hypothesis that individuals who vary genetically in their capacity to learn (or to adapt developmentally; Ref. [9]) will leave most descendants because they will have the greater capacity to adapt.

In a short and insightful paper that appeared in 1987, Hinton and Nowlan [11] developed a simple computational model based on an extended version of genetic algorithm to demonstrate the magnitude of what was now being called the ‘Baldwin effect’. Their simulation, suggesting ‘how learning can guide evolution’, shows straightforwardly that creatures that are genetically predisposed to learn (in their oversimplified mode) by guessing the solution to a given environmental obstacle, by virtue of having correct settings on all the hardwired alleles, are on average more fit than those who cannot guess the solution. Moreover, their model demonstrated that, without ‘learnable alleles’, pure evolutionary search is completely blind and exceedingly slow.

The Berkeley biochemist, Wilson [12], who in the 1960s introduced the concept of a ‘molecular clock’ (based on genetic mutations that accumulate since they parted from a common ancestor) in evolutionary biology, predicted in 1985 that the presence of cultural factors may create a selective pressure for the ability to learn itself. Based on his early results on quantitative molecular evolution, he developed the concept of a ‘cultural drive’, through which the time required for a population to fix a mutation that complements a new behavior is shorter if the new behavior spreads quickly not only to offspring (vertically) but also to other members of the population (horizontally). His example of this cultural drive was the rise of agriculture that imposed new selection pressures, leading to swift genetic changes in human populations. He then considered the well-known example [13] of the introduction of milk sugar (lactose) into the diet of adults as the result of the invention and social propagation of dairy farming (pastoralism). In the relatively short period of 5000 years, genes conferring the ability to absorb lactose reached a level of 90% in populations dependent heavily on dairy farming, while, in contrast, the level of these genes is virtually zero in human populations that do not drink milk and in all other mammalian species tested. Analyzing the same phenomenon, the correlation of a genetic variation and a cultural trait, Feldman, Cavalli-Sforza, and Zhivotovsky [13] described it as ‘gene-culture coevolution’.

The edge-of-chaos regime is the optimal condition to be in a constantly changing environment, because from there one can always explore the patterns of order that are available and try them out for their appropriateness to the current condition. What is not necessary at all is to get stuck in a state of order, which is bound sooner or later to become obsolete. In that way, complex social systems that can evolve will always be near the transition region, poised for that creative leap into novelty and innovation, which is the essence of the evolutionary process.

‘life is evolution at the edge of disorder’.

In 1987, Modelski (Ref.[19], Chap. 5) proposed that the rise and decline of world powers (known also as the long cycle, the constitutive process of world politics) are best understood as a learning process, and in 1991 [20] described it as “evolutionary learning”. In 1996, he presented the evolution of global politics as a complex system situated at the border between order and chaos (Ref. [21], pp. 331-332)… Modelski and Perry [23] argued that, in the perspective of centuries, democratization is the process by which the human species is learning how to cooperate, and demonstrated that the rise in the proportion of the world’s population living in democracies (now exceeding 50%) is best described as a logistic process of the diffusion of a strategic social innovation.

The pace of the process, and hence the duration of the K-wave, is determined by the two biological control parameters already discussed: the cognitive (the collective learning rate), driving the rate of exchanging and processing information at the microlevel, and the generational, constraining the rate of transfer of knowledge (information integrated into a context) between successive generations at the macrolevel.

typical values for the diffusion learning rate of basic innovations are 16-17%, corresponding to typical time spans of about 25-30 years [generational turnover] for the spread of these radical innovations.

We note, first, the multilevel (or hierarchical) character of this evolutionary analysis. It posits that social evolution is not a singular process with one simple trajectory but an entire cascade across a number of levels—agent, institutional, species-wide—and those evolutionary processes occur or proceed at each of these levels [recall panarchy]. That accords broadly with the position of Gould, described by him as the “hierarchical theory of selection” (Ref. [34], Chap. 8). Contrary to the conventional Darwinian argument, that selection operates solely at the organismic level, and which has recently been expanded to the level of the genes (in Richard Dawkins’ ‘gene selection’), Gould argues that “Darwinian individuals” (those with a reproductive potentiality, hence evolution-capable) may be found across an entire biological hierarchy, beginning with genes and cells, to organism, deme, and species, and it is the last level that is of interest for the present analysis. [how about ecosystem?]

The phenomenon at hand (the cascade of world evolutionary processes) is then a cascade of scale-invariant, interdependent, and structure-transforming processes at several levels of organization of the self-organizing complex world system. In other words, such structure-transforming processes come to existence through the innovation process occurring at the several levels of the cybernetic hierarchy [systems higher in the order are relatively high in information while those lower down are relatively high in energy. That is, in effect, information controls energy (via communication).] and at the several scales of world organization (local, national, regional, and global). But innovations must diffuse in and be learned by society, and the adaptive mechanism of learning is paramount in giving the pace of change at each level.

In as much as “information controls energy”, the cybernetic hierarchy might be seen as the expression of the requirements of learning. This is why, thirdly, each of the four world system processes can be described as an algorithmic (Dennett [37], Chap. 2) learning process, because each might be seen as four-phased, and the phases are ways in which information is transformed into energetic solutions. The phases of a social learning process are generally seen to be (1) developing a variety of information; (2) mobilizing support; (3) choosing and/or deciding; and (4) implementing. Most notably, this concept of learning also comprises the essential elements of Darwinian evolution, namely (1) variation, (2) cooperation, (3) selection, and (4) amplification (differential survival) [9]. This evolutionary concept can also be rephrased as specifying a set of simple rules whose application brings about complex systems. These rules are (1) generate variation; (2) mobilize (and generalize); (3) select; and (4) amplify/reinforce. [recall four phases of adaptive cycle]

our postulated cascade of world system processes: social systems may self-tune their structure to a poised regime between order and chaos (as if by an invisible hand, in Adam Smith’s felicitous phrase, and as Kauffman has pointed out), with a power law distribution of breakthrough events, or in other words, of innovations.

what is seen as self-organization might more precisely also be systemic learning. In more general terms, ours might be recognized as a “learning civilization”. It is good to know, too, that world history might be the unfolding of a millennial learning process. If, as Gould (Ref. [34], p. 1055) maintained, “most evolutionists. . . are historians at heart”, then maybe the reverse could also come to be true.

from this paper.

The concept can be understood by looking at its Table 2: Cascade of modern evolutionary processes.