Memory and Consciousness (via Audition)

Print Friendly
Mnemosyne by Dante Gabriel Rossetti (1881)

Mnemosyne (1881), Dante Gabriel Rossetti

I’d like to explore the following strong claim: Without memory, consciousness is not possible. Put another way: If you had no memory at all, you would not be conscious. I suspect this claim to be true, even if we take the minimum requirement for consciousness to be experience (of any sort).

I’ll explore this by thinking about the relationship between memory and audition, starting with a thought experiment:

Imagine having an extremely short memory while listening to a melody. So short, that there would be no melody, and instead an unrecognized series of disparate, unrelated pitch experiences. Now shorten the memory even more. Now imagine having no memory at all. Suppose the note A above middle C is sounded on a piano. Strings now vibrate at 440 cycles per second (excluding overtones), and in turn excite the air molecules in the room to vibrate at that same frequency. These in turn are picked up by your auditory faculties, which endeavor to produce in your mind the sensory impression of the pitch A. Without any memory at all, however, you would not experience the pitch. In fact, you wouldn’t be able to experience even one cycle of oscillation (several of which would be needed in order to experience a pitch). For one cycle, your mind must take in and store, over the duration of 1/440th of a second (about 2.27 ms), the entirety of an oscillation, the beginning of which will have been forgotten by the time it reaches its end.

My suggestion here is that, without memory, you cannot connect (i.e., by way of experience) the beginning of a single unit of oscillation, of which there are 440 per second, with its middle or end. Your mind thus becomes an experience sieve. It all just passes through.

(Is it an experience or stimuli sieve, or both? Perhaps that question can be answered later.)

This line of thinking invites a lot of questions, philosophical and empirical. I’ll attempt at some point (not today) to address the latter empirically—e.g., with audio software. Many of the theoretically empirical questions, however, will be difficult, if not impossible, to test. Typically, those are the ones I find most interesting.

How this line of questioning addresses the central claim—consciousness requires memory—will hopefully become apparent as I go along. I’m not always sure of how to order these thoughts, as they constitute a system rather than a line.

This is a speculative exploration done for the sake of exploration. If you get bored, skim or skip.

Reflections on the relation between memory, time, and consciousness by way of audition:

(1) One cycle (or oscillation or period or what I’ll usually call unit) of A-440 (i.e., 440 Hz) is 1/440th of a second, or about 2.27 ms. This is too short to register as a pitch in human audition, though it does register as a click. In other words, though short, it is long enough to stimulate media (e.g., speaker cones, air molecules, biological auditory mechanisms…) that will transduce the event so that it may finally be translated by the human brain into an auditory experience.

If an event is sufficiently short or weak, it won’t be translated into experience. A question that will arise here is whether a single unit of A-440 is itself an aggregate of several smaller events—the smallest of which I’ll categorize as atomic stimuli—that, on their own would not result in auditory experience, but, in close succession, somehow manifest in a biological aggregate in the human brain in such a way that results in neurons firing etc. and, ultimately, the experience of sound.

This conception depends, I presume, on atomic stimuli being individual things (events, or what have you) in the world in themselves that, like molecules, exhibit some significant properties of the experience-inducing stimulus that they help constitute, yet must aggregate in order to induce in order to result in that constitution; i.e., in order to result in experience (e.g., by analogy: you won’t taste one molecule of butter on your toast). (Below, I will distinguish atomic and molecular stimuli.)

Note that even though I distinguish mind and brain, and (like pretty much everybody) I generally talk as though they are separate, I’m not claiming that the physical event of stimuli-translation happens and then auditory experience happens over and beyond biology; auditory experience must correspond to some simultaneous biological event. I will not, however, refer to the biological event and its corresponding phenomenological experience as identical; because they’re not.

More on all this below, with some refinement (or, more likely, with further complication).

(2) Several continuous (or close enough to continuous) units of A-440 will register in one’s experience as a pitch. Apparently, the duration needs to be at least 100 ms for this to happen. (Required duration may change depending on things like wave type [e.g., sine vs. sawtooth] and frequency [lower notes need to last longer than high ones]; these details aren’t important here.) So, even several units of A-440 will not register as a pitch.

(3) Suppose 100 ms is the precise, smallest duration of A-440 that will register as pitch. Suppose now that a series of A-440 (in sine wave form) units hits your ear continuously for at least 100 ms. This will not register as a series of clicks, but as a sustained pitch. Do you not experience a series of clicks because each unit is being translated into sound as it would given an individual unit, but when those experiences are run together in experience, they result in the, let’s say, more vivid (or colorful, distinctive, “bigger,” etc.) experience of a pitch? Call this the Aggregate Theory of Pitch, or ATP.

Or is it rather the case that the stimuli result in a kind of biological buildup that results in a very different sort of translation process, so that the pitch is created not as a concatenation of smaller parts (i.e., experiences), but as a different sort of indivisible phenomenological (i.e., mental or experiential) event altogether? An event that even in its smallest iteration is still the full pitch experience, so that, for example, the experience of A may be just as short as a click, yet still be experienced as the pitch A. Call this the Gunky Theory of Pitch, or GTP.

(Note that gunk is a generic term for any substance that is not made up of proper parts. If you cut gunk up into the physically smallest parts possible [so that to divide the part any further would be to annihilate it entirely: there’s no matter, energy, wave, or whatever left], each of those parts is gunk, just as much as the pre-cut thing was gunk. I don’t think of pitch as being made of physical substance, but gunk analogizes nicely here given that a three-second pitch experience would not be divisible into parts [i.e., into time segments] smaller than would allow for an experience; if the division is so small that there’s no experience, the pitch has been annihilated).

(4) Suppose one unit of A-440 hits your ear. One millionth of that unit would register no sound at all. Nor would the next millionth, nor the next. But the complete unit does; it results in a click. Indeed, perhaps half, or even a quarter, of the unit would register as a click. (We could test this by playing back, for example, roughly 1/880th of a second of a recording of a unit of 440).

(5) Suppose one’s memory lasts 1/440th of a second. Suppose now that a sustained series of A-440 units hits your ear for some duration greater than 100 ms. Will you experience this stimulus as a series of clicks? I can only speculate.

As the series of stimuli (or potential stimuli) hits the ear, the brain is doing its job as translator, which, once the first instance of experience has been initiated, at any following instant results in some experience. If the stimulus hits the ear for one second, you’ll be given one second worth of experience(s). But you’ll only remember anything about the experience in increments of 1/440th of a second. This does not seem to be long enough to register as a pitch, and so it seems that one’s experience would be that of a click.

What’s interesting about this question is that, in attempting to answer it, the line between experience and memory blurs. In this scenario, it seems to me that you have 440 distinct experiences (i.e., mental events) per second (though perhaps these may be further divided as well), rather than experiencing 1/440th of a larger mental event each 1/440th of a second.

In other words, I think we should avoid the tempting view that the pitch is streaming by (which actually means the pitch experience itself is streaming by), as it were, fully formed, with memory functioning as a kind of window that opens up in 2.27-ms increments to allow some sort of “mind’s ear” to access—and thus to effect the realization of—the stored passing pitch (which, again, is itself an experience). This would fall in line, I think, with the GTP view.

There are points against this account. For one thing, it presupposes the occurrence of the experience, yet before any experience has occurred (i.e., there is no pitch without experience: the full ontological content of the experience just is the pitch [or, more precisely, the qualities that make up the pitch, including if its being high or low, its timbre, amplitude, and so on]).

Furthermore, the account commits a homunculus fallacy, in which we now need to account for the memory and experience of that “mind’s ear,” as distinct from the memory and experience of the person in whom the “mind’s ear” is present.

I’ll expand on all this a it’ll. It’s not as if there’s a literal pitch being produced in the brain or mind that’s hanging out in, or passing through, the mind to be either experienced or not experienced. Rather, there is a biological process going on that translates a given stimulus into experience, and that experience just is the click or pitch (of whatever given duration).

As that biological process is going on, memory is perhaps involved in collecting the incremental results of that biological process so that they may be experienced as a single event (over time); if there’s anything “streaming by,” it is that biological, mechanical process that, when experience results, amounts to a translation of stimuli into experience. What I am trying to make sense of is the relationship between that translation process and, simply put, consciousness: i.e., the resulting experience(s), the duration of a given singular experience (if indeed it makes sense to speak of such a thing), and memory.

These observations so far tempt me to reject the GTP view, but I will leave it on the table, if only because I really don’t know what the answer is.

(6) Space and perhaps time, it’s often said these days, are of finite stuff (discretely fine grained or granular, not infinitely divisible and continuous). This strikes me as obvious: just because we can cut a number in half forever, it doesn’t mean that the same is true of anything that isn’t a number; indeed, it seems to me that this idea is only applicable to the concepts we call numbers.

Mental events aren’t made up of anything. But they do happen over time, or perhaps their nature as a repetitive series of events that are similar enough—are perhaps qualitatively identical—so as to create the impression of a single extended event, with a beginning and middle (and sometimes an end) is what gives us a sense of time to start with—or better still, just is what we (informally) mean by time (roughly formalized as a kind of directionless relation between events; not as an independently flowing entity per se). I won’t explore that here, but will consider time an important feature of experience (e.g., if we were to think of pitch experience as being made up of parts, time—roughly characterized by music theorists as duration—would be one of them); and I will consider it possible to chop up time in precisely the way we do when we refer to minutes, seconds, half a second, etc.

(7) I’ll distinguish here between two sorts of memory: analog and phenomenal. By analog memory, I mean the storage—more precisely, the accumulation—of information in some physical medium. Such memory can take multiple forms. Here I’m concerned with something (something) like what happens when, for example, a spring is wound up or when you slowly push your hand into clay. The introduction of information amounts to an accumulation of small state changes that directly correspond to—or perhaps are equal to—the information itself: these states are largely defined or formed by each bit of information as its introduced, plus whatever information/states that preceded it; in the end, you have the sum of these smaller state changes: a wound spring (or, if you like, a store of of energy), and a hand impression.

Similarly, I imagine that information being introduced to one’s auditory faculties must be held and must build up to some threshold before its potential as a stimulus is met so that it may be translated into experience. In other words, I presume that, when a sustained series of continuous A-440 units hits your ear, there occurs a kind of analog buildup of matter (e.g., in the auditory cortex), and when this material accumulation hits a certain threshold, the sum result is activity (e.g., neural activation/firings) that produces a pitch experience.

This notion of a threshold is important: Some small fraction of an oscillation will not meet the threshold for initiating the production of an experience. I am presuming that there is an event that is strong (or long) enough to register as a, let’s say, potential stimuli—i.e., to produce a brain state—but that does not result in experience until that event has been built upon with relevant subsequent stimuli spaced sufficiently close together (you don’t get to a fully wound-up spring if you release the spring after each incremental winding).

(Note: It’d be interesting to examine what goes on with habituation: the stimuli are hitting your auditory faculties in more than sufficient robustness to result in experience, but no experience occurs.)

For a crude analogue, consider what happens when a wooden door is hit by a bullet. The first molecule-sized bullet slice that hits the wood would have little effect on its own. The second slice hits, then the third. We could slice the bullet into very small parts. Accumulatively, as these parts hit in quick succession, there’s damage done to the wood. (If we put a 10-second break between each slice, I presume no damage would occur.) 1

By phenomenal memory, I mean the memory of an experience; which is to say the experience of remembering. (This has an analog/biological basis, but the relationship is unclear; thus the aptly called hard problem of consciousness.) For example: I call your name, and some moments later you remember the sound of my voice uttering your name.2 The role of phenomenal memory here is: to recognize that you are hearing a melody, you must remember the first note—or at least that there were some previous notes and roughly how were related to one another—while hearing later notes.

(8) When I refer to “no memory,” I mean no phenomenal memory of any sort whatsoever, though this may still allow for a kind of analog memory (i.e., even in the absence of experience).

(9) How to (sort of) test the thought experiment in real life: Try spacing apart the notes (they can be of normal or extended duration) of a melody by a day, week, month, year, etc.3

(10) To emphasize the distinction between physical and phenomenal events—i.e., between stimuli and sound—consider the old thought experiment: If a tree falls in the woods and nobody is around to hear it, does it make a sound? Given what I mean here by sound (which is what we usually mean by that word), the answer is: No, the tree does not make a sound.

Sound arises when vibrating matter of some elastic sort (e.g., air molecules) makes its way to your auditory receptors, which sympathetically vibrate at the same rate, etc., setting off a chain of events that includes neurons firing, etc., ultimately resulting in an experience of the sort we call sound. The vibrating air molecules constitute the stimulus and sound is the term we broadly apply to the auditory experience that follows such a stimulus.

If a tree falls in the woods and no one is around who can hear, the falling tree will kick off vibrating air molecules, but the chain of events described above will not happen, resulting in no experience of sound. So: Vibrating matter? Yes. Sound? No.

To make this more vivid, suppose you’re deaf. You see a tree fall. You feel the ground shake and the displaced air wash over you. Your body vibrates in sympathy with these events, which might initiate some of the physical—e.g., inner ear and neurological—processes that occur in people with audition, but for whatever reason, none of this turns into sound. The vibrations may be useful to you and may even be (aesthetically, physically…) pleasing in some way. But you hear nothing. There is no sound.

(11) As noted above, one unit (or period) of A-440 will produce in its hearer a click. Several of these continuously concatenated will produce the pitch experience of A. Here’s a graph I made with Desmos of three continuous units of A-440:

Three units of A-440 continuously concatenated. The x-axis tracks time. Repeated indefinitely (or at least 10 times longer than it is here), this will just sound like the sustained A of a pitch fork, guitar harmonic, sine wave on a synthesizer, etc. Any given unit (here color coded) on its own—and indeed this entire three-unit sample given its brevity—would sound like a click rather than a pitch.


Here’s the simple math I used to graph the above in Desmos. 440 is easy to spot in each case. I share this because it is just as much a model of what I’m discussing as is the above graph, and I wish to emphasize that just because something can be represented with such models, doesn’t mean it directly corresponds to real-world phenomena.

To be clear, the above units could be discontinuous, with very small holes between each unit, and a (human) hearer wouldn’t notice; which is to say that it would result in the same experience as when there are no holes. Make the silent gaps wider, and eventually it will sound like a series of clicks. At what point does that happen?4

The graph and math I use here are meant to stoke reflection on how stimuli and experience might be parsed out, as well as difficulties in talking about such things. See point (12).

(12) Before going further: Note that it’s important not to confuse the models we use to represent stimuli or experience—e.g., with graphs and with ways of speaking about such events as though they were singular, whole objects in themselves—with what they actually are: a collection of many events (a term I use vaguely, as there’s no strict way to strip the collection down to clearly distinct events that cannot again be similarly stripped down) that are significantly similar to one another, overlapping and following one another over some duration of time, etc.

There is simply no way at all to talk about such things in the world without employing models. It’s also fair to say that our experiential representation of those things are themselves models (e.g., that allow us to navigate the world, or, by [bad] luck perhaps, to enjoy or suffer from it; with roots in some evolutionary benefit or another, even if now vestigially so) of a first-order kind; our talking about it and making diagrams of it being of a second-order kind of model—designed in large part to bridge the gap that exists between our minds: to give us a precise way to talk to one another about our private experience of stimuli as pitches (and colors and pain and tastes…), even if we cannot know for sure that the experiences themselves are the same for each of us (in many cases they clearly are not).

To be clear, I consider the experience of a pitch, pain, and so on to be things that exist in themselves, and so that are not models in a special way that, say, graphic representations do not exist (i.e., their primary function is to be a model). Of course, however, experiences may be conceived of as models (often referred to as representations) when considering their use for navigating the world—for navigating independent reality (whatever that means). This is another reason, I think, we endeavor to make formal models (mathematical, graphical, theoretical, programming languages) for talking about the world in itself (e.g., stimuli) that do a better job than those informal models that are provided to us through natural language: it seems that the formal models are stimuli-oriented, the informal models experience-oriented. Natural language is a way of externalizing internal experience, to get those experiences into the bodies, as it were, of others.5

The formal is precise, strives for objectivity, is operationally defined (e.g., when a doctor measures levels of consciousness in a comatose patient, or asks a patient to rate pain on a scale of 1 to 10), mathematically consistent; the informal is vague, personal,subjective, mathematically inconsistent (e.g., informal, or intuitive, probability models often involve assigning greater than 1 to the summed probabilities of several mutually exclusive events).

Developing this division in model-making is, in fact, a major feature of what defines the early Scientific Revolution, and is why behavioral psychology beat out cognitive psychological for so long as a serious science (until cognitive psychology could develop convincing empirically workable, operationally defined methods, etc.). There’s much more to explore here, such as the increasing role (it seems to me) of personal experience as positive grounds for epistemic justification following WWII (making it possible, for example, for Gettier’s 1963 paper to be more persuasive to us than it would have been to, say, Descartes or Newton; later having an important part to play in feminist epistemology; serving as a basis for formal probability models that deal in subjectivity [i.e., opinion] rather than frequency; etc.). I’ll explore these things elsewhere.

(13) Now suppose I were to concatenate a string of qualitatively indistinct frequency units. For example, 440, 440.00001, 440.0001, and 440.001. (From here on, I won’t write “Hz.”) I (safely) predict that a single pitch would be experienced. I imagine that this is closer to what happens in the real world, in fact, when we are exposed to vibrating air molecules (or whatever medium). By qualitatively indistinct, I mean when stimuli that are oscillating at different rates are rounded off, as it were, by your auditory mechanisms so that those stimuli produce identical pitch experiences. In general, the word qualitative (and its variants) here roughly refers to the character of an experience.

Note that this is different from, say, experiencing a note from a cello that is being played with vibrato, which amounts to a slight, but distinguishable, variation in pitch. We might have a hard time precisely distinguishing those variations, but the fact that we experience vibrato indicates the detection of qualitative differences. When qualitatively indistinct stimuli are presented, they may waver between various frequencies, but the result is a single, steady, unwavering pitch experience.

Generally speaking, the experience given to us by our senses is an estimate of what’s happening in the world. Otherwise, when you look at a patch of grass you might see millions or even trillions of distinct colors.

We might say that stimuli that are similar enough to produce identical sense experiences in humans must actually be physically similar. I think it’s an open question whether these things actually are, in any objective sense, similar. Our math tells us that there is an infinite range of possibility between any two frequencies, even 440.01 and 440.001. So perhaps many frequencies that are qualitatively indistinguishable are in fact, outside of the crude precision of human experience, dissimilar—indeed, they may be considered “nearly infinitely” apart (in our modeled conception of them, at any rate, and within the context of the math we humans have developed; depending on our perspective, of course, those same two frequencies may be considered infinitesimally close as well, so long as the distance between them is not zero).

On the other hand, perhaps there is something meaningful to draw from the fact that brains of the sort that give rise to human consciousness could only have come about in an environment of stimuli regularity. Perhaps things that strike the human mind as similar (in some significant way; a soda can and some lava may both appear red, but they are not significantly similar), through direct perception or through formal models, are indeed similar; perhaps, even, possessing that feature—that sort of appeal to human sensibilities—is precisely what turns out to be the basic meaning of the word similar whenever we use it. And that feature seems to correspond perfectly to the phenomenon of stimuli regularity: of series constituted by significantly similar events. I’ll use the word from here on with this in mind.

(How deep does this run? Your friend today is recognized and accepted as still your friend from yesterday due to certain shared similarities between what you perceive of your friend from day to day. Interestingly, similarity need not always be invoked to explain treating something today as if it were identical[-ish] to a thing from yesterday. Suppose a nonhuman animal is wired to treat an entity with a certain collection of attributes in a certain way. If its offspring from yesterday continues to have the relevant attributes today, it will treat that entity in the relevant way.)

I won’t further explore the above observation about math here—I’m working on another essay about that (namely challenging an assertion I often heard as a young musician: “There is an infinite amount of notes between any two keys on the piano”). The observation about the evolution of consciousness, I will touch on again below.

At any rate, we see that a collection of stimuli need to be similar enough in order to produce a meaningful stream of experience. Human experience is not very fine grained, so it seems the stimuli need only be roughly similar (if not in our experiences, where they are identical, then in our formal models that attempt to quantify the physical bases for those experiences). If you’d like to experience this for yourself, here’s a sine wave generator you can play with, which I’ve set to 440.01 as a starting point:

Notice that even 440 and 441 are qualitatively similar—I suspect to many listeners they will be qualitatively identical; well-trained musicians will likely hear a difference right away. No human, however, could distinguish 440.00001 and 440.000001 (if you find such a human, insert more zeros). I presume that the corresponding biological responses, at least on the back end (i.e., neurons firing) to such similar stimuli will be identical.6

Similarly, if a tone generator claimed to produce 440.00000000000000000000000000000001, how would I verify that my computer speaker is actually vibrating at that rate? Do we have the tools to measure this? A standard digital tuner would register 440. Could lasers be used to count the oscillations of a speaker cone made of some incredibly fine material, capable of oscillating at extremely fine grains?

More on this in the aforementioned article about the notes between piano keys. For now I will say that, at first glance, it may seem that there’s something arbitrary—anthropocentrically so—about thinking that 440.00000000001 would produce the same mechanical responses in nature as 440 (or something closer to 440, like 440.0001). This may be true, but perhaps not quite in the way it first seems. Consider also the bridge twisting in sympathy with the frequency of the wind. I presume that it’s a frequency range, not some strikingly precise frequency, say to 100 decimal places. Increasingly, this sort of observation convinces me of the arbitrariness of our real number system (as well as its implications such as their being different sizes of infinity).

There is also the fact that very fine grained frequency distinctions sound identical to humans (and surely other animals). Our labels for the external stimuli corresponding to our experiences will be as convenient as we can make them. Just as we tend to deal in relatively small natural numbers in our daily lives, and we center the origin of the Cartesian plane at (0,0) rather than at (0.005001, 7.00020305), and we express significant temperatures in easy numbers, etc.

It’s plausible that there are a finite number of vibrational frequencies possible, particularly for larger objects. Our auditory machinery may very well cover this range, but evolution takes the easiest path, which seems to have been to translate large segments of this range into identical experiences (indeed, experiencing a trillion colors in a small patch of foliage might make it difficult to make out any single blade of grass [if such a thing could have any meaning in that context], much less the tiger hiding behind the bushes). It seems that perceiving creatures started with simpler experiences that got only as rich and complicated as they needed to be.

The point is: Even if some tiny particle manages to vibrate at 440.000000000000000000000000000001 (or insert another 1,000 zeros), it seems plausible to me that this is outside the range of larger objects (e.g., air molecules), and this restriction is reflected (though with even greater restriction) in our experience, which is the initial basis upon which we have labeled the external stimuli associated with experiences in the first place.

(14) A more interesting question is what happens when qualitatively distinct, though not wildly distinct, pitches are cut into units and continuously concatenated. For example (within the same octave): A+C+D+A#+C#+B+D# and so on.

Continuous concatenation of one unit each of three frequencies: A-440 + C-261.6 + D-293.665.


The simple math I used to graph this, expressed so that each frequency (440, 261.6, 293.665) is obvious.

Sustained indefinitely, and with a sufficient variation of pitches spaced sufficiently apart in time, I presume this will sound like a series of clicks. And I imagine that if we start bringing the frequencies closer and closer together, somewhere between the clicks transforming into the experience of a chord (see the next paragraph) and then the experience of a single wavering pitch (and then finally a settled, sustain pitch), you might get an experience that corresponds to something like an average of frequencies that are contiguous in the chain, perhaps weighted in favor of whatever comes earliest (would this build up so that, for example, you get the average of ([Freq-1] and [Freq-2]), and then the average of ([the average of Freq-1 and Freq-2] and [Freq-3]), and so on?). We could also multiply the number of instances of each unit—i.e., lengthen each unit—until we would simply hear a melody.

Note that were we to repeat a small number of distinct frequency units in close succession to one another, it’ll sound like a chord (e.g., were we to alternate units of A, C#, A, C#… for several seconds, it would sound like the notes A and C# occurring simultaneously).

(15) Suppose we now take fractions of each unit, say 1/16th. That would be a little more than a tenth of a millisecond at 440—about .142 milliseconds. Think of it this way. Imagine I’ve recorded a Minimoog playing a sine wave at 440. I have this as an audio file. I play back 1/10th of a millisecond. Supposing the speaker registers and transduces the event, my hunch is that it would be too short to count as a stimulus. If not, make it shorter.

(16) Now, suppose we continuously concatenate the first 16th of 440, the second 16th of 261.6 (or we could keep a constant duration, so it would be the second .142 ms segment; whatever best results in continuity), and so on. There would be produced a sustained physical event that resembles a repeating sine wave (though the period would not be regular). And one’s auditory machinery would be hit with that sustained physical event, which, it seems to me, would produce some sort of analog buildup, but it would be inconsistent.

I predicted that stringing sufficiently distant frequency units together would result in a series of clicks. If that’s so, what would happen here? It seems it should be filtered out as physical noise (i.e., things happening in one’s environment that aren’t regular or strong enough to spend bio-phenomenal resources on), resulting in no sound at all. Or, rather, would the auditory machinery, again, take some (weighted?) average, rounding things out as best it can given the sporadic stimuli, resulting in the experience of something like a sustained (or wavering or wobbling) pitch or chord? Or would it just register as a kind of meaningless noise—I’m imagining the faint, staticky crackle of errant electrical impulses.

If filtered out, the result would be like the series of units that result in clicks. But instead of clicks, you get repeating silences. In other words: no sound at all. This would be a very strange result.

(17) I’ve been using examples that manipulate stimuli, but what I’m ultimately interested in is what’s required for experience. The point of these Frankenstein cases is to emphasize the need for duration and regularity (to complete the conditions for similarity?) of stimuli in order to get meaningful experience. From there, I’m interested in the experience itself. More on that next, starting with examining more closely the distinction between what I refer to as atomic stimuli and atomic experiences.

(18)  Atoms of stimulus and atoms of experience are distinct kinds of things.

An atom of stimulus could be whatever is the smallest instance of an event that has the relevant properties of a stimulant, though that event may not itself result in experience. (We might here also designate a molecular stimulus: the smallest event that counts as a stimulus of a certain kind [e.g., a molecule of a click may be even shorter than a unit—whatever is precisely big enough to result in an experience of a certain sort; an atom will be the smallest event that recognizably has the same features as the molecule, but is too small to result in experience of any sort; the molecule and the atom may turn out to be precisely the same events, however].

When I talk about an atom of experience, this is partly a statement about time. What counts as a stimulus might be a matter of multiple events that have built up, resulting in an accumulative release or transfer of energy amounting to the thing that we refer to broadly as the stimulus. But experience itself happens over time, and I can’t conceive of any other way to break down anything that we’d think of as a singular experience in any other way than over time.

Atomic experience is the experience that derives from an atomic stimulus. Presumably, it’s the shortest phenomenological event of a certain kind. (I wouldn’t say, “the shortest phenomenological event one can be aware of it,” because if you’re not aware of it, it’s not a phenomenological event—it’s not experience).

Atomic experiences occurring over time is not difficult to conceive of. We might think they have a synchronic dimension as well: we might simultaneously experience a sustained pitch of A-440 while smelling gasoline and having an itchy leg, etc. We might say that, at any given moment, there are several points one could isolate as an atomic experience, each point being a constitutive part of one of those distinct streams of experiences. That’s easy to conceive of. Harder to make sense of is the ontological nature of the experience A-440 produced by, say, a distorted electric guitar. The experience includes distinct overtones that one will be consciously aware of—will hear as sonic events per se—to a greater or lesser degree depending on the acuity of one’s ear; and it will have timbre and amplitude. These parts may be considered separately, just as the red and roundness of an apple may be. I wouldn’t try to break experience down atomically around such features, neither literally nor for the sake of talking about them with greater precision. Rather, a vaguer conception of atomic experience is sufficient for the sort of discussion I’d like to have, though awareness of this vagueness—of the limitations of this conception—is important.

(19) Consciousness constructs—or indeed just is (i.e., consciousness is not a vessel for content; it is the content)—models, which either amount directly or indirectly to narrative. Narrative seems to play a critical role in human understanding. Direct narrative (which corresponds to the first-order model account I gave above) would simply be whatever one experiences directly; for example, the experience of a pitch. This is gained from being exposed to the appropriate stimulus (which may be internal, as with auditory hallucination, dreaming, imagination). Indirect narrative (corresponds to second-order models) is when formal models of more involved events are made for the purpose of constructing narratives; for example, when we make a system for graphing stimuli (such as sine waves) for the sake of telling a story more complicated or comprehensive than what we can glean from direct experience.

An interesting way to conceive of consciousness-as-narrative is to conceive of atomic experiences as micro-narratives: cognitive devices that help us make sense, over time—indeed, to impart a sense or feature of time to—stimuli (at this level, atomic stimuli). These are in turn pieced together into narratives: i.e., the experience of a sustained pitch, or even an entire melody. These are in turn collected into a macro-narrative: a full composition, or even something of greater ontological significance, such as a person, social group, or national history.

Here again, memory is integral. Consciousness just is the content of experience, is the narrative. With no memory, there is no assemblage, no narrative, no consciousness, etc. Consciousness just is the assemblage of atomic experiences. Notice, though, that no matter how brief an experience is, it can be divided to exactly the same extent that its duration over time can be; if it lasts 1 millisecond, it can be cut in half (so long as the duration of time in question makes sense).

(20) Time, I noted above, I conceive here of time of as a certain kind of relation between simultaneous and non-simultaneous events; experience—i.e., linear ordering of sets of events (including: several things happening simultaneously could be viewed as a set of events that itself is an event in the linear narrative)—amounts to the construction of a narrative or ordering of these events, which, as a whole, is consciousness. Narrative is ordered representation: it is isolating sets of events as having significant relations among themselves; these relations are arbitrary, and thus not meaningful in the world (i.e., to themselves, of themselves, etc.).

I would say that the narrative—the experience—is not a function of time, but rather a function of arbitrary (i.e., useful to humans) relations of a certain kind between events; time, in turn, is a generalized function of that narrative, which has been robustly developed as a feature of math and physics, etc.

(21) This does not invoke a Zeno’s Paradox of the Achilles and the Tortoise sort translated to experience. My concern is not with how, in order to experience the entire click, we must first experience the first half of the click. My concern, rather, is that, lacking memory, when you experience the final 1/1000th of the click, you’ve forgotten the first 1/1000th of the click, and thus do not experience the click—perhaps do not experience anything.

It may, however, be related to a problem in math of the following sort. A single point on a plain has no area (we can think of such a point as a degenerate circle: a circle with a radius, and thus an area, of zero). Yet if we aggregate several of these points, we will get an area greater than zero. Phenomena in nature, of course, are not prone to these conceptual problems. It is our attempts to describe and conceptualize nature, rather, that are problematic. There’s a lot to say about this. I’ll restrict myself to: What I find interesting here is that how we carve up the parts that make up stimuli will differ from how we carve up the experiences following from those stimuli. How much of my conceptualizing of these objects is influenced by my conception of math as a world-descriptor; my studies in psychology, psychoacoustics, neuroscience, audio engineering, and philosophy; and my introspective reflection as an experiencer—is unclear to me. (In other words, I don’t know what the borders are between reality and models I inhabit of the formal [purposefully constructed by individuals and groups (e.g., through social interaction; to some degree this may be better viewed as quasi-formal, such as with cultural notions about race), indirect, second-order] and informal [naturally constructed by the brain, direct perceptual experience, first-order] kind).

(22) When speaking of consciousness, I’m not relying on arbitrary definitions such as those used by, say, an anesthesiologist or neuroscientist who must quantify and operationalize (e.g., in degrees) aspects of cognitive activity as a function of a particular job they need to do (e.g., regulating pain, determining whether and to what extent a comatose patient is internally aware). A relevant question here, then, is not “How are you parsing or operationalizing or quantifying consciousness?”—but: “What is the thing you’re parsing?” That thing is simply the presence of experience. It is striking to me how difficult it is to get some people to understand what is meant by experience. The best I can do is give an example: When you touch a piece of ice and you feel the texture, weight, cold of the ice, you are experiencing; the ice, however, while being touched by you, experiences nothing.

(23) If you were born and continued through life with a memory that lasts, say, only 1/100th of a second, nothing meaningful could ever cross your mind. But suppose you’ve lived from birth to adulthood with a normal memory, and come to a time when your memory has been reduced to 1/100th of a second. Imagine two scenarios. (A) One in which you maintain your past memories, and (B) one in which even those are gone.

(A) I don’t think you can have any meaningful thoughts. If you access your past memories, you cannot reformulate them. That is, there is a biological basis—e.g., engrams—from which memories may be retrieved, or more appropriately reconstructed (better: re-membered), but if anything at all can be remembered, there will be no meaningful narrative to follow. Whatever impressions are generated, there will be no meaningful association between them: like watching a single frame of a film flash by. I suspect even this is more than what we should expect. We are active participants in the reconstruction of memories: I remember X, and then Z, and then notice that Y needs to be inserted in between. This makes memories highly prone to error (often in the service of a certain kind of narrative or another), to the point of there being nothing at all should memory be sufficiently short.

(B) There would be no meaningful distinction here with (A). Sense impressions will arise and fade as a series (though not experienced as a series) of flashes. It would be no different than suddenly coming into and then out of existence as an experiencing human being over the time of 1/100th of a second.

These existences don’t allow for reflection or the formation of meaningful thoughts, but they do still involve experience, and thus such an entity counts as being conscious. Such a being still possesses micro-narratives, which are generally not meaningful in themselves, though there could still be a better or worse way to exist in this state: suppose you could choose between being in horrible pain or euphoric pleasure as such an entity.

What we need to imagine next is a further reduction in memory, to a having no memory at all. I claim that this means a lack even of any micro-narrative.

(24) The narrative view is distinct from another, and I think unsatisfying, way of viewing perceptual experience; i.e., as the answer to the question: What, roughly, happened in the world barely an instant ago? Indeed, we are looking for a story, something more than an instant, more than a representation of an atomic stimulus (such small events are not featured in our awareness anyway, though for the cognitive faculties that assemble small events into bigger ones, the above question is being answered, so to speak).

This view is also problematic because the time it takes for one’s senses to turn stimuli into experiences is regulated—manipulated—in order to construct something like a temporally veridical narrative (i.e., to arrange the experiences so that they sufficiently match the relations of their corresponding events in the world). If you are touched on the ankle and the neck at the same time, the nerve impulses from the neck reach your brain before those from the ankle do, but are held onto until the signal from the ankle is ready to be turned into experience: the experience is of simultaneous touching.

(25) Reflect on an inability to recognize that events in a series are similar. Suppose your brain were suddenly endowed with super-discrimination in detecting differences between vibration cycles, so that a given air molecule seemed to be vibrating wildly to you, while to my (typical) brain, that air molecule seemed to be consistently right at 440 (imagine the brain as like a coin counter; yours lets everything in as is; mine only lets in U.S. currency, even if it has to resize and reshape the material in question). Suppose this led your brain to not give you any sound (or at least nothing intelligible, as you’re getting hundreds of distinct experiences per second), while I experienced the pitch A.

This relates to the following question: Would consciousness have evolved/arisen without an environment full of events that are similar enough so as to be perceived as similar and thus as repeating. Imagine, for example, an environment where the manifestation of physical laws is so different from our own that were you to, say, pluck a guitar string, it would vibrate at wildly different sets of frequencies from one instant to the next, as would the air molecules that it excites. In such an environment, I can’t imagine that experience could evolve, unless its phenomenological estimation of the environment was so imprecise that everything would be the same color, shape, pitch, etc. This is unintelligible in evolutionary terms, because it rules out a capacity for preference among sources of stimuli.

Perhaps an entirely different set of senses would arise, but this is impossible to conceive of, as we cannot imagine experiencing a sense we don’t have, whether the environment is thoroughly alien to our own or not.

At any rate, it seems plausible to me that rounding off stimuli in experience is useful, as it founds our notion of similarity, which contributes to a kind of subjective probability about whether a series of events happening in the world are issuing from a particular region of space, or along a continuous trajectory. In other words, can be viewed as a single thing that’s happening (a speaking voice, a song, a tree that continues [sufficiently] identical to itself from moment to moment, a running tiger, etc.).

Perhaps easier to imagine is an environment where the manifestation of physical laws are the same as in ours, but events are too subtle, short, sporadic—e.g., due to constant changes in the environment due to collision of whatever bodies are present—for any repeating series to occur. What’s difficult to imagine here, however, is that life would form. Perhaps, then, given their shared reliance on stability, life and consciousness go hand in hand, though my hunch is that the latter requires more stability than the former.

(26) An implication: While rounding off the world’s complexity seems to be evolutionarily successful, it lands us in the problem of being incapable of conceiving of just how complex the world really is. We need models in order to deal with complexity; our second-order models are often made with this in mind (are indeed often made explicitly to deal with complexity), but our natural state is to expect a relatively simple, linear narrative to explain everything from why the world exists to why someone prefers one song or one flavor ice cream to another.

This perceptual rounding-off presents us with objects: a table is presented to us as a thing in itself, rather than as billions of particles. (Interestingly, it seems that a much, much smaller organism would not need to round quite so radically.) And just as we experience some set of events (with certain relations etc.) as a table, we experience some set of events as a song, melody, pitch, or click.

Perhaps consciousness is the most efficient, cost-effective (though until just now I’ve thought of consciousness as being more costly than the same behavior without consciousness) way to represent the trillions and trillions and trillions of relations around oneself at any moment as objects—as macro-narratives, so that those objects may be engaged with and navigated. In particular, significantly similar quick successions and groupings (i.e., relations of a relatively stable sort, at least for a long enough time): the particles involved in that pitch; the particles involved in that bag of chips; the particles involved in your spouse. It’s certainly an ingenious way to do so (at the risk of anthropomorphizing evolution; a common risk).

We then expand on and generalize this way of engaging our immediate present to our past (e.g., statistics; informally or formally) and future (e.g., probability; informally or formally) models. These models take different forms based on the concept in question: the self, for example, as a specially indexed instance of personhood (which also has informal and formal iterations: the former being simply my sense of self; the latter being the entity with a social security number and who pays taxes… in older times, who would have been punished or rewarded by God).

(27) Further reading.

Might be of interest in this context; add it to the pile of readings on theories of consciousness I haven’t read yet: The Remembered Present (Gerald Edelman, 1990).

This article claims to challenge “the traditional perception/memory distinction”:

This 2012 article collection from  been on my wish list for a while now: The Human Auditory Cortex

(28) Rather than stating an explicit conclusion to this tangled system of ideas, I leave it off with the following impressions:

without memory, there is no experience; without an environment with series of events that may count as similar and repeating, capacity for experience (i.e., perceptual faculties) will not evolve; nor will experience occur without similarity and repetition even once perceptual faculties and memory are in place; our models (graphic, mathematical, linguistic) for making sense of all this are limited and may even be misleading; experience rounds off the world to emphasize sufficient similarity between events (the corollary of which is when things are sufficiently dissimilar to warrant attention; consider dishabituation); if we can get a better handle on understanding whether experience is of the aggregate (e.g., ATP) or gunky (e.g., GTP) sort, we might be better able to get a handle of the sort of stuff (as it were) that experience per se is, and of how it relates to physical phenomena (stimuli, biology, etc.).



  1. If you have the expertise to better describe this process, please share!
  2. In the short term your memory is clearer and vivider thanks to short-term auditory—i.e., echoic—memory. This is distinct, in several ways, from longterm memory. I’m not going to complicate the bigger picture here by carefully distinguishing these memory types, but will say that the domain I’m mostly dwelling in here is that of short-term memory. And as I shorten that memory, and I’m blurring the line between short-term memory and direct perception.
  3. This superficially implicates memory as having a role in our experience of music, but I think memory has a deeper role to play, namely in our aesthetic experience of music. Catchy hit songs—which may be quickly tiresome or timeless—tend to use a small number of notes, often only three or four. Short melodic hooks make it easy to connect the notes on first listen. A complicated work, however, might take several listens to fully sink in; after many listens, you know how what you’re hearing at the beginning relates to what you you’ll be hearing later (and vice versa).
  4. With longer durations, there can be relatively wide gaps and it still register to a hearer as a pitch, perhaps even a single pitch being broken up (a kind of auditory confabulation; this is commonly used as a musical effect). The effect will no doubt be quite different when putting silent gaps between units rather than between, say, half-second bursts.
  5. There is another deep direction in which to go from natural language as well, to that of artistic expression. Visual art, poetry, music (the engagement with which is based, it seems, in the same areas of the brain as is language), and so on are, each in and their own and varied ways, means for a kind of formal expression that goes where natural language cannot.
  6. In general, I hold that it is not possible for events to repeat: any event is only identical to itself. I have written before about this (Pulse: You Are a Series of Non-Repeating Events), and I definitely still hold this view when it comes to physical events; what this means for phenomenal experiences in themselves, I need to give more thought.

Dan Jacob Wallace

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.