Critical Realist Agent-Based Approach
How can social science research truly empower agency for systems change—without reinforcing the very systems we seek to transform?
Many popular systems approaches claim to address complexity by mapping leverage points, coordinating stakeholders, and fostering shared vision. However, when grounded in positivist assumptions, these methods risk reducing open, living systems into mechanistic closed systems — frozen in time, fixated on KPIs, and blind to emergent, relational, or epistemic harms.
In practice, once a “leverage point” becomes institutionalized (e.g., increasing college attendance), the system locks itself into a frozen assumption. Over time, the blind spots of that assumption become breeding grounds for new systemic harm.
This alternative approach combines:
Critical realism’s stratified ontology and morphogenetic analysis
Insights from complex adaptive systems — especially Derek Cabrera et. al.’s Agent-Based Approach
Together, they allow researchers to surface generative behavioral rules that agents follow across seemingly different domains.
🔍 Core Principles:
Wicked problems are sustained by recurring agent-level rules, not just by static structures or surface conditions.
Agents self-organize around rules that are simple, transfactual, and often unconscious.
Causality lies not in linear events or fixed variables, but in the activation of underlying generative mechanisms.
✳ Example:
Consider the rule:
“Redeem allocated resources that I depend on by fitting into allocation measures.” This encapsulates Allocation Dependence, a behavioral rule observable across education, international relations, and even human-animal dynamics. It persists even when surface phenomena (e.g., particular policy) appear to change.
By contrast, a rule like:
“Take care of the world around me through my self-actualization, and take feedback as regenerative input,” describes a relational, prosocial rule — as seen in Learning by Caring — that fosters emergence through care rather than conformity.
🧠 A simplified walk-through:
Identify a wicked problem that resists intervention.
Trace the patterns of action agents continue to take to cope or survive — despite those interventions.
Infer the behavioral rules or generative regularities at play.
Test whether the same rule exists in other contexts or systems — to determine its transfactual nature.
Prototype low-barrier, agent-level interventions that shift the behavioral rule itself — not just surface conditions or structures.
This approach de-centers the myth of top-down systems control and instead empowers transformative agency from within — wherever an individual, group, or initiative may be situated. It invites social science to be not just analytical, but deeply enabling.
Read more
See also
Last updated