Figure 1.

An illustration of two element systems with increasing values of differentiation (note that the system elements displayed here are only a subset of a larger set of elements). Each quadrant of the figure shows the state transitions and differentiation values of the corresponding system, as well as the constrained and unconstrained cause-effect repertoire and cause-effect information of the mechanism composed of both elements, in the state m = (0, 0). System (A) has no differentiation, system (B) has low values of both $\mathcal{D}$₁ and $\mathcal{D}$₂, system (C) has low $\mathcal{D}$₁ and high $\mathcal{D}$₂, system (D) has high values of $\mathcal{D}$₁ and $\mathcal{D}$₂. As $\mathcal{D}$₁ increases (systems A→B→D), so does the cause information of the mechanism. This change happens by making the mechanism's cause repertoire more selective without changing the unconstrained cause repertoire (reducing convergence). As $\mathcal{D}$₂ increases (systems A→B→C), so does the effect information of the mechanism. This happens by increasing the entropy of the unconstrained effect repertoire without changing the effect repertoire (reducing divergence).