Thursday, June 28, 2018

Sub-experiences and Minimal Duration

Suppose that our conscious experiences have a certain minimum duration, say 100 ms.  Take a subject experiencing a second of pleasure, and let 't1' denote the first 50 ms time period, 't2' the next 50 ms, and so on through to 't20'.  So the subject experiences pleasure from t1 - t20.  Do they experience pleasure at t1 (and accrue a proportionate momentary welfare boost at this time)?

I'm inclined to answer 'yes'.  But this may seem to entail that both whether you're experiencing pleasure at a time and whether you accrue positive momentary welfare can be extrinsic, not fixed by the intrinsic properties of the moment.  After all, if the agent had been knocked unconscious after t1, then they would not have experienced any pleasure during this period due to the associated neural activity lasting for less than the minimum experiential duration.  Their neural activity at t1 will only get to (partly) constitute a pleasant experience if it continues on for at least another 50 ms.  This gets especially puzzling if one posits an open future.  It might then be indeterminate at t1 whether the agent is currently experiencing pleasure -- the facts about the agent's t1-experiences would not be settled until a later time (perhaps at t3 they get retroactively 'fixed').  That seems weird.

So, if you're inclined to view facts about what you're experiencing (and about how this affects your momentary welfare) as temporally intrinsic, you might instead be led to the negative conclusion that you mustn't experience pleasure at t1 after all (or at any other such short moments, though you may, of course, experience pleasure over longer durations that happen to include t1 and/or other moments as parts).  The puzzling thing about this view is the awkward implication that you can experience pleasure from t1 - t20 without experiencing pleasure at any of t1, t2, ... , or t20.

I'd be tempted to go the extrinsic route rather than be stuck with this disconnect between what you're experiencing over a period vs at all of the moments that make it up.  But perhaps we can avoid both problematic views.  After all, it seems that the puzzle depends upon a simplistic form of materialism that identifies pleasure with associated neural activity of the right sort (and duration).  That's why, when we stipulate that pleasure has a minimum duration but the underlying neural activity does not, we get puzzled about how to describe matters.  But suppose that rather than being constituted by (or identical to) appropriate underlying neural activity, pleasant experiences are instead separate mental entities that are generated or caused by appropriate neural activity.  There would then be no automatic presumption that they must temporally coincide.

Perhaps no conscious experience is generated until after the minimum duration of neural activity has been established (100 ms in our imagined example).  On the simplest such view, appropriate neural activity from t1 - t20 might then give rise to pleasure experienced from t3 - t22.  (What if the subject is killed at t20?  If qualia generation is delayed in the manner suggested above then perhaps destroying their body/brain doesn't actually end them, the subject of experience, until a couple of moments later.)  Each moment t3, t4, ... t22 is one in which the subject experiences pleasure.  And, moreover, there is nothing (metaphysically) temporally extrinsic or potentially indeterminate about this. (The fact that t3 contains pleasure is merely causally dependent on the neural activity at t1 - t2. The pleasure is not constituted by anything temporally extrinsic.)

So this dualistic interpretation seems to get us the best of both worlds.  Would anyone prefer to defend either of the first two options?  Or is there a better option yet for how to think about moments of experiences on the assumption of a minimum duration?


  1. My feeling is that consciousness should be disentangled from linear temporality; I don't know if it's the predominant academic model of consciousness, but the one I'm most convinced by is that it's is a retro-active narrative being spun by multiple spawning/competing/merging top-level processes based on input from below. Combined with memories working along various scales, fuzziness and malleability, it seems absurd to try to pin down a single persistent thread of consciousness with definite states at discrete moments/spans.

    Perhaps we can start from a fine-grained physical timeline and say it has an abstract relation to the multiple hierarchical processes that eventually condense into a subjective timeline only loosely related to the physical one; its minimal durations may be of varying length, subject to retroactive revision. Perhaps the minimal duration isn't even contiguous, but can have both gaps and overlap.

  2. "The puzzling thing about this view is the awkward implication that you can experience pleasure from t1 - t20 without experiencing pleasure at any of t1, t2, ... , or t20."

    Is that so puzzling? If we began with the assumption that conscious experiences have a minimum duration, it seems like the obvious conclusion to draw. And it doesn't seem strange in principle: plenty of wholes have properties that their parts do not.

    1. Hmm, well it seems odd to me at least. It's true that many wholes have properties that aren't inherited by their parts. But the particular case of experience over time seems rather special. It seems to me that if you are having a certain sort of uninterrupted constant experience over a period of time, then you should be having that experience at each of the moments that make up the period. Minimum duration doesn't make it "obvious" that this is false, since that merely establishes a counterfactual requirement: that had you not had the longer period of experience (extending beyond just this moment) you wouldn't have had this experience at this moment (inherited, as it is, from the larger whole).

  3. You are absolutely right. If we consider the physical changes that must be undergone within the neural circuitry; the closing and opening of synapses, the transmission of particular specialized chemical carriers to particular receptors, and various sorts of other high-level assemblages of these fundamental inputs - the experience is functionally a cumulative process of varying feedback-based cyclical sub-processes intertwining with one another and so even if your head was beheaded, for the remaining duration of time for which the state of your physical brain was in tact with all momentum of all states of all particles being retained, you could theoretically be experiencing pleasure 0.000001 nanoseconds after you were beheaded.

    There are often stories of people having 'experiences' even after supposed 'death' or irreversible processes (to the current state of knowledge) - what this implicates is that the experience itself is independent of any event outside of those particular momentary transitive periods of neural processing.

    For example, someone who is unable to process emotions or have a self-awareness of it can still have bodily reactions/physiological states that change due to external stimuli causing interior systems to automatically react in a certain way, and one can notice this acutely but not be able to comprehend the nature of origin of such changes.

    The physical state change can happen, but it can be not perceived.
    It can be perceived, but the physical state change might not happen at all.
    Essentially, it must be perceived with the physical state change (neurological) occurring but not necessarily other physical changes (physiological, biological, reflexes) for it to constitute an experience.

    There is certainly a minimal amount of time required for it to be an experience. Take anyone with an IQ of 70 and you will readily see their deficiencies in their memory buffer capacity; everything is a reaction, and they are blessed with permanent amnesia of their emotional states and experiences, with no cause and effect relations at all. So even if someone does experience an effect, the extent to which it is experienced varies in variation of the ability of processing of simultaneous inputs, and to store said information into a retrievable state, and permeate it sufficiently as to retrieve it (Long-term potentiation). This is why you always see people with high intelligence being overexcitable, because they are able to process many inputs at once; the latest research tells us that these individuals are better at pruning irrelevant inputs and are better suited to structurally altering their circuitry to communicate across all domains of their brain in the most efficient manner as to heighten the amount of information processed, and hence experience 'intensity' or 'quality'.

    Nevertheless, if a sodium atom cannot cross a gated-ion channel, you will never have x experience. Even worse so, if the transmission of multi-layered feedback loop cascades don't start. So yes, there is a minimum time of processes for everything, and no it does not follow a cumulative density probability distribution of accumulation of experience, it is time-independent; just compare an example of an exposed cut with air to a needle being jabbed into your peripheral arteries.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)

Note: only a member of this blog may post a comment.