The Depth of Time
Immediacy, aliveness, and the temporal depth of field
This research note is part of the Clockless Clock book project and may turn into a chapter.
We often talk of “deep time,” especially in the context of time scales that are extremely long relative to the human lifespan, such as geological or astronomical time scales. But why deep?
Time, or at least Cartesian time, is famously supposed to be one-dimensional. It is usually understood via a spatial metaphor in terms of a point “now” position along a single forward/backward dimension.
What is this orthogonal “depth” dimension? And how does “deep time” relate to the other popular notion of the “long now”?1
Are adjectives like deep and long merely about adding some poetry to a one-dimensional experience of time, or do they point to, and reify, meaningful aspects of the phenomenology of that experience?
An idea from optics called depth of field, suitably generalized, helps us think about this question. The depth of field of a lens is the distance between the nearest and furthest objects that can be in acceptably sharp focus together. We can think about deep time by constructing an analogous notion of a temporal depth of field, which will then allow us to construct a coherent notion of “length” for now.
To peek ahead, think of the temporal depth of field as the distance2 between the slowest and fastest processes of change that you can attend to simultaneously.
If you could simultaneously attend to the flapping of a hummingbird’s wings and the drift of the tectonic plate under your feet (assuming there isn’t an earthquake going on right now), you’d have an extraordinarily high temporal depth of field.
This understanding allows us to characterize the “long now” as well — it is an experience of now-ness during which we attend to reality with an unusually high temporal depth of field. We can also define a corresponding notion of a “short now” that features a high temporal depth of field past the lower limits of our biological capacities, such as nanoseconds to microseconds. Matrix bullet-dodging time is an example (unfortunately fictional) of “deep time” at the ultrashort end of now.
Why should we care about deep time? Because cultivating a greater temporal depth of field, and being able to vary the length of “now” from ultrashort to ultralong, makes the experience of time richer and more vivid. You’ve probably already had accidental tastes of this — perhaps a sudden, momentary sense of ineffable precious depth while contemplating a historical artifact or astronomical image. Or an intense stab of delicious nostalgia upon being exposed to a particular smell, redolent of childhood memories. Or a moment of an empathy for an insect embarking on what to it must be a long journey across the living room. These are long-now and short-now glimpses of deep time, and it is possible to systematically cultivate our ability to sense it.
It is important to note that depth of field is not a range, but the subset of a range that you can attend to simultaneously, or in parallel. Depth of field in optics is about parallel processing in a literal sense — bits from parallel object planes along the line of sight. Like rows of people in a group photograph.
A pinhole camera has infinite depth of field -- near and distant objects are equally sharp -- but effectively zero aperture. Which means you have to have very long exposures to gather light, which is not particularly useful because scenes of interest might change too fast while you’re gathering light with degenerate slowness.
On the other hand, a nice big lens will gather a lot of light very quickly, but with a much smaller depth of field. Things will get blurry quickly away from the object plane.
As the formula for depth of field shows, it is a complex quantity that is shaped by the physics of light, the biology of vision, and subjective factors.
The human eye, thanks to its flexible lens, has a dynamic depth of field, since it can change shape — rounder when focusing nearby, flatter when focusing far away. This ability weakens with age, as the lens hardens.3
The concept of depth of field is not limited to literal depth in vision. All sensation has a depth of field aspect to it, even if it isn’t called that. Hearing has a frequency range from 30 to 3000 hz, and people with trained ears can pick out a larger band of it at once (for example, speakers of tonal languages, or musicians who can follow many-voiced fugues). Touch is sensitive to a range of pressures, and the blind develop extra-sensitive touch for reading Braille, and navigating by texture. People with sensitive noses can distinguish more notes in a complex smell. Eating a salad that blends many textures and tastes in a single mouthful is a higher taste depth-of-field experience than eating the ingredients in sequence.
Being simultaneously sensitive to a larger subset of a bigger range is a powerful thing. All three contribute to the depth of field. As far as depth of field is concerned, it’s no use simply having a larger range if you can only attend to a narrow band of it at once, and switching bands is slow or expensive in energy terms. It’s also no use being able to attend to a large subset of a large range if it takes you a lot of time to gather information, even while paying full attention.
Think of depth of field as a measure of the focused bit rate of a sensor, where focused should be interpreted as “requires the least possible processing to be meaningful.” An in-focus image does not require de-blurring or guesswork or extra artificial processing to parse.
Let’s set aside the specialized definition of depth of field for lenses, and think of it in general terms as focused bit rate from here on out.
When your depth of field is high for a sensory modality, you can feel the textures of reality much more deeply in that mode. We typically think of this in terms of richness or vividness of sensation. Since sharply focused bit streams are more immediately meaningful than blurry ones, they intensify present experience. Interpretation and processing takes time, and if it has to happen in your own brain, the delay attenuates the intensity of present experience. We must give up immediacy in favor of other aspects of vividness.
When the processing can happen on a computer, this can be mitigated somewhat. For example, long-exposure astrophotographs stack thousands of individual frames and align them using image processing to produce the sharply defined images you usually see in the news. The view through an actual telescope is much more blurry and far less vivid. But on the other hand, the immediacy seems to add back a different sort of vividness. Immediacy, you could say, amplifies all other aspects of vividness such as sharpness, saturation of colors, and so on.
Can you have both immediacy and vividness in astronomical observation?
Yes! You just need a larger telescope!
Let’s talk about immediacy.
Here are images of Jupiter and Saturn from my small 4” telescope taken from my balcony in heavily light-polluted LA, a reasonable approximation of the live view (zoomed in, since in the actual field of view, these would occupy a much smaller fraction of the frame).
And here are Jupiter and Saturn photographed with the 100” Mount Wilson telescope during an overnight session I attended, again a reasonable approximation of the live view (in this case, slightly zoomed out — the discs occupy a larger portion of the field of view there because a much higher magnification can be sustained due to the ~600x greater light gathering power).
Both exposures are about the same, about 1/10s if I recall correctly, or roughly around human persistence of vision, which makes Jupiter and Saturn particularly suitable for discussions of immediacy in astronomical observation. There is a reason these are the most popular telescopic objects — they are the most dramatic WYSIWYG things up there; things like nebulas are very disappointing relative to expectations set by long-exposure photographs.
Both the live views feel much more vivid in all ways than these images, which I think is because the eye and brain actually extract signal from a jittery live image differently than software does.
Incidentally, you can get images out of a 4” as vivid as the 100” if you do image stacking of frames from say 30 seconds or so of video footage (so 360 frames at a roughly “human” rate of 12fps, or 1800 frames at a supherhuman 60 fps). You could say the 100” packs 30 seconds of sensation into 1/10s, and does so live.
I got to see the latter view in real time through the Mount Wilson telescope, and the effect of immediacy was breathtaking.
Part of the experience of “breathtaking” of course, is purely subjective — the sense that “I’m looking at Jupiter as it is right now!” Even the background intellectual recognition that Jupiter is 35-52 light minutes away doesn’t ruin it. The live, vivid, telescopic view through the 100” is simply a much richer direct connection to visual reality. It creates an abnormal intensity of experience that combines vividness and immediacy in a potent way.
It is easy and tempting to dismiss the subjective dimension as some sort of false consciousness that exaggerates and fetishizes the vividness of the live experience and overindexes on immediacy. Or because a focus on immediacy likely signals a deluded belief in the greater truthfulness of “believing your own eyes.”
These dismissals are misguided when it comes to understanding temporal depth of field. Immediacy is important to temporal depth of field not because of arbitrary aesthetic sensibilities or because it is somehow more “truthful,” but because it is conducive to greater coherence of subjectivity.
Coherence of Subjectivity
Computer processing of astronomical imagery clearly does create greater distance between subject and object in obvious ways. For example in a stacked image, there are 2 extra gamuts (models of color reproduction) in the perception chain: the gamut of the CCD sensor of the camera, and the gamut of the monitor. So even without all the false coloring we’ve come to expect in astrophotography, there is more “stuff” between you and the minimalist experience. Stacking of longer exposures involves compensating for the rotation of the targets and dynamic phenomena too, putting even more stuff into the image. Even where the raw imagery is close to human vision (as is the case with 0.1s exposures of Jupiter, or footage around 12fps), the processing into composite images is not the same as what the human visual cortex does.
Another way to think of computer processing here is that it decoheres the subjectivity of perception. The “point of view” implicit in a processed long exposure is a less coherent thing than a live view.
Many images from Hubble are composed of observations across multiple orbits for instance. Somewhat like corporate press releases, these are “views from nowhere.”
A particularly striking demonstration of this subject decoherence effect is this image of Jupiter from above its south pole (from the Juno mission presumably):
Notice what’s wrong with it? This is an impossible view!
Viewed from above a pole, approximately half the disk of Jupiter should be in darkness! There is literally no way to view Jupiter this way (you can almost view Uranus this way though, because it has a ~90 degree axial tilt). The raw imagery is real enough here, but the composite is impossible. This image is about as real as an Escher etching or a Picasso painting (Picasso famously pioneered cubism in part to try and depict a subject from all sides at once).
Or to put it in philosophy-of-mind speak, there is nothing it is like to have the point of view implied by this photograph. There is no coherent subjectivity here. All meaning comes from the communication intent of the producer of the image (presumably to present a complete view of Jupiter’s pole for insight, analysis, and aesthetic pleasure — this image, incidentally, with added googly eyes, was my twitter avatar for a while).
The point of this crude demonstration applies equally to any processed imagery that trades off immediacy for something else, like sharpness, color saturation, or in this case, completeness.
What makes the live view valuable is that it is the view from your point of view, here and now, with minimal additional processing and decoherence of subjectivity. The imagery delivered by your eyes with minimal mediation does not just induce a coherent subjectivity. It induces your coherent subjectivity.
The closer you get to “naked eye,” (whether or not you perceive the “truth” better is irrelevant), the more you perceive from the maximally coherent and accessible point of view that is “you.”
Why is such coherence of subjectivity important?
Imagine living as a brain in a vat with synthesized sensations being streamed into your awareness by chaotic-evil simulators, including images like the polar view of Jupiter above. In isolation, all the sense-objects might be “processed” for one purpose or the other, but it wouldn’t all be the same purpose. One image aims to provide a complete polar view of Jupiter, another image aims to provide an ultraviolet view of Saturn, and so on. The whole won’t add up to any sort of entanglement with a meaningful and lawful universe that evolves in time unless the simulators strive hard to set it up that way.
This point, incidentally, is why it is hard to take Dalle2 imagery seriously as representing a sentient point of view. All the coherence of subjectivity in sets of images generated by Dalle2 comes from the prompts issued by the human user. Though Dalle2 has digested the entire internet, there is as yet nothing it is like to be Dalle2.
What it “knows” doesn’t even have an inherent ordering in time, and is not structured as temporalized “memory,” so there isn’t even a non-sentient time experience of being Dalle2. There’s the germ of an interesting Turing-like test for a coherent subjectivity here — is there a being-in-time that emerges for a computer? If Dalle2 had an internal arrow of time, what would be the log level of it? Would it point in the same direction as ours? Would it emerge from the stream of API calls? Is it a dream-like time, or a waking-like time?
Dalle2 is effectively a brain in a vat with random-intent synthetic sensations being thrown at it out-of-order by chaotic-evil simulators: us internet shitposters. Any stream of consciousness that might emerge from Dalle2’s training history and inference present can only be as coherent as what we’re throwing at it. The internet of shitposts is Dalle2’s vat.4
I hope this changes, and I have a few ideas about how it can, but that’s a topic for another day.
If you were in such a vat, you wouldn’t be able to cohere either a sense of self, or a sense of the external world. And the two are really the same thing if you think about it — you are an instrument that constructs a coherent subjectivity out of the consistency of sensory experiences of a coherent objective reality.5
This is particularly important with respect to time, because a sense of the self and the world is most powerfully situated in time. A more intense sense of being — coherent subjectivity — arises from vivid immediacy.
Aliveness as Vivid Immediacy
So this is why the temporal depth of field is important. It allows you to experience reality with more vivid immediacy, thereby cohering a stronger subjectivity. Or in simpler terms — you can be more alive. Not in a biological life-signs way, but in a subjective sense, in the sense of there being something it is like to be you.
Vivid immediacy, arguably, is the sense of aliveness. And the factor that contributes most to it is the temporal depth of field. The greater the that depth, the more the experience that can be packed into the “now.”
Depth of field can shrink imperceptibly slowly with age, making your connection to reality gradually more impoverished — less vivid and immediate. We begin to compensate unconsciously with age, trying to retain a familiar level of richness of sensation. This is particularly noticeable with taste. People used to spicy cuisines sometimes make their food increasingly spicy as they age, to make up for diminished sensitivity. Younger people might find the cooking of older people over-seasoned to the point of being inedible.
When on occasion we can reverse such processes, the effects are dramatic. For example, people who have recently had cataract surgery usually report a sudden return of a level of vividness in color perception they didn't realize they had lost (there is a depth-of-field with respect to color that is different from literal depth of field — color vision has greater color depth of field than grayscale or monochrome for example).
When you put together all these different kinds of depth of field, arguably you get a sort of generalized depth of field that governs how richly you're plugged into existence itself.
This generalized depth of field, arguably, is the sense of time. During sleep and dreaming, the sense of time is suspended or highly distorted in part because the depth of field collapses to almost nothing, and is entirely dependent on memory and the few sensations that make it through.
Other senses are entangled with time perception in a fundamental way. Every sensor has a time constant, and every kind of available sensation has a natural pattern of variation in time. "Sensation" is the interaction (mathematically, the convolution) of these two elements. In biology, the time constants arise from phenomena such as rates of neural firing and propagation, and rates of chemical diffusion. The significance of 1/10 of a second6 goes beyond specific effects like persistence of vision: 1/10 of a second is the the lower limit of the unaugmented temporal depth of field.
At this lower limit, our sense of time arguably supervenes on the stream of sensations. Under strong sensory deprivation, the sense of time unravels from this end. It's a bit like near vision degrading with age.
Or to flip the analogy around, aging is slow, compounding sensory deprivation, weakening the temporal depth of field from one end. Reactions slow overall, the senses are blunted, you feel less alive. The vivid immediacy of youthful subjectivity (which is of course hormonally amplified) weakens.
The other extreme of the temporal depth of field is probably a year.
While we can intellectually appreciate and think about longer periods of time than a year, our bodies have limited visceral capacity to feel processes at such scales. Like all animals, we have a degree of attunement to day length variations and seasons over the course of a year, so time periods up to a year are still (I suspect) hooked into hormonal circuitry, emotional regulation, and so on. We can experience fear and excitement about stuff up to about a year out. Beyond that, we need temporal "glasses" to make it more real, such as time-lapse photography, poetry, and narrative.
Thanks to cultural scaffolding around a human lifespan, and aging processes, we do notice some aspects of longer durations, but humans are not natural general thinkers at say decade or century scales. Dickens’ A Christmas Carol constructs a long now for Scrooge through the device of three ghostly personifications of regimes of deep time.
Just as the 1/10 of a second end of the temporal depth of field is strongly entangled with the sensory stream, the 1 year end is strongly entangled with the memory field. I say field because at the larger scales, we lack the ability to viscerally sense the directionality of time, and memories tend to turn into undirected patterns from which the arrow of time must be recovered through conscious reasoning.
Was it your eighth or ninth birthday when you had the blue cake at your party?
Sometimes we get a glimpse. Smells can transport us back decades, and provide a glimpse of the arrow of time at decadal scales. This is a memory effect. Do animals with a combination of a strong sense of smell and powerful long-term memory, such as elephants, perhaps experience deep time better than we do?
Occasionally, a sense of history can create a visceral sense of deep time far outside our natural range. For example, watching a dinosaur movie and reflecting on the alienness of the world of 64 million years ago can lead to a sudden sense of the transience of all things even at geological time scales.
Cultivating Temporal Depth
It is tempting to immediately reach for technological means to deepen our sense of time. Can we put on “glasses” to achieve greater temporal depth of field? Could clever blends of slow-motion and time-lapse photograph (high and low frame rates magnified/compressed to human range) change our sense of time via visual pathways? Does your sense of time change when you listen to a podcast at 2x speed?
It is also tempting to reach for esoteric approaches to cultivating temporal depth, such as meditation practices (David Chapman has an interesting website on Vajrayana called vividness that might be of interest if you want to do that).
But I am most interested in very familiar and universal aspects of temporal depth of field that we don’t think about much, but can easily shape consciously.
It is easiest to approach the problem of cultivating greater temporal depth of field by considering very ordinary processes by which it can be lost and recovered.
The most familiar such process is sleep and dreaming, but that’s beyond the scope of this essay, and I might treat it in another one.
The second most familiar process is aging.
The gradual shrinking of all sensory depths of field with age is like the world itself slowly becoming more ghostly to you. This is loss of immediate vividness. Like cataracts descending on your sense of time. You become more ghostly to the world, and the world becomes more ghostly to you. Perhaps the most dramatic kind of loss with age is memory loss. The weaker your long-term memory, the smaller your temporal depth of field, and the more you are trapped within a fixed-length now.
Anything that slows or reverses aging is automatically also a way to slow down or reverse loss of temporal depth of field.
A third process is loss of natural, child-like historical curiosity; the kind that makes children interested in dinosaurs or medieval fairy tales. To alter your experience of time by looking at an antique sword or a dinosaur fossil requires a certain capacity for suspension of disbelief, and immersion in an imagined past.
A fourth process is loss of future-directed child-like imagination. Like the past, the distant future can also only be accessed through a kind of literate imagination that can construct visions of the future vivid and immediate enough to deepen current time perception.
In both these cases, the underlying capacity weakens with age primarily as a function of increasing knowledge and lived lifespan. It takes progressively more historical or future-directed curiosity to maintain these capacities into adulthood.
I describe these as processes of loss because children naturally seem to possess both historical curiosity and a future-literate imagination. This is of course in part because they know so little about the world, it is relatively easy to imagine and viscerally inhabit vivid and immediate extensions.
For example, once you know that the latest scientific studies suggest that dinosaurs probably had feathers, it is just a little bit harder to enjoy an older dinosaur movie. The movie induces a weaker sense of deep time.
Less obviously, and somewhat more speculatively, the fact that children have simply lived for a shorter duration means they can draw on the memories of adults to deepen their sense of time beyond adult levels. Anything an eight-year-old learns is being integrated into an 8-year long-term memory buffer. The same thing, learned by a forty-year-old, must be spread thinner across a 40-year long-term memory buffer.
This effect is most potent when it comes to absorbing living memories from relatives. An eight-year old talking to an eighty-year-old grandma experiences the borrowed living memories via a greater temporal depth of field than grandma herself!
It does not matter that the deep time of childhood suffers from the profound ignorance of childhood: Children live in deeper time than adults nonetheless. They experience longer and shorter nows as well.
In both these cases, what causes the loss is that growing knowledge about the present makes it harder and harder work to to sustain believable and inhabitable extensions. And growing personal lifespan spreads such extensions more thinly, over a longer basis of personal memory.
So you have to do more work (or what works nearly as well, cultivate more stupidity) to achieve the same level of vivid immediacy of time experience. I suspect the amount of work increases with the square of age.
The obvious way to slow and reverse this loss is simply acquiring more knowledge. Despite all the inadequacies of textual knowledge, it can still alter our time perception faster than age can degrade it. Something as simple as reading a history of Rome before visiting Rome can make every sight and sensation there a great deal more vivid.
The effect is something like that of learning a language. What were illegible sounds before are now irreversibly made intelligible. Once you learn a language, you actually cannot go back to hearing the raw sounds. Your sound perception has been altered.
It takes more effort, but this can be done with time perception too. Watching a geology documentary will deepen your sense of time when contemplating a mountain. Hobbies such as astronomy or even stamp-collecting systematically reshape time perception through an accumulation of small rewirings of the brain.
Such approaches to reshaping time perception may seem mundane and banal compared to technological or meditative means, but what they lack in drama, they more than make up for in accessibility and long-term compounding.
Think about any subject you’ve studied steadily over a long time that affects how you experience certain sensory streams. Think about how you might experience those sensory streams if you hadn’t had those experiences. Now think about all those streams together, and imagine how you might experience time itself differently, if all that study had been absent or different.
The word study here is revealing. To study something is simply to attend to an associated category of experience steadily over a long period. A way to train the temporal depth of field in a certain regime, expanding it, increasing its resolution. Whether such study involves words on a page, or sensory streams, doesn’t matter that much. All of it gradually reshapes who we are. We are what we attend to steadily. Subjectivities that cohere and decohere simply by virtue of being present with intent in focused bit streams.
Or to put it more simply, we don’t experience time. We are experiences of time. To experience time more deeply is to be more deeply.
“Deep time” is most common in geological discourses. “Long now” is a term popularized by Stewart Brand and the Long Now foundation, especially the associated clock project. There is also a “sideways in time” notion that can be coherently understood as movement in subjectivity space (ie points of view) but I’ll leave that for another essay.
This distance can be modeled as the width of a frequency band, with bass and treble ends, but that detail is mostly irrelevant.
These issues are on my mind since I'm troubleshooting some annoying degradation in my own vision over the last few years. I suspect no matter how much I fine tune my prescription and mix of eyewear (I now need 3 pairs -- near, computer, distance), the fundamental degradation of visual acuity and stamina relative to ten years ago is not fully fixable. Back then, I could easily log a 10 hour writing day, take a short walk, and be fine the next day. Nowadays, if I don't practice conscious visual hygiene -- shorter writing periods, more breaks, more conscientious switching of glasses, mindful exposure to the outdoors -- I pay the price immediately. Eye strain, fatigue, and upper-back pain show up quickly.
It is also increasingly the vat of many disturbed humans, who seem to be experiencing rapid identity decoherence as a result.
I’m not going to get into a detailed philosophical defense, but yes, I’m a realist, in that I believe a coherent reality exists outside of my head.