The First to Fall

Human Self-Domestication…selection against autonomy, not hot heads.

Richard Wrangham frames selection against reactive aggression (he uses the term “hot heads”) as the driver of human self-domestication and argues that our level of proactive aggression largely remained the same. He describes these as distinct evolutionary strategies, each with different adaptive costs and benefits.

To be clear, reactive aggression is impulsive, emotionally-driven violence in response to provocation or frustration (e.g. bar fights, chimpanzee dominance squabbles, etc.). Proactive aggression is calculated, planned violence deployed strategically for advantage (e.g. ambushes, executions, coordinated raids).

Wrangham’s central point is that self-domestication arises when reactive aggression is consistently punished (and culled), while proactive aggression not only persists but is sometimes institutionalized (authorities get a monopoly on violence).

His reasoning is as follows.

In small-scale societies, reactive aggressors were costly to group stability. They disrupted cooperation, created unpredictability, and risked alienating allies. With language and coalitionary power, groups gained the ability to collectively punish or kill these “hot heads.” Over many generations, this reduced the frequency of impulsively aggressive temperaments in the gene pool. The result is a calmer, more tolerant baseline disposition in humans compared to chimpanzees…one of the classic “domestication syndrome” traits.

What rubs me the wrong way is how quickly Wrangham assumes, out of all the traits that make up domestication syndrome, that reactive aggression is what was being selected for. Why wouldn’t the selection pressure be for proactive aggression, for example? Wrangham admits that proactive aggression was reinforced in human evolution. We became better at planned violence (executions, warfare, conquest) than any other primate. Crucially, proactive aggression is socially sanctioned…it’s framed as justice, punishment, or defense of the group. That makes it evolutionarily advantageous, not disadvantageous. In Wrangham’s model, the ability to conspire and kill reactively aggressive individuals is itself an expression of proactive aggression, and therefore part of what made us more cooperative at scale.

This hypothesis feels reductive to me. Domestication in other species involves selection for predictability, docility, and compliance, not just low reactivity. By centering only on reactive aggression, Wrangham treats self-domestication as a paradoxical success story…calmer humans enabled cooperation, and cooperation enabled civilization. It leaves out what civilization actually does…the flattening of error landscapes, where any form of reactivity (not just aggression) becomes maladaptive in large, controlled groups.

I’ve been thinking seriously about whether an argument could be made, just as strong or stronger than Wrangham’s, that selection for proactive aggression was the real driver in the human domestication story.

Large-scale violence is a consistent theme in the emergence of complex societies…from the mass graves of the Neolithic to the conquest states of Mesopotamia, Mesoamerica, and beyond. Warfare, conquest, and raiding were not incidental to civilization. They were the engines of state formation, with proactive aggression (planned and coordinated violence) clearly rewarded at both the genetic and cultural level.

Take the Y-chromosome bottleneck (5,000-7,000 years ago). It shows that ~90-95% of male lineages were extinguished, leaving only a few dominant bloodlines. This is genetic evidence of the real pattern of civilizational “coordination”: violent conquest and reproductive monopoly by elite men. Where in civilization’s history is Wrangham’s “peaceful coalitionary suppression of “bad apples”?” I just don’t see it. “Super-ancestor” events (e.g. Genghis Khan’s lineage) show the same thing in miniature. Proactive, organized aggression yields massive reproductive skew.

In fact, let’s turn to reproductive skew and polygyny. Even convention historical narratives tell a story of high-status males (kings, chiefs, emperors, warlords) with harems, concubines, and multiple wives. these are outcomes made possible by proactive aggression…conquest, enslavement, and the monopolization of resources. Lower-status men were excluded from reproduction, not because they were “too reactive” (though those certainly would have been excluded as well), but because they lost wars, were enslaved, or killed.

Proactive aggression isn’t just violence. It’s long-term planning, coalition-building, deception, and symbolic justification (myths, laws, and religions sanctifying violence makes up most of the human history book). These are precisely the traits that expand during the civilizing process…organizational capacity, abstract rule-following, and symbolic reasoning, all in service of controlling large groups.

I have a few thoughts on why Wrangham favors the other story (selection against reactive aggression). It links directly to his bonobo analogy (their lower reactivity compared to chimps). And it fits with domestication syndrome traits (softer faces and reduced baseline violence), of course. But these seem weak to me. What it comes down to, I believe, is Wrangham gravitating toward an age-old optimistic narrative…humans becoming more cooperative (from the “less hot-headed” angle), writing poems, and painting the Sistine Chapel. To me this is yet another just-so story tilted toward optimism. Real, documented human history (and the present, to a large extent) reads like selection for manipulative, proactive violence. Those who excel at strategic violence and symbolic control reproduce disproportionately. Full stop. This fits much better with what we see in the pages of history…runaway systems of control, hierarchies, and narrative manipulations that still structure our domesticated condition. These are better explained as the costs of selecting for proactive aggression than as some sort of “goodness paradox”.

In fact, it might be a silly thought experiment, but who’s to say that if were possible to actively select for proactive aggression in other species, that domestication traits wouldn’t appear?

To me, domestication syndrome (floppy ears, smaller brains, prolonged neoteny, pigmentation changes, altered reproductive cycles) arises because selection pressure narrows the error landscape of a species. The mechanism most often discussed is neural crest cell changes…but the reason for those changes could be any number of selection pressures. In foxes, it was tameness toward humans. In humans (Wrangham), he says it was lower reactive aggression. But it could also plausibly be selection for predictability, planning, and controlled aggression if that’s what the system demanded (and did, and does!).

The core idea is if you reduce the payoff for being “unpredictably reactive” and increase the payoff for being “strategically compliant,” the biology shifts. The neural, hormonal, and developmental systems adapt to reward that niche. The syndrome may look similar (the smaller brains, juvenilization, etc.) because what’s really being selected for is attenuation of wild-type reactivity in general.

Let’s move away from what I see as Wrangham’s too-narrow focus and broaden this narrative a bit.

Let’s look at the human story from a predictive coding lens, and consider scarcity as a selector. In times of ecological stress, groups face more prediction errors (crops fail, animals migrate, rivers dry up). Some individuals resolve error by updating their model (adjusting expectations, moving). Others resolve error by updating the world…forcing it into alignment with their model. The latter is the logic of domestication…bend plants, animals, landscapes, and people into predictability.

From here, we can see proactive aggression as control in action. On the ground, this isn’t abstract. Pull up wild plants and keep only the docile grains. Cull the fence-jumping sheep and reactive roosters…breed the calm ones. Raid nearby villages, enslave, execute dissenters, and reward compliance. This is proactive aggression. Planned, systemic, future-oriented control. It’s violence as policy.

This makes me think of how Robert Kelly frames humanity’s cultural revolution. He proposes that symbolic thought makes it possible to imagine not just “what is,” but “what should be.” And “what should be” becomes a shared prior (model of the world) that groups can coordinate around…even if it doesn’t match reality. Once you can coordinate around a model, you can impose it, and enforce conformity within the group. To me, that’s proactive aggression (if we’re still calling it that) elevated…control not only of bodies now, but of perception and imagination.

What disappears under a system like that? Well, for one, reactive aggression clearly becomes intolerable. It represents autonomous feedback (an individual saying “no” in the moment). In control systems, that kind of unpredictable resistance is punished most severely. You know that. Slaves who rebel are killed. Chickens that cause problems are culled. Men who resist capture are killed first. The system slowly culls “reactors” and favors the predictable (those who update their selves rather than the system).

This is what I see as the flattening of the civilizing process…the rock tumbler effect. Proactive aggression is the abrasive force that flattens everything…landscapes, genetic diversity, behavioral variation, etc., etc. Reactive aggression is just one of the first “edges” to be ground away. A byproduct of selecting for proactive control. A footprint of the real selection pressure. And what remains is a domesticated phenotype…more compliant, less volatile, more predictable.

Not convinced? Try this experiment.

Write these two hypotheses out on a sheet of paper:

  1. “Coalitions punish hot-heads -> reactive aggression selected against -> cooperative, domesticated humans emerge.”
  2. “Coalitions punish (in- and out-group) resistors to group control -> resistance (often expressed as reactive aggression…rebellion, resistance to domination) -> compliant, predictable humans emerge.

Now read as many history books as you can, testing these as you go. Take notes.

Only one of these hypotheses explains why proactive aggression thrives where reactive aggression doesn’t. There is no paradox. Proactive aggression isn’t punished because it aligns with group objectives, and what disappears isn’t “bad tempers” but unmanaged defiance. Resistors are killed. Compliant captives are taken. Rebels are executed. Compliant laborers survive.

This is selection against the unpredictable expression of autonomy that disrupts control.

Written by

Ken

in

Richard Wrangham frames selection against reactive aggression (he uses the term “hot heads”) as the driver of human self-domestication and argues that our level of proactive aggression largely remained the same. He describes these as distinct evolutionary strategies, each with different adaptive costs and benefits.

To be clear, reactive aggression is impulsive, emotionally-driven violence in response to provocation or frustration (e.g. bar fights, chimpanzee dominance squabbles, etc.). Proactive aggression is calculated, planned violence deployed strategically for advantage (e.g. ambushes, executions, coordinated raids).

Wrangham’s central point is that self-domestication arises when reactive aggression is consistently punished (and culled), while proactive aggression not only persists but is sometimes institutionalized (authorities get a monopoly on violence).

His reasoning is as follows.

In small-scale societies, reactive aggressors were costly to group stability. They disrupted cooperation, created unpredictability, and risked alienating allies. With language and coalitionary power, groups gained the ability to collectively punish or kill these “hot heads.” Over many generations, this reduced the frequency of impulsively aggressive temperaments in the gene pool. The result is a calmer, more tolerant baseline disposition in humans compared to chimpanzees…one of the classic “domestication syndrome” traits.

What rubs me the wrong way is how quickly Wrangham assumes, out of all the traits that make up domestication syndrome, that reactive aggression is what was being selected for. Why wouldn’t the selection pressure be for proactive aggression, for example? Wrangham admits that proactive aggression was reinforced in human evolution. We became better at planned violence (executions, warfare, conquest) than any other primate. Crucially, proactive aggression is socially sanctioned…it’s framed as justice, punishment, or defense of the group. That makes it evolutionarily advantageous, not disadvantageous. In Wrangham’s model, the ability to conspire and kill reactively aggressive individuals is itself an expression of proactive aggression, and therefore part of what made us more cooperative at scale.

This hypothesis feels reductive to me. Domestication in other species involves selection for predictability, docility, and compliance, not just low reactivity. By centering only on reactive aggression, Wrangham treats self-domestication as a paradoxical success story…calmer humans enabled cooperation, and cooperation enabled civilization. It leaves out what civilization actually does…the flattening of error landscapes, where any form of reactivity (not just aggression) becomes maladaptive in large, controlled groups.

I’ve been thinking seriously about whether an argument could be made, just as strong or stronger than Wrangham’s, that selection for proactive aggression was the real driver in the human domestication story.

Large-scale violence is a consistent theme in the emergence of complex societies…from the mass graves of the Neolithic to the conquest states of Mesopotamia, Mesoamerica, and beyond. Warfare, conquest, and raiding were not incidental to civilization. They were the engines of state formation, with proactive aggression (planned and coordinated violence) clearly rewarded at both the genetic and cultural level.

Take the Y-chromosome bottleneck (5,000-7,000 years ago). It shows that ~90-95% of male lineages were extinguished, leaving only a few dominant bloodlines. This is genetic evidence of the real pattern of civilizational “coordination”: violent conquest and reproductive monopoly by elite men. Where in civilization’s history is Wrangham’s “peaceful coalitionary suppression of “bad apples”?” I just don’t see it. “Super-ancestor” events (e.g. Genghis Khan’s lineage) show the same thing in miniature. Proactive, organized aggression yields massive reproductive skew.

In fact, let’s turn to reproductive skew and polygyny. Even convention historical narratives tell a story of high-status males (kings, chiefs, emperors, warlords) with harems, concubines, and multiple wives. these are outcomes made possible by proactive aggression…conquest, enslavement, and the monopolization of resources. Lower-status men were excluded from reproduction, not because they were “too reactive” (though those certainly would have been excluded as well), but because they lost wars, were enslaved, or killed.

Proactive aggression isn’t just violence. It’s long-term planning, coalition-building, deception, and symbolic justification (myths, laws, and religions sanctifying violence makes up most of the human history book). These are precisely the traits that expand during the civilizing process…organizational capacity, abstract rule-following, and symbolic reasoning, all in service of controlling large groups.

I have a few thoughts on why Wrangham favors the other story (selection against reactive aggression). It links directly to his bonobo analogy (their lower reactivity compared to chimps). And it fits with domestication syndrome traits (softer faces and reduced baseline violence), of course. But these seem weak to me. What it comes down to, I believe, is Wrangham gravitating toward an age-old optimistic narrative…humans becoming more cooperative (from the “less hot-headed” angle), writing poems, and painting the Sistine Chapel. To me this is yet another just-so story tilted toward optimism. Real, documented human history (and the present, to a large extent) reads like selection for manipulative, proactive violence. Those who excel at strategic violence and symbolic control reproduce disproportionately. Full stop. This fits much better with what we see in the pages of history…runaway systems of control, hierarchies, and narrative manipulations that still structure our domesticated condition. These are better explained as the costs of selecting for proactive aggression than as some sort of “goodness paradox”.

In fact, it might be a silly thought experiment, but who’s to say that if were possible to actively select for proactive aggression in other species, that domestication traits wouldn’t appear?

To me, domestication syndrome (floppy ears, smaller brains, prolonged neoteny, pigmentation changes, altered reproductive cycles) arises because selection pressure narrows the error landscape of a species. The mechanism most often discussed is neural crest cell changes…but the reason for those changes could be any number of selection pressures. In foxes, it was tameness toward humans. In humans (Wrangham), he says it was lower reactive aggression. But it could also plausibly be selection for predictability, planning, and controlled aggression if that’s what the system demanded (and did, and does!).

The core idea is if you reduce the payoff for being “unpredictably reactive” and increase the payoff for being “strategically compliant,” the biology shifts. The neural, hormonal, and developmental systems adapt to reward that niche. The syndrome may look similar (the smaller brains, juvenilization, etc.) because what’s really being selected for is attenuation of wild-type reactivity in general.

Let’s move away from what I see as Wrangham’s too-narrow focus and broaden this narrative a bit.

Let’s look at the human story from a predictive coding lens, and consider scarcity as a selector. In times of ecological stress, groups face more prediction errors (crops fail, animals migrate, rivers dry up). Some individuals resolve error by updating their model (adjusting expectations, moving). Others resolve error by updating the world…forcing it into alignment with their model. The latter is the logic of domestication…bend plants, animals, landscapes, and people into predictability.

From here, we can see proactive aggression as control in action. On the ground, this isn’t abstract. Pull up wild plants and keep only the docile grains. Cull the fence-jumping sheep and reactive roosters…breed the calm ones. Raid nearby villages, enslave, execute dissenters, and reward compliance. This is proactive aggression. Planned, systemic, future-oriented control. It’s violence as policy.

This makes me think of how Robert Kelly frames humanity’s cultural revolution. He proposes that symbolic thought makes it possible to imagine not just “what is,” but “what should be.” And “what should be” becomes a shared prior (model of the world) that groups can coordinate around…even if it doesn’t match reality. Once you can coordinate around a model, you can impose it, and enforce conformity within the group. To me, that’s proactive aggression (if we’re still calling it that) elevated…control not only of bodies now, but of perception and imagination.

What disappears under a system like that? Well, for one, reactive aggression clearly becomes intolerable. It represents autonomous feedback (an individual saying “no” in the moment). In control systems, that kind of unpredictable resistance is punished most severely. You know that. Slaves who rebel are killed. Chickens that cause problems are culled. Men who resist capture are killed first. The system slowly culls “reactors” and favors the predictable (those who update their selves rather than the system).

This is what I see as the flattening of the civilizing process…the rock tumbler effect. Proactive aggression is the abrasive force that flattens everything…landscapes, genetic diversity, behavioral variation, etc., etc. Reactive aggression is just one of the first “edges” to be ground away. A byproduct of selecting for proactive control. A footprint of the real selection pressure. And what remains is a domesticated phenotype…more compliant, less volatile, more predictable.

Not convinced? Try this experiment.

Write these two hypotheses out on a sheet of paper:

  1. “Coalitions punish hot-heads -> reactive aggression selected against -> cooperative, domesticated humans emerge.”
  2. “Coalitions punish (in- and out-group) resistors to group control -> resistance (often expressed as reactive aggression…rebellion, resistance to domination) -> compliant, predictable humans emerge.

Now read as many history books as you can, testing these as you go. Take notes.

Only one of these hypotheses explains why proactive aggression thrives where reactive aggression doesn’t. There is no paradox. Proactive aggression isn’t punished because it aligns with group objectives, and what disappears isn’t “bad tempers” but unmanaged defiance. Resistors are killed. Compliant captives are taken. Rebels are executed. Compliant laborers survive.

This is selection against the unpredictable expression of autonomy that disrupts control.

Comments

Leave a comment

More posts