Beyond Western Nodes: Modeling the Multidimensional Sociality of sub-Saharan Africa
Capturing Sub-Saharan African Sociality in Social Networks to Generate a Culturally Realistic Population
The paper introduces a framework for generating culturally realistic virtual populations in sub-Saharan Africa using multidimensional social networks. It aims to support the SICOMORES system for simulating how communication and influence actions from peacekeeping forces propagate through complex, non-Western social structures.
TL;DR
Peacekeeping missions often fail not due to lack of force, but due to a lack of cultural resonance in communication. This paper presents a methodology for building culturally realistic virtual populations by mapping the unique "sociality" of sub-Saharan Africa into a multidimensional social network. By modeling layers like extended family lineages and "partisan undertakings," the system (SICOMORES) simulates how messages propagate in fractured, war-torn environments.
The Problem: The "Flat" Network Fallacy
Most information propagation models use a "one-dimension" network approach. While this works for simple social media analysis, it fails in the complex human terrain of sub-Saharan Africa. Previous systems like COMPOEX or StateSim often aggregated populations into broad ethno-political groups or used overly simplistic Hofstede cultural dimensions.
The authors argue that Sociality—the way individuals organize their society—is not a monolith. In Africa, an individual isn't just a "node"; they are simultaneously a member of a nuclear household, a participant in a village lineage, and a follower of a religious leader. To ignore these distinct layers is to ignore the actual "pipelines" of information.
Methodology: The Multidimensional Reading Grid
The core innovation is the use of a Multidimensional Network, where every individual node exists on multiple layers simultaneously.
1. Primary Sociality Layers
- The Family Dimension: Unlike Western nuclear models, the authors include Extended and Enlarged families. They specifically account for polygamy and lineage groups, where "Heads of Family" act as critical opinion leaders connecting different households.
- The Neighborhood Dimension: Rooted in village or urban district identities. Even in urbanized settings, the "community principle" remains the primary resource for solidarity and resource exchange.
- The Friendship Dimension: Based on "elective affinities," age groupings, and associations (e.g., tontines).
2. The Partisan Dimensions (The Religion-Politics Blur)
A key anthropological insight here is the porosity between religion and politics in Africa. The authors treat these as "Partisan Undertakings." In many sub-Saharan states, religious leaders hold political sway and political elites use religious symbols to communicate. The model reflects this by creating highly dense clusters around "Leaders" that integrate local branches into broader regional structures.
Simulating War-Torn Realities: Communitarianism
A critical feature for peacekeeping training is the Social Cohesion Indicator (0-100).
In a "failed state" scenario, social cohesion drops. The model translates this sociological state into network physics:
- Link Weakening/Deactivation: When cohesion is low, links between different ethnic or religious groups are "severed" or weakened, simulating discriminatory sociality.
- The War-Time Dimension: Conflict creates new, intense links based purely on shared grievances or ethnic survival. These links are only activated when the social cohesion threshold is breached.
Critical Insight: Why This Matters
The value of this research lies in its Inductive Bias. By forcing the network generator to respect specific African social structures (like the lineage-based connectivity), the resulting simulations of "message spreading" are naturally more accurate than a generic random-walk model.
However, a notable limitation is the current lack of geographical grounding. In actual peacekeeping, physical distance often dictates which media (loudspeakers vs. tracts) are effective. While the authors mention "Temporary Links" for events, a full integration of spatial GIS data would be the logical next step.
Conclusion: A Generic Framework for Specific Cultures
While the case study focuses on Africa, the methodology is a template for "Cultural Network Engineering." By identifying the specific "primary groups" and "partisan undertakings" of any region, we can generate virtual populations that reflect the true social fabric of the human terrain.
Takeaway: In complex social systems, the type of link is as important as the existence of the link.
