Is there a way to determine chaos exists, as in a self-referential chaos, without knowing the equation behind it?

My guess is no. No matter how far you scan what appears to be a fractal, you'll never know if it's truly chaotic without knowing the fueling equation because it's always possible it will repeat.

But what methods of scanning for chaos are there, and is there a way to determine a relative degree of certainty that a system is possibly chaotic?

I ask because it has ramifications when scanning a system of axioms for self-replicacy (at the very least, in a finite number of steps). Say, for example, trying to determine a general set of rules for scanning a primordial-soup simulator for DNA-analogues.