A(Life and Death): A Computational Model of Terror Management Theory

Overview

This repository contains the codebase for the self-organising multi-agent simulation described in the A(Life and Death) paper. The simulation is designed to model the effects of mortality salience on behviour, based on Terror Management Theory (Greenberg, Pyszczynski, & Solomon, 1986). Agents vary by attachment style and make decisions to self-sacrifice based on internal drives for worldview validation, relationship validation, and survival.

Core Features

• Attachment Styles: Secure, Dismissive, Preoccupied, Fearful — based on Bartholomew and Horowitz (1991).
• ASP Modules: Self-sacrifice decisions are driven by:
  • Mortality Salience
  • Worldview Validation
  • Relationship Validation
• Elimination Logic:
  • Natural death via aging (telomere).
  • Voluntary self-sacrifice.
  • Involuntary elimination if minimum eliminations not met.

Project Structure

TMT_Attachment
├── agents/             # Agent implementations for each attachment style
├── config/             # Simulation configuration
├── gameRecorder/       # Logging utilities for recording simulation data
├── infra/              # Core types, interfaces, and grid logic
├── JSONlogs/           # JSON files containing simulation logs per turn
├── plots/              # Python scripts for analyzing and visualizing results
├── server/             # Main server loop and elimination mechanics
├── grid_visualiser.py  # Agent grid visualisation tool
├── main.go             # Entry point for running the simulation

Running the Simulation

Ensure prerequisites are installed, then run:

go run main.go

From the root directory.

JSON logs will be output to the JSONlogs/ folder.

Plotting and Visualisation

Python plotting scripts are provided in the plots/ directory:

• population_over_time.py: Plots the total agent count over time.
• attachment_distribution.py: Tracks the proportions of each attachment type per turn.
• volunteers_vs_required.py: Compares voluntary self-sacrifice counts to required eliminations.

Visualisation of the agent simulation on the grid environment is also provided in the Python script grid_visualiser.py.