Herein we consider various concepts of entropy as measures of the complexity of phenomena and in so doing encounter a fundamental problem in physics that affects how we understand the nature of reality. In essence the difficulty has to do with our understanding of randomness, irreversibility and unpredictability using physical theory, and these in turn undermine our certainty regarding what we can and what we cannot know about complex phenomena in general. The sources of complexity examined herein appear to be channels for the amplification of naturally occurring randomness in the physical world. Our analysis suggests that when the conditions for the renormalization group apply, this spontaneous randomness, which is not a reflection of our limited knowledge, but a genuine property of nature, does not realize the conventional thermodynamic state, and a new condition, intermediate between the dynamic and the thermodynamic state, emerges. We argue that with this vision of complexity, life, which with ordinary statistical mechanics seems to be foreign to physics, becomes a natural consequence of dynamical processes.
I. INTRODUCTIONWhy do things get more complicated with the passage of time? While it may not be a mathematical theorem, it certainly seems clear that as cultures, technologies, biological species and indeed most large-scale systems, those with many interacting components, evolve over time either they become more complex or they die out. The goal of understanding the mechanisms by which evolution favors increased complexity over time is too ambitious an undertaking for us here. We do not explore these mechanisms in part because they are phenomena-specific and we are concerned with only the universal properties of complexity even though it has no clear definition. More importantly, however, complexity is very often self-generating. Herein we discuss many of the problems and paradoxes that are entangled in the concept of complexity in the restricted domain of the physical sciences. This is done because if complexity does have universal properties they should be independent of the phenomena being studied and therefore we choose the simplest context possible. For example the generation and dissipation of fluctuations involved in complex phenomena are examined through the concepts of irreversibility and randomness, but more importantly through the objectivity principle, see for example Monod [1]. The science of complexity is multidisciplinary and so it might by argued that the schema we construct, based on the paradigm of physics, is incomplete. On the other hand, we believe that the principle of objectiveness gives "hard" sciences like physics an important advantage in addressing the difficult task of understanding complex systems. As a matter of fact, we should not ignore centuries of philosophy of science and, more generally speaking, epistemology, when we refer to the concept of objectiveness. It is right in this case indeed that science and philosophy benefit by mutual exchanges. We can accept the fact tha...