An analytical framework that explains how social, technical and material effects emerge from networks composed of heterogeneous 'actors'—human and nonhuman—that enroll, translate and stabilize relations; it treats agency as the capacity to affect and be affected within a relational network rather than as an intrinsic property of individuals alone.

Principle

Collective outcomes are produced and changed by reconfiguration of relations among heterogeneous actors; intervening in material or discursive mediators can alter those outcomes without appealing to prior social essences.

Demonstration

Illustrative scenario → A city plans a bike‑sharing program. Recognition: analysts identify actors (users, bikes, docks, payment software, municipal regulations, repair shops). Action: planners modify docking locations and the payment interface to better enroll commuters and vendors. Consequence: usage patterns shift and new local repair economies emerge because changes to nonhuman mediators reconfigured relations among actors.

Misapplication

Interpreting the framework as attributing humanlike intentions to objects. The semantic error is conflating 'agency as capacity to affect' with literal intentionality; ANT attributes causal effectiveness to nonhumans without claiming they possess human mental states.

Consequence

Shifts research and intervention from solely studying individuals or institutions to tracing material and discursive networks; practical consequence is that policy and design may target mediators (technologies, texts, infrastructures) to produce social effects.

Reversal

When large‑scale statistical regularities or macrostructural constraints (e.g., aggregate economic forces) are the primary explananda, network tracing may be insufficient; ANT is less useful where explanation requires population‑level generalization or probabilistic causation rather than relational description.

Boundary

Clearly within: analyses of socio‑technical change where objects, texts and practices materially shape outcomes. Boundary case: institutional rules that operate through both networked mediators and durable legal categories—analysis may require both ANT tracing and institutional theory. Clearly outside: explanations that locate causation exclusively in individual psychology without reference to material or relational mediators.

Semantic Tension

Agency ↔ Structure — ANT's relational notion of agency can conflict with analyses that posit enduring social structures as primary causal bearers.

Synthesis

The key insight is that what acts in the world is a networked assemblage of humans and nonhumans; to change outcomes, identify and reconfigure the mediators and relations that assemble those networks.