Epistemic status: Non-validated conjecture served with a pinch of salt
Category: In-depth analysis of everyday things

Dall-e: “A person with an exploding head, depicted as a nuclear mushroom cloud, digital art“

Do you want to understand why it makes sense to threaten civilian lives with horrific nuclear weapons? Then I can recommend the Game Theory classic “The strategy of conflict[pdf-link]. The book was written way back in 1981, by Thomas C. Schelling. The lessons are as relevant as ever, including ideas on salary negotiations, hostage situations and traffic.[1]

In this post, I’ll apply concepts from the book on inner conflicts, using the Internal Family Systems framework. We’ll use ideas and concepts from the book to explore things like procrastination, not going to bed on time, and the nature of dessert.

The text will end with some practical ideas for how to circumvent the need for draining inner conflicts. Enjoy!

Part 1: The Strategy of Inner Conflict

Context: Internal Family Systems (IFS)

“Internal family systems” (IFS) is a school of thought within psychotherapy. In internal family systems, you picture internal conflicts as a struggle between different parts[2] of yourself. This conflict can then be resolved using techniques from group counselling.

A smoker might have a short-term oriented part that wants to cope with stress by smoking cigarettes, and a long-term oriented part that wants to avoid dying of lung cancer. The back-and-forth most smokers do while quitting can be visualized as a battle raging between these parts,[3] with power changing hands over time.

And with battle comes strategy. Let’s go through some ideas from “The Strategy of Conflict” to see how we can apply them to inner conflict.


Idea 1: Schelling Points

“The Strategy of Conflict” introduces the concept of a “Schelling Point”.[4] Schelling points are natural/obvious places for "negotiations" to end up. To illustrate the point, let's start with a very special kind of negotiation: negotiation without communication.

Example - without communication

You are separated from your friend while shopping at Megamart. Your friend is mobile-free for obscure contrarian reasons. Communication is blocked, but “negotiation” is not; you’re able to “agree” on where to go by trying to pick an “obvious” meeting point. After some thinking, you both head to the checkout line, where you reunite.

In this example, the Schelling point is the checkout line; an obvious place to end up.[5]

Exercise - without communication

You are paired up with an unknown person that lives nearby. You can’t communicate with this person in any way. You are both asked to name an amount of money. If you both name the same amount, you get paid that amount. If you name different amounts, you get NOTHING! What amount of money should you choose?

Answer: The correct solution can be found here. Why is it the solution? 1. It’s a lot of money. 2. It’s an “obvious” amount of money, a natural Schelling point from your shared cultural context.

Perhaps unexpectedly, Schelling points have a large impact on the outcome of “conflicts” with a normal level of communication.

Example 1 - with communication

You and your friends are getting hungry. One of your friends create a Schelling point by saying “I’d love to go to the new Thai place, but I’m open for pretty much anything”. Unless someone has a strong preference and manages to sway the group, you will end up at the new Thai place.

Example 2 - with communication

You are sad about how much your country has shrunk since the days of empire. To remedy the situation you go to therapy invade a neighbouring country. Since this is generally regarded as a dick move, you try to improve optics by establishing a Schelling point: the neighbouring country is actually yours, and you're just reclaiming it. Unfortunately, the neighbouring country has a stronger border-placement Schelling point: Current borders. Nevertheless, your Schelling point will allow your friends and allies to support the invasion without admitting that they are being dicks.[6]


Schelling Points in Inner Conflicts

(note: internal Schelling points are sometimes referred to as “Schelling fences”)

Where are the Schelling points of inner conflicts? Everywhere!

In addiction: Addiction tends to escalate over time. In order to stop this gradual process from going overboard, people use Schelling points to separate “reasonable substance use” from “dangerous substance use”. Examples: limiting smoking to parties, only eating sugary stuff right after dinner,[7] restricting candy to Saturdays.

Other examples: A bedtime that splits the evening into “awake time” and “sleep time”. A budget that decides what spending is ok. Only having two beers.

In all these examples, your long-term parts use Schelling points to create a “defensible point”, where gradual escalations are kept in check. Or rather, this is the intended effect. Most people let their short-term parts overstep repeatedly. More on that later.


Idea 2: Weakness as strength

Back to “The Strategy of Conflict”. In the book, Schelling describes how weakness, in some situations, can be a strength.

Example 1

You are going to an interview for a new job. You are really bad at salary negotiations. Knowing that you will accept their offer without protest if prompted, you decide to make a promise to your partner: “I will not accept a salary without talking to you about it”.

When the prospective employer pressures you to accept an offer, you can calmly tell them that you’ve promised to talk it through with your partner, and can’t commit. You place a restriction on yourself to increase your bargaining power. Clever move.

Example 2

You are taken hostage by an armed stranger. As soon as possible, you ask the kidnapper for earplugs and a blindfold. By blinding yourself, you reduce your ability to testify against the kidnapper in court. This way, you are less of a threat, lowering the risk of the kidnapper killing you to preserve anonymity.

Example 3

You peeked under the blindfold. As you enter the witness stand, you swear a holy oath on your favourite holy book. With the oath, you face the risk of being found guilty of perjury, should it become apparent that you’ve made false statements on the witness stand. By limiting your ability to lie, you increase the weight of your testimony, giving you more power over the hapless kidnapper.

Weakness as Strength in Inner Conflicts

Where does this show up in inner conflicts? Again; everywhere!

By posting your running metrics on Facebook, you risk losing face if you stop running. This benefits your pro-running parts to the detriment of your pro-sofa parts.

By setting a deadline on a project, you increase the consequences of procrastination, until your “let’s-do-stuff” parts overcome the “let’s-play-around” parts.

Recovering alcoholic? Grab a sobriety token. Going to the gym? Sign a long-term contract that’s hard to get out of. Want to eat healthily? Brag to everyone about how healthy you are.

There are even services for enforcing self-punishment.[8]

Part 2: Avoiding Inner Conflict

I don’t like most of the dynamics I’ve written about thus far. They all imply an antagonistic relationship between your different parts. This antagonism creates unnecessary suffering and friction. It’s an unskillful way to live.

Let’s analyse this friction from a game theoretical perspective.[9]


Consequences of inner conflict

Unilateral Action → Rebellion

If a group bases its decisions on informed agreements, it’s much easier to find solutions that benefit all parts. If someone chose to force their chosen path on everyone else, that’s usually to the detriment of the group. Without a shared agreement, the expected outcome is worse for most (if not all) participants.

As such, it makes sense to rebel against tyranny, and punish tyrants. This punishment makes it costly to engage in unilateral action, increasing the attractiveness of seeking agreement.

Here, the oppressor can choose two paths: either settle for agreement or take the hard stance. In the hard stance, the oppressor tries to increase the cost of dissent through threats and punishments. This dynamic can form self-reinforcing spirals, where the parts escalate punishments over time.

Here is a concrete example of this dynamic:

An IFS part makes a unilateral decision to enrol in an extra university course. 
Other parts respond by rebellion through procrastination 
the procrastination is met with a deadline, making it costlier 
the procrastination intensifies, leading to a failed course 
The pro-course part responds with self-shaming thoughts, hurting all the parts. Self-shaming gives the pro-course part a strategic advantage: In future negotiations, it will be able to credibly threaten self-shaming in order to dissuade procrastination.

For more info on the nature of threats, read page 126 of “The Strategy of Conflict”.


Coercion → Opportunistic power-grabs

Another natural response to competitive situations is to strive for more, particularly when an opportunity presents itself. Opportunities form when opponents’ bargaining positions get weaker. If you are used to coercing parts of yourself into submission, you are likely to face opportunistic power grabs, including a tendency I call “escalating oversteps”.

A part engaging in an escalating overstep engages in motivated reasoning similar to:
“Now that I’ve already X, I might as well X+1”.

“Now that I’m already awake past my bedtime, I might as well watch another video.” 
”Now that I’ve already eaten a cookie, I might as well eat another one.”
”Now that I haven’t been to the gym in a week, I might as well skip another workout.”


My instincts tell me that there’s more to be explored when it comes to the nature and consequences of inner conflict. Regardless, I’m fairly sure that the verdict is clear: inner conflict = bad vibes.

I’m also rather sure about the way to get out of internal conflicts: de-escalating the situation. How do we de-escalate inner conflicts? Well, how would you go about de-escalating a normal conflict?

Maybe stop threatening and punishing? Maybe listen to the needs and joys of the other? Gather together, talk, and find a solution that is pleasant to everyone involved!

If you keep missing your bedtime, figure out what your late-night part wants. Maybe it’s sick of being thrown between “shoulds”, and tries to sneak in a period of relaxation. If this is the case, how can you make going to bed more relaxing?

First iteration: an hour before you want to fall asleep, turn off your computer and brew yourself a cup of caffeine-free tea. Sit down and relax, slowly drinking the tea. Then, lie down in bed and listen to a guided yoga Nidra meditation. Iterate as needed.

The key idea here is that your late-night part is probably tired and perhaps not very conscientious. By making a plan that takes your late-night part’s perspective into account, you can satisfy it more thoroughly/efficiently than if it has to haphazardly patch something together while in the middle of rebelling.

Finally, you ARE your parts. By making life better for all versions of yourself you will have a nicer time. Make yourself an offer you don’t want to refuse.

  1. ^

    “I can block your car in the road by placing my car in your way; my deterrent threat is passive, the decision to collide is up to you. If you, however, find me in your way and threaten to collide unless I move, you enjoy no such advantage: the decision to collide is still yours, and I enjoy deterrence. You have to arrange to have to collide unless I move, and that is a degree more complicated.” - Thomas C. Schelling

  2. ^

    These parts correspond fairly well to “shards”, see “shard theory of human values”

  3. ^

    This mental model is the basis of motivational interviewing.

  4. ^

    I’m pretty sure the term was added post-hoc, but who cares

  5. ^

    There is a risk of one person going to the car in the parking lot, or to the “lost & found”, but the strategy is the same. You are unlikely to find your friend waiting by the cereals.

  6. ^

    If any of my readers find themselves in this situation, I recommend therapy over invasions. They are much more cost-efficient, for one.

  7. ^

    HA! Told you I’d get to desserts

  8. ^

    I imagine an alien species scratching their heads, very confused. 

    n.b. The start of this footnote is a trope:y way of indicating an outside-observer perspective, not a literal belief about aliens. I think the nature of aliens is best explained by https://grabbyaliens.com/paper. Hence, I think speculations about silent observer aliens are misconstrued.

  9. ^

    If you are disappointed at the lack of payoff matrices, here’s one for you:

    A lot of this post is about getting to Ii, rather than using threats to try getting IIi (ending up on IIii due to rebellion). Note that this payoff matrix is short-term, I expect there to be a gradual payoff increase (similar to compound interest), which would improve the top row outcomes over time, as more time is spent on the top row. The inverse is true for the bottom row.

    As such, I expect Ii to provide more value over time; the resulting increased amount of top-row time will make the compound interest act in your favour. For more info on how to read payoff matrices, check out this link

New to LessWrong?

New Comment
4 comments, sorted by Click to highlight new comments since: Today at 2:11 PM

Ooh, I think there's a lot of implicit Beeminder criticism here that I'm eager to understand better. Thanks for writing this up! 

We previously argued against similar claims -- https://blog.beeminder.com/blackmail/ -- and said that the "just get the different parts of yourself to get along" school of thought was insufficiently specific about how to do that. But here you've suggested some smart, specific ideas and they sound good!

My other Beeminder defense is that there are certain bare minimums that you know it would be irrational to fall below. So I recommend having the Beeminder goal as insurance and then also implementing all the strategies you describe. If those strategies work and it's easy-peasy to stay well above Beeminder's bright red line, then wonderful. Conflict avoided. If those strategies happen to fail, Beeminder will catch you. (Also you get a nice graph of your progress, quantified-self-style.)

PS: More recently we had a post about how compatible Beeminder turns out to be with CBT which I think also argues against the dichotomy you're implying here with Conflict vs Cooperation. https://blog.beeminder.com/cbt/ 

Didn't expect this reply, thanks for taking your time. I do mention Beeminder briefly at one point, and yes, a lot of the post is about how beeminder-esque motivational strategies tend to backfire.

To start with: I have friends that thrive on coercive motivational strategies. I'm pretty sure my claims aren't universally applicable. However, coercive approaches seems to be a strong cultural norm, and a lot of people use coercive strategies in unskillful ways (leading to procrastination etc). These people might find a lot of value in trying out non-coercive motivational strategies.

Reading your linked pages, I start thinking about what makes coercive motivations (or "self-discipline", as you write on your page) a good fit for some and a bad fit for others. Might write up something about that on my substack in the future, I'll link it to LW if I remember. Also, I'm curious is there a pre/trans dynamic here, where non-coercion after coercison is different to non-coercion from the beginning.

As for your concrete claims:

What are the "smart, specific ideas" I suggested? In this post I mainly attempted to describe what not to do, and ended with some basic non-coercion. I'm curious what you found valuable.

Re: bare minimum that would be irrational to fall below/insurance. Maybe this is correct! I think I would find it hard to mix strategies in this way, since coersion vs non-coersion are pretty far apart as paradigms. A lot of the difference is about how you view yourself. I'm concerned that the coercion might "leak" through, if you keep it as a plan B. But then again, I haven't thought about this, so take it with a pinch of salt :)

Re: CBT & "Conflict vs Cooperation" (I interpret as coersion vs non-coersion). This feedback really tickled my nerd-spot. I'm a practicing stoic, and CBT is basically stoicism without the ontologies/eudaimonia. In my mind, CBT/Stoicism is about shifting personality traits and behavior patterns through changing actions, judgements and thought patterns. These are interconnected, in just the way you're saying, and I agree that it's possible to bootstrap new thought patterns by changing one's actions.

However, this is orthogonal to my post. I'm not claiming that coercive motivational strategies are bad because they are "shallow", I'm claiming that they are bad because they lead to unnecessary friction, and might be full-out counter-productive since it's easy to misuse and act in unskillful ways. The "it doesn't affect fundamental things, we need to be holistic" is a common critique of CBT therapy as well, and I always find it ironic. I find it ironic because the critique assumes it's possible to shift actions without affecting personality, which is a non-holistic perspective on the psyche. Hoisted by their own petard.

The Schelling/Shelling substitution is a bit distracting throughout

I've fixed the spelling, thanks for the correction