plausible view is that consciousness contains both an “explicit” and “implicit” awareness of temporally extended features of the world, which perhaps involves a departure from the Husserlian view as Dan describes it.

    I’m still puzzled by how all this connects up with the FMRI data. Even if consciousness does have something like an Husserlian structure, and these neural structures are keeping track of elapsed time in default mode, I feel myself wanting more evidence that these things are connected. At any time I am presumably keeping track of the time elapsed since many recent and less recent events – for example, or how long its been since breakfast. But it would be implausible to suggest that to characterize my conscious life at any moment, we have to mention a sense of how long its been since breakfast. Why think that the representation of how long it has been since the last task isn’t like that? (I guess this is a bit like the difference between the perceptual presence of the back of an object, and your knowledge of how the world extends spatially beyond your visual field – your knowledge of the layout of rooms, buildings, streets around you. Perhaps there is no way in principle of drawing line between these two forms of awareness).

    thanks!

    Geoff

  15. Thank you, Goeff. Yes, your characterization of the Husserlian idea is right on and nicely analogous to the spatial cases you mention. On the last point — briefly, as the sun sets on COnsline: Husserl argued that both protention and retention had horizons, and (also like spatial perception) as time passed the acuity of temporal awareness diminished, as if events are seen from afar. The impression he gives is that these horizons are measured in seconds rather than minutes/hours. I conjecture that the boundaries of temporal awareness are not at all sharp, but nonetheless there may be an empirical way to explore this further. A summer project, I think.
    Thanks to all of you; this virtuality has been interesting, informative, and kind of metaphysical!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s