Shibboleth: An agent-based model of signalling mimicry

Area of Science:

• Evolutionary Biology
• Social Dynamics
• Computational Modeling

Background:

• Mimicry is a key strategy in competitive co-evolutionary dynamics.
• Detecting mimicked group-identity signals is crucial for preventing harm and may drive linguistic diversity.

Purpose of the Study:

• To investigate the role of mimicry detection sensitivity versus mimicry cost in the fixation of group identity signals.
• To model intergroup interactions using agent-based simulations inspired by the Shibboleth story.

Main Methods:

• Five agent-based models simulating multi-generational interactions between in-group and out-group individuals.
• Models varied parameters including mimicry quality, detection quality, and cost of detection.
• Incorporation of group-level behaviors like altruistic punishment in later model variations.

Main Results:

• Group members' sensitivity to mimicry (detection quality) was a stronger predictor of signal fixation than the cost of mimicry detection.
• The likelihood of being detected as a mimic is more critical than the costs incurred by detected mimics.
• This suggests a greater emphasis on detection likelihood in theoretical models of biological signaling.

Conclusions:

• Sensitivity to mimicry is more influential than mimicry cost in establishing group identity signals.
• The ability to imitate and evade detection is fundamental for gaining group membership through linguistic cues.
• Findings have implications for understanding linguistic evolution and social cohesion.