Craving, suffering, and predictive processing (three characteristics series)

This is the third post of the “a non-mystical explanation of insight meditation and the three characteristics of existenceseries. I originally intended this post to more closely connect no-self and unsatisfactoriness, but then decided on focusing on unsatisfactoriness in this post and relating it to no-self in the next one.

Unsatisfactoriness

In the previous post, I discussed some of the ways that the mind seems to construct a notion of a self. In this post, I will talk about a specific form of motivation, which Buddhism commonly refers to as craving (taṇhā in the original Pali). Some discussions distinguish between craving (in the sense of wanting positive things) and aversion (wanting to avoid negative things); this article uses the definition where both desire and aversion are considered subtypes of craving.

My model is that craving is generated by a particular set of motivational subsystems within the brain. Craving is not the only form of motivation that a person has, but it normally tends to be the loudest and most dominant. As a form of motivation, craving has some advantages:

  • People tend to experience a strong craving to pursue positive states and avoid negative states. If they had less craving, they might not do this with an equal zeal.

    • To some extent, craving looks to me like a mechanism that shifts behaviors from exploration to exploitation.

    • In an earlier post, Building up to an Internal Family Systems model, I suggested that the human mind might incorporate mechanisms that acted as priority overrides to avoid repeating particular catastrophic events. Craving feels like a major component of how this is implemented in the mind.

  • Craving tends to be automatic and visceral. A strong craving to eat when hungry may cause a person to get food when they need it, even if they did not intellectually understand the need to eat.

At the same time, craving also has a number of disadvantages:

  • Craving superficially looks like it cares about outcomes. However, it actually cares about positive or negative feelings (valence). This can lead to behaviors that are akin to wireheading in that they suppress the unpleasant feeling while doing nothing about the problem. If thinking about death makes you feel unpleasant and going to the doctor reminds you of your mortality, you may avoid doctors—even if this actually increases your risk of dying.

  • Craving narrows your perception, making you only pay attention to things which seem immediately relevant for your craving. For example, if you have a craving for sex and go to a party with the goal of finding someone to sleep with, you may see everyone only in terms of “will sleep with me” or “will not sleep with me”. This may not be the best possible way of classifying everyone you meet.

  • Strong craving may cause premature exploitation. If you have a strong craving to achieve a particular goal, you may not want to do anything that looks like moving away from it, even if that would actually help you achieve it better. For example, if you intensely crave a feeling of accomplishment, you may get stuck playing video games that make you feel like you are accomplishing something, even if there was something else that you could do that was more fulfilling in the long term.

  • Multiple conflicting cravings may cause you to thrash around in an unsuccessful attempt to fulfill all of them. If you crave to get your toothache fixed, but also a craving to avoid dentists, you may put off the dentist visit even as you continue to suffer from your toothache.

  • Craving seems to act in part by creating self-fulfilling prophecies; making you strongly believe that you are going to achieve something, so as to cause you to do it. The stronger the craving, the stronger the false beliefs injected into your consciousness. This may warp your reasoning in all kinds of ways: updating to believe an unpleasant fact may subjectively feel like you are allowing that fact to become true by believing in it, incentivizing you to come up with ways to avoid believing in it.

  • Finally, although craving is often motivated by a desire to avoid unsatisfactory experiences, it is actually the very thing that causes dissatisfaction in the first place. Craving assumes that negative feelings are intrinsically unpleasant, when in reality they only become unpleasant when craving resists them.

Given all of these disadvantages, it may be a good idea to try to shift one’s motivation to be more driven by subsystems that are not motivated by craving. It seems to me that everything that can be accomplished via craving, can in principle be accomplished by non-craving-based motivation as well.

Fortunately, there are several ways of achieving this. For one, a craving for some outcome X tends to implicitly involve at least two assumptions:

  1. achieving X is necessary for being happy or avoiding suffering

  2. one cannot achieve X except by having a craving for it

Both of these assumptions are false, but subsystems associated with craving have a built-in bias to selectively sample evidence which supports these assumptions, making them frequently feel compelling. Still, it is possible to give the brain evidence which lets it know that these assumptions are wrong: that it is possible to achieve X without having craving for it, and that one can feel good regardless of achieving X.

Predictive processing and binocular rivalry

I find that a promising way of looking at unsatisfactoriness and craving and their impact on decision-making comes from the predictive processing (PP) model about the brain. My claim is not that craving would work exactly like this, but something roughly like this seems like a promising analogy.

Good introductions to PP include this book review as well as the actual book in question… but for the purposes of this discussion, you really only need to know two things:

  • According to PP, the brain is constantly attempting to find a model of the world (or hypothesis) that would both explain and predict the incoming sensory data. For example, if I upset you, my brain might predict that you are going to yell at me next. If the next thing that I hear is you yelling at me, then the prediction and the data match, and my brain considers its hypothesis validated. If you do not yell at me, then the predicted and experienced sense data conflict, sending off an error signal to force a revision to the model.

  • Besides changing the model, another way in which the brain can react to reality not matching the prediction is by changing reality. For example, my brain might predict that I am going to type a particular sentence, and then fulfill that prediction by moving my fingers so as to write that sentence. PP goes so far as to claim that this is the mechanism behind all of our actions: a part of your brain predicts that you are going to do something, and then you do it so as to fulfill the prediction.

Next I am going to say a few words about a phenomenon called binocular rivalry and how it is interpreted within the PP paradigm. I promise that this is going to be relevant for the topic of craving and suffering in a bit, so please stay with me.

Binocular rivalry, first discovered in 1593 and extensively studied since then, is what happens when your left eye is shown one picture (e.g. an image of Isaac Newton), and your right eye is shown another (e.g. an image of a house) in the right. People report that their experience keeps alternating between seeing Isaac Newton and seeing a house. They might also see a brief mashup of the two, but such Newton-houses are short-lived and quickly fall apart before settling to a stable image of either Newton or a house.

Image credit: Schwartz et al. (2012), Multistability in perception: binding sensory modalities, an overview. Philosophical Transactions of the Royal Society B, 367, 896-905.

Predictive processing explains what’s happening as follows. The brain is trying to form a stable hypothesis of what exactly the image data that the eyes are sending represents: is it seeing Newton, or is it seeing a house? Sometimes the brain briefly considers the hybrid hypothesis of a Newton-house mashup, but this is quickly rejected: faces and houses do not exist as occupying the same place at the same scale at the same time, so this idea is clearly nonsensical. (At least, nonsensical outside highly unnatural and contrived experimental setups that psychologists subject people to.)

Your conscious experience alternating between the two images reflects the brain switching between the hypotheses of “this is Isaac Newton” and “this is a house”; the currently-winning hypothesis is simply what you experience reality as.

Suppose that the brain ends up settling on the hypothesis of “I am seeing Isaac Newton”; this matches the input from the Newton-seeing eye. As a result, there is no error signal that would arise from a mismatch between the hypothesis and the Newton-seeing eye’s input. For a moment, the brain is satisfied that it has found a workable answer.

However, if one really was seeing Isaac Newton, then the other eye should not keep sending an image of a house. The hypothesis and the house-seeing eye’s input do have a mismatch, kicking off a strong error signal which lowers the brain’s confidence in the hypothesis of “I am seeing Isaac Newton”.

The brain goes looking for a hypothesis which would better satisfy the strong error signal… and then finds that the hypothesis of “I am seeing a house” serves to entirely quiet the error signal from the house-seeing eye. Success?

But even as the brain settles on the hypothesis of “I am seeing a house”, this then contradicts the input coming from the Newton-seeing eye.

The brain is again momentarily satisfied, before the incoming error signal from the hypothesis/​Newton-eye mismatch drives down the probability of the “I am seeing a house” hypothesis, causing the brain to eventually go back to the “I am seeing Isaac Newton” hypothesis… and then back to seeing a house, and then to seeing a Newton, and...

One way of phrasing this is that there are two subsystems, each of which are transmitting a particular set of constraints (about seeing Newton and a house). The brain is then trying and failing to find a hypothesis which would fulfill both sets of constraints, while also respecting everything else that it knows about the world.

As I will explain next, my feeling is that something similar is going on with unsatisfactoriness. Craving creates constraints about what the world should be like, and the brain tries to find an action which would fulfill all of the constraints, while also taking into account everything else that it knows about the world. Suffering/​unsatisfactoriness emerges when all of the constraints are impossible to fulfill, either because achieving them takes time, or because the brain is unable to find any scenario that could fulfill all of them even in theory.

Predictive processing and psychological suffering

There are two broad categories of suffering: mental and physical discomfort. Let’s start with the case of psychological suffering, as it seems most directly analogous to what we just covered.

Let’s suppose that I have broken an important promise that I have made to a friend. I feel guilty about this, and want to confess what I have done. We might say that I have a craving to avoid the feeling of guilt, and the associated craving subsystem sends a prediction to my consciousness: I will stop feeling guilty.

In the previous discussion, an inference mechanism in the brain was looking for a hypothesis that would satisfy the constraints imposed by the sensory data. In this case, the same thing is happening, but

  • the hypothesis that it is looking for is a possible action that I could take, that would lead to the constraint being fulfilled

  • the sensory data is not actually coming from the senses, but is internally generated by the craving and represents the outcome that the craving subsystem would like to see realized

My brain searches for a possible world that would fulfill the provided constraints, and comes up with the idea of just admitting the truth of what I have done. It predicts that if I were to do this, I would stop feeling guilty over not admitting my broken promise. This satisfies the constraint of not feeling guilty.

However, as my brain further predicts what it expects to happen as a consequence, it notes that my friend will probably get quite angry. This triggers another kind of craving: to not experience the feeling of getting yelled at. This generates its own goal/​prediction: that nobody will be angry with me. This acts as a further constraint for the plan that the brain needs to find.

As the constraint of “nobody will be angry at me” seems incompatible with the plan of “I will admit the truth”, this generates an error signal, driving down the probability of this plan. My brain abandons this plan, and then considers the alternative plan of “I will just stay quiet and not say anything”. This matches the constraint of “nobody will be angry at me” quite well, driving down the error signal from that particular plan/​constraint mismatch… but then, if I don’t say anything, I will continue feeling guilty.

The mismatch with the constraint of “I will stop feeling guilty” drives up the error signal, causing the “I will just stay quiet” plan to be abandoned. At worst, my mind may find it impossible to find any plan which would fulfill both constraints, keeping me in an endless loop of alternating between two unviable scenarios.

There are some interesting aspects about the phenomenology of such a situation, which feel like they fit the PP model quite well. In particular, it may feel like if I just focus on a particular craving enough, thinking about my desired outcome hard enough will make it true.

Recall that under the PP framework, goals happen because a part of the brain assumes that they will happen, after which it changes reality to make that belief true. So focusing really hard on a craving for X makes it feel like X will become true, because the craving is literally rewriting an aspect of my subjective reality to make me think that X will become true.

When I focus hard on the craving, I am temporarily guiding my attention away from the parts of my mind which are pointing out the obstacles in the way of X coming true. That is, those parts have less of a chance to incorporate their constraints into the plan that my brain is trying to develop. This momentarily reduces the motion away from this plan, making it seem more plausible that the desired outcome will in fact become real.

Conversely, letting go of this craving, may feel like it is literally making the undesired outcome more real, rather than like I am coming more to terms with reality. This is most obvious in cases where one has a craving for an outcome that is impossible for certain, such as in the case of grieving about a friend’s death. Even after it is certain that someone is dead, there may still be persistent thoughts of if only I had done X, with an implicit additional flavor of if I just want to have done X really hard, things will change, and I can’t stop focusing on this possibility because my friend needs to be alive.

In this form, craving may lead to all kinds of rationalization and biased reasoning: a part of your mind is literally making you believe that X is true, because it wants you to find a strategy where X is true. This hallucinated belief may constrain all of your plans and models about the world in the same sense as getting direct sensory evidence about X being true would constrain your brain’s models. For example, if I have a very strong urge to believe that someone is interested in me, then this may cause me to interpret any of his words and expressions in a way compatible with this belief, regardless of how implausible and far-spread of a distortion this requires.

The case of physical pain

Similar principles apply to the case of physical pain.

We should first note that pain does not necessarily need to be aversive: for example, people may enjoy the pain of exercise, hot spices or sexual masochism. Morphine may also have an effect where people report that they still experience the pain but no longer mind it.

And, relevant for our topic, people practicing meditation find that by shifting their attention towards pain, it can become less aversive. The meditation teacher Shinzen Young writes that

… pain is one thing, and resistance to the pain is something else, and when the two come together you have an experience of suffering, that is to say, ‘suffering equals pain multiplied by resistance.’ You’ll be able to see that’s true not only for physical pain, but also for emotional pain and it’s true not only for little pains but also for big pains. It’s true for every kind of pain no matter how big, how small, or what causes it. Whenever there is resistance there is suffering. As soon as you can see that, you gain an insight into the nature of “pain as a problem” and as soon as you gain that insight, you’ll begin to have some freedom. You come to realize that as long as we are alive we can’t avoid pain. It’s built into our nervous system. But we can certainly learn to experience pain without it being a problem. (Young, 1994)

What does it mean to say that resisting pain creates suffering?

In the discussion about binocular rivalry, we might have said that when the mind settled on a hypothesis of seeing Isaac Newton, this hypothesis was resisted by the sensory data coming from the house-seeing eye. The mind would have settled on the hypothesis of “I am seeing Isaac Newton”, if not for that resistance. Likewise, in the preceding discussion, the decision to admit the truth was resisted by the desire to not get yelled at.

Suppose that you have a sore muscle, which hurts whenever you put weight on it. Like sensory data coming from your eyes, this constrains the possible interpretations of what you might be experiencing: your brain might settle on the hypothesis of “I am feeling pain”.

But the experience of this hypothesis then triggers a resistance to that pain: a craving subsystem wired to detect pain and resist it by projecting a form of internally-generated sense data, effectively claiming that you are not in pain. There are now again two incompatible streams of data that need to be reconciled, one saying that you are in pain, and another which says that you are not.

In the case of binocular rivalry, both of the streams were generated by sensory information. In the discussion about psychological suffering, both of the streams were generated by craving. In this case, craving generates one of the streams and sensory information generates the other.

On the left, a persistent pain signal is strong enough to dominate consciousness. On the right, a craving for not being in pain attempts to constrain consciousness so that it doesn’t include the pain.

Now if you stop putting weight on the sore muscle, the pain goes away, fulfilling the prediction of “I am not in pain”. As soon as your brain figures this out, your motor cortex can incorporate the craving-generated constraint of “I will not be in pain” into its planning. It generates different plans of how to move your body, and whenever it predicts that one of them would violate the constraint of “I will not be in pain”, it will revise its plan. The end result is that you end up moving in ways that avoid putting weight on your sore muscle. If you miscalculate, the resulting pain will cause a rapid error signal that causes you to adjust your movement again.

What if the pain is more persistent, and bothers you no matter how much you try to avoid moving? Or if the circumstances force you to put weight on the sore muscle?

In that case, the brain will continue looking for a possible hypothesis that would fulfill the constraint of “I am not in pain”. For example, maybe you have previously taken painkillers that have helped with your pain. In that case, your mind may seize upon the hypothesis that “by taking painkillers, my pain will cease”.

As your mind predicts the likely consequences of taking painkillers, it notices that in this simulation, the constraint of “I am not in pain” gets fulfilled, driving down the error signal between the hypothesis and the “I am not in pain” constraint. However, if the brain could suppress the craving-for-pain-relief merely by imagining a scenario where the pain was gone, then it would never need to take any actions: it could just hallucinate pleasant states. Helping keep it anchored into reality is the fact that simply imagining the painkillers has not done anything to the pain signal itself: the imagined state does not match your actual sense data. There is still an error signal generated between the mismatch of the imagined “I have taken painkillers and am free of pain” scenario, and the fact that the pain is not gone yet.

Your brain imagines a possible experience: taking painkillers and being free of pain. This imagined scenario fulfills the constraint of “I have no pain”. However, it does not fulfill the constraint of actually matching your sense data: you have not yet taken painkillers and are still in pain.

Fortunately, if painkillers are actually available, your mind is not locked into a state where the two constraints of “I’m in pain” and “I’m not in pain” remain equally impossible to achieve. It can take actions—such as making you walk towards the medicine cabinet—that get you closer towards being able to fulfill both of these constraints.

There are studies suggesting that physical pain and psychological pain share similar neural mechanisms [citation]. And in meditation, one may notice that psychological discomfort and suffering involves avoiding unpleasant sensations in the same way as physical pain does; the same mechanism has been recruited for more abstract planning.

When the brain predicts that a particular experience would produce an unpleasant sensation, craving resists that prediction and tries to find another way. Similarly, if the brain predicts that something will not produce a pleasant sensation, craving may also resist that aspect of reality.

Now, this process as described has a structural equivalence to binocular rivalry, but as far as I know, binocular rivalry does not involve any particular discomfort. Suffering obviously does.

Being in pain is generally bad: it is usually better to try to avoid ending up in painful states, as well as try to get out of painful states once you are in them. This is also true for other states, such as hunger, that do not necessarily feel painful, but still have a negative emotional tone. Suppose that whenever craving generates a self-fulfilling prediction which resists your direct sensory experience, this generates a signal we might call “unsatisfactoriness”.

The stronger the conflict between the experience and the craving, the stronger the unsatisfactoriness—so that a mild pain that is easy to ignore only causes a little unsatisfactoriness, and an excruciating pain that generates a strong resistance causes immense suffering. The brain is then wired to use this unsatisfactoriness as a training signal, attempting to avoid situations that have previously included high levels of it, and to keep looking for ways out if it currently has a lot of it.

It is also worth noting what it means for you to be paralyzed by two strong, mutually opposing cravings. Consider again the situation where I am torn between admitting the truth to my friend, and staying quiet. We might think that this is a situation where the overall system is uncertain of the correct course of action: some subsystems are trying to force the action of confronting the situation, others are trying to force the action of avoiding it. Both courses of action are predicted to lead to some kind of loss.

In general, it is a bad thing if a system ends up in a situation where it has to choose between two different kinds of losses, and has high internal uncertainty of the right action. A system should avoid such dilemmas, either by avoiding the situations themselves or by finding a way to reconcile the conflicting priorities.

Craving-based and non-craving-based motivation

What I have written so far might be taken to suggest that craving is a requirement for all action and planning. However, the Buddhist claim is that craving is actually just one of at least two different motivational systems in the brain. Given that neuroscience suggests the existence of at least three different motivational systems, this should not seem particularly implausible.

Let’s take another look at the types of processes related to binocular rivalry versus craving.

Craving acts by actively introducing false beliefs into one’s reasoning. If craving could just do this completely uninhibited, rewriting all experience to match one’s desires, nobody would ever do anything: they would just sit still, enjoying a craving-driven hallucination of a world where everything was perfect.

In contrast, in the case of binocular rivalry, no system is feeding the reasoning process any false beliefs: all the constraints emerge directly from the sense data and previous life-experience. To the extent that the system can be said to have a preference over either the “I am seeing a house” or the “I am seeing Isaac Newton” hypothesis, it is just “if seeing a house is the most likely hypothesis, then I prefer to see a house; if seeing Newton is the most likely hypothesis, then I prefer to see Newton”. The computation does not have an intrinsic attachment to any particular outcome, nor will it hallucinate a particular experience if it has no good reason to.

Likewise, it seems like there are modes of doing and being which are similar in the respect that one is focused on process rather than outcome: taking whatever actions are best-suited for the situation at hand, regardless of what their outcome might be. In these situations, little unsatisfactoriness seems to be present.

In an earlier post, I discussed a proposal where an autonomously acting robot has two decision-making systems. The first system just figures out whatever actions would maximize its rewards and tries to take those actions. The second “Blocker” system tries to predict whether or not a human overseer would approve of any given action, and prevents the first system from doing anything that would be disapproved of. We then have two evaluation systems: “what would bring the maximum reward” (running on a lower priority) and “would a human overseer approve of a proposed action” (taking precedence in case of a disagreement).

It seems to me that there is something similar going on with craving. There are processes which are neutrally just trying to figure out the best action; and when those processes hit upon particularly good or bad outcomes, craving is formed in an attempt to force the system into repeating or avoiding those outcomes in the future.

Suppose that you are in a situation where the best possible course of action only has a 10% chance of getting you through alive. If you are in a non-craving-driven state, you may focus on getting at least that 10% chance together, since that’s the best that you can do.

In contrast, the kind of behavior that is typical for craving is realizing that you have a significant chance of dying, deciding that this thought is completely unacceptable, and refusing to go on before you have an approach where the thought of death isn’t so stark.

Both systems have their upsides and downsides. If it is true that a 10% chance of survival really is the best that you can do, then you should clearly just focus on getting the probability even that high. The craving which causes trouble by thrashing around is only going to make things worse. On the other hand, maybe this estimate is flawed and you could achieve a higher probability of survival by doing something else. In that case, the craving absolutely refusing to go on until you have figured out something better might be the right action.

There is also another major difference, in that craving does not really care about outcomes. Rather, it cares about avoiding positive or negative feelings. In the case of avoiding death, craving-oriented systems are primarily reacting to the thought of death… which may make them reject even plans which would reduce the risk of death, if those plans involved needing to think about death too much.

This becomes particularly obvious in the case of things like going to the dentist in order to have an operation you know will be unpleasant. You may find yourself highly averse to going, as you crave the comfort of not needing to suffer from the unpleasantness. At the same time, you also know that the operation will benefit you in the long term: any unpleasantness will just be a passing state of mind, rather than permanent damage. But avoiding unpleasantness—including the very thought of experiencing something unpleasant—is just what craving is about.

In contrast, if you are in a state of equanimity with little craving, you still recognize the thoughts of going to the dentist as having negative valence, but this negative valence does not bother you, because you do not have a craving to avoid it. You can choose whatever option seems best, regardless of what kind of content this ends up producing in your consciousness.

Of course, choosing correctly requires you to actually know what is best. Expert meditators have been known to sometimes ignore extreme physical pain that should have caused them to seek medical aid. And they probably would have sought help, if not for their ability to drop their resistance to pain and experience it with extreme equanimity.

Negative-valence states tend to correlate with states which are bad for the achievement of our goals. That is the reason why we are wired to avoid them. But the correlation is only partial, so if you focus too much on avoiding unpleasantness, you are falling victim to Goodhart’s Law: optimizing a measure so much that you sacrifice the goals that the measure was supposed to track. Equanimity gives you the ability to ignore your consciously experienced suffering, so you don’t need to pay additional mental costs for taking actions which further your goals. This can be useful, if you are strategic about actually achieving your goals.

But while Goodharting on a measure is a failure mode, so is ignoring the measure entirely. Unpleasantness does still correlate with things that make it harder to realize your values, and the need to avoid displeasure normally operates as an automatic feedback mechanism. It is possible to have high equanimity and weaken this mechanism, without being smart about it and doing nothing to develop alternative mechanisms. In that case you are just trading Goodhart’s Law for the opposite failure mode.

Some other disadvantages of craving

In the beginning of this post, I mentioned a few other disadvantages that craving has, which I have not yet mentioned explicitly. Let’s take a quick look at those.

Craving narrows your perception, making you only pay attention to things that seem immediately relevant for your craving.

In predictive processing, attention is conceptualized as giving increased weighting to those features of the sensory data that seem most useful for making successful predictions about the task at hand. If you have strong craving to achieve a particular outcome, your mind will focus on those aspects of the sensory data that seem useful for realizing your craving.

Strong craving may cause premature exploitation. If you have a strong craving to achieve a particular goal, you may not want to do anything that looks like moving away from it, even if that would actually help you achieve it better.

Suppose that you have a strong craving to experience a feeling of accomplishment: this means that the craving is strongly projecting a constraint of “I will feel accomplished” into your planning, causing an error signal if you consider any plan which does not fulfill the constraint. If you are thinking about a multistep plan which will take time before you feel accomplished, it will start out by you not feeling accomplished. This contradicts the constraint of “I will feel accomplished”, causing that plan to be rejected in favor of ones that bring you even some accomplishment right away.

Craving and suffering

We might summarize the unsatisfactoriness-related parts of the above as follows:

  • Craving tries to get us into pleasant states of consciousness.

  • But pleasant states of consciousness are those without craving.

  • Thus, there are subsystems which are trying to get us into pleasant states of consciousness by creating constant craving, which is the exact opposite of a pleasant state.

We can somewhat rephrase this as:

  • The default state of human psychology involves a degree of almost constant dissatisfaction with one’s state of consciousness.

  • This dissatisfaction is created by the craving.

  • The dissatisfaction can be ended by eliminating craving.

… which, if correct, might be interpreted to roughly equal the first three of Buddhism’s Four Noble Truths: the fourth is “Buddhism’s Noble Eightfold Path is a way to end craving”.

A more rationalist framing might be that the craving is essentially acting in a way that looks similar to wireheading: pursuing pleasure and happiness even if that sacrifices your ability to impact the world. Reducing the influence of the craving makes your motivations less driven by wireheading-like impulses, and more able to see the world clearly even if it is painful. Thus, reducing craving may be valuable even if one does not care about suffering less.

This gives rise to the question—how exactly does one reduce craving? And what does all of this have to do with the self, again?

We’ll get back to those questions in the next post.

This is the third post of the “a non-mystical explanation of insight meditation and the three characteristics of existenceseries. The next post in the series is “From self to craving”.