Why human cooperation tracks the volatility of the payoff rather than the virtue of the players.
The Moral Attribution Error
When a team hoards credit, silos its work, or refuses to share pipeline leads, leadership almost always diagnoses a culture problem. We treat cooperation and selfishness as fixed internal dispositions—moral traits that people bring into the building and that HR can screen for at the door. Yet personality and evolutionary psychology suggest that much of what we label as "culture" is not a transmitted set of values at all. It is evoked culture: universal psychological mechanisms switched on or off by the mathematical structure of the local environment.
In their foundational work on cognitive adaptations for social exchange, Leda Cosmides and John Tooby pointed out that cooperative sharing is governed by a strict environmental variable: resource variance. Not how much of a resource exists, but how predictably a single actor's effort converts into a yield on any given day.
Hunting vs. Gathering
Consider the field data collected by Hillard Kaplan and Kim Hill (1985) among the Ache foraging society of Paraguay. The Ache acquire calories through two distinct channels: hunting game and gathering plant foods. Meat from hunting is a classic high-variance resource. On any given day, the probability that an Ache hunter returns empty-handed is roughly 40 percent. Even a skilled hunter routinely fails due to luck, weather, or animal movement. Gathered food, by contrast, is a low-variance resource: yield scales directly and predictably with the hours and skill an individual invests.
If generosity were a global personality trait or an overarching cultural norm, the Ache would treat both resources identically. They do not. Meat is deposited with a central distributor and divided communally across the entire band according to family size. Gathered food is kept strictly within the immediate family. The same human beings are radical collectivists with protein and guarded individualists with carbohydrates, on the very same afternoon.
Elizabeth Cashdan (1980) documented the exact same pattern across San bands in the Kalahari Desert. Among the !Kung San, where food and water availability fluctuate wildly, fierce egalitarian norms dominate; being called "stingy" is a severe social offense met with immediate sanction. Among the Gana San, whose environment provides a lower-variance, more predictable food supply, sharing rarely extends beyond the immediate family, and substantial wealth inequality is tolerated without moral outrage.
Risk-Pooling as a Cognitive Switch
The mechanism driving this split is straightforward actuary logic baked into human cognition. Under high-variance conditions—where a windfall exceeds what one person can consume before it spoils, and tomorrow's failure is guaranteed by probability—sharing functions as an interpersonal storage bank. You give away surplus meat today to buy a claim on your neighbor's catch next week. Under low-variance conditions, where outcomes track individual effort rather than luck, mandatory sharing ceases to be insurance and becomes a tax on competence.
What looks like a clash of values is actually a difference in variance:
- High-variance regimes (early-stage R&D, hit-driven creative pitches, enterprise prospecting) naturally evoke communal risk-pooling—provided the upside of a rare hit is shared and the Spoilage Rule applies.
- Low-variance regimes (billable-hour execution, routine maintenance, quota-grinding) naturally evoke territoriality and strict reciprocity accounting, because output is a direct function of individual sweat.
Auditing the Environment, Not the Persona
When we model human behavior—whether we are diagnosing a dysfunctional agency floor or engineering persuasive narratives in NeuraScene—treating cooperation as a static character trait is a category error. You cannot sermonize a group into high-variance generosity while measuring them on low-variance individual metrics.
Before you design an intervention to make an audience, a user base, or an engineering team more collaborative, look past their stated values and audit the payoff matrix. If every actor's daily return is tightly bound to their solo effort, they will guard their territory. If you want them to share the kill, change the variance of the hunt.
Method note. This piece was built in NeuraScene's Advertising Mode using an explicit psychological frame:
- Persuasion trigger: Reciprocity
- Target affect: Flow State
- Gestalt principle: Common Region
Reciprocity directly operationalizes the article's core mechanism—cooperative risk-pooling under high variance is a delayed reciprocity ledger. Common Region visually binds the two otherwise rival researchers inside a single sharp boundary of floor light, making their shared risk pool immediately legible to the eye while their intense absorption reflects Flow State.
