🧵Empathy is a biased algorithm 🧵
Empathy is often romanticised as the pinnacle of moral development. But in evolutionary terms, it is a bounded adaptation - an affective mechanism fine-tuned to stabilise cooperation within coalitions, not to promote indiscriminate benevolence.
Empathy is often romanticised as the pinnacle of moral development. But in evolutionary terms, it is a bounded adaptation - an affective mechanism fine-tuned to stabilise cooperation within coalitions, not to promote indiscriminate benevolence.
From an adaptive standpoint, empathy functions as an in-group reinforcement system. Its evolutionary utility lies in detecting distress or need in individuals whose welfare reciprocally benefits one’s own genetic or cultural fitness.
In essence it is a neurochemical expression of inclusive fitness theory (Hamilton, 1964).
The neurobiology of empathy reveals its selectivity. fMRI studies show that when subjects view in-group members in pain, the anterior insula and anterior cingulate cortex exhibit strong activation.
When viewing out-group members, those same regions exhibit attenuated or even inverted responses (Hein et al., Neuron, 2010; Xu et al., PLoS ONE, 2009).
These effects are measurable down to milliseconds as event-related potentials show differential P3 amplitudes depending on perceived group membership. Empathy, neurologically, is not egalitarian, it is parochial pattern recognition shaped by millennia of coalitionary conflict.
The proximate mechanism is oxytocin, often miscast as the “love hormone.” In truth, oxytocin amplifies in-group trust and cooperation while simultaneously increasing out-group defensiveness (De Dreu et al., Science, 2011).
It does not make us universally kind, instead it seems to makes us selectively loyal.
Empathy behaves like a weighted function of relatedness and familiarity as the more genetically or culturally similar someone feels, the stronger our empathic pull. When those cues are absent, the emotional response decays, unless overridden by reason or institutional norms.
Historically, this bias was clearly adaptive. In small-scale societies, unbounded empathy would have produced maladaptive energy expenditure and vulnerability to free-riders.
Selective compassion ensured the survival of cohesive, kin-structured groups.
Selective compassion ensured the survival of cohesive, kin-structured groups.
However, the modern world has outgrown its evolutionary scaffolding.
Globalisation demands moral circles far larger than those empathy evolved to sustain.
We now inhabit social networks orders of magnitude beyond our Pleistocene brain’s tribal limit (Dunbar’s number ≈ 150).
Globalisation demands moral circles far larger than those empathy evolved to sustain.
We now inhabit social networks orders of magnitude beyond our Pleistocene brain’s tribal limit (Dunbar’s number ≈ 150).
The result is compassion fatigue and scope neglect, a phenomena described by Slovic (2007) showing that emotional engagement decreases as the number of victims increases.
Our affective architecture was never designed to handle abstraction, scale, or anonymity.
Our affective architecture was never designed to handle abstraction, scale, or anonymity.
This is why empathy often misfires in moral decision-making. It privileges the vivid over the statistical, the proximate over the distant, the familiar over the foreign.
As Greene’s work on moral cognition shows, moral progress depends on overriding our tribal intuitions with cognitive control or reasoned impartiality over affective bias.
Empathy bias explains political polarisation, humanitarian inconsistency, and moral tribalism. Neural imaging confirms that ideological partisans exhibit mirror-neuron suppression when witnessing out-group suffering (Bruneau et al., PNAS, 2012).
Our moral circuits are not universalist, they seem to be factional.
The implication can be unsettling, as empathy is not a moral virtue but a neurobiological heuristic.
It was selected for fitness, not fairness. Its function is not to make us humane, but to keep us cohesive.
The implication can be unsettling, as empathy is not a moral virtue but a neurobiological heuristic.
It was selected for fitness, not fairness. Its function is not to make us humane, but to keep us cohesive.
Civilisation, therefore, represents an ongoing act of cognitive rebellion against biology.
Institutions, ethics, and cosmopolitan norms exist to correct for the biases of an ancestral brain optimised for kin altruism and tribal survival.
Institutions, ethics, and cosmopolitan norms exist to correct for the biases of an ancestral brain optimised for kin altruism and tribal survival.
The future of morality depends on algorithmic augmentation, by designing educational, cultural, and technological systems capable of expanding the effective moral circle beyond the emotional radius of our evolved empathy.
To survive as a global species, we must replace intuitive empathy with principled impartiality or a morality grounded not in affective contagion, but in reasoned abstraction.
Empathy clearly evolved to favour us, not all.
Civilisation begins the moment we realise that goodness is not a feeling, it’s merely an override.
🧵END🧵
Civilisation begins the moment we realise that goodness is not a feeling, it’s merely an override.
🧵END🧵
• • •
Missing some Tweet in this thread? You can try to
force a refresh
