Evolution of Egalitarian Social Policing

Do you have questions or comments about this model? Ask them here! (You'll first need to log in.)

WHAT IS IT?

This model is associated with the MS/article 'Modelling egalitarian social policing of prestige-based status hierarchies in human social evolution' by Ben Gleeson. It examines evolved human socio-behavioural dynamics with a view to explaining paradoxically high levels of contemporary male violence and social status. It is based on a conceptual synthesis of ‘the challenge hypothesis’ (Wingfield, 2017; Wingfield et al., 1990) and ‘biosocial model’ hypothesis (Mazur, 1985; Mazur & Booth, 1998) of testosterone, together with theorised ‘social selection’ against domineering ancestral males (Boehm, 1993, 2017; Gintis et al., 2019; Knauft, 1991; Sarkar & Wrangham, 2023; Wrangham, 2021b, 2021a), and inferred transitions from dominance-based male status competition to complex and socially-mediated ‘prestige-based’ forms (Erdal & Whiten, 1994; Henrich & Gil-White, 2001; Zeng et al., 2021).

The model depicts a generic hominin social environment where prestige-based status competition allows success in culturally constructed social hierarchies, but impulsive (‘reactive’) aggression is punished. The exact sociocultural context and modes of status competition are unspecified, meaning that these basic model dynamics may apply to any hypothetical cultural group, population scale, or period of human social evolution. Similarly, the model does not prescribe specific behaviours that would constitute a ‘reactive’ response. These might include any hostile social defection prompting the individual to leave the group, or be ostracised, including physical violence, egregious insult, refusal to comply with authority, destruction of property, or other ‘hot-headed’ social transgression.

HOW IT WORKS

Each agent within the model has two primary traits representing the implications of focal hypotheses. Their status potential within a given group hierarchy is determined by their ‘Status Score’ (SS); and their chance of reacting with anti-social aggression, leading to group ostracism or defection, is determined by their ‘Reactive Aggression’ (RA). At the beginning of each model run, the operator selects the total number of agents (between 10 and 1500) and the mean-RA (1 to 10) for that population. Each agent is then randomly assigned an RA score from a normal distribution based on the determined mean, with a standard deviation set at 15% of that mean. Agent SS is also randomly assigned from a normal distribution with a standard deviation of 15%; but, because social stratification is entirely relative, the exact numerical mean is irrelevant. Given this, mean SS is arbitrarily set to 3, with a standard deviation of 0.45.

Agents begin each model run as an isolated individual, randomly located within the model environment. Because they are not yet part of a social group, their assigned ‘status level’ is zero. As the run progresses, agents move and interact with nearby others to form social groups. When an agent joins a group, they automatically rise to status level 1, and move to cluster around the shifting mean central point of their group. Within groups, agents ‘compete’ in status hierarchies by regularly comparing their own SS to other group members of the same status level. If they have the highest SS at that level, and there is room in the status level above, they rise to the next highest level. Each agent at a given status level above 1 requires three agents in the level below, following the roughly 3:1 scaling found in many real world social hierarchies (Dunbar & Sosis, 2018; Hamilton et al., 2007). This scaling ensures hierarchical social complexity (the number of differentiated status levels in a given group) is a direct function of group size; in effect, larger model populations lead to more stratified social hierarchies.

At each time-step of a model run, if an individual does not have the highest SS for their current status level, they are deemed to lose a status contest, and will either stay at their current level, or will fall to a lower level if another agent (rising from below) displaces them. Based on continuous SS comparisons, agents rise and fall between status levels. However, each time they lose a status contest, or are demoted, there is a chance they will respond reactively and defect (or be ostracised), thus leaving the group. Agents that defect become ungrouped individuals and revert to a status score of zero, moving randomly until they encounter another agent or group. Whether a losing agent responds ‘reactively’ and defects, or not, depends on a randomly generated number between 1 and 20 being lower than the agent’s RA score; so, the lower their reactivity (RA score), the less likely each agent is to defect. In addition, to represent social policing effects, the current status level (1-7) of the highest-ranked individual in a given group is deducted from each agent’s RA score before it performs a reactivity test. In effect, therefore, the larger and more stratified the social group, the lower the chance of group members reactively defecting after losing a status contest.

HOW TO USE IT

The basic form of the model allows the user to select a population size (10-1500) and population 'mean RA' (1-10), click 'set up' and 'go', then monitor aspects of agent social interaction, including Maximum group size, Max Status-level, Mean Status-level, and to see graphed Status-level distributions.

With the 'Sex-bimodal?' switch set to 'on', the user may also select different ratios of Female to Male RA and SS. Adjusting the 'F:M-Reactive-Aggression' slider determines the mean of the normal distribution of female Reactivity by applying the selected percentage ratio of the set ‘Mean-Reactive-Aggression’ for that specific population. The 'F:M-Status-Score' slider determines the mean of the normal distribution of female SS by applying the selected percentage ratio of the mean SS (which is always 3). The user may also then monitor the percentages of Male and Female agents that are members of groups, as opposed to wandering defectors, and the relative 'Status-level' of Male and Female agents.

THINGS TO NOTICE

With the 'Sex-bimodal?' switch set to on, and 'F:M-Reactive-Aggression' and 'F:M-Status-striving' both set to 1, the F:M Status-level ratio typically finishes somewhere around 1. Predictably, lowering the 'F:M-Status-score' slider leads to a relative reduction in Female mean status. While the overall 'Mean-Reactive-Aggression' slider is set to low levels (1-3), decreasing the 'F:M-Reactive-Aggression' ratio does not significantly improve this F:M status-level outcome. However, at higher 'Mean-Reactive-Aggression' levels (>5), a lower 'F:M-Reactive-Aggression' ratio (meaning Females are relatively less reactively aggressive) increases female Status-level outcomes, largely because the proportion of male defectors (ungrouped loners with 0 social status) increases.

As an arbitrary example, in a population of 500, with 'Mean-Reactive-Aggression' of 10 (i.e., maximum RA setting), if the 'F:M-Reactive-Aggression' level is set to 0.33 and the 'F:M-Status-score' is 0.66, most defectors are Male, most group members are Female, Females have a higher mean Status-level than Males, BUT a small number of Males still tend to occupy the top-most Status-level ranks. These outcomes, though extreme, resemble many real-world social patterns.

THINGS TO TRY

Start with the 'Sex-bimodal?' switch turned off and shift the 'population-size' and 'mean-reactive-aggression' sliders to get a feel for the basic workings of the model. Compare higher and lower levels of Population Mean-RA with consistent population sizes to see how the pattern of grouping behaviour and status levels change as the average level of aggressive reactivity declines over time. Then try fixing mean RA and see how grouping behaviour changes with different population levels. Next, try turning on the 'Sex-bimodal?' switch and adjusting the two F:M ratios of SS and RA to see how bimodal distributions affect grouping behaviour and status differences between the sexes.

EXTENDING THE MODEL

Further iterations of this model are encouraged and may prove useful to improve accuracy, or relevance to specific human social contexts. The current model assumes humans automatically and randomly group together, and provides no limit to this behaviour. But real-world group sizes vary according to cultural traditions and ecological constraints. In addition, individual decisions to join a group may be influenced by familial ties, personalities, and past social interaction. Modelled groups also currently coalesce on contact, but smaller real-world societies with the capacity to persist in isolation may do so for a range of reasons, including preference for self-determination, or to avoid relatively low status in neighbouring social hierarchies due to ethnic or cultural bias.

Also, modelled agents only ever defect from the group; they are not executed, and nor is reproduction occurring. As such, this is not a model of evolutionary change via selection and differential fitness. Instead, each model run provides a snapshot of hierarchical social complexity and status distributions for a population of a given size and mean reactivity. Comparison of different sizes and reactivity levels imply effects from varied population densities and theorised change in reactivity over evolutionary timescales. However, future iterations of the model might mimic dynamic population-level evolution by incorporating births and deaths, and tracing differential fitness from varied reactivity and social status. Another expansion could incorporate intergroup competition, whether through physical hostility, or via fitness differentials from socioculturally enhanced cooperation and sharing.

COPYRIGHT AND LICENSES

The MIT License (MIT)

Copyright (c) 2025 Ben Gleeson.

All rights reserved.

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

CREDITS AND REFERENCES

The model was constructed with Netlogo 6.1.0 (Wilensky, 1999) reformatted for Netlogo 7.0.2 and is available on the Netlogo Modelling Commons. For further description, including code, and ODD protocol (following Grimm et al., 2020), see Supplementary Information available at https://github.com/Ben-Gleeson71/MedMaleConAgg

Boehm, C. (1993). Egalitarian Behavior and Reverse Dominance Hierarchy. Current Anthropology, 34(3), 227–254.

Boehm, C. (2017). Ancestral Precursors, Social Control, and Social Selection in the Evolution of Morals. In M. N. Muller, R. W. Wrangham, & D. R. Pilbeam (Eds.), Chimpanzees and Human Evolution (pp. 746–790). Harvard University Press. https://doi.org/10.4159/9780674982642-020

Dunbar, R. I. M., & Sosis, R. (2018). Optimising human community sizes. Evolution and Human Behavior, 39(1), 106–111. https://doi.org/10.1016/j.evolhumbehav.2017.11.001 Erdal, D., & Whiten, A. (1994). On Human Egalitarianism: An Evolutionary Product of Machiavellian Status Escalation? Current Anthropology, 35(2), 175–183. https://doi.org/10.1086/204255

Gintis, H., van Schaik, C., & Boehm, C. (2019). Zoon politikon: The evolutionary origins of human socio-political systems. Behavioural Processes, 161, 17–30. https://doi.org/10.1016/j.beproc.2018.01.007

Grimm, V., Railsback, S. F., Vincenot, C. E., Berger, U., Gallagher, C., DeAngelis, D. L., Edmonds, B., Ge, J., Giske, J., Groeneveld, J., Johnston, A. S. A., Milles, A., Nabe-Nielsen, J., Polhill, J. G., Radchuk, V., Rohwäder, M.-S., Stillman, R. A., Thiele, J. C., & AyllÃ3n, D. (2020). The ODD Protocol for Describing Agent-Based and Other Simulation Models: A Second Update to Improve Clarity, Replication, and Structural Realism. Journal of Artificial Societies and Social Simulation, 23(2), 1–7.

Hamilton, M. J., Milne, B. T., Walker, R. S., Burger, O., & Brown, J. H. (2007). The complex structure of hunter–gatherer social networks. Proceedings of the Royal Society B: Biological Sciences, 274(1622), 2195–2203. https://doi.org/10.1098/rspb.2007.0564

Henrich, J., & Gil-White, F. J. (2001). The evolution of prestige: Freely conferred deference as a mechanism for enhancing the benefits of cultural transmission. Evolution and Human Behavior, 22(3), 165–196. https://doi.org/10.1016/S1090-5138(00)00071-4

Knauft, B. M. (1991). Violence and Sociality in Human Evolution. Current Anthropology, 32(4), 391–428.

Mazur, A. (1985). A Biosocial Model of Status in Face-to-Face Primate Groups. Social Forces, 64(2), 377–402. https://doi.org/10.2307/2578647

Mazur, A., & Booth, A. (1998). Testosterone and dominance in men. Behavioral and Brain Sciences, 21(3), 353–363. https://doi.org/10.1017/S0140525X98001228

Sarkar, A., & Wrangham, R. W. (2023). Evolutionary and neuroendocrine foundations of human aggression. Trends in Cognitive Sciences. https://doi.org/10.1016/j.tics.2023.02.003

Wilensky, U. (1999). Netlogo (Version 6.1.0) [Netlogo]. Centre for Connected Learning and Computer-Based Modeling, Northwestern University. ccl.northwestern.edu/netlogo

Wingfield, J. C. (2017). The challenge hypothesis: Where it began and relevance to humans. Hormones and Behavior, 92, 9–12. https://doi.org/10.1016/j.yhbeh.2016.11.008

Wingfield, J. C., Hegner, R. E., Dufty, A. M., & Ball, G. F. (1990). The “Challenge Hypothesis”: Theoretical Implications for Patterns of Testosterone Secretion, Mating Systems, and Breeding Strategies. The American Naturalist, 136(6), 829–846.

Wrangham, R. W. (2021a). Targeted conspiratorial killing, human self-domestication and the evolution of groupishness. Evolutionary Human Sciences, 3. https://doi.org/10.1017/ehs.2021.20

Wrangham, R. W. (2021b). The Execution Hypothesis for the Evolution of a Morality of Fairness. Etica & Politica / Ethics & Politics. https://doi.org/10.13137/1825-5167/32385

Zeng, T. C., Cheng, J. T., & Henrich, J. (2021). Dominance in Humans [Working Paper]. https://henrich.fas.harvard.edu/publications/dominance-humans

There is only one version of this model, created 4 days ago by Ben Gleeson.

Attached files

This model does not have any ancestors.

This model does not have any descendants.