Consciousness and Time, Part II

2,900 words

Header image taken from here.

Sign up for the Blank Horizons mailing list here.


Three levels of consciousness

In defending panpsychism, the view that everything is conscious, philosophers have often distinguished between two kinds of consciousness: (1) proto-consciousness and (2) the consciousness that humans (and likely animals) are familiar with. (From here on out, I’ll refer to (2) as “familiar consciousness.”) Proto-consciousness is the kind of consciousness possessed by electrons, forks, cars, and all other varieties of inanimate matter. (Footnote 1)

In part 1 of this essay series, I essentially argued that proto-consciousness is the experience of the passage of time. I’d like to make a small, but nevertheless crucial, revision to that claim: proto-consciousness is participation in the passage of time, whereas the experience of the flow of time is a necessary precondition for familiar consciousness. (Note that I am not equating familiar consciousness with the experience of the flow of time, since familiar consciousness likely has many other defining features that the new “science of consciousness” will hopefully reveal over the coming decades.)

What does this distinction amount to? Proto-conscious entities have a succession of experiences, while entities with familiar consciousness have an experience of succession. (2) That is, each moment of experience for a proto-conscious entity is completely separate from the previous one; there is no experience of continuity from one moment to the next. However, entities with familiar consciousness are capable of unifying successive moments of time into a single experience. Human beings have one stream of consciousness that moves forward in time, observing the progression of moments, rather than a new consciousness at each moment of time. (3) Insofar as it has any awareness at all, a rock is aware of only the present, but human beings (and supposedly other animals) are aware of the present as flowing forward from the immediate past. Hence, rocks can’t experience flow – and, consequently, change – whereas we can.

As I mentioned at the end of part 1, there are highly altered states of consciousness, typically induced by large doses of psychedelics, breakthroughs in meditation, mystical experiences, and life-threatening emergencies, in which the experience of the flow of time totally ceases. To accommodate these “exotic” experiences of time, I claim that there is a third level of consciousness that, at the risk of sounding like a woo-woo New Age mystic, I will tentatively call transcendental consciousness. Here, the past, the present, and the future may feel like they are all unfolding simultaneously in a single, eternal moment; time may literally stop passing; or time may feel like a completely meaningless concept.

So, to summarize, there are three levels of consciousness in my framework:

  • Proto-consciousness: succession of experiences
  • Familiar consciousness: experience of succession
  • Transcendental consciousness: experience of no succession


Experience of succession

What is it that creates the experience of succession in entities that have familiar consciousness? This is a very difficult question, but one that we have to answer in order to successfully explain “phenomenal time,” or the subjective experience of time. Indeed, for ordinary states of consciousness, I can think of no single feature that is more intrinsic to phenomenal time than the experience of succession. Time is experienced as something that flows; to treat it as though it were static, or as though it jumps or skips from one separate moment to another, would entirely rob it of its phenomenal character.

Most theories of subjective time completely miss this point. They implicitly characterize the stream of consciousness as a succession of experiences rather than an experience of succession. Prevailing models in psychology and neuroscience claim that we can understand how we experience time if we figure out how the brain measures the length of time between moments. In other words, modeling the brain as though it were a timer or a clock will shed light on the underlying “mechanisms” of phenomenal time. However, this view is philosophically misguided. There is nothing in a timer or a clock that is capable of unifying multiple (timed) moments into a single experience of the passage of time. Indeed, the brain could discern that an event lasts, say, 10 seconds without necessarily perceiving those seconds as consecutively flowing into one another. Each second could be experienced as a discrete “frame” that is entirely separate from the frames that precede and follow it.

If we accept that flow is not an illusion (4), then how do we experience the present as flowing forward from the immediate past? There are two potential answers to this question. Either (a) moments of experience are stretched out over an interval of time (say, 2-3 seconds, which is the standard in psychology) that encompasses both the physical present and the near past, or (b) present moments of experience contain “echoes” or representations of preceding moments. In the philosophical literature on temporal consciousness, (a) is known as the extensionalist model and (b) is called the retentionalist model.

Which one is right? Extensionalism presumes that moments of experience are continuous, not discrete, but it turns out that this is an incoherent premise. If our consciousness were to be spread out over a period of two seconds, then within any time interval of two seconds, our conscious experience does not change (or if it does change, then the change is not substantial enough to give rise to a new moment of experience). However, we can easily apply transitivity to demonstrate that our conscious experience would never change under this framework (5). The theory implies that our experience at moment t = 0 (call this E0) is identical to our experience not only at moment t = 2 (call this E2), but also at every moment in between. Therefore, E1 = E0 = E2. However, E1 is also equal to E3, so E0 = E3. Extending this conclusion, we find that each conscious experience is identical to every other one, if we accept the baseline assumptions of extensionalism.

Retentionalism, on the other hand, supposes that moments of experience are discrete; that is, at any given moment, we are conscious of exactly one physical “time-slice” that has either an infinitesimal or very short duration. While there are fixed boundaries that divide moments of experience, the present and the immediate past can nevertheless share some, though not all, of the same experiential content. In other words, the present can “retain” aspects of prior moments.

Philosophers don’t seem to be in agreement about what exactly these retentions are. (Then again, when do philosophers ever agree about anything?) Edmund Husserl, a philosopher who wrote extensively about the relationship between consciousness and time, proposed that there is a mode of experience called retentional consciousness. It seems like Husserl spilled a great deal of ink on the fundamental difference between this kind of consciousness and the consciousness that perceives reality in the present moment, but he says comparatively little about the nature of retentional consciousness itself. He seems to equate it with a particular type of memory, but he doesn’t really elaborate on what it is like to experience it. Franz Brentano referred to retentions as “proteraestheses,” which seems like a fancy but rather empty word for earlier acts of experience; Brentano doesn’t explain how we access them in the present moment. William James apparently thought that retentions are essentially after-images of the immediate past. This last view is the simplest account of retentions, and also the one that accords the most with our everyday experience of the world. When any object moves quickly through our visual field, we see in a single moment not just one image of the object but instead multiple layers of after-images. We actually observe these after-images all the time; however, when there isn’t very much motion in our surrounding environment, they are so faint that they are virtually imperceptible.


The pseudo-time arrow

The pseudo-time arrow, a hypothesis developed by Andrés Gomez Emilsson of the Qualia Research Institute, is a (partially) retentionalist theory of phenomenal time that seeks to mathematically model the layering of after-images, as well as the connectivity between after-images, in moments of experience. (Full disclosure: I was an intern at the Qualia Research Institute for the last six weeks.) The theory posits a “time-dependent qualia decay function (TDQDF),” which determines the rate at which after-images fade from our perception. Under the influence of psychedelics, we tend to see many more after-images, so the tail of the function becomes much longer.

How do we compute the TDQDF? The pseudo-time arrow hypothesis states that it can be calculated from the network of local binding relationships between qualia (subjective experiences). In particular, if we assume that this network can be encoded in a graph structure in which nodes represent qualia and edges represent local binding, then the TDQDF corresponds to the probability that a node or an edge will disappear at each physical moment. (6) (Although it is inaccurate to say that the TDQDF is this probability, I will, for the sake of convenience, be equating the two in this essay.)

If you don’t have knowledge of graph theory or the concept of “local binding,” this last paragraph might have sounded like it was written in a foreign language. I’ll back up a bit: a graph is a mathematical abstraction that models connections between pairs of objects. The objects in the graph are called “nodes,” and the connections between them are “edges.” A straightforward example of a graph is a social network, where nodes represent people and edges correspond to the relationships between them.


Fig 1. A very simple graph. If this graph depicted a social network, then the letter in each node may correspond to the first letter in the name of each person in the network. Image taken from here.

Local binding refers to the binding of qualia into unitary experiential objects. For instance, when I look at the table in front of me, the quale of “whiteness” gets bound together with the quale of “squareness,” along with the quale of “smoothness” and so on. Binding may appear to be a trivial phenomenon, but only if we assume the seemingly incontrovertible but in fact deeply objectionable premise that we perceive the world exactly as it is. Of course I see the table as simultaneously white, square, and smooth, one might say, because the table possesses all of those characteristics and I perceive the table itself when I observe it with my senses. In reality, what we see is an internally generated image of the table, stitched together from information encoded in a huge multitude of photons. Somehow, the neurons in the brain that detect color, shape, and texture have to communicate very rapidly with one another so that the brain can construct a unitary experience of the table. If these neurons didn’t share information with one another, then we would still observe whiteness, squareness, and smoothness, but each quale would be experienced separately from all the other ones. That is, we would perceive multiple, seemingly independent qualia rather than a single table.

The graph depicting the local binding between qualia in a single moment of experience would look something like this, in ordinary states of consciousness:


Fig 2. Image taken from here.

where vertical layers of nodes correspond to the qualia in a single after-image, and horizontal edges represent the local binding relationships between separate after-images. Note that the plurality of vertical layers reflects the fact that there are multiple after-images in a single moment of experience. The more recent after-images are farther to the right of the graph. In normal, sober states, the after-images and the binding between them fade away very quickly. The TDQDF is quite high.

On psychedelics, however, the TDQDF decreases substantially. Consequently, there are more vertical layers in the graph (see below); nodes and edges that are farther in the past are less likely to disappear from the present moment of experience.


Fig 3. Image taken from here.

In ordinary states of consciousness, the arrow of subjective time is the same direction as the gradient of “brightness” of the after-images in a moment of experience (this idea was not discussed in Emilsson’s essay on the pseudo-time arrow). For instance, in a moment of experience that contains five after-images of a tennis ball that is moving through the air, the temporal order of the after-images will correspond exactly to the gradient of brightness; the after-image that is most recent will be the brightest one, the one that is furthest in the past will be the least bright, and so on. It is important to note that “brightness,” in this case, is not referring to phenomenal brightness but instead “reality brightness.” If the tennis ball were lit on fire and then extinguished in the middle of its trajectory, the after-image of the ball following the dousing of the fire will still have a greater “reality brightness” value than prior after-images, even though the ball is more visually bright when it’s on fire.

Normally, the arrow of subjective time lines up fairly well with the arrow of physical time. While there are occasions when time seems to speed up or slow down, time generally flows forward in a linear fashion, and the vast majority of people never get to experience time in any other way. However, in highly altered states of consciousness, the direction of phenomenal time can actually change. Under the influence of psychedelics, time may appear to be flowing backwards; time might repeat itself in a seemingly never-ending loop; multiple, mutually exclusive timelines may be experienced simultaneously; and, as I mentioned earlier, the passage of time may come to a complete stop. Most of these “exotic experiences of time” are discussed in Emilsson’s essay, but they haven’t been analyzed in any of the academic literature on psychedelics, which is quite sparse as it is. One of Emilsson’s primary motivations for developing the pseudo-time arrow hypothesis was to explain these exotic experiences. If the pseudo-time arrow turns out to be correct, then, as far as I am aware, it will be the only mathematical theory of the third level of consciousness that I outlined earlier, i.e. “transcendental consciousness.”

In Emilsson’s theory, the arrow of phenomenal time in non-ordinary states of consciousness can be inferred from the patterns of connectivity between layers of after-images in the graph of local binding relationships between qualia. For instance, while they generally point in the forward direction in sober experiences of time, pairs of edges will have opposite directions in experiences of time looping, such that the overall structure of the graph resembles a set of interlinked circles (Fig 4). In so-called “moments of eternity” when the flow of time ceases, edges from separate layers of after-images converge onto a single center, reflecting the feeling that multiple moments of time are unfolding simultaneously in the present (Fig 5).

Fig 4 (left) and Fig 5 (right). Both images taken from here.


A physical implementation of the pseudo-time arrow?

In his book Vision, the late neuroscientist David Marr distinguishes between three levels of analyzing an information processing system: (1) the computational level, which addresses the objective of the system; (2) the algorithmic level, which describes the algorithms that the system uses in order to accomplish its objective; and (3) the implementational level, which is concerned with the physical realization of these algorithms in the system’s hardware.

As it stands, the pseudo-time arrow hypothesis is a purely algorithmic account of phenomenal time. Emilsson’s essay does not mention any particular ways in which the pseudo-time arrow might be implemented in the brain. Over the past few weeks, I’ve been trying to find an implementation-level description of the pseudo-time arrow. While I’m still too early in my investigations to report any firm conclusions, I suspect that the synchronization of many networks of oscillators in the brain may play a crucial role in executing the pseudo-time arrow. I plan to discuss this idea in much greater detail in Part 3, assuming that I make progress this coming month on determining whether it is true.


(1) Plants and cells may be proto-conscious, or they might have some kind of consciousness in between proto-consciousness and familiar consciousness. We know too little about consciousness to make any conclusions.

(2) I am borrowing the italicized terms from the entry on temporal consciousness in the Stanford Encyclopedia of Philosophy.

(3) Some philosophers may dispute the notion that we have a one stream of consciousness for our entire lives, so I’ll qualify this statement by claiming that our stream of consciousness is unitary for any period of time in which our consciousness is uninterrupted by sleep, coma, anesthesia, or anything else that would disrupt it.

(4) I actually don’t have very strong arguments against the notion that flow is an illusion created by the brain in order to “stitch together” separate moments of experience. It seems plainly evident to us that we directly experience flow in the present through our raw sense perceptions, but I’m open to the idea that flow is a mind-made impression that arises from comparing frozen snapshots of the present to ones of the immediate past.

(5) This idea should be credited to Andrés Gomez Emilsson of the Qualia Research Institute, who communicated it to me in person.

(6) The graph structure presupposes qualia formalism, the philosophical view that there is a mathematical object whose features are isomorphic to the features of qualia.

Leave a Reply

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

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

Facebook photo

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

Connecting to %s