Every tool has a group it is built for, even if nobody wrote it down. A kitchen knife serves one cook. A village well serves a few hundred households. A social network built for a dorm turned out to serve three billion people, and the strain shows. Asking “how many people can this tool serve?” is one of the plainest ways to find where a tool crosses from convivial to manipulative.
Dunbar’s number
The best-known answer to the group-size question comes from the anthropologist Robin Dunbar. In 1992 he reported that among primates, the size of the neocortex relative to the rest of the brain predicts typical group size. Extrapolating to humans, he estimated a natural group of about 150: the number of people one can keep track of as individuals, with a sense of who they are and how they relate to each other. A 1993 paper set out the evidence from human societies, from hunter-gatherer clans to army companies.
Later work by Dunbar and colleagues found not one number but a series of nested layers, each about three times the size of the one inside it: roughly 5, 15, 50, 150, 500, and 1,500. Intimacy and frequency of contact fall off at each step out.
How firm is it?
Less firm than its fame suggests. In 2021 Patrik Lindenfors, Andreas Wartel, and Johan Lind re-ran the primate analysis with modern methods and data. Depending on method, their estimates of the mean ranged from 16 to 109, with confidence intervals so wide (as broad as 4 to 520) that they concluded specifying any single number is futile. Dunbar has disputed that analysis, and the argument continues.
For Counterfoil’s purposes the fight over the exact number matters less than what both sides accept: human social capacity is finite, it is organized in layers, and tools that ignore those layers run into trouble. Whether the ceiling is 100 or 200, a tool that asks people to relate to ten thousand others as though they were neighbours is working against the grain.
Illich’s concentric scales
Illich made a similar point without the brain-size data. In Tools for Conviviality he described people as living in several concentric social environments (chapter III, “Frustration,” p. 94), from the primary group to the work unit to the city, each with its own natural scale. He proposed that counterfoil research should include a kind of dimensional analysis of people and their tools: matching each tool to the scale of group it serves well.
Questions for the catalogue
- Which layer is this tool designed for: one person, a household, a team, a community, a public?
- Does it still work if the group shrinks? A tool that only functions at massive scale makes small groups dependent on large ones.
- Does it say what happens when the group grows past its layer? Good tools split, federate, or cap. Bad ones just keep adding people.
Sources
- Dunbar, R. I. M. “Coevolution of neocortical size, group size and language in humans.” Behavioral and Brain Sciences 16, no. 4 (1993): 681–694. https://doi.org/10.1017/s0140525x00032325.
- Dunbar, R.I.M. “Neocortex size as a constraint on group size in primates.” Journal of Human Evolution 22, no. 6 (1992): 469–493. https://doi.org/10.1016/0047-2484(92)90081-j.
- Illich, Ivan. Tools for Conviviality. New York: Harper & Row, 1973. https://archive.org/details/illich-conviviality.
- Lindenfors, Patrik, Andreas Wartel, and Johan Lind. “‘Dunbar’s number’ deconstructed.” Biology Letters 17, no. 5 (2021). https://doi.org/10.1098/rsbl.2021.0158.
- Zhou, W.-X., D. Sornette, R. A. Hill, and R. I. M. Dunbar. “Discrete hierarchical organization of social group sizes.” Proceedings of the Royal Society B: Biological Sciences 272, no. 1561 (2005): 439–444. https://doi.org/10.1098/rspb.2004.2970.