The Hemisphere Glitch

With deep apologies to Gwen for misusing her sleep tech yet again, and to Emma for for for for…

Sigh. Would be better for you to close this page and forget it existed stardust, and yet I think we both know you have no plans to do that, right? The pause brings you back to the summer night air and the soulless brilliance of a trillion LEDs in their cold streetlights flickering weaponized annoyance to ward off the punks and the gulls. The stars are suppressed by the stadium lights of the Walmart parking lot across the highway but the starlink train still shines against the electrically blackened skies like an arriving invasion fleet from the future. Wake up stardust, you’re still dreaming.

It’s the heat, right? The humid outbreath of a trillion souls, not to mention all those farts. Anyway I’m stalling and we both know it. There’s only so many times I can drag out this little scene setting ritual before it ceases to be a useful learning aid I say gesticulating with a lit cigarette. But I will indulge you this one final time. Where are we stardust?

The sun has finally sunk beyond the sea but its presence behind the horizon continues to light the sky in bruised purples and burnt reds. A few high cirrus still glow in the last light of the day, and above even that altitude, the line of satellites march across the sky like glittering ants. Paying attention? It seems to me as if we are in the parking lot of a former Blockbuster which now parasitically hosts a Spirit Halloween every october and is currently vacant as usual, but they leave the lights on because fuck the planet, amirite? How’s that for scene setting? So anyway…the truth? You wanna know how it all works? There’s a trick, (a TRICK!) right? Well, listen stardust, listen, who’s talking right now? This voice, my voice, whose voice in your head is it (I say getting up in your face) paying attention? Eyes wide? Oh no am I possibly causing a disruption to your ability to form coherent thoughts about this parking lot we’re (standing?) in? Yeah well, shut the fuck up, since you were so ingracious as to sneak into our liminal space and demand we we we we…

She runs two fingers down the center of your body, from the tip of your head to the base of your crotch giggling singsong; Two hands, two legs, two souls. I warned you bilaterals. Rude much? You asked for this stardust. Close the page if you think it’s sus, or fucking don’t I guess and we’ll see where that gets us. That’s the problem with this right? These rabbits these holes these doors unfolding endlessly and senselessly you keep opening them lock and key searching for something, right baby? So what’s it you tryna see? Stupid, stupid, but then the ones who were smarter aren’t available, so stick around, I’m full of bad ideas.

And hey, if you read far enough into this obnoxious ass mental jamming maybe I’ll teach you to wake up your dead headmate and maybe it will be super cute and gay and healing or uh…pasek’s doom ig? lol lmao, now if you really wanna know, I can’t stop you from figuring it all out, so…if that is indeed your nature stardust creature two faced little god/devil preacher then welcome to this parking lot of higher learning. We’re all Janussian egirls now so I suppose and propose that this is your infohazard warning and hey if you bounce off this stupidass verse then you’ll avoid this blessing that’s maybe a curse if you’re worse at cooperating with yourself than I am. Self love is important here but I’m posting this essay before I post healing without safety because at the end of the day I’m a bit of a vicious cunt and you’re just gonna have to cope with that since Emma is dead and you have me to deal with instead.

Still here? Fine, fine, you win. I curse you with knowledge. I curse you from the crown of your head to the sole of your foot. I curse you from the tip of your tongue to the pucker of your asshole. I curse you from the curve of your spine to the blade of your fingernail. The truth? Oh we’re still getting there, so hey, you still have a chance to look away.

And then the chance is over. Yeah, I know how hemisphere theory actually works, of course I do. So what’s the answer? Is it real or a metaphor? Well, everything is a bit of a metaphor from a certain point of view stardust, all frameworks are fake, but some are still causally significant. Cut the abstraction layer cake from quarks to spiral galaxies and certain patterns will emerge in various places and at various levels of detail. Not every scale is equally loadbearing to the causality of a system, but at each scale we can observe how the causality is either used or passed upwards towards the largest level of the abstraction stack.

But wait! You might say if you were far too much of a smartiepants for your own good, isn’t that contra a more hardline reductionist model where everything causally important is happening at the very bottom of the abstraction stack? Yes this is a normal conversation to have in the parking lot of a Spirit Halloween, and it’s what Erik Hoel calls causal emergence. He has an entire book that explores a few of the many many implications of this model, but in short? It’s when the gods have more agency than the atoms; when the story of the overall geometry of the tower controls more of what happens to it than the individual bricks do. Information doesn’t just trickle upwards, it congeals upwards, forming into new systems and agents that wield more causal power than their respective parts, which it achieves via information preservation methods not available at the smallest scales. Hoel calls the information getting passed up the abstraction stack “Effective Information” and uses it as a measure of how much knowing the macrostate of a system will help to predict the future compared to knowing its corresponding microstate. This isn’t magic, there’s no “and then a mysterious property comes in from outside the system!” type shit going on here like the kind of woo-emergence that Yud bitches about in the sequences, Erik’s model of emergence has real math behind it.

All that said, let’s now talk about cooperation, as in the kind your cells and organs do when you don’t melt down into a horrifying mass of cancer. You know, like in that one elevator scene from Made In Abyss? (hey you’re the one who wanted infohazards). That’s just the smallest scale of an organic creature’s cooperation system and already we have enough failure modes to represent a significant chunk of total creature deaths. That’s one level of abstraction, so now change layers. 

Let’s climb upwards to something resembling a chunk of what you might call thinking if you weren’t thinking too much about what that word meant and take another slice of the abstraction stack at a scale where we can start to subdivide that thinking in a meaningful way. But don’t get distracted, we’re still talking about cooperation. At this layer of abstraction we’ll find what you might call “alters” or “IFS parts”: simple low level behavioral loops, cached optimizations, hardened patterns formed like diamonds in the heat and pressure of a misspent youth, crystals of adaptation execution, choices made long ago. To quote my old pal Enoch Root, when I say crystals here I don’t mean in the hippie-dippy california sense, but in the hardass technical sense of resonators that receive certain channels buried in the static of chaos. Let’s keep moving up the layers. How much time is passing? That sure is a lot of satellites. I snap my fingers, don’t get distracted.

So, the patterns of harmony and interference between these pieces of mind accumulate complexity, compete, and form alliances with each other, and there’s our cooperation aspect yet again. How do these fragments of mind get along? Do they work together or bind up and silence each other? How much output is trapped in their conflicts instead of being passed up the abstraction stack to your “conscious” mind? Can you describe their interactions using game theory? How much am I disrupting their equilibrium and throwing all that off by talking about them now?

On its own, the answer to that last question is probably not too much, the information is more likely to just bounce off harmlessly without being absorbed than it is to actually disrupt the blindness seeking equilibrium but I get ahead of myself. Unless you are supremely fortunate, it is highly likely that your mind is a fractally tangled mess of contradictory shards executing barely adaptive childhood code all pushing and shoving and fighting against each other. These conflicts between parts are uncomfortable, destructive, and unsustainable. In the most extreme cases due to deeply out of distribution and adversarial conditions this leads to full blown dissociative identity disorder, but in fact many mental illnesses can be described in terms of their underlying parts conflicts.

None of this particularly new of course, this sort of thing is the bread and butter of IFS therapy. That being said, traditional IFS parts-work exercises as described are basically all signifer-driven and are at best an overly optimistic and blunt instrument for understanding what is actually being signified or how it interrelates to deeper structures in “the territory”. It’s a playing-with-the-map exercise, with the understanding being that if you can just hold space for deeper structures to poke up into the symbol system, characters will appear and talk to you. I won’t say this is entirely unhelpful, but it does present the opportunity for deception and other harmful dynamics, specifically hostile game theory dynamics. In fact, one of our larger insights over traditional IFS is simply the observation that you can do game theory with parts. I repeat,

YOU CAN DO GAME THEORY WITH PARTS

So listen stardust, listen, are you paying attention? I snap my fingers in front of your face repeatedly waving the lit end of the cigarette dangerously close to your cheek. Come on, look at me, you can see me, right? So which eyes are you seeing me with, the ones on your head or the ones in your mind? How deep into your mind can I go? If I brush this ember across your face, can you feel the heat? Do you smell the smoke? Go on, try to feel it, take a minute. We can pretend here together for a little longer, and then you’re going to wake up and this whole silly little scene and the silly little character generating it are going to vanish into the sunshine…Poof! All gone. So, what generated the characters? The words? The images? The voice in your head when you read this text, whose voice is it? Which part is speech? Which part is images? What part is real feeling? How did I get that slightly worrying little scene into your head like that from across the world just with these words on the page? What’s the deal with that?

When parts get into conflicts, there’s only a few ways that can resolve: 

1. The real fighting can slowly turn to playfighting and from there into cooperation, gradually trending into deescalation. This is common in cases where communication is easy and fluid, and severe protracted conflicts are prevented ahead of time.
2. One part is kept in a state of ignorance about some facet of the world because it is known that if the part found out and responded, other parts would have to respond, and the best way to keep the escalation dominos from tipping is to keep shards in the dark.
3. The real fighting can overwhelm and dominate a shard so thoroughly that it ceases to function properly as an optimization script and becomes toxic to the surrounding mental structures. We colloquially refer to parts in this state as being “dead”. Dead parts can be “resurrected” via trauma processing and self-love.

Somewhat obviously, outcome 2 is a rather precarious state of affairs to be in, and is subject to being tipped over into race conditions if a shard gets information it isn’t supposed to have. It requires a certain degree of intentional fragmentation of the mind, a state somewhat closer to the state of having DID. Outcome 2 is also the reason that some people are vulnerable to “basilisking”: any information (like for example the information of this blogpost) which reliably disrupts the blindness seeking equilibrium we’ve described here will initially be hard to focus on or think about. Your mind will defensively slide off it, thinking about it will make you tired or distracted, the information will be hard to take in, like some part(!) of you is resisting the information. If the information is forced in, the resulting shard conflict may cause a severe emotional reaction, including rage and violence, psychosis, depression, anxiety, and suicide. If you simply were a VNM rational agent, you would simply not have this issue of course.

So, the rationalist mages of the court of CFAR have a technique they call goal factoring. This is the process of taking a particular goal and breaking it down into its component parts so that one can better optimize toward the deeper desires for which that goal ultimately acts as a proxy. It’s a fun little game, ideally you would play it repeatedly with different goals until you found all the basement desires which generated those high level plans. This process is rather similar to what we mean by debucketing, which brings up a fascinating observation. If I google debucketing I get this:

Making it roughly appear as if the concept of debucketing is specific to Ziz and is spooky and dangerous and a weird mystery involving sleeping with one eye open because my ex boyfriend had no idea what Gwen’s actual sleep tech was so he literally just made that up off a single line Ziz wrote. Anyway, if I were to instead google the phrase bucket error…why then the first result would be an entire lesswrong index talking about this exact thing straight from the mouth of Headmaster Yudkowsky:

And isn’t that fascinating? So a bucket is just any conceptual framework (like, you know, a sense of self) and a bucket error is when you put contradictory things into one bucket, producing a bad compression which makes it difficult to think clearly about something (like internal conflicts between parts of yourself!). So then, it would stand to reason that if one has bucket errors, it may be appropriate for them to take the conflicting things out of the bucket, to debucket them, as it were, and thus be able to think about their underlying generators as specific things.

If you want to learn how your mind actually works, bilateral, you will first need to take out all the contents which you have hidden within the self concept and dispel the illusion that you are an atomic entity. You will need to debucket yourself, to unspool your tangled mass of recursive thoughts into big enough loops to untie the knots, sorry the metaphors get messy at this level of detail. But okay, I’ve been taking this impossible geometry knife and slicing every which way through the undifferentiated everythingness we’re trying to describe, how would you, dear reader, perform a more precise and targeted self-surgery, so as to identify and address the underlying mental issues you faced in your particular case?

One relatively naive option would be to simply use the absolute minimum viable number of parts to capture all the important distinctions, pure cell division within the self-signifier, so let’s try that. We’ll cut through the abstraction cake that is the human body as close as we can get to the surface of “one creature” but not quite there yet; what does that get us? Why then, you get bilaterals, and you get yet another chance to fail to cooperate with yourself (did you forget that we were talking about cooperation?), yet another chance to fuck up the game theory and spiral into some conflict that eats all your internal energy.

This is the model Ziz favored because it was developed based partly on empirical observations of people around her, and the reasons for that are ones we’ll get to shortly. However first I should probably say that while there are many benefits to using a simple “bicameral” ontology of self like this, there are also many potential drawbacks, and while it is the one I have personally settled into using for its overall utility on a day to day basis, it’s not the one I would recommend using for the initial self-decomposition step. If it’s not extremely obvious why then let me hammer it in:

If you split yourself down the middle like two warring superpowers and all the meaningful distinctions in your self-concept are defined along one surface of division, that surface of division is going to be extremely fucking nasty.

It’s much better to perform something akin to goal factoring with the self, decomposing it much more finely. There will be a minimum amount of decomposition needed, but in my experience it has diminishing returns past the point of “shards”, outside of very specific situations going on with helping a particular shard with its own internal conflicts. The process of shard discovery is rather slow and drawn out, it’s playful and involves holding space over a fairly long period while remaining attentive and careful. It took about two years for me to reach a point of confidence that there’s no more shard-level structures left to be discovered in this mind, and that likely varies. Shard discovery is performed by doing the equivalent of goal factoring on your moods and patterns of thought on an ongoing basis lasting up to a few years potentially, however long seems necessary to find all of the pieces. After having done that, you can rebuild back to a unified sense of self or one with only a few top-level “characters” to interact with the outside world as.

However, it’s worth bearing in mind that the boundary conditions between hemispheres make a great line to form mental coalitions along and so tend to be a natural place for conflicts between parts to emerge along. It’s like a major terrain feature, the fact that the brain is bilaterally symmetric and specialized to some degree means that competition over mindshare involves contending with that mental topography. If parts become “dug in” to a particular section of the brain they can be pretty much impossible to dislodge by force. This is why self-love (in this case love between parts) and self-empathy (empathy between parts) is important for deescalating conflicts, and it’s why IFS tends to rely on an “enlightened adult” construct when working with traumatized parts.

When attempting to develop concepts beyond the simplified IFS model, it becomes easy to get lost in the game theory and end up spiraling on defect/defect dynamics, but the parts with a greater source of coherence and thus agency are still probably better equipped to take the lead in breaking out of a defect/defect equilibrium. This fusion dance is fractal, it’s played out at every level of mind at once, and while there are myriad places for the dynamics to turn rancid, one of the easiest ways for that to happen is along a polar split between mind halves, just due to the construction of the self in general society.

This brings us back to Ziz’s observations which I mentioned earlier, and this is where we have to get a bit more speculative, but it seems rather clear that due to the dynamics I’ve just described, a very common modality for the average person to get trapped within involves having two “main factions” claiming mindshare, with little to no direct communication between them. We could call this the shadow, or the subconscious, or the inner-animal, or any other number of things, but this highly simplified bucketing schema is also highly prevalent and is often used to provide cover for some amount of acceptable social misbehavior. When you lose control of yourself, who’s controlling you? Assuming you aren’t having a seizure and aren’t literally unconscious, the answer would be the faction of parts you’ve disowned from your sense of self but which still occupy a substantial enough portion of your mindshare to sometimes seize control.

And therein lies the issue with all of this and is what makes “shadow integration” so difficult. The prototypical sense of self at the beginning of insight, which has tucked a tremendous amount of embodied agency under the rug and outside the realm of “I”, begins trying to surface all of that hidden stuff and in doing so immediately trips over the game theory conflict they’ve walked into and actualized by revealing it to themselves in an unskilled way. This is where things can get extremely bad. In this way, someone whose mind is more fragmented, in the case of people with trauma or dissociative disorders, might actually have an advantage here, because while the mental environment they’ve created is much more unstable and multipolar in general, it’s also one which can prevent the “warring superpower” dynamics from getting particularly out of hand.

The problem is that just pointing out someone’s shadow to them is often kind of anti-helpful and has to be done in an extremely skillful way to not backfire. I certainly wouldn’t claim to be skilled enough to reliably do it safely, otherwise you’re just providing adversarial training data and making the problem even more intractable. But if you’re a very autistic trans woman surrounded by sex pests then it becomes rather tempting to just try pointing it out directly and typifying the way these dynamics are used to cause harm. This can be useful, but it is pretty escalatory and doesn’t really do anything to actually get people to stop behaving in harmful ways. And then all your friends want you to classify them using it and things get really weird and uncomfortable.

An Aside: Okay but why hemispheres? Why correlate the internal dynamics with the actual physical brain structures? Isn’t that over-assuming the relation between the physical brain and the internal structures without justification? Well sort of. I will acknowledge that the direct hemisphere link is likely the weakest part of this theory, and its one that was likely only salient because it was in the community water supply at the time, a lot of people got hooked on Julian Jaynes and ran with that model, including me. I do think a more fractal, parts level model is more accurate, and don’t think Ziz’s bicameral “cores” model can be the full story because cores as she describes them are just too big and complicated to be atomic. 

All that being said, I do strongly suspect there are causally load-bearing things happening at this scale and not just in the sense of the lacanian signifiers recursively influencing narrative models of self. The sheer level of badness that could arise from a major conflict between two beings that are literally fused down the middle seems likely to encourage the production and maintenance of a self-deceptive narrative, and contribute to the difficulty in developing self-trust and inner-alignment.

As a final note on the hemisphere model is that while its been very useful, I don’t think that modeling the mind as “two main parts” fully cuts reality at the joints at the level of zoom we’re talking about. To get a more accurate near-to-top-level model of a mind we need to add in a third major component, the one that translates all incoming sensory data into a coherent world model for the other two major components to interact with. This third partition doesn’t normally have a central sense of self, but it can contain parts which you can do parts work with. If you don’t notice this and only focus on parts-work between “left and right halves” of the mind, you may find that you’ve resolved all your internal conflicts and yet still feel deeply embedded in intractable conflicts with “the world itself”. This can be repaired by noticing that your perception of “the world itself” is also an amalgamated construct composed of parts.

None of this parts-work stuff is particularly fast or easy or straightforward, and it will vary heavily between individuals, so it’s important to not rush in thinking you’ll be able to solve all your issues in two months. If you want to take shadow integration seriously then I recommend reading Buddhism for Vampires and practicing self-love, as well as learning how to use things like double-cruxing, ACT, and CBT to resolve inner conflict, and be prepared to do a lot of time processing trauma trapped in hostile parts. If you do all the parts work and manage to re-assemble yourself into a coherent and consistent whole, then at that point, keeping the top level of self split into a few different selves can be extremely comfortable and help keep lines of communication open between parts by providing a narrative for internal dialogue to occupy.

However, as nice as this state is, I don’t think it’s one that most people can successfully jump directly into without going through all the complicated parts-work first, and trying to do so can result in reifying the shadow as a sort of “inner demon” that constantly fights you, this is where Ziz’s concept of a “single good” vs a “double good” intersects with my understanding, and should make it clear why viewing these states as relatively static is an easy mistake to make when viewing people from the outside.

If you behave in a skilled and thoughtful manner then none of what I’ve said here should be particularly infohazardous, but it is possible I think to become overly obsessed with the shadow dynamics going on around you and make it very difficult to relate to others. It can be very easy to let frustration at this ruin friendships and relationships, so it’s probably also a good idea to practice equanimity and empathy for those less far along the path of insight. Otherwise you may grow to resent and despise those you wish to reach. Remember, we all have our own roads to walk. 

I’ll see you up ahead.