Authors

Ivan Briz i Godino¹, Virginia Ahedo², Myrian Álvarez³, José Ignacio Santos², José Manuel Galán²

¹ ERAAUB, IAUB, Universitat de Barcelona (Spain) & CADIC-CONICET (Argentina).
² Departamento de Ingeniería de Organización, Universidad de Burgos (Spain).
³ Departamento de Historia y Arqueología, Universitat de Barcelona (Spain) & CADIC-CONICET (Argentina).

Abstract

The Kawésqar society, native to the western Patagonian channels, thrived as a hunter-fisher-gatherer (HFG) group until the XX century. Renowned for their salient nautical technology, as well as a high degree of mobility and dispersion, the Kawésqar spanned circa 1000 km along the Pacific Coast in Southern Patagonia and Tierra del Fuego.

As a HFG group, their activity was circumscribed to two main areas: (i) inner channels and fjords and (ii) outer channels, islands and rocks (open waters). Notably, the inner channels provided them with the ‘usual’ resources (fish, molluscs, sea lions, etc.), along with the occasional stranding of sick, injured, or old whales that ventured into calmer waters. In turn, the outer channels offered not only heightened availability of the usual resources but also enhanced accessibility to marine mammals, particularly healthy whales, albeit with a notable increase in hunting risk.

Collective whale hunting required the cooperation of approximately 20 canoes with roughly 50 men, gathering around the massive animal to wound it with harpoons. This undertaking was perilous, especially in open waters, often resulting in serious injuries or even death (Duplessis, 2003; Gusinde, 1974). According to Gusinde (1974), the Kawésqar people did not face food scarcity periods, as resources from the inner channels were generally sufficient for survival, and the outer channels provided a resourceful alternative in case of ecological niche saturation. Consequently, the motivations behind the pursuit of unnecessary hazardous scenarios in general, and participation in collaborative whale hunting in particular, remain unclear.

One of the main hypotheses is related to the societal benefits they might have derived from engaging in such endeavours. More precisely, the Kawésqar people followed the mandates of Xólas (their supreme being), which strongly enforced a diligent altruism and a constant willingness to assist anyone in need and/or danger. Noteworthily, social control mechanisms were in place to ensure adherence to the mandates of Xólas, with admonitions, sanctions or even community expulsion being the consequences of non-compliance (Gusinde, 1974). Indeed, the hazardous way of life of these Fuegians resulted in a high occurrence of widows, orphans, invalids, and people weakened by age (approximately one person from every three families, as indicated by Gusinde). Crucially, all these individuals received abundant resources through a mechanism of recirculating food surplus based on network reciprocity, along with support and care within their social circle (including both family and non-relatives).

Alternative hypotheses propose that the driving force to engage in cooperative whale hunting could have stemmed from a weighted equilibrium between accessing resource surplus and the assurance of future mutual social support (Briz i Godino et al., 2014).

Inspired by the above narrative on the Kawésqar society, we have developed a stylised agent-based model (ABM) that is general enough to explore different artificial societies (including —but not limited to— the Kawésqar), driven by two main individual choices: (i) whether or not to take part in cooperative and risky quests for food, and (ii) whether or not to help others. Our main interest is to understand the different dynamics and equilibria that may emerge under the different parametrisations. Note that we opted for modelling since it enables us to investigate different scenarios, and ABM is commonly used in both evolutionary game theory to study cooperation (Adami et al., 2016; Roca et al., 2009) and in archaeological studies (Lake, 2014; Romanowska et al., 2021). In the following paragraphs, although our model is general and can represent many different societies, we describe it by occasionally making explicit references to the Kawésqar and their behaviour. This is done to help facilitate its understanding.

Specifically, our model has a finite population of N agents that represent individuals. The number of agents remains constant during simulation. Each agent has two variables that determine her behaviour: proactivity (her probability to participate in cooperative whale hunting, named whale-hunt in the model) and cooperation (her probability to share food resources with others, named help-others).

Regarding process variables, there are three of them: (i) resources, (ii) social capital and (iii) fitness (a Cobb-Douglas function that weights resources and social capital). Resources quantifies the amount of food that the agent has obtained, either individually and/or through the generosity of the group. As for social capital, it quantifies the social reputation of each agent resulting from her decisions to participate in cooperative hunting and/or to assist others. Besides, we created one binary state variable: abled/disabled, according to which each agent will be either fit or unfit to seek food resources.

As for the exogenous variables that define the simulation scenario, these encompass log-success-prob (the probability of whale hunting being successful, derived from a logistic distribution); whaling-harm-prob (the probability of being injured during hunting); theta, (a parameter that determines the exponents of the fitness function); and mutation-prob (agents imitate other strategies through random tournament and mutate with the probability defined by mutation-prob).

Eventually, the decisions taken by the agents determine their value of resources and social capital. Regarding resources, the agents choosing to cooperate in whale hunting obtain resources in the form of a high payoff, contingent upon the probability of success (log-success-prob); those who do not cooperate or cooperate but fail to succeed, obtain resources in the form of a low payoff. The agents that decide to help those in need contribute with their resources to a pool of resources, which is subsequently distributed equally among the members of the community. As for social capital, those agents engaging in cooperative whale hunting and helping others, obtain a very high social reputation in the form of a high payoff; the agents that either participate in whale hunting or help others, get an intermediate social reputation (intermediate payoff); and, eventually, those who neither cooperate in whale hunting nor help those in need, obtain social reputation in the form of a low payoff.

The simulation of our model following a rigorous design of experiments and the subsequent analysis of the results obtained will offer valuable insights into the behaviour of different societies in general, and of the Kawésqar in particular. It will shed light on the relationship between managing hazardous contexts and cooperation, highlighting the significance of these collective attitudes in confronting emerging critical scenarios.

References

Adami, C., Schossau, J., Hintze, A., 2016. Evolutionary game theory using agent-based methods. Phys. Life Rev. 19, 1–26. htps://doi.org/10.1016/j.plrev.2016.08.015

Briz i Godino, I., Santos, J.I., Galán, J.M., Caro, J., Álvarez, M., Zurro, D., 2014. Social Cooperation and Resource Management Dynamics Among Late Hunter-Fisher-Gatherer Societies in Tierra del Fuego (South America). J. Archaeol. Method Theory 21, 343–363. htps://doi.org/10.1007/s10816-013-9194-3

Duplessis, 2003. Périple de Beauchesne à la Terre de Feu (1698-1701). Une expédition mandatée par Louis XIV. Transboréal, Paris.

Gusinde, M., 1974. Die Feuerlandindianer, III, 1. Die Halakwulup. Mödling: Anthropos.

Lake, M.W., 2014. Trends in Archaeological Simulation. J. Archaeol. Method Theory 21, 258–287. htps://doi.org/10.1007/s10816-013-9188-1

Roca, C.P., Cuesta, J.A., Sánchez, A., 2009. Evolutionary game theory: Temporal and spatial effects beyond replicator dynamics. Phys. Life Rev. 6, 208–249. htps://doi.org/10.1016/j.plrev.2009.08.001

Romanowska, I., Wren, C., Crabtree, S., 2021. Agent-based modeling for archaeology: simulating the complexity of societies. The Santa Fe Institute Press, Santa Fe.