Evolutionary Game Theory Modelling of Guilt

Luı́s Moniz Pereira, Tom Lenaerts, Luis A. Martínez-Vaquero, The Anh HanTeesRep 2017

Abstract

Inspired by psychological and evolutionary studies, we
\npresent two theoretical models wherein agents have the potential to
\nexpress guilt, with the ambition to study the role of this emotion in
\nthe promotion of pro-social behaviour.We show that the inclusion of
\nthe emotion of guilt, in the sense arising from actual harm done to
\nothers from inappropriate action or inaction, is worthwhile to incorporate
\nin evolutionary game theory models of cooperation, for it can
\nincrease cooperation by correcting and inhibiting defection. The abstract
\nstudy thereof profitably transpires to concrete considerations
\nin the design of artificial multi-agent populations. To achieve this
\ngoal, analytical and numerical methods from evolutionary game theory
\nhave been employed, but not shown in too fine detail here, to
\nidentify that reasonable conditions exist for which enhanced cooperation
\nemerges within the context of the iterated prisoners dilemma.
\nGuilt is modelled explicitly as two features, i.e. a counter that keeps
\ntrack of the number of transgressions and a threshold that dictates
\nwhen alleviation (through for instance apology and self-punishment)
\nis required for an emotional agent. Such alleviation introduces an
\neffect on the payoff of the agent experiencing guilt. We show that
\nwhen the system consists of agents that resolve their own guilt without
\nconsidering the co-player’s attitude towards guilt alleviation then
\ncooperation does not emerge. In that case, agents expressing no guilt
\nor having no incentive to alleviate the guilt they experience easily
\ndominate the guilt prone ones. On the other hand, when the guilt
\nprone focal agent requires that guilt only needs to be alleviated when
\nguilt alleviation is also manifested by a defecting co-player, then cooperation
\nmay thrive. This observation proves consistent in a generalised
\nmodel discussed in this article. In summary, our analysis provides
\nimportant insights into the design of multi-agent and cognitive
\nagent systems, wherein the inclusion of guilt modelling can improve
\nagents’ cooperative behaviour and overall benefit.
BibTeX
@inproceedings{pereira2017evolutionary,
  author = {Luı́s Moniz Pereira and Tom Lenaerts and Luis A. Martínez-Vaquero and The Anh Han},
  title = {{Evolutionary Game Theory Modelling of Guilt}},
  booktitle = {TeesRep},
  year = {2017},
}

Loading related content...

Figure