Ricard SOLÉ & Brian GOODWIN
Signs of Life — (2000)
p.13 : Unpredictability comes in 2 forms: the first arises from the sensitivity to initial conditions that is characteristic of strange attractors. The second one relates to emergent phenomena. Here the issue is whether one can predict the behavior of a system of many interacting components when the properties of the individual components themselves are understood.
p.17 : Reductionism takes us from a complex phenomenon to more elementary properties of its components, giving us consistent explanations between the descriptive levels. But rarely can we go from the properties of the constituent parts to a description of the whole. Understanding the properties of H2 and O2 does not allow us to predict the properties of H2O; understanding the molecular properties of H2O does not allow us to derive the Navier-Stokes equations; having the Navier-Stokes equations does not give us a prediction and description of Bénard cells. In fact, unanticipated consequences of the Navier-Stokes equations are being discovered all the time. Self-organizing behavior emerges unpredictably in systems at different levels. We make it intelligible by recognizing how it is consistent with lower-level properties and by finding appropriate mathematical descriptors. But in doing this we do not reducea whole to the properties of its parts and their interactions. Although there are cases where this has been achieved, in general there is a causal gap between one level of description and the next, which is covered by a mathematical relationship producing consistency between levels. Emergent properties provide the recognition that nature can be creative while denying the occurrence of miracles or inconsistencies.
Clarifying these issues has been a major task of the sciences of complexity, the study of nonlinear systems whose behavior, though regular and repeatable for any particular conditions, changes in unpredictable ways when circumstances are changed. It is relevant to look briefly at some of the different viewpoints in this discussion, for there is as yet no consensus about emergent properties.
There is a powerful school of scientific thought that regards all apparently emergent properties as epiphenomena, having no explanatory significance or causal efficacy. Emergent properties in this view do not influence lower-level behavior in ways that cannot be accounted for by the lower-level properties themselves. This position holds that reality is as described in classical mechanics, and reductionism is the only mode of satisfactory explanation: wholes can always be described in terms of the properties and interactions of their parts.
It is generally acknowledged that quantum mechanics is not subject to this kind of reductionism; it is fundamentally holistic, with emergent properties of wholes that cannot be described in terms of parts. Examples of this are legion… superconductivity, superfluidity, ferromagnetism, crystals, lasers. Nonlocal connectedness… Symmetry-breaking in quantum systems. Initial symmetry- breaking in cosmic evolution cannot be explained in terms of the interactions of parts, because at that stage of the cosmic drama there were no parts.
The occurrence of emergent, irreducible properties in quantum systems does not mean necessarily that they also occur in macroscopic systems. It is worth noting, however, that such emergent phenomena as superfluidity and superconductivity are macroscopic, so emergence is not restricted to microstates. Thus the nature of physical reality is not that described by classical physics, and there is no reason to assume a priori that all irreducible macroscopic emergent properties are necessarily derived from quantum properties. Many aspects of macroscopic systems have the appearance of irreducible emergent properties but may not be describable by quanum mechanics. Michael SIlberstein uses the term « radical emergence » to describe the position that there are properties of macroscopic systems, living or otherwise, that are intrinsically irreducible to interactions among component parts…
We believe that reductionism is inadequate as the primary explanatory framework of science. Progress in understanding natural phenomena requires more than a study of parts in interaction. It often involves grasping relevant aspects of whole systems and finding appropriate mathematical descriptors that capture these properties… This process requires constant movement from phenomenon to model and back again. This aspect of scientific understanding, this continuous conversation between parts and wholes, between models and reality, is a major theme of this book.
The observer is crucial in all this. At any time a consensus may develop that a particular phenomenon, originally considered irreducible to a lower level of description, has been made reducible by the discovery of an appropriate lower-level description and the relevant pattern of interactions among components. This is part of the creative process of science. Only the process of scientific exploration and analysis can reveal which phenomena are « radically emergent » and which will turn out to be reducible. However,… « the unpredictability of emergents will always stay one step ahead of the groun won by prediction… As a result, it seems that emergence is now here to stay. » All the examples presented in this book have required high-level concepts to describe and make them understandable, involving a comparative study and analysis of higher and lower levels in their resolution. In every case, pure reductionism was an inadequate strategy, irrespective of whether observers later reached a consensus that the phenomenon can be understood in terms of appropriate components and interactions.
Thus we see parallels between chaos and emergent properties. Both involve unpredictability, though arising from different sources. With chaos, it is sensitivity to initial conditions that makes the dynamics unpredictable. With emergent properties, it is the general inability of observers to predict the behavior of nonlinear systems from an understanding of their parts and interactions. Once an emergent property is observed, we must find the appropriate descriptors that capture its intelligible dynamic essence and then show how this is consistent with lower-level properties. As James Critchfield has put it, « It is rarely, if ever, the case that the appropriate notion of pattern is extracted from the phenomenon itself using minimally biased procedures. Briefly stated, in the realm of pattern formation ‘patterns’ are guessed and then verified ». This is the creative work of scientists that parallels the intrinsic creativity of nature.
François KEPES 