3.3. Complexity and self-organization

In the introductory section, we have mentioned two features of the cellular automata model: the self-organization and the complexity. Depending on the context, different authors looked at cellular automata from both perspectives: as complex systems that, starting from a random initial state, manifest the self-organization property, or as simple, regular systems that exhibit a very complex, random behavior [1, 4, 7, 8].

In [19], a possible reconciliation of the two perspectives includes them both in the concept of "apparent complexity," understood as a "complex phenomenal appearance backed by a structural simple generative rule" (in order to be efficiently used, the term "apparent complexity" should be further, more clearly, defined). The same evolution or pattern may appear complex or simple, depending on the perspective of the analysis. In the particular case of linear cellular automata, of one does not know the rule and the initial state, it is difficult to infer them by analyzing the behavior. We recognize here again the problem of synthesis, as this issue is directly related to the previous one.

This dichotomy complexity—self-organization is the third paradox of cellular automata.
