The Abstractionist’s Papers · Part II · Chapter 5 of 13
A caterpillar goes into a chrysalis and a butterfly comes out. Nothing about the caterpillar tells you what’s coming.
That’s what we mean by emergence, and it isn’t only caterpillars. Minds do it. So do languages and companies.
The work happens inside the process, where nothing shows. Processes reorganize internally long before anything surfaces, so once we notice the new behavior the change is already done. All we get is the butterfly.
The missing piece was always there, hidden by the blindfold. Selection adapts what exists and creates what never existed.
Every process weighs keeping its current internal state against reorganizing entirely. When staying costs more than rebuilding, emergence follows.
Emergence is the only show in town.
Contents
5.1 Darwin’s Natural Selection
5.1.1 Limitations
5.1.2 Toward a Broader Framework
5.2 A Theory of General Selection
5.3 Emergence
5.3.1 Revisiting Natural Reality
5.3.2 Beyond Space and Time
5.3.3 Through Orthogonality
5.3.4 Seeing Natural Spaces
5.3.5 Gödel, Darwin, and the Blindfold
5.4 Demystifying Emergence
5.5 Across Contexts
5.6 The Economics of Emergence
5.6.1 Attachment
5.6.2 Registering a Problem
5.6.3 The Costs of Staying
5.6.4 The Costs of Going
5.6.5 The Crossing
5.6.6 The Same Accounting at Scale
5.6.7 Using the Model
5.6.8 Where Disruption Comes From
5.7 Closing Remarks
5.1 Darwin’s Natural Selection
Charles Darwin’s theory of natural selection is still how we explain evolution. In On the Origin of Species (1859), he showed that populations change as traits that improve survival or reproduction grow more common. Variability introduces traits that persist when they align with environmental demands, while less effective traits disappear.
Natural selection explains adaptation by showing how species adjust to changing conditions. Antibiotic resistance in bacteria shows this. Random mutations introduce differences, and selective pressures determine which strains survive and spread. Fossil records support this model, showing how traits accumulate over millions of years, marking transitions between species.
5.1.1 Limitations
Natural selection leaves two things out:
- Scope: natural selection deals with biological evolution, but similar selection processes influence other systems like technology, culture, and economies. They show adaptation and persistence too, beyond the reach of Darwin’s model.
- Emergence: natural selection explains each new trait by the advantage it carries. Each step, feathers for insulation, feathers angled for gliding, feathers built for powered flight, carries its own advantage, and the same case-by-case story covers the move from single-celled to multicellular life. Darwin’s model tells that story one domain at a time. The reorganization itself, the same mechanism whether it’s a mind changing its own logic, a market changing its own rules, or a species changing its own body, is a different question.
5.1.2 Toward a Broader Framework
Darwin captured part of the picture. Natural intelligence extends beyond biology, happening wherever patterns persist and remake themselves. Brains and consciousness aren’t required. The capacity is built into how things work.
Selection operates in two domains. Processes interact in the Blue Space, and meaning gets assigned in Red Space. A vertical axis, which we’ll call the Causation Axis, shows how processes resist the rules governing them. As they accumulate Incoherence, their impedance changes, and they reorganize or take on new functions.
Selection works in two directions. One adapts within current constraints. The other breaks free to create new possibilities.
5.2 A Theory of General Selection
General Selection explains how processes change in every context. The same mechanisms handle gradual adaptation and disruptive breaks.
From inside a Red Space, reality seems to evolve incrementally, with changes that look continuous and gradual. Evolution happens through causal mechanisms. Underneath, in the Blue Space we can’t see, processes are adapting, and the gradual progress in our Red Space reflects it. Each vertical transformation is a change in causal impedance, carrying the process up the Causation Axis.
General Selection works in a continuous loop:
Each cycle fits a process a little better to its surroundings and leaves room for small change and for complete transformation.
(1) Interaction
The loop begins with interaction, where processes engage with their environment or other processes. These interactions create the conditions for change.
(2) Variability
Variability comes from interactions, introducing new possibilities. This phase includes both small adjustments and larger, more disruptive changes. Variability creates the raw material for selection.
(3) Selection
Selection determines which variations persist and which disappear. It works the same way for small improvements and for larger breaks. Variations the surroundings support are kept, and the rest are filtered out.
(4) Accumulation
Accumulation layers retained changes over successive cycles. Small adaptations build on one another and refine the process steadily. Larger breaks change the process itself.
Each step feeds the next, so the loop keeps turning. It fine-tunes a process toward its environment and introduces the disruptive variability that can take it somewhere else, using the same four moves for both jobs.
Note: Natural Reality is about natural processes, and the examples from here on include languages, markets, scientific fields, and whole industries, none of which is one on its own. Each is a distributed reality, built from the natural processes participating in it. A market is what accumulates when every person buying or selling adds up. A language is what a population of speakers sustains between them. The symbols can outlast the people, and the meaning can’t. General Selection still applies, because every node in that distribution is a natural process in its own right, free to change its relationship to the rules it operates under.
5.3 Emergence
General Selection functions in two directions. Horizontal selection adapts processes within existing constraints. Vertical selection breaks those constraints entirely, producing new possibilities.
The vertical dimension is the Causation Axis. Along it, processes move orthogonally beyond their current boundaries.
This capacity is everywhere. Building from difference and harmonizing Incoherence is something natural processes do.
5.3.1 Revisiting Natural Reality
Variability introduced during interactions influences causal impedance, defining how constraints act upon a process.
Under typical conditions, change (δ) operates within the same plane (n) as decay (λ), remaining aligned with the process’s current Natural Space, where its causal impedance (ZΨ) is fixed. Whatever its angle to decay, change on that plane pays its decay. When variability introduces Δ, δ turns out of the plane entirely, and ZΨ moves, allowing the process to sidestep prior limitations. The decay stays behind on the plane.
The process builds new causal connections. Emergence follows. Variability adjusts ZΨ, changing how strongly Ψ reaches the process. As the process’s impedance climbs, Ψ gets enforced less, and new pathways open up.
This move is Incoherence (Δ), the jump in impedance from one plane to the next.
The bigger the departure, the bigger the Δ.
5.3.2 Beyond Space and Time
Seen from above, the loop appears as a flat circle. Natural Reality adds another dimension:
Viewed at an angle, the loop shows the Causation Axis rising perpendicular to the plane of space and time, a plane of meaning.
Along the Causation Axis, Incoherence modulates causal impedance, enabling processes to bypass constraints and reorganize. From this perspective, the loop becomes a spiral, where each cycle expands the process into higher levels of complexity.
A snail’s shell grows this way, each turn laid on the last. A loop that only repeats draws a circle. A loop that accumulates while it repeats draws a spiral, and nature is full of processes doing both at once.
Emergence, then, is orthogonal. Emergence operates in the Blue Space, where we can’t watch it. What we can watch is our own Red Space, which blends the physical, the biological and the cognitive into one flat picture, and that’s what traditional models of evolution are describing.
Without recognizing Incoherence as an active mechanism, we see only the outcomes of emergence and miss the process that drives it.
5.3.3 Through Orthogonality
Emergence spirals up the Causation Axis, and complex numbers already describe two components at right angles.
A complex number has a real part and an imaginary one, and neither reduces to the other. Euler’s equation relates them.
eiθ = cos θ + i sin θ
θ is how far around you’ve turned from the real axis toward the imaginary one. Natural Reality has the same arrangement, with the Blue Space as the real component and the Red Space as the imaginary one. Light, where they meet, is the Natural Reality Axis. Every process has both.
Selection drives change within a plane and accumulates Incoherence along the Causation Axis. That changes the process’s impedance and makes emergence possible.
5.3.4 Seeing Natural Spaces
Imagine transparent boxes stacked along the Causation Axis.
From above, all Natural Spaces appear flattened into a single plane. In Figure 28, the Causation Axis becomes visible, showing that what seemed like a flat loop belongs to a larger reality. When viewed from the side, each box appears as a distinct layer:
The Causation Axis shown above sits on the real side of Figure 27, the side of the Blue Space. Like the layers, it belongs to our model, and we use it to account for causal impedance.
Processes move along the Causation Axis in the Blue Space, where vertical selection determines how they resist or adapt to change, and our model keeps the account.
The yin-yang symbol compresses a continuous process into opposing forces. Our perception does the same thing to emergence, flattening an open spiral into a closed loop. We watch Natural Spaces from inside our own Red Space, which blends them all into one flat picture, and the Causation Axis is what gets lost.
5.3.5 Gödel, Darwin, and the Blindfold
In 1931, Kurt Gödel proved that any consistent formal system capable of expressing arithmetic contains true statements it can’t prove. Such a system, if consistent, is always incomplete, and it can’t establish its own consistency from within.
Natural Reality reads the same limit in natural processes. No process can settle everything from inside its own rules, and what its current rule can’t handle accumulates as pressure to reorganize.
Darwin showed how species improve within their constraints, one selected trait at a time. Gödel showed that a formal system’s constraints have edges it can’t see past from inside. Neither explained how a process gets past those edges.
Natural Reality treats incompleteness as a driver. When a process meets something its current rule can’t hold, paradox accumulates. The process either stays stuck in repetition or reorganizes its internal model to hold what it couldn’t hold before.
Each cycle of selection meets new incompleteness. The process stagnates or changes how it interprets what it meets. When a new perspective holds what was Incoherent before, it persists. The model stretches, and the loop returns at a higher level of organization. Where that happens, circular repetition turns out to be an ascending spiral.
What a process can’t settle from inside its own rules becomes the pressure that carries it past them. That pressure is how incompleteness drives emergence.
5.4 Demystifying Emergence
Bend a paperclip back and forth. For a while nothing seems to change. Then one bend, no different from the others, and it snaps.
The last bend didn’t break it. Every bend before it left something behind, a fault too small to see, and the faults piled up inside while the outside continued to behave like a paperclip. Engineers call it fatigue. The fracture was building inside all along.
We save the word emergence for butterflies and flight, and for the moments we change our own minds. The paperclip does the same thing. So does everything else. We’ve only ever accounted for one inside, our own, which is why everything else seems to change out of nowhere.
What a process becomes depends on where it can go.
You’re standing on a train track with a train coming. You can run toward it and die sooner. You can run down the track away from it and die later. You can jump up and down and nothing changes. All three are coherent interactions, moves that follow the rules of the space you’re in, and not one of them changes your relationship to a train on a track.
Step off the track.
The train goes past and you watch it go. Nothing about the train changed. The rule about trains and the things in front of them stopped applying to you. That’s an incoherent interaction.
Whether it works depends on what’s beside the track.
On a bridge there’s nothing there, and stepping off ends you faster than the train would. If there’s a platform, you’re standing on it while the train goes by, in a place with its own rules. That’s Harmonized Incoherence, and you’re rarely the first to find it. Others stepped off before you and the platform is where they went.
When several people end up on the platform, they start interacting under whatever rules hold there. Whether that lasts depends on how it sits with everything around it.
Running faster is evolutionary. It improves the outcome and leaves the rules alone. Stepping off is revolutionary, and it works only when the surroundings support it. Only stepping off is emergence, and processes need both.
5.5 Across Contexts
General Selection reaches well beyond biology.
Language evolves as new words, slang, and grammar change communication patterns. If a word makes speech clearer, it gains ground against outdated forms. Learning any skill works the same way. You test different approaches and the ones that work join your repertoire. Small experiments accumulate into major transformations.
Thinking about thinking is its own evolutionary process. People test different ways of solving problems, managing emotions, making decisions. Effective strategies persist and change how we reflect and adapt. The mind is its own laboratory.
The same capacity operates in bacteria developing resistance and in humans learning new skills. Processes sense their environment, try new approaches, keep what works, and build on what they kept. Small differences either harmonize with what came before or break away into something new, and selection determines which.
If every mind builds its own reality, why do human lives rhyme?
People across cultures and centuries move through recognizably similar changes. Struggles with identity. Grief that hijacks a life. Regret. The long search for meaning. Carl Jung saw the patterns and called them archetypes, placing them in a shared layer of the psyche passed down across generations.
The patterns are real. They don’t need a shared layer to explain them. Minds operate with comparable machinery, holding what they expected up against what arrived, looping when the comparison fails, reorganizing when the loop holds long enough. Repeat the same operation across billions of people and you get familiar arcs.
The hero’s story works this way. A life reaches a crisis, hits an impasse, and begins again in a new form. The caterpillar does the same. It meets a limit, dissolves, reforms, and returns with new capacities. Transformation follows recognizable patterns because the same mechanism runs underneath, wherever it happens.
5.6 The Economics of Emergence
Emergence has a trigger.
Most of what’s written about emergence describes it after the fact. Something new appeared, and here is the story of how. The trigger goes missing. Without it you can’t see the change coming, can’t bring it on, and can’t hold it back.
The trigger is economic. A process keeps its current organization for as long as keeping it costs less than replacing it. When that stops being true, it reorganizes.
A process sits in a position defined by two things, how strongly it resists what doesn’t fit its rule, and how much it has riding on that rule being right. When the rule stops delivering what the process expected, paradox accumulates inside. Every paradox costs something to hold, and the more the process has invested in it, the more it costs. The loops get expensive. Sooner or later maintaining them costs more than rebuilding, and the process rebuilds. In a mind that might be one belief or the whole architecture around it. In a flower, an organization, or a galaxy, the scale changes and the arithmetic doesn’t.
All of that happens where we can’t look. It’s why the paperclip seemed to snap on an ordinary bend. The accumulation is internal, the tipping is internal, and once anything reaches us the potential has already turned into flow. We meet the result and never the buildup.
John Boyd came close in 1976. His paper Destruction and Creation describes a mind that has to keep taking its own concepts apart and putting new ones together, through what he called destructive deduction and creative induction. Arguing from Gödel, Heisenberg, and the Second Law, he showed that effort turned inward to improve the match between a concept and the world only widens the mismatch. Disorder rises until the old concept is easy enough to shatter, and you shatter it and synthesize the pieces into something broader. Structure, unstructure, restructure, without end. The paper never says where the threshold sits for any particular turn.
Our model needs two quantities. We already have impedance. Attachment comes next.
5.6.1 Attachment
Impedance is how strongly a process resists what doesn’t fit its rule. Attachment is how much it has riding on the rule being right.
The boy at the lemonade stand expected the even split to settle things. He had built that expectation over years of it working, and he defended it every week it failed. Attachment is the commitment a process keeps to its own expectations. The daughter whose father left never let go of the rule that a parent who leaves a marriage leaves the children. The institutional trader stakes a career on the belief that fundamentals drive price.
Attachment makes expectations expensive to revise, which is why processes defend them long past the point the evidence turned. Between them, impedance and attachment fix where a process stands. Standing there costs it.
5.6.2 Registering a Problem
A disturbance costs a process nothing by itself. It costs something when the process registers it as a problem, and processes differ enormously in what they can register.
The daughter’s father called, visited, and explained himself for years. None of it reached her as what it was, because her logic converted every approach into confirmation. When he showed up, he shouldn’t have left. When he stayed away, see, he left you. When he explained, excuses don’t change what happened.
Twenty years of a man trying, and almost none of it registered as a problem. Her rule had no opening for it. The order is fixed. Something happens outside, the process picks up what it is built to pick up, that passes through the model already in place, and only then does anything land as a problem.
So the same event bankrupts one process and costs another nothing. A process that reads a problem accurately aims its fix better, and a fix that lands costs less than one that misses.
5.6.3 The Costs of Staying
Staying put has a cost, and a process pays for it twice.
The first cost is upkeep. Holding impedance and attachment in place takes effort, and the more extreme the position, the more effort it takes. The daughter spent twenty years defending a rule against a father who kept calling. Every call was work.
The second cost is patching. When something goes wrong, a process fixes it the cheap way, adjusting either its impedance or its attachment, whichever is easier that day. The child explained, then argued, then made an exception. Blockbuster forgave late fees, launched a mail service, tried kiosks. Every one was the cheapest thing available at that moment, and not one of them touched the rule underneath.
Patches get more expensive as you go. A rule already bent resists bending again, so the tenth fix costs more than the first. And the earlier ones don’t go away. The process is still paying for all ten.
Patch by patch, staying looks cheap. Added up, it isn’t. A process feels the price of the next fix and never the running total, which is why the crossing arrives without warning.
5.6.4 The Costs of Going
Going costs a lot, all at once. You tear down what you were built on, build something you don’t have yet, and make the new thing work while the old one is still paying the bills.
Netflix paid all three. It licensed content for digital delivery, built server capacity, and wrote recommendation software, then made streaming pay while DVDs kept the lights on. Reorganization costs enough that processes avoid it, which is why patching goes on as long as it does.
The payoff comes afterward. A process built around the rules its environment enforces stops spending to resist them, and it has somewhere to be when the old arrangement stops working. Netflix had a business in 2010. Blockbuster had stores.
5.6.5 The Crossing
Put the two next to each other. Staying costs upkeep plus every patch so far. Going costs the rebuild. As long as staying is cheaper, the process stays and keeps patching. The moment it isn’t, the process goes.
The child, the daughter and Blockbuster each came out differently.
The child crossed fast because his patches were expensive and frequent. Every week the friend skipped meant another patch, each costing more resentment than the last. A child’s fairness rule is cheap to rebuild.
The daughter took twenty years because so little reached her, and rebuilding would have cost her everything. The rule was load-bearing for who she took herself to be. It let little through, so the bill climbed at almost nothing a year against a price she couldn’t pay. Then she held her own child, and something got through at last. The crossing came in an afternoon.
Blockbuster never crossed because its patches stayed cheap. Late fee forgiveness, a mail service, kiosks, each affordable, each within the existing organization. Cheap patching kept the running total below a reorganization cost that stayed enormous, since going meant abandoning the stores that were the company. The math never tipped. Blockbuster went under making sensible decisions the whole way down.
How long it takes has nothing to do with it. Patches pile up, and a process can burn through them in a season or take a generation.
It happens in three stages. Early on the problems are small and cheap to fix, so staying wins easily. Then the problems stop going away. Fixes pile up, each costing more than the last, and upkeep climbs as the process gets pushed toward extremes. Staying still wins, by less each round. Then it doesn’t win, and the process goes.
Anyone watching from outside sees only the third phase and reads the whole change as sudden.
5.6.6 The Same Accounting at Scale
GameStop, January 2021. Institutional processes carried high impedance to social sentiment and high attachment to fundamental analysis. Retail coordination on Reddit created a problem the models couldn’t absorb. For the first weeks the cheap patch was trimming attachment, hedge, wait for fundamentals to reassert. Then the problem grew faster than patching could close it. Impedance fell, attachment collapsed, and the bill overtook the cost of reorganizing. Social sentiment became something institutions watch.
Climate science, 1979 to 1995. The field carried high impedance to attributing observed warming to carbon dioxide, since natural variability could account for much of the same signal. The prediction came early. Pinning the warming already happening on human causes was the harder claim to earn. Through the 1980s each new finding forced a patch, a revised model, a narrowed claim, a qualified statement about what the data could and couldn’t show. By the early 1990s the patches had grown numerous and expensive to maintain. The 1995 IPCC report crossed, finding that the balance of evidence pointed to a discernible human influence on global climate, and attribution studies became normal science from there.
Financial markets carry high coordination costs and cascade in weeks. Scientific consensus carries lower patching costs and takes decades. The economics tip the same way in both.
5.6.7 Using the Model
The model gives you three places to push, whether you want the crossing sooner or later. Lower the strain of upkeep and staying gets cheaper, so the crossing recedes. Raise what reorganizing costs and staying wins longer even as the pressure builds. Make cheap fixes easy to reach and the running total climbs more slowly. Blockbuster shows the third lever working against a company. An affordable patch was available at every turn, and taking them kept the total under the threshold until there was nothing left worth rebuilding.
The costs stay hidden, and the patching doesn’t. A company announcing its third fix in two years for the same problem is telling you something. So is a person explaining the same position more insistently with each failure. We can’t see the bill. We can count the patches, and patches coming faster around one unresolved thing is the surface that internal accumulation leaves.
Nobody has measured any of this. Doing it would mean finding stand-ins for each quantity. Impedance might show up as how long a process takes to respond, measured against some baseline. Attachment might show up as how much it spends defending what it already believes. The problem itself would need some normalized measure of surprise. Then someone would have to work out how patching costs climb as impedance and attachment move, calibrate the rebuild cost against crossings we can watch, and test the result on a case that didn’t inspire it.
It would be hard. Pick a different stand-in and you get a different answer, so nothing counts until it holds up across several. Problems both cause and follow from changes in impedance and attachment, which makes it hard to say what drove what. And rebuild costs move for reasons nobody watched. None of that is solved here.
The pattern keeps showing up. Test it against a crossing you watched happen.
5.6.8 Where Disruption Comes From
Reorganization inside one process becomes disruption for the processes around it. When a process crosses its threshold and reorganizes, it creates new behavior in the Blue Space. If disruptive enough, this behavior becomes selection pressure that filters surrounding processes. Disruption also arrives from outside the population entirely, through Blue Space changes unrelated to any internal reorganization. Both sources operate through the same General Selection mechanism. Netflix and Blockbuster show both.
Netflix was the first source. It never had stores to defend, and through the DVD years it built what streaming would need.
Then external disruption arrived. Bandwidth infrastructure improved between 2005 and 2007, built by telecom companies outside the media industry, making streaming technically viable at scale. It came from outside the population being selected and had nothing to do with the industry’s own reorganization.
Neither source would have done it alone. Without the bandwidth, Netflix’s streaming capacity had nowhere to run. Without that capacity, the bandwidth would have reached a company with nothing ready to use it. In 2007 the two met, and the cost of going fell below the cost of staying. Netflix launched streaming and crossed a threshold it had spent years preparing for, while discs still paid the bills.
For Blockbuster, the same change arrived as selection pressure. Its stores and its discs were the rules Netflix had stopped obeying, and every patch it reached for left both in place. Blockbuster filed for bankruptcy in 2010. One process reorganized, the world around it changed, and the population was filtered.
5.7 Closing Remarks
Change happens through General Selection. Every process cycles through interaction, variability, selection, and accumulation, adapting within existing constraints and transforming beyond them when the economics tip. The same loop that turns a bacterial strain resistant to antibiotics turns a mind toward a new understanding of fairness and turns an industry toward a new way of delivering content.
From the inside, it’s the struggle you know. Every birth and rebirth, every passage through fire, was this mechanism underneath. That strain is the pressure that precedes transformation. The same process that built the layers of reality builds perspective from contradiction in a human mind.
Inside the chrysalis, the caterpillar takes itself apart along a path its lineage paid for over millions of years. Then it rebuilds. All we ever see is the butterfly.
General Selection determines what changes. Light determines how influence propagates.
Chapter 5 in the Chapter Manual · Study questions