The ground principle here is one of flexibility: that your model of behavior should allow for context dependence, not that it need be entirely context driven. A

The ground principle here is one of flexibility: that your model of behavior should allow for context dependence, not that it need be entirely context driven. As an analogy, consider a ground principle that your statistical model shouldn’t assume normal distributions or equal variance.