Conversation

I have a dubious theory that historical change is always at least an exponential process (even if very slow for a long time) because it’s driven by some unbalanced dynamic that’s at least a first-order positive feedback loop (dx/dt = ax) Can anyone think of counterexamples?