The First to Fall

What Wrangham Gets Wrong About Human Domestication

(Hint: 900,000 cows are slaughtered daily. They shit where they eat and wouldn’t have a hope in hell at surviving without human care. But they’re nice.)

In The Goodness Paradox, Richard Wrangham argues that the main selection pressure in human (self-)domestication was the weeding out of reactive aggression. It’s a nice story that makes the net gain of human domestication harder to argue against. But, to me, it’s clear that selection against reactivity in general (or unpredictability) is the bigger, truer story, of which the reduction of “reactive aggression” is simply the most visible (and PR-friendly) chapter. Taken as a whole, and across species, the domestication package is clearly a general downshift in arousal/reactivity with a re-tuning of social expectations…not just the loss of hair-trigger violence.

Let’s look at domestication again while entertaining this broader (and inconveniently less moralistic (duller, rather than nicer humans) selection pressure.

For one thing, physiology moves first…and it’s general. In classic domestication lines (e.g. Belyaev’s foxes), selection for tameness blunts the HPA axis and stress hormones overall…fewer and fewer cortisol spikes, calmer baselines. That’s not “anti-aggression” specifically; it’s lower stress reactivity across contexts. Brain monoamines shift too (e.g. higher serotonin). That’s a whole-system calm that would make any behavior less jumpy (including but not limited to aggression).

Developmental mechanism also points to a broader retune. The “domestication syndrome” is plausibly tied to mild neural-crest hypofunction, a developmental lever that touches pigmentation, craniofacial shape, adrenal medulla, and stress circuitry. In humans, BAZ1B (a neural-crest regulator) is linked to the “modern” face and is part of the self-domestication story. None of that is news…but if you tweak this lever, you clearly soften the whole reactivity profile…not just aggression. And my guess is that whoever’s fucking with the lever has his eye on the “compliance” dial more than any other.

Comparative signals align, too. Genomic work finds overlaps between human selective sweeps and domestication-candidate genes across species…showing a syndrome-level process rather than some sort of single behavioral knob. Craniofacial “feminization” over time in H. Sapiens fits reduced androgenic/reactive profiles, too.

Domesticated behavior tracks a “global calm.” Domesticated animals are less fearful, less erratic, and more socially tolerant than their wild counterparts. Your dog’s tendency to “look back” to you in unsolvable tasks is a manifestation of that…when arousal is lower and social cues are trusted, help-seeking beats reactive persistence. That’s a broad predictability play (that has nothing to do with aggression).

Obviously, Wrangham’s focus still matters. His key point, the decoupling of reactive vs proactive aggression in humans (we got tamer in the heat-of-the-moment sense, but remained capable of planned, coalitionary violence), is real and important to explain. It’s part of the story, but not the whole story. As general reactivity is reduced, strategic (planned) aggression is preserved…because strategic aggression isn’t a startle reflex; it rides on executive control and group coordination. But selection against reactive aggression isn’t the driver in this story. It’s just one behavioral readout of a deeper arousal/volatility downshift. A nice part (maybe) of an otherwise quite shitty story (from life’s vantage point). The beef industry might point out how nice the cows are, but I don’t think even they would try to argue that “nice” is what it’s aiming for. Dull. Compliant. And so it goes with all domestication. There is an objective in the domestication process, and any and all traits that impede progress toward that objective are pruned. (adding “self-” to domestication when it comes to humans, while accurate in the sense that the domesticating agent was of the same species, gives it a voluntary flavor that has no evidence in history…the domestication of humans was driven by systemic enslavement and reproductive control just as it was for all domesticates)

Why is it so important to me to find the driver of human domestication at all? Why not just start from the broadly-accepted premise that we are a domesticated species and go from there? Because I need to know what’s truly going on in the brain during this domestication process. How do we get to the brain we call “typical” now? What was it selected for? Was it selected for something broadly adaptive? Or is it more like runaway selection? An overfitting?

To me, cognitively, domestication looks like a down-weighting of volatility and a reallocation of precision (in predictive-coding terms). Brains with lower expected volatility (that have “the world is less jumpy” as a hyperprior…fewer LC-NE-style alarm bursts…a calmer autonomic tone), higher precision on social priors (human signals are treated as the most trustworthy ones…ecological “noise” gets less weight), and policy canalization (high confidence in proximity/compliance/help-seeking policies).

I think that human self-domestication primarily targeted behavioral and physiological volatility (a population-level reduction in phasic arousal and unpredictability) of which lower reactive aggression is a salient subset. And that the result is down-tuned HPA/LC reactivity, strengthened social priors, and canalized, low-variance action policies. Think of what happened as some sort of reactivity pruning (where reactive aggression was one prominent branch that got lopped off).

What is the domesticated brain? Zoomed out, it’s clearly an instrument that’s been made dull. One that exhibits blunted responses to non-social unpredictability (startle, sensory oddballs, metabolic stressors), not just to dominance threats. And anti-aggression alone doesn’t suppress those.

If I’m reading the studies properly, there are signatures of what I’m talking about in stress-regulatory and neuromodulatory pathways (HPA, serotonin, vasopressin) and neural-crest development…not just androgenic or specifically aggression-linked loci. Recent multispecies work pointing at vasopressin receptors and neural-crest regulators certainly seems consistent with this.

Wrangham’s story doesn’t account for lower intra-individual variance in exploratory/avoidant switches and faster convergence on socially scaffolded policies (like help-seeking) across types of tasks (anti-aggression predicts biggest effects only in conflict contexts). It doesn’t explain the psychotic consensus reality holding everyone in, as it rolls off a cliff.

(In fact, I question how much of the reactive aggression branch got lopped off…surely, not nearly as much as we think. What self-domestication mostly did was gate when, where, and how the majority of people show reactivity. When accountability and real-world consequences are high, most people keep a lid on it. When consequences drop (anonymity, distance, no eye contact, no immediate cost), the lid starts to rattle…online, in cars, in fan mobs, in comment sections. I don’t think reactive aggression was bred out so much as trained into context…and how well you do in that context will largely determine the story you tell. Harvard professors are clearly doing quite well in the civilizational context and consequently have pretty stories to tell.)

Written by

Ken

in

(Hint: 900,000 cows are slaughtered daily. They shit where they eat and wouldn’t have a hope in hell at surviving without human care. But they’re nice.)

In The Goodness Paradox, Richard Wrangham argues that the main selection pressure in human (self-)domestication was the weeding out of reactive aggression. It’s a nice story that makes the net gain of human domestication harder to argue against. But, to me, it’s clear that selection against reactivity in general (or unpredictability) is the bigger, truer story, of which the reduction of “reactive aggression” is simply the most visible (and PR-friendly) chapter. Taken as a whole, and across species, the domestication package is clearly a general downshift in arousal/reactivity with a re-tuning of social expectations…not just the loss of hair-trigger violence.

Let’s look at domestication again while entertaining this broader (and inconveniently less moralistic (duller, rather than nicer humans) selection pressure.

For one thing, physiology moves first…and it’s general. In classic domestication lines (e.g. Belyaev’s foxes), selection for tameness blunts the HPA axis and stress hormones overall…fewer and fewer cortisol spikes, calmer baselines. That’s not “anti-aggression” specifically; it’s lower stress reactivity across contexts. Brain monoamines shift too (e.g. higher serotonin). That’s a whole-system calm that would make any behavior less jumpy (including but not limited to aggression).

Developmental mechanism also points to a broader retune. The “domestication syndrome” is plausibly tied to mild neural-crest hypofunction, a developmental lever that touches pigmentation, craniofacial shape, adrenal medulla, and stress circuitry. In humans, BAZ1B (a neural-crest regulator) is linked to the “modern” face and is part of the self-domestication story. None of that is news…but if you tweak this lever, you clearly soften the whole reactivity profile…not just aggression. And my guess is that whoever’s fucking with the lever has his eye on the “compliance” dial more than any other.

Comparative signals align, too. Genomic work finds overlaps between human selective sweeps and domestication-candidate genes across species…showing a syndrome-level process rather than some sort of single behavioral knob. Craniofacial “feminization” over time in H. Sapiens fits reduced androgenic/reactive profiles, too.

Domesticated behavior tracks a “global calm.” Domesticated animals are less fearful, less erratic, and more socially tolerant than their wild counterparts. Your dog’s tendency to “look back” to you in unsolvable tasks is a manifestation of that…when arousal is lower and social cues are trusted, help-seeking beats reactive persistence. That’s a broad predictability play (that has nothing to do with aggression).

Obviously, Wrangham’s focus still matters. His key point, the decoupling of reactive vs proactive aggression in humans (we got tamer in the heat-of-the-moment sense, but remained capable of planned, coalitionary violence), is real and important to explain. It’s part of the story, but not the whole story. As general reactivity is reduced, strategic (planned) aggression is preserved…because strategic aggression isn’t a startle reflex; it rides on executive control and group coordination. But selection against reactive aggression isn’t the driver in this story. It’s just one behavioral readout of a deeper arousal/volatility downshift. A nice part (maybe) of an otherwise quite shitty story (from life’s vantage point). The beef industry might point out how nice the cows are, but I don’t think even they would try to argue that “nice” is what it’s aiming for. Dull. Compliant. And so it goes with all domestication. There is an objective in the domestication process, and any and all traits that impede progress toward that objective are pruned. (adding “self-” to domestication when it comes to humans, while accurate in the sense that the domesticating agent was of the same species, gives it a voluntary flavor that has no evidence in history…the domestication of humans was driven by systemic enslavement and reproductive control just as it was for all domesticates)

Why is it so important to me to find the driver of human domestication at all? Why not just start from the broadly-accepted premise that we are a domesticated species and go from there? Because I need to know what’s truly going on in the brain during this domestication process. How do we get to the brain we call “typical” now? What was it selected for? Was it selected for something broadly adaptive? Or is it more like runaway selection? An overfitting?

To me, cognitively, domestication looks like a down-weighting of volatility and a reallocation of precision (in predictive-coding terms). Brains with lower expected volatility (that have “the world is less jumpy” as a hyperprior…fewer LC-NE-style alarm bursts…a calmer autonomic tone), higher precision on social priors (human signals are treated as the most trustworthy ones…ecological “noise” gets less weight), and policy canalization (high confidence in proximity/compliance/help-seeking policies).

I think that human self-domestication primarily targeted behavioral and physiological volatility (a population-level reduction in phasic arousal and unpredictability) of which lower reactive aggression is a salient subset. And that the result is down-tuned HPA/LC reactivity, strengthened social priors, and canalized, low-variance action policies. Think of what happened as some sort of reactivity pruning (where reactive aggression was one prominent branch that got lopped off).

What is the domesticated brain? Zoomed out, it’s clearly an instrument that’s been made dull. One that exhibits blunted responses to non-social unpredictability (startle, sensory oddballs, metabolic stressors), not just to dominance threats. And anti-aggression alone doesn’t suppress those.

If I’m reading the studies properly, there are signatures of what I’m talking about in stress-regulatory and neuromodulatory pathways (HPA, serotonin, vasopressin) and neural-crest development…not just androgenic or specifically aggression-linked loci. Recent multispecies work pointing at vasopressin receptors and neural-crest regulators certainly seems consistent with this.

Wrangham’s story doesn’t account for lower intra-individual variance in exploratory/avoidant switches and faster convergence on socially scaffolded policies (like help-seeking) across types of tasks (anti-aggression predicts biggest effects only in conflict contexts). It doesn’t explain the psychotic consensus reality holding everyone in, as it rolls off a cliff.

(In fact, I question how much of the reactive aggression branch got lopped off…surely, not nearly as much as we think. What self-domestication mostly did was gate when, where, and how the majority of people show reactivity. When accountability and real-world consequences are high, most people keep a lid on it. When consequences drop (anonymity, distance, no eye contact, no immediate cost), the lid starts to rattle…online, in cars, in fan mobs, in comment sections. I don’t think reactive aggression was bred out so much as trained into context…and how well you do in that context will largely determine the story you tell. Harvard professors are clearly doing quite well in the civilizational context and consequently have pretty stories to tell.)

Comments

Leave a comment

More posts