Meditation has been claimed to have all kinds of transformative effects on the psyche, such as improving concentration ability, healing trauma, cleaning up delusions, allowing one to track their subconscious strategies, and making one’s nervous system more efficient. However, an explanation for why and how exactly this would happen has typically been lacking. This makes people reasonably skeptical of such claims.

In this post, I want to offer an explanation for one kind of a mechanism: meditation increasing the degree of a person’s introspective awareness, and thus leading to increasing psychological unity as internal conflicts are detected and resolved.

Note that this post does not discuss “enlightenment”. That is a related but separate topic. It is possible to pursue meditation mainly for its ordinary psychological benefits while being uninterested in enlightenment, and vice versa.

What is introspective awareness?

In an earlier post on introspective awareness, I distinguished between being aware of something, and being aware of having been aware of something. My example involved that of a robot whose consciousness contains one mental object at a time, and which is aware of different things at different times:

Robot’s thought at time 1: It’s raining outside
Robot’s thought at time 2: Battery low
Robot’s thought at time 3: Technological unemployment protestors are outside
Robot’s thought at time 4: Battery low
Robot’s thought at time 5: I’m now recharging my battery

At times 2-5, the robot has no awareness of the fact that it was thinking about rain at time 1. As soon as something else captures its attention, it has no idea of this earlier conscious content - unless a particular subsystem happens to record the fact, and can later re-present the content in an appropriately tagged form:

Time 6: At time 1, there was the thought that [It’s raining outside]

I said that at time 6, the robot had a moment of introspective awareness: a mental object containing a summary of its previous thoughts, which can then be separately examined and acted upon.

Humans are not robots. But I previously summarized the neuroscience book Consciousness and the Brain, and its global neuronal workspace (GNW) model of consciousness. According to this model, the contents of consciousness correspond to what is being represented in a particular network of neurons - the global workspace - that connects different parts of the brain. Different systems are constantly competing to get their contents into the global workspace, which can only hold one piece of content at a time. Thus, like robots, we too are only aware of one thing at a time, and tend to lose awareness of our earlier thoughts - unless something reminds us of them.

In what follows, I will suggest that like robots, humans also have a type of conscious content that we might call introspective awareness, which allows us to be more aware of our previous mental activity. (I am borrowing the term from the meditation book The Mind Illuminated, which distinguishes between introspective attention, introspective awareness, and metacognitive introspective awareness. I am eliding these differences for the sake of simplicity.)

I will also explore the idea that introspective awareness is a sensory channel in a similar sense as vision and sound are. The experience of sight or sound is produced by subsystems which send information to consciousness; likewise, introspective awareness is produced by a subsystem which captures information in the brain and then sends it (back) to consciousness.

We can train our other senses to become more accurate and detailed. Gilbert, Sigman & Crist (2001), reviewing the neuroscience on sensory training, list a number of ways in which discrimination can be increased in a variety of sensory modalities: among other things, "visual acuity, somatosensory spatial resolution, discrimination of hue, estimation of weight, and discrimination of acoustical pitch all show improvement with practice"; even the spatial resolution of the visual system can be deliberately increased by training.

If introspective awareness is a sensory channel, can it also be practiced to improve the number of details it will pick up on? One may feel that I am stretching the metaphor here. But in fact, Consciousness and the Brain suggests that all sensory training is in a sense training in introspection. The additional information that we get by training our senses has always been collected by our brain, but that information has remained isolated at lower levels of processing. To make it conscious, one needs to grow new neural circuits which extract the lower-level information and re-encode it in a format which can be sent to consciousness.

Thus, the brain already has the ability to take normally unavailable subconscious information and make it consciously available by practice. What is needed is a way to point that learning process at the kind of information that we would normally consider "introspective", rather than on an external information source.

From Consciousness and the Brain:

... a fourth way in which neural information can remain unconscious, according to workspace theory, is to be diluted into a complex pattern of firing. To take a concrete example, consider a visual grating that is so finely spaced, or that flickers so fast (50 hertz and above), that you cannot see it. Although you perceive only a uniform gray, experiments show that the grating is actually encoded inside your brain: distinct groups of visual neurons fire for different orientations of the grating. Why can’t this pattern of neuronal activity be brought to consciousness? Probably because it makes use of an extremely tangled spatiotemporal pattern of firing in the primary visual area, a neural cipher too complex to be explicitly recognized by global workspace neurons higher up in the cortex. Although we do not yet fully understand the neural code, we believe that, in order to become conscious, a piece of information first has to be re-encoded in an explicit form by a compact assembly of neurons. The anterior regions of the visual cortex must dedicate specific neurons to meaningful visual inputs, before their own activity can be amplified and cause a global workspace ignition that brings the information into awareness. If the information remains diluted in the firing of myriad unrelated neurons, then it cannot be made conscious.
Any face that we see, any word that we hear, begins in this unconscious manner, as an absurdly contorted spatiotemporal train of spikes in millions of neurons, each sensing only a minuscule part of the overall scene. Each of these input patterns contains virtually infinite amounts of information about the speaker, message, emotion, room size . . . if only we could decode it—but we can’t. We become aware of this latent information only once our higher-level brain areas categorize it into meaningful bins. Making the message explicit is an essential role of the hierarchical pyramid of sensory neurons that successively extract increasingly abstract features of our sensations. Sensory training makes us aware of faint sights or sounds because, at all levels, neurons reorient their properties to amplify these sensory messages. Prior to learning, a neuronal message was already present in our sensory areas, but only implicitly, in the form of a diluted firing pattern inaccessible to our awareness.

Richard’s therapy session

We saw an example of introspective awareness in my post on the book Unlocking the Emotional Brain. In the transcript, a man named Richard has been suffering from severe self-doubt, and is asked to imagine how it would feel like if he made confident comments in a work meeting. The following conversation follows:

Richard: Now I’m feeling really uncomfortable, but-it’s in a different way.
Therapist: OK, let yourself feel it - this different discomfort. [Pause.] See if any words come along with this uncomfortable feeling.
Richard: [Pause.] Now they hate me.

The therapist is asking Richard to focus his attention on the feeling of discomfort, generating moments of introspective awareness about the discomfort. Notice that Richard becomes more thoughtful and less reactive to the anxiety as he does so. My guess of what is happening is something like this:

When Richard is feeling anxious, this means that a mental object encoding something like “the feeling of anxiety” is being represented in the workspace. This activates neural rules which trigger the kinds of responses that anxiety has evolved to produce. For example, a system may be triggered which attempts to plan how to escape the situation causing the anxiety. This system’s intentions are then injected into the workspace, producing a state of mind where the feeling of anxiety alternates with thoughts of how to get away.

Introspective awareness is its own type of mental object, produced by a different subsystem which takes inputs from the global workspace, re-encodes them in a format which highlights particular aspects of that data, and outputs that back into the workspace. When a representation of an anxious state of mind is created, that representation does not by itself trigger the same rules as the original anxiety did.

As a result, as representations of the anxiety begin to alternate together with the anxiety, there are proportionately less moments of anxiety. This in turn triggers fewer of the subsystems attempting to escape the situation, making it easier to reflect on the anxiety without being bothered by it.

When Richard’s therapist asks him to feel the anxiety and to see if any words come along with it, the subsystem for introspective awareness was primed to look for any content that could be re-presented in verbal form. As Richard’s anxiety had been produced by an emotional schema including a prediction that being confident makes you hated, some of that information had passed through the workspace and been available for the awareness subsystem to capture. This brought up the verbalization of what the schema predicted would happen if Richard was confident - “now they hate me”.

Therapist: “Now they hate me.” Good. Keep going: See if this really uncomfortable feeling can also tell you why they hate you now.

According to the GNW model, when a particular piece of content is maintained as the center of attention, it strengthens the activation of any structures associated with it. As Richard’s therapist guides him to focus on the verbal content, more information related to it is broadcast into the workspace. The further prompt guides the awareness subsystem to look for patterns that feel like the reason for the hate.

Richard: [Pause.] Hnh. Wow. It’s because… now I’m… an arrogant asshole… like my father… a totally self-centered, totally insensitive know-it-all.

The therapist then takes a pattern which Richard has brought up and helps crystallize it further, and throws it back to Richard for verification.

Therapist: Do you mean that having a feeling of confidence as you speak turns you into an arrogant asshole, like Dad?
Richard: Yeah, exactly. Wow.

In this example, we saw that having more moments of introspective awareness was beneficial for Richard. As aspects of his moment-to-moment consciousness were made available for other subsystems to examine, the emotional schema causing the anxiety was identified and its contents extracted into a format which could be fed into other subsystems. Later on, when Richard’s co-worker displayed confidence which others approved of, a contradiction-detection mechanism noticed a discrepancy between reality and the prediction that confidence makes you hated, allowing the prediction to be revised.

Under this model, the system which produces moments of introspective awareness is a subsystem like any other in the brain. This means that it will be activated when the right cues trigger it, and its outputs compete with the outputs of other systems submitting content to consciousness. The circumstances under which the system triggers, and its probability of successfully making its contents conscious, are modified by reinforcement learning. Just as practicing a skill such as arithmetic eventually causes various subsystems to manipulate the content of consciousness in the right order, practicing a skill which benefits from introspective awareness will cause the subsystem generating introspective awareness to activate more often.

Meditation as a technique for generating moments of introspective awareness

Just as there are different forms and styles of therapy, there are also different forms and styles of meditation. All of them involve introspective awareness to at least some degree, but they differ in what that awareness is then used for.

In the example with Richard, his therapist asked him to imagine being confident and to then bring his awareness to why that felt uncomfortable. In contrast, a more behaviorally oriented therapist might not have examined the reason behind the discomfort. Rather, they might have taught Richard to notice his reaction to the discomfort, and then use that as a cue for implementing an opposite reaction. Both kinds of therapists would ask their clients to generate some introspective awareness, but aiming that awareness at different kinds of features, and using the awareness to trigger different kinds of strategies. The results would correspondingly be very different.

Likewise, systems of meditation differ in how much introspective awareness they produce, what kinds of features the awareness-producing subsystem is trained to extract, and what that awareness is then used for. For this article, I have chosen to use the example of the system in The Mind Illuminated (TMI), as it is clearly explained and explicitly phrased in these terms. (Again, TMI has a more precise distinction between introspective attention and introspective awareness, which I am eliding for the sake of simplicity.)

In TMI’s system, as in many others, you start with trying to keep your attention on your breath. In terms of our model, this means that you want to keep sensory outputs corresponding to your breath as the main thing in your consciousness.

The problem with this goal is that there is no subsystem which can just unilaterally decide what to maintain as the center of attention. At any given moment, many different subsystems are competing to make their content conscious. So one system might have the intention to follow the breath, and you do it for a while, but then a planning system kicks in with its intention to think about dinner. Such planning has tended to feel rewarding, so it wins out and the intent to meditate is forgotten until five minutes later, when you decide what you want for dinner and then suddenly remember the thing about following your breath.

TMI calls this mind-wandering from forgetting, and the first step of practice is just to notice it whenever it happens, congratulate yourself for having noticed it, and then return to the breath. Being able to notice forgetting requires having a moment of introspective awareness which points out the fact that you had not been following your breath. When you take satisfaction in having noticed this, your awareness-producing subsystem gets assigned a reward and becomes slightly more likely to activate in the future. “Have I remembered to follow my breath or not?” acts a feedback mechanism that you can explicitly train on.

As the awareness-producing system starts to activate more often and ping you if you have forgotten to meditate, periods of mind-wandering grow shorter.

Now, even if you stop getting entirely lost in thought, you still have distraction: content from other subsystems that is in consciousness together with the sensations of the breath and the intention to focus on the breath. For example, you might be having stray thoughts, hearing sounds from your environment, and experiencing sensations from your body.

To more exclusively focus on the breath, you are instructed to maintain the intent to both attend to it and also to be aware of any distractions. The subsystems which output mental content can, and normally do, operate independently of each other. This means that the following may happen:

Subsystem 1: I’m meditating well!
Subsystem 2: Hmm, what’s that smell.
Subsystem 1: I’m meditating well! No distractions.
Subsystem 2: Smells kinda like cookies.
Subsystem 2: Mmm, cookies.
Subsystem 1: Continuing to meditate well!
Subsystem 2: Say, what’s for dinner?

That is, a system which tracks the breath can continue to repeatedly find the breath, and report that your meditation is proceeding well and with no distractions… all the while the content of your consciousness continues to alternate with distracted thoughts, which the breath-tracking subsystem is failing to notice (because it is tracking the breath, not the presence of other thoughts). Worse, since you may find it rewarding to just think that you are meditating well, that thought may start to become rewarded, and you may find yourself just thinking that you are meditating well… even as that thought has become self-sustaining and no longer connected to whether you are following the breath or not!

There are all kinds of subtle traps like this, and reducing the amount of distraction requires you to first have better awareness of the distraction. This means more moments of introspective awareness which are tracking what’s actually happening in your mind:

Subsystem 1: I’m meditating well!
Subsystem 2: Hmm, what’s that smell.
Subsystem 1: I’m meditating well! No distractions.
Subsystem 2: Smells kinda like cookies.
Subsystem 2: Mmm, cookies.
Awareness subsystem: Wait, one train of thought keeps saying that it’s meditating well, but another is totally getting into the thought of food.
Subsystem 1: Oh. Better refocus that attention on the breath, and spend less time thinking about the concept of following the breath.

This kind of a process also teaches you to pay attention to patterns of cause and effect in your mind. In this example, the smell of cookies caused you to think of cookies, which in turn made you think of dinner, which could have ultimately led to forgetting and mind-wandering.

Catching the train of thought after “mmm, cookies” meant that three “processing steps” had passed before you noticed it. If you practice tracing back trains of thought in your mind, you seem to teach your awareness-system to collect and store data from a longer period, even when it is not actively outputting it. This means that at the “mmm, cookies” stage, you can query your awareness to get a trace of the immediately preceding thought chain.

You notice that you started to get distracted starting from the smell of the cookie and can then use this as further input to your awareness system. You are essentially taking the re-presented smell of the cookie which the system output, and feeding it back in, asking it to pay more attention to detecting “things like this”. The next time that you notice a smell, your introspective awareness may flag it right away, letting you catch the distraction at the very first stage and before it turns into an extended train of thought.

Note that there is nothing particularly mysterious or unusual about any of this. You are employing essentially the same process used in learning any skill. In learning to ride a bike, for example, attempting to keep the bike balanced involves adjusting your movements in response to feedback. When you do so, your brain becomes better at detecting things like “tilting towards the right” in the sense data, increasing your ability to apply the right correction. After you have learned to identify tilting-a-lot-but-not-quite-falling, your brain learns to backtrace to the preceding state of tilting-a-little-less, and apply the right correction there. Once its precision has been honed to identify that state, you can further detect an even subtler tilt, until you automatically apply the right corrections to keep you balanced.

Essentially the kind of a learning algorithm is being applied here. Increased sensory precision leads to improvements in skill which allow for increased sensory precision. (See also this article, which goes into more detail about TMI as a form of deliberate practice.)

Uses for moments of introspective awareness

I should again emphasize that the preceding explanation is only looking at one particular meditation system. There are other systems which work very differently, but they all use or develop introspective awareness to some extent. For example:

  • In Shinzen Young's formulation of “do nothing” practice, you have just two basic instructions: let whatever happens, happen and when you notice an intention to control your attention, drop that intention. This trains introspective awareness to notice when one is trying to control their attention… but it is also a very different system, since maintaining an intention to notice when that happens would also be an attempt to control attention! Thus, one is instructed to drop intentions if one spontaneously notices them, but not to actively look for them.
  • In noting practice, you are trying to consciously name or notice everything that happens in your consciousness. Introspective awareness is trained to very rapidly distinguish between everything that happens, but is not trained to maintain attention on any particular thing.
  • In visualization practice, you might create a visual image in your mind, then use introspective awareness to examine the mental object that you’ve created and compare it to what a real image would look like. This gives the subsystem creating the visualization feedback, and helps slowly develop a more realistic image.

Going back to TMI-style introspective awareness, once you get it trained up, you can use it for various purposes. In particular, once you learn to maintain it during your daily life - and not just on the meditation couch - it will bring up more assumptions in your various schemas and mental models. Think of Richard paying attention to the assumptions behind his unwanted reactions and making them explicit, but as something that happens on a regular basis as the reactions come up.

Romeo Stevens described what he called “the core loop of Buddhism”:

So, what is the core loop?
It's basically cognitive behavioral therapy, supercharged with a mental state more intense than most pharmaceuticals.
There are two categories of practice, one for cultivating the useful mental state, the other uses that mental state to investigate the causal linkages between various parts of your perception (physical sensations, emotional tones, and mental reactions) which leads to clearing out of old linkages that weren't constructed well.
You have physical sensations in the course of life. Your nervous system reacts to these sensations with high or low valence (positive, negative, neutral) and arousal (sympathetic and parasympathetic nervous system activation), your mind reacts to these now-emotion-laden sensations with activity (mental image, mental talk) out of which you then build stories to make sense of your situation.
The key insight that drives everything is the knowledge (and later, direct experience) that this system isn't wired up efficiently. Importantly: I don't mean this in a normative way. Like you should wire it the way I say just because, but in the 'this type of circuit only needs 20 nand gates, why are there 60 and why is it shunting excess voltage into the anger circuits over there that have nothing to do with this computation?' way. Regardless of possible arguments over an ultimately 'correct' way to wire everything, there are very low hanging fruit in terms of improvements that will help you effectively pursue *any* other goal you set your mind to.

Again, we saw an example of this with Richard. He had experienced his father as acting confident and as causing suffering to Richard and others; sensations which his mind has classified as negative. In order to avoid them, a model (story) was constructed saying that confidence is horrible, and behaviors (e.g. negative self-talk) were created to avoid appearing horrible.

Now, this caused problems down the line, making him motivated to try to appear more confident… meaning that there was now a mechanism in his brain trying to prevent him from appearing confident, and another which considered this a problem and tried to make him more confident, in opposition to the first system. See what Romeo means when talking about circuits that only need 20 gates but are implemented using 60?

The article “tune your motor cortex” makes the following claims about muscle movement:

Your motor cortex automatically learns to execute complex movements by putting together simpler ones, all the way down to control of individual muscles.
Because the process of learning happens organically, the resulting architecture of neural connections (you can think of them as "hidden layers" in machine learning terms) is not always perfectly suited to the task.
Some local optima of those neural configurations are hard to get out of, and constantly reinforced by using them.
There is some pressure for muscle control to be efficient, and the motor cortex is doing a "good enough" job at it, but tends to stop a fair bit from perfection.
By repeating certain movements and positions over and over again (e.g. during sitting work), we involuntarily strengthen connections between movements and muscles that don't make much sense lumped together.
E.g. control of shoulders might become spuriously wired together with control of thighs (both are often tense during sitting).

There are various mental motions which are learned in basically the same way as physical motions are:

  • You learn to calculate 12*13 by a technique such as first multiplying 10*13, keeping the result in your memory, calculating 2*13, and then adding the intermediate results together.
  • You learn that a particular memory makes you feel slightly unpleasant, and that flinching away from anything that would remind you of it takes the pain away.
  • You learn that this also works on uncomfortable chores, teaching you to keep pushing the thought of them away.
  • You learn that your father’s behavior is painful to you, and that any confidence reminds you of that, so you learn negative self-talk which blocks you from acting confident.
  • You learn that saying “no” to people reminds you of being punished for saying “no” to your parents, but that saying “yes” too often means that you are constantly fulfilling promises to other people - so you learn to avoid situations where you would be asked anything.
  • You learn that there’s something you can do in your mind to stop feeling upset, so you start ignoring your emotions and any information they might have.
  • You learn that if you feel bad about not getting the respect you want, thinking “if only I was good enough at persuasion, I would get what I want” gives you a sense of control - even though this pattern also makes you feel personally at fault when you don’t get what you want.
  • You learn that it’s rewarding to punish people who have wronged, so you always want to punish someone when something goes wrong - even if there is nobody but reality to punish.
  • You learn that it feels good to mentally punish someone who is munching too loud, but actually complaining about it would feel petty, and you’ve learned that pettiness is frowned upon. So you also learn to block the impulse to say anything out loud, but continue to get increasingly angry about the sound, causing an escalating circle of both the annoyance and the blocking ramping up in intensity.

As with physical movements, these can form local optima that are hard to get out of. Many of them are learned in childhood, when your understanding of the world is limited. But new behaviors continue to build on top of them, so you will eventually end up with a system which could use a lot of optimization.

If you have more introspective awareness of the exact processes that are happening in your mind, you can make more implicit assumptions conscious, causing your brain’s built-in contradiction detector to notice when they contradict your later learning. Also, getting more feedback about what exactly is happening in your mind allows you to notice more wasted motion in general.

One particular effect is that, as Unlocking the Emotional Brain notes, the mind often makes trade-offs where it causes itself some minor suffering in order to avoid a perceived greater suffering. For example, someone may feel guilt in order to motivate themselves, or experience self-doubt to avoid appearing too confident. By employing greater introspective awareness, one may find ways to achieve their goals without needing to experience any suffering in order to do so.

Of course, Buddhist meditation is not the only way to achieve this. Various therapies and techniques such as Focusing, Internal Family Systems, Internal Double Crux, and so on, are also methods which use introspective awareness to reveal and refactor various assumptions. Increased introspective awareness from meditation tends to also boost the effectiveness of related techniques, as well as reveal more situations where they can be employed.

If introspective awareness is so great, why don’t we have it naturally?

As with anything, there are tradeoffs involved. Having more introspective awareness can help fix a lot of issues… but it also comes with risks, which I assume is the reason why we have not evolved to have a lot of it all the time.

First, it’s worth noting that even for experienced meditators, intense emotional reactions tend to shut down introspective awareness. If one of the functions of e.g. fear and anxiety is to cause a rapid response, then excessive amounts of introspective awareness would slow down that response by reducing cognitive fusion. Many emotions seem to inhibit many competing processes from accessing consciousness, so that you can deal with the situation at hand.

Another consideration involves traumatic memories. In the beginning of the article, I suggested that anxiety is a special kind of mental object which activates particular behaviors. In general, different emotional states have specific kinds of behaviors and activities associated with them - meaning that if you have some memories which are really painful, they can become overwhelming, making it necessary to block them in order to carry on with your normal life. Meditation can be helpful for working through your trauma, but it can also bring it up before you are ready for it, to the point of requiring professional psychotherapy to get through. If you are better at noticing all kinds of subtle details in your mind, it also becomes easier to notice anything that would remind you of things you don’t want to remember. A decrease in introspective awareness seems to be a common trauma symptom, as this helps block the unpleasant memories from being too easily triggered.

I have also heard advanced meditators mention that increased introspective awareness makes it difficult to push away pangs of conscience that they would otherwise have ignored, causing practical problems. For example, people have said that they are no longer able to eat animal products or tell white lies.

On the other hand, extended concentration practice can also make it easier to block things which you would be better off not blocking.

So far, this article has mostly focused on using introspective awareness to notice the content of your thoughts. But you can also use it to notice the structure of the higher-level processes generating your thoughts. Part of how you develop concentration ability is by maintaining introspective awareness of the fact that being able to concentrate on just one thing feels more pleasant than having your attention jump between many different things. This can give you an improved ability to choose what you are concentrating on… but also to selectively exclude anything unpleasant from your mind.

For example, there was an occasion when I needed to do some work, but also had intense anxiety about not wanting to; intense enough that it would normally have made it impossible for me to focus on it. So then I tried to work, and let my introspective awareness observe the feeling of head-splitting agony from my attention alternating between the work and the desire not to… and to also notice that whenever my attention was on the work, I felt temporarily better.

After a while of this, the anxiety started to get excluded from my consciousness, until it suddenly dropped away completely - as if some deeper process had judged it useless and revoked its access to consciousness. And while this allowed me to do the work that I needed to, it also felt internally violent, and like it would be too easy to repress any unpleasant thoughts using it. I still use this kind of technique on occasion when I need to concentrate on something, but I try to be cautious about it.

The negative side of being able to get better feedback about your mental processes, is that you can also get better feedback on exactly how pleasant wireheading feels. If you like to imagine pleasant things, you can get better and better at imagining pleasant things, and excluding any worries about it from your consciousness. Meditation teacher Daniel Ingram warns:

Strong insight and concentration practice, even when that practice wasn’t dedicated to the powers, can make people go temporarily or permanently (or for the rest of that lifetime) psychotic. The more the practice involves creating experiences that diverge significantly from what I will crudely term “consensus reality”, and the longer one engages in these practices, the more likely prolonged difficulties are. It is of note that a significant number of the primary propagators of the Western magickal traditions became moderately nuts towards the ends of their lives.
As one Burmese man said to Kenneth, “My brother does concentration practice. You know, sometimes they go a little mad!” He was talking about what can sometimes happen when people get into the powers. [...]
I remember a letter from a friend who was on a long retreat in Burma and was supposed to be doing insight practices but had slipped into playing with these sorts of experiences. He was now fascinated by his ability to see spirit animals and other supernormal beings and was having regular conversations with some sort of low-level god that kept telling him that he was making excellent progress in his insight practice—that is, exactly what he wanted to hear. However, the fact that he was having stable visionary experiences and was buying into their content made it abundantly clear that he wasn’t doing insight practices at all, but was lost in and being fooled by these.

Now, it should be pointed out that “being able to exclude anything unpleasant from your consciousness” is only going to be a worry for advanced practitioners who spend a lot of time on the kind of practice that inclines you towards these kinds of risks. Before you get to the point of something like this being a risk, you will get to resolve a lot of internal conflicts and old issues first.

Here is Culadasa, the author of The Mind Illuminated, being interviewed about this kind of a “first you resolve a lot of issues, but then you can get the ability to push down the rest” dynamic:

Michael Taft: … and you’re using the meditation practice to help work with your stuff. But what about the other case that we both know of where people have reached very high levels of meditative capacity, they’ve got a lot of insight, maybe they’re at some level of awakening, and they seem to have, in a way, missed a whole pocket of material, or several pockets of material. It’s like they think they’re doing fine, but maybe everyone around them is aware that they’ve got these behavior patterns that do not seem awake at all. And yet the meditation has somehow missed that.
Culadasa: Yes, yes. [...] ... there seems to be a certain level of the stuff that we’re talking about that it’s necessary to deal with to achieve awakening, but it’s sort of a minimal level. [...] What I think that is indicative of is that if that hasn’t been sufficiently dealt with earlier, it has to get dealt with in one way or another at that point. That doesn’t necessarily mean that it’s going to get resolved; it may just get reburied a little more deeply.
Michael Taft: Pushed out of the way.
Culadasa: Yeah, pushed out of the way, or bypassed in some way. That allows a person to go ahead and [progress] and it’s unrealistic to think that everything has been resolved. [...] a lot of the things that change [...] actually help to push these things aside, to bypass them in one way or another, whereas before somebody has [made as much progress] these would have been sufficiently problematic in their life that, in one way or another, they would be aware of them, whether or not they did anything about them or were at a place of just taking for granted that I have these, quote, “personality characteristics” that are a bit difficult.

I used to be very enthusiastic about TMI’s meditation system. I still consider it important and useful to make progress on, but am slightly more guarded after some of my own experiences, hearing about the experience of a friend who reached a high level in it, reading some critiques of its tendency to emphasize awareness of positive experiences [1 2], and considering both the interview quoted above and Culadasa’s subsequent actions. (That said, the focus on positive experiences can be a useful counterbalance for people who start off with an overall negative stance towards life.)

I continue to practice it, and would generally find it safe until you get to around the sixth or so of its ten stages, at which point I would suggest starting to exercise some caution. Off the couch, I mostly don’t do much concentration practice (except in a context where I would need to concentrate anyway). Rather I try to focus my introspective awareness towards just observing my mind without actively interfering with it, Internal Family Systems -style practice, and other activities that do not seem to risk excluding too much unpleasant material.

Finally, developing too much awareness into your mind may cause you to start noticing contradictions between how you thought it worked, and how it actually works. I suspect that a part of how our brains have evolved to operate, relies on those differences going unnoticed. This gets us to the topic of enlightenment, which I have not yet discussed, but will do in my next post.

Thanks to Maija Haavisto, Lumi Pakkanen and Romeo Stevens for comments on an earlier draft.

New to LessWrong?

New Comment
11 comments, sorted by Click to highlight new comments since: Today at 3:37 PM

As someone who sometimes has a hard time coming up with concrete examples, I really like the list of mental motions.

Kaj, first off, huge thank you for writing up all these posts on meditation. I don't have time to read a lot of these books, but it's good literature, so it's perfect to get the summary from someone I can trust.

"Introspective awareness" sounds like the right object. Or, more specifically, it definitely feels like it's describing my own experience. And my own, homegrown hypothesis was something like: consciousness is like an echo or picture-in-picture. We can get glimpses of "ourselves" because we can look at / load partial concepts of ourselves into the working memory.

Introspective awareness is its own type of mental object...

Wow, yes, yes, yes! My original fear with starting meditation was that it would remove certain experiences from my life. It felt like I'd lose something. And in the 7+ years I've done it, that just hasn't happened. And now I have the precise language to describe it. That is exactly what happens for me: like, I can still see the suffering and all that, but it's like it's wrapped in a bubble, which still allows me to see the emotion/sensation exactly like I would before, but it loses the ability to instantly propagate its agenda to the rest of my system.

This kind of a process also teaches you to pay attention to patterns of cause and effect in your mind.

Yeah, I have seen this emphasized in various teachings, but I never actually practiced it. Probably some low hanging fruit here for me. I'll try it out.

“first you resolve a lot of issues, but then you can get the ability to push down the rest” dynamic

One, I think the foremost goal for any type of meditation should be to learn to see what is there. (Just like with rationality.) So there's no pushing or pulling or trying to change anything. And sometimes with bad feelings or headaches, I've noticed that simply paying attention to it (instead of flinching away) resolves it. It doesn't make it "go away" or "block it", it actually just unties the knot, so to speak.

Two, I think there are two orthogonal skills that get bucketed into "spiritual development." One skill is insight (see MCTB for a good definition), but it's basically the ability to see what's going on in your mind. And that skill will eventually take you to enlightenment. The other skill is morality. And while some practices do empathize morality practice as well, I think in the classical western tradition (MCTB is a good example of this) we found paths to get to enlightenment without all other "unnecessary" stuff... like morality training. So you end up with people who are technically enlightened, but it doesn't automatically make them good human beings. (I think this also resolves the confusion in one of the recent SSC posts.)

Thank you very much for your kind words. :)

"Introspective awareness" sounds like the right object. Or, more specifically, it definitely feels like it's describing my own experience. And my own, homegrown hypothesis was something like: consciousness is like an echo or picture-in-picture. We can get glimpses of "ourselves" because we can look at / load partial concepts of ourselves into the working memory.

Yeah, that sounds like it's talking about the same thing. I also quite liked the way Mark Lippmann talks about "afterimages", which seems to also be very closely related:

When you look, see, or notice something, there’s a very predictable pattern that then occurs. First, there is your contact with the actual sensory experience. This is very, very brief. Almost immediately, your mind moves to phase two.

In this phase, you are no longer paying attention to the actual sensory experience, but you are instead paying attention to a sort of “afterimage” of the experience. This is what your mind actually collects and takes away from the outside world, and this is what you actually think about, make sense of, and reason about. [...]

Another way to get a sense of afterimages is to generate a short sound or some other sensory experience and then ask how you know it happened. For example, snap your fingers. Ok. How do you know you just snapped your fingers? You remember you did, right?

Unless you wait too long, part of the experience of that memory is the afterimage of you snapping your fingers. And, there’s often a special property of afterimages that you can play with: You can access the afterimage to more fully replay the experience that led to the afterimage. A replay is not available for some experiences and you might lose the replay for the experience if you wait too long before accessing it. Finally, even if a replay isn’t available, the afterimage may still contain some detail that you can inspect.

So, being aware of and using afterimages is one way that you can inspect subtle phenomena, especially phenomena that goes by very fast.

When doing so, there are some caveats to be aware of.

First, it’s good to remember that the afterimage is not a perfect replica of the experience. It is a “tag” that the experience happened, that may contain or evoke some of the structure or phenomenology of the original experience. If you’re using afterimages to investigate experience, you have to make some effort to to separate out what the experience of the afterimage is versus what remains of the original experience.

Second, it’s important to note that afterimages will always have some conceptual contamination. Afterimages are part top down and part bottom up. That is, afterimages are partially composed of what you expect to see. That’s why you can be positive you just saw a bug skitter across the flow but when you look closely it was just some very suggestive dust caught in a draft. The afterimage is what your reflexes and emotions actually react to, and the afterimage is not the same thing as what was actually there. The way to partially get around this is to try to not have preconceptions and to try to take lots of careful observations of the phenomena.

Finally, there’s a subtler point, here. It seems to be the case that you may be able to “take” or “get” an afterimage only if you already have some inkling of what you’re looking for. That is, if you already have some hint of an idea or concept of what’s there. That doesn’t mean you have to have a name for the experience. And, it doesn’t mean that you’ve had to explicitly reflect, before, on some prior occasion, on having those sorts of experiences. I just means that somewhere in your mind there has to be some sort of… familiarity for the experience before you go looking or paying attention in general.

So, how do you get that initial experience, if you can only have the experience if you’ve had the experience? It seems to “bootstrap” slowly, by simply paying attention in the vicinity of what you’re looking for. You brain eventually, faintly discerns a pattern on the edge of experience, and you gain a creeping sense of familiarity that becomes clearer and clearer, until finally you can put your finger on it, haltingly describe it with great difficulty, and maybe finally name it as a thing or break it down into further parts.

Hey Kaj, I was introduced to your content on craving a week ago and have blown through a number of your posts since then. Thanks for taking the time to compose such informative write-ups!

As someone who is currently around Stage 4-6 of the TMI roadmap, I'd be interested in hearing you elaborate on why you consider TMI post-Stage 6 to be dangerous or risky. If it's dangerous because of suppressing undesired content, what do you think about jhana practices?

Thank you! I'm happy to hear that.

Mostly because of the possibility of suppressing undesired content, and because around Stage 6 was where I noticed myself picking up the ability to just completely exclude unwanted parts of my mind from consciousness. A friend who did TMI all the way to stage 10 also said that roughly this stage was the region where it might get tricky.

To be clear, it doesn't seem to me that TMI past this stage would automatically cause harm: it's just something to be aware of, in case you notice yourself suppressing subminds rather than taking their concerns into account. A TMI teacher who I spoke with said that the most typical outcome of trying to use the system to suppress some of your desires is that you just stop making progress or revert to an earlier stage. The mind does have its defenses against this kind of a thing.

Not sure what you're asking about jhana practices, but I have very little personal experience with them, and I haven't heard them described as having the same problems. (Especially since people typically don't do strong jhanas off the couch, and suppressing subminds in your daily life is the part that seems most risky to me. If it's just that you concentrate really deep during formal meditation, that's probably not an issue.)

This was a great post!

You included some speculation about the stages of TMI. May I ask you what stage of TMI you were in at the time when you wrote this? (If you've been doing enough TMI practice that the question makes sense.)

And in case you've mastered more of the stages since the time you wrote this, do you still believe your speculation in this article was correct?


I was around stage 6, touching 7 on a really good day if I recall correctly. At some point after writing the post, concentration practices stopped working for me (I have some guesses of the reason but no clear answers) and since then I've been at stage 2-3 whenever I've given it another try, sometimes very briefly on 4.

Interesting! Have you written about this experience - i.e., where concentration practices stopped working? (If it's something personal I won't pry, but if you've written about it before I am very curious to read it.)

Very interesting read. Thank you.

Would "pure awareness" technique (like Vipassana or MBSR) have the same dangers?

As the focus is exclusively on real sensation seems like the danger of getting into imagination is if not avoided is significantly reduced.

Would "pure awareness" technique (like Vipassana or MBSR) have the same dangers?

Not these particular dangers, I would think.