Fig 1.
Physical technology landscape, V(x), in which the agents of the community of knowledge learn.
Each point x represents a particular theory/technology. Points to the left of the valley minimum xb represent theories/technologies which belong to the old paradigm, which achieves maximum value at the local optimum x0, and points to the right represent theories/technologies which belong to the new paradigm, which achieves maximum value at the global optimum x1. Initial conditions involve n agents at the current paradigm x0.
Fig 2.
One realization of Eq (1) for different values of α.
(a): α = 0, (b): α = 0.15, (c): α = 0.7, (d): α = 1.
Fig 3.
Mean time to paradigm shift, as a function of α (the social learning parameter), for different values of n (population size).
Three different regimes can be observed for all values of n. For low values of α, which means that most learning is individual, an increase in α reduces the mean time to paradigm shift, until a minimum is achieved. For intermediate values of α, an increase in α increases the mean time to paradigm shift, until a maximum is achieved. Finally, for high values of α, which means that most learning is social, an increase in α decreases the mean time to paradigm shift.