Is domain formation decided before or after the transition?

There is increasing evidence that causality provides useful bounds in determining the domain structure after a continuous transition. In devising their scaling laws for domain size after such a transition, Zurek and Kibble presented arguments in which causality is important both before and after the time at which the transition begins to be implemented. Using numerical simulations of kinks in $1+1$ dimensions, we explain how the domain structure is determined exclusively by what happens after the transition, even though the correlation length freezes in before the transition.