Chapter 3. Evo Devo Foresight: Unpredictable and Predictable Futures

The RISVC Model: Self-Organization via Evolutionary Development

Virtually every interesting complex adaptive system we know of within our universe, from solar systems to cells, undergoes some form of ReplicationInheritance, Selection, and uses mechanisms of Variation and Convergence on future forms and functions to build its complexity.

We can call this a RISVC cycle. It is a “five-stage” process that all complex adaptive systems engage in. We can call it a cycle, as it is based on replication, but these five processes are not sequential, but occur in parallel simultaneously, often in different parts of the same system. For example, as you are maturing to an adult, parts of your body and mind are getting ready for or are already engaged in replication (of your genes, your ideas, your behaviors), parts of you are a product of your inheritance (genes, family conditions, ideas), and of environmental selection (on your thoughts, feelings, behaviors, and organism), parts of you varying (creating), driven to find things and information that is truly different and new, and parts you are converging on a variety of future structures, physical and mental, as you develop. Here again is the a graphical depiction of the model (picture right).

The VRISC Model of Self-Organization (Evolutionary Development)

The RISVC Model of Self-Organization via Evolutionary Development

Recall that standard evolutionary theory, as first described by Darwin, is often simplified to “Variation, Inheritance, and Selection” over long periods of Time, summarized as the VIST acronym. Historically, this approach treated all Inheritance as an evolutionary process. It took the emergence of evo-devo biologists in the 1990s to remind us that it is critical to recognize that certain types of inheritance (developmental genes and factors) vary far less in space and time than others (evolutionary genes and factors). The VIST model also ignores Convergence. Convergence at cultural, ecological, and universal scales is one of the topics still missing from the standard neo-Darwinian synthesis: a recognition of universal development.

Our model assumes that Replication of some complex system, whose physical and informational boundaries we must define, is always the root source of self-organization. All self-organization, or “order that apparently is free”, happens because a system, its seeds, and its environment have been involved in past replication. A developmental approach adds Convergence (to environmental optimality, and to allow future replication), as two key ways to understand long-range developmental change. Development is guided both by carefully-tuned initial replicators (developmental genes), and by organismically and environmentally-guided convergence on predictable future form and function in the selection environment.

This convergence process is called convergent evolution by Darwinists, but it is better understood as ecosystem, biogeographic, cultural, technological, planetary, or in the broadest analysis, universal development, as we have described. In an evo devo universe both universal evolution and universal development are simultaneously occurring. Darwin’s Inheritance factor (the Seed) is split in the EDU model between varying inherited characteristics (the vast majority of genes, in the 95/5 rule of genetics), which we consider under Variation, and stable inherited characteristics (developmental genes and factors) which we consider under Convergence. Both are inputs to Replication (Organism form and life cycle). One of the unique benefits of evo devo thinking is that it requires us to pay attention both to the things that are likely to change, cycle by cycle, and the things that are likely to stay the same.

Environmental Selection, in turn, is also both an evolutionary and developmental process, and we color it purple to indicate that it is a hybrid of both evo and devo processes, along with the Replicator (Organism) and its mechanisms of Inheritance (Seed). Evolutionary selection creates unpredictable variation (divergence), replicators running adaptive experiments, while developmental selection converges predictably on future form and function that will protect the replicator and the replication process. Both are necessary for survival of the replicator in a complex and often hostile world.

Thus, using our own slightly altered definitions of evolution and development offered in A Meta-Darwinian Model, we can identify evolutionary, developmental, and evo devo features in replicating complex systems at virtually every scale.

A few of countless examples of replicating, inheriting, selecting, varying, and converging (RISVC) complex systems include:

  • Stars, which have advanced from the primitive Population III stars to the far more complex Population I solar systems, like ourSun and its complex rocky planets, over galactic time.
  • Prebiotic chemicals, which have built up their complexity in these special solar systems to create cells, over billions of years.
  • Cells, which created multicellular life with nervous systems, again over billions of years.
  • Nervous systems, which went from the Cambrian explosion to hominids, over roughly 500 million years.
  • Languages, ideas, and behaviors in hominid brains which birthed nonbiological computing systems, over roughly 5 million years.
  • Computing and robotics systems, whose replication is presently aided by human culture, may soon (within the next few decades, it seems) be able to replicate, evolve, and develop autonomously, bringing an even more complex adaptive system to “life”.

Identifying the RISVC life cycle of any replicative system, whether it be chemicals in a dish, stars in the universe, species on Earth, ideas and behaviors in brains and bodies, or technologies in societies, can tell us quite a lot about its future, independent of its adaptive environment.

This replication-centric view of the universe doesn’t fit every complex adaptive system. For example critics have pointed out that Galaxies, and the Universe itself, have not replicated over the last 13 billion years. That is an important point.

But since it is easy to argue that nearly every other interesting complex system in the universe has undergone some form of variation, replication, inheritance, selection, and convergence to build their complexity, it is parsimonious (conceptually the simplest theoretical model) to suspect this is how both galaxies and the universe built up their own adaptive internal complexity as well, via a long chain of prior replications in a hypothetical environment that physicists call the multiverse.

We can also ask another question: adaptation for what purpose? To what extent can we understand adaptation in biological evo devo as a teleological process (driven by evolutionary and developmental “goals” or “ends”) and apply this insight to our universe as a potentially evo devo system?

Like living organisms, our universe may have a developmental life cycle.

Like living organisms, our universe may have a developmental life cycle.

This brings us to the work of physicists like Lee Smolin, who presents evidence and argument in The Life of the Cosmos (1999) that our universe may be chained to a developmental life cycle, engaged in a process he calls Cosmological Natural Selection, with black holes acting as the replication environments. Smolin’s model has stood up to a number of criticisms since. But whether Smolin’s or other cyclic models in cosmology are ultimately validated is perhaps less important, at this stage in our science, than the observation that our universe looks like it is engaged in a RISVC life cycle, like all the other interesting complex systems within it.

We now think our universe had a definite birth (a Big Bang, 13.7 billion years ago) and it has passed through several developmental stages of growth, one of which may well be, according to astrobiologists, the development of life throughout the universe, with high probability. See Darling’s Life Everywhere (2007) for some estimates of how broadly life may exist in our universe. A growing number of scientists now propose that the creation of intelligent life is another of these developmental stages, either highly probable or inevitable on Earth-like planets.

We’ve also known for roughly 160 years that our universe must eventually die a “heat death”. The second law of thermodynamics, on which the heat death concept is based, is one of the most fundamental and widely validated laws of physics. See Adams’ The Five Ages of the Universe (2000) for a variety of speculations on how our universe will die. As our universe grows islands of accelerating local order and intelligence in a sea of ever-increasing entropy, thermodynamics tells us this process cannot continue forever. The universe’s “body” is aging, and will end in heat death, or perhaps even earlier in a big rip, or some similar fate. But if our universe is also a replicating complex adaptive system that engages in both evolution and development, as it grows older it must package its intelligence into some kind of reproductive system, a key feature of development, so that its complexity can survive its inevitable systemic death and begin again.

Perhaps the process that remains most obscure at present is how the universe converges on Replication, and what role intelligence, and future human civilizations, might play in that replication. Complexity scholar James Gardner offers one hypothetical model for that in his book Biocosm (2003). I offer another in Evo Devo Universe? (2008). While it is too early to know which if any of these models are more likely, the idea that the universe is a replicating system, and that its past replication explains a good deal of its internal complexity, and that one key role of intelligence is to have a positive influence on the replication of the universe are all increasingly popular ideas that need better scientific exploration.

This brings us to the concept of self-organization, which can be defined as a process of development that happens to complex systems and their environment as they go through multiple replicating cycles to build up adaptive information. In evo-devo biology, and in an evo devo universe, any physical system that has both evolutionary (variation, experimentation) and developmental (replication, convergence) features, and operates in a selective environment, will self-organize its own adaptive complexity as its replication proceeds.

A textbook example of self-organization is what happens to the molecules of a virus, or many supramolecular assemblies taken from inside cells, when you cut them up and place them in a petri dish. We are surprised to see these molecules partly self-assemble again “on their own” (or “by themselves”), with “no apparent help” from us. This self-organization happens only because the shapes and charges of these disaggregated molecules exploit the persistent physical laws of their environment to re-assemble historical form and function, which is encoded in the information of their present structure. They learned and internalized this evolutionary developmental information, as special shape-charge relationships, over countless past RISVC cycles. They also tuned a lot of that information into their genes, which express that information as three dimensional proteins in the cell, in a way that is developmentally robust to the noisy and chaotic conditions in typical cellular aqueous environments.

In exactly the same manner, if we live in an evo devo universe, certain physical and informational processes happening around us will not just evolve, they will self-organize, or engage in evolutionary development. Because of many past cyclings of our universe in a multiverse, certain ordering processes are implicit in our universe’s “genes” (fundamental parameters and other initial conditions, symmetries, and boundary conditions) and the surrounding metastable multiversal environment, so these processes don’t just evolve, they also develop.

From the perspective of the RISVC model, any scholar who talks about self-organization, is also talking about evolutionary development. The “e-word” is frequently found in self-organization literature, but the “d-word” is much more rarely used. Very few scholars consider the idea that development, at, many different system scales, may be occurring.

Consider the following quotation:

“Natural selection is important, but it has not labored alone… self-organization is the root source of order. The order of the biological world, I have come to believe, is not merely tinkered, but arises naturally and spontaneously because of these principles of self-organization – laws of complexity that we are just beginning to uncover and understand…” —Stuart Kauffman

Kauffman 1996

Kauffman 1996

Stuart Kauffman is a theoretical biologist, and a pioneer in the study of self-organization (evolutionary development). His first nontechnical book, At Home in the Universe: The Search for the Laws of Self-Organization and Complexity (1996) is the best introduction I’ve seen yet, for general readers, to the many ways that order arises spontaneously from chaos in the universe.

Like many scholars of self-organization, Kauffman studies the origin of life on Earth, one of most amazing examples we know of self-organization. There are other nearly as amazing self-organizing processes, such as the origin of Galaxies, which involve far vaster space-time and energy-mass scales. So too the origin of consciousness, which involves far smaller scales.

We know of several replicating chemical systems will self-organize (evolve and develop, or stably replicate) to produce order, and some of these have similarities to living cells. Our future evo devo information theory must explain why good, stable replicators and strong self-organizers inevitably outcompete systems that are only weakly good at these things in their use of local space, time, energy, and matter (STEM) resources. In evo devo terms, there must be some adaptive value in doing so, for even these nonliving replicators.

Pross (2014)

Pross (2014)

Chemist and origin of life scholar Addy Pross, author of What is Life? (2014), offers one such model in his concept of dynamic kinetic stability (DKS). Restating his work in evo devo terms, DKS argues that over time, replicating systems that can best harness energy to produce both evolutionary diversity and developmental stability, and which encode evo devo information in their self-organizing structure and environment, will outcompete, be more adaptive, than replicating systems that are less efficient at these things.

We start with replicating chemicals, and end up with resilient, compartmentalized cells. I look forward to further advances in DKS and other origin of life research leading us to a more generalized understanding of evo devo information theory. Self-organized criticality is another feature of dynamical systems that may describe how they spontaneously produce complexity.

Biochemist Nick Lane’s The Vital Question: Energy, Evolution, and the Origins of Life? (2015) may offer more clues to this mystery. Lane focuses on the special environment of alkaline hydrothermal vents at the bottom of the oceans on Earthlike planets. These undersea vents continually provide the necessary precursors of a special kind of rock, water, heat, hydrogen and carbon dioxide, and contain iron and sulfur-rich compounds which can catalyze primitive biochemistry, and offer an electrochemical gradient similar to life’s hydrogen ion gradient. If Lane’s hypothesis for the formation of the first organisms is true, given that our current Kepler-based estimate argues there are 40 billion Earth-like planets in the Milky Way alone, we can expect simple life (bacteria) to be exceedingly common in our universe.

Lane does a beautiful energetics analysis in which he estimates that eukaryotes are able to produce up to 200,000 times more energy per gene than prokaryotes, once they enter an endosymbiotic relationship with archaea, which become mitochondria in all complex cells. This is a fantastic example of STEM compression (energy densification) the natural developmental process we find behind all accelerating change, which we discussed in Chapter 2.

Unfortunately, Lane also makes a significant mistake in his book, in my view. On page 288, he proposes that one of the next critical steps in life’s complexification, the merger of bacteria and archaea to create eukaryotes, is likely to be quite rare in the universe. Perhaps he concludes this because it took 2.5 billion years for eukaryotes to emerge on Earth after simple life. But from a Cosmic Calendar perspective, 2.5 billion years sounds like the right amount of time for the incredible advance of symbiosis to be discovered. It is still just a fraction of the universe’s 13.5 billion year lifespan so far, and all the complexification events before it took much longer, depending on where you draw the lines. The formation of our solar system and of Earth, with its special hydrothermal vents, took perhaps five billion years.

To support this view, Lane cites the likelihood of “innate conflict” between the host and the endosymbiont genes and their differing drives as lowering the odds for symbiosis to occur. I am unimpressed with this argument, as I think it ignores emerging evidence for processes of universal development. I expect that we’ll soon discover funnels, or portal pathways in all this biochemistry, which naturally lead prokaryotes and archaea toward eukaryotes in a variety of ways, over the 2.5 billion years it too for the merger to occur.

By arguing that complex life must be rare, Lane falls into the trap of the astrobiologists who advanced the Rare Earth hypothesis who argued that both simple and complex life are likely to be rare. That hypothesis has been largely discredited by recent Kepler data with respect to simple life, and I fully expect them to be proven wrong with respect to complex life, once we can better simulate the biochemistry of the endosymbiosis portal to complex life.

Bottom line: we need a physical and information theory that tells us how replicating systems densify, experiment, and adapt, and why the accumulation of evolutionary and developmental learning, under special circumstances, eventually becomes life.

Again, self-organization is central to evo devo systems. If you cut up a virus, or many biological macromolecules, they will self-assemble again. They are just chemicals, but they almost seem alive. Their shapes and charges are adapted to each other, they find their energy minimum, as we see in the protein-folding funnel. They can vary their states and shapes (they are soft, floppy) yet they also develop, into predictable future states. The reason they do is because they’ve been selected, over many previous replications, to maintain a critical balance between evolving and developing abilities.

These pre-living chemicals use three physical systems to maintain their adaptive complexity, a replicating informational Seed (a molecular template, which eventually became DNA), a developing physical Organism (a protocell), and an informationally rich, selective, and protective Environment. This “SOE partitioning,” the distribution of growing intelligence between the initial parameters of the replicating Seed, the capabilities of the Organism and the Environment seems likely to be a critical way of understanding the origin of life, and it will likely be critical to maximizing the performance of naturally intelligent evo devo computers as well.

Many folks in the nanotechnology community are exploring self-assembly of nanostructures, as a way to fabricate useful computational and physical devices. Unfortunately, we humans have a very low understanding of how to do this. We can create environments where certain 2D structures will turn into 3D ones, we can catalyze particular reactions, but we don’t know how to control well for natural error or entropy, or to create dynamically stable, replicative structures.

We may need to be able to simulate how nature does self-organization, bottom up, using SOE partitioning, and run those evo devo simulations vastly faster in virtual space than they would occur in physical space, in order to find better strategies for creating smarter, more self-improving computers in the lab.

Now that we’ve considered the RISVC cycle, let’s claim a few very interesting things about the intersection of acceleration (Chapter 2) and development.

  • Oleksiy Teselkin

    Yes, self-assembling virus particles almost seem alive – put they are not, since all the information exhibit is in their physicochemical “properties.” True life – on top of this – is such because of another kind of information: information for other agents to multiply itself.

Share your Feedback

Better Wording? References? Data? Images? Quotes? Mistakes?

Thanks for helping us make the Guide the best intro to foresight on the web.