The most recent and refined version of the evolutionary worldview that was first presented in Evolutionís Arrow can be found in the 34 page document The Evolutionary Manifesto which is here
Chapter 8. Smarter Cooperation
A manager that is in full control of an organisation has the potential to establish a wide range of cooperative activities. In principle, such a manager is capable of putting together any pattern of cooperative specialisation and division of labour. But what particular cooperative activities should it organise? Of all the alternatives the manager could support, how does it discover those that are best for itself and the organisation? Unless the manager is able to discover the most productive forms of cooperation, it will not be able to fully exploit the benefits of cooperation. It is not enough that a manager is able to organise cooperation. It must also be smart. It must be able to discover the cooperation that is best.
Early cells, the first multicellular organisms, and human tribes all had the potential to construct an enormous variety of cooperative relationships between their members. But they had to discover those that were best, and adapt them as circumstances changed. And this did not come easily. Millions of years passed before the cells within multicellular organisms discovered how to specialise and cooperate to form an effective eye. Many more millions of years passed before the cells discovered how to cooperate to produce the complex brain found in humans and other mammals. And it was a long time before humans discovered how to cooperate together to build nuclear power stations, and to send people to the moon.
Organisations that are superior at discovering new and better cooperative adaptations amongst their members will have an evolutionary advantage. Their greater evolvability will enable them to exploit more effectively the immense potential benefits of cooperation. In turn, these benefits will reward improvements in evolvability. So the potential benefits of cooperation drive more than just the evolution of managed organisations of greater and greater scale. The benefits will also drive the evolution of managed organisations that are better at evolving. As evolution unfolds, living processes will get better at evolving, smarter at searching out the best ways to cooperate, and more innovative and creative at adapting their cooperation as conditions change.
In the next four Chapters, we will look at how evolution has progressively improved the adaptability and evolvability of living processes on this planet, and how it will continue to do so in the future. We will see that the existence of this progressive sequence of improvements raises significant questions for each of us: as individuals, where are we located in the evolutionary sequence? How much room for improvement is there in our adaptability and evolvability? Will we need new psychological capacities if we are to evolve in whatever directions are necessary for us to contribute to the future evolution of life in the universe? Can our knowledge of the direction in which evolvability improves point to how we must evolve psychologically if we are to contribute to future evolution?
The existence of an evolutionary progression in evolvability also raises important questions for humanity as a whole: how much room for improvement is there in the evolvability of human society? How could the evolvability of human society be enhanced? How could we improve the ability of our systems of government to search out new and better ways to manage our societies? Multicellular organisms eventually evolved brains and nervous systems that are far superior at adapting and evolving than the individual cells that form them. Must human society do the same? Must our societies evolve supra-individual adaptive processes that will be far smarter than the humans they contain? If so, how can we construct these supra-individual adaptive processes?
We will see that an understanding of the direction in which evolvability evolves is critically important for future human evolution. Such an understanding will enable us to locate ourselves within the evolutionary sequence and to see what improvements in our adaptive capacities will be necessary in the future. It will point to what we have to do to improve our evolvability, as individuals and collectively. Becoming aware of the direction in which evolvability improves is an important step in the evolution of improved evolvability.
We will begin by looking in detail at how evolvability has improved during the past evolution of life. But before we start to trace this evolutionary sequence, we need first to develop a thorough understanding of the basic process that is used by living things to evolve and discover new adaptations. This will help us to see how the simplest version of the process could be progressively improved by evolution.
The principle that underlies this basic adaptive process is surprisingly simple. Living processes search for better adaptation by trying out changes. These changes can be made within the organism during its life, or in its offspring. The usefulness of the changes is tested by evaluating their effects on the organism in which the change is made. So better adaptation is discovered by trying out changes and then testing their effects to see whether or not they improve adaptation. Trial-and-error is at the heart of the basic adaptive process.
Importantly, this change-and-test process can work well even if it does not use any intelligence to decide the changes that are to be tested. The basic process can discover complex adaptations even if the changes that are tested are chosen randomly. It is the testing that sorts through these changes and discovers any that are better. Even if the changes are made randomly, some may be improvements, and these will be discovered when the changes are put to the test. Without any knowledge or insight into what might improve adaptation, a change-and-test process can discover better adaptations.
A number of examples will show how powerful this simple change-and-test process can be, and how widespread it is in living organisms. We will see that it is not only the basis of evolutionary mechanisms that discover adaptations and pass them from generation to generation. It also underpins physiological and other non-evolutionary adaptive systems that adapt organisms or groups of organisms only during their life, and do not produce change across the generations.
The genetic evolutionary mechanism itself is one of the simplest examples. In genetic evolution, changes are generated when an organism produces offspring. The offspring generally vary in a small number of ways from their parents and from each other due to genetic changes. The genetic changes are tested by their effects on the offspring that carry them. If a change causes its carrier to do better, the carrier will produce more offspring, and the numbers of individuals who carry the change will increase in the population. Eventually, all members of the population will carry the change, and it will be established as an adaptation. The genetic evolutionary mechanism will have discovered a better adaptation. When the environment changes, a different change might then do better. Once the new change has spread throughout the population, the mechanism will have discovered a better adaptation to the new conditions.
For an example, consider a hypothetical population of snow hares. The genetic evolutionary mechanism will tend to produce hares that have the thickness and length of fur that is best for the environmental temperatures met by the population. It will do this by trying out offspring with a variety of types of fur. Those with fur that is best suited to the conditions will be the most competitive. But if temperatures change significantly, offspring with different fur will do better. The length and thickness of fur in the population will change as the conditions met by the population change. In this way, the genetic evolutionary mechanism will adapt the type of fur in the population to track changes in environmental temperatures.
In contrast to genetic adaptation, individual organisms adapt physiologically by trying out changes within their bodies during their life. The changes are tested on the basis of whether or not they produce a useful effect within the organism. For example, warm-blooded organisms use such a process to discover adaptations that keep their body temperatures constant despite changes in their environment. The animal tries out changes that influence the amount of heat produced within the organism, and the rate at which this heat is lost to the environment. It might change its metabolic rate, its level of movement and other general activity, the amount of its food intake, its posture, its rate of panting and sweating, the amount of blood flowing to its extremities, and the extent to which its hair or feathers are raised and lowered. The animal does not know in advance what pattern of changes are needed to maintain its temperature at the best level in the face of change in its external environment. This pattern is discovered by trial-and-error, by testing changes against their effects on the animals temperature. In this way, patterns of adaptive change are made within the animal during its life to track changes in external temperatures.
Some animals are also able to adapt to varying external temperatures by trying out changes in behaviours that can affect heat gain or loss. The animal may move in or out of the sun and change its skin colour in the search for changes that will help maintain the desired temperature. Humans may try out different types and amounts of clothing to adapt to changes in temperature. Many animals also use simple change-and-test mechanisms for other adaptive challenges. Most complex multicellular organisms are able to use change-and-test processes to adapt other aspects of their behaviour, and to discover new behaviours that are better at meeting their adaptive goals.
Societies of humans and other animals also adapt using variants of the basic change-and-test process. A colony of honeybees can maintain the temperature of its nursery around 34 degrees C (93 degrees F) despite large changes in temperature outside the hive. This is the best temperature for hatching eggs and rearing young. Bees can increase the temperature of the nursery by clustering more tightly around it, raising the rate at which they metabolise sugars, and by flexing their muscles more often. They can reduce the temperature by fanning their wings to improve the ventilation of the hive, and by increasing the amount of water that is evaporated within the hive. Nothing in the hive knows the particular pattern of these behaviours that will maintain the temperature of the nursery at the ideal. The pattern that will maintain the best temperature is discovered by testing changes in these behaviours against their effects on the temperature in the nursery.
Human economic markets are an example of a process within human societies that uses a basic change-and-test mechanism to achieve adaptation. For example, markets can adapt the level of production of particular goods to the needs and preferences of consumers. If insufficient goods of a certain type are being produced, manufacturers who increase production will be rewarded with higher profitability. This mechanism will increase production even if individual manufacturers are completely unaware that there is an emerging shortage of the product, or have no idea what is causing it. It will work even if manufacturers use only simple trial-and-error to decide their level of production. In this way, an economic market can adapt the level of production of warm clothing to track changes in demand as environmental temperatures vary over a series of winters.
These examples can also help us to see how the basic change-and-test process can be combined with more complex arrangements to improve the ability of the process to discover adaptations. The basic process searches for adaptation by trying out changes. It can discover improvements even if the changes are chosen randomly. But the search will be more efficient if the changes are chosen so that they have a better than random chance of proving adaptive. Change-and-test processes will do better if the changes they try out are non-random and are instead targeted at the types of changes that are likely to prove adaptively useful.
So the genetic evolutionary mechanism will be more effective if the genetic changes that are tested are non-random, and instead are more likely to be adaptive. We will see in the next Chapter that evolution has indeed established genetic systems that produce targeted genetic changes. And the physiological systems that adapt warm-blooded animals to differing external temperatures do not test out random changes within the organism. In high temperatures they test out changes that will cool the animal and reduce its internal production of heat. In low temperatures they test changes targeted at doing the opposite. The same is the case for the changes that are tested within beehives in the search for adaptation to varying external temperatures. They still use trial-and-error to discover the best pattern of changes, but do so far more efficiently by targeting the changes.
In some cases the arrangements that target physiological changes are established by the genetic evolutionary mechanism. Genes that cause physiological systems to target changes will do better than alternative genes that do not. But targeting can also be established by learning within the organism. If an organism discovers by trial-and-error that a behaviour pays off in particular circumstances, it will do better if it can learn to immediately try out this behaviour whenever the circumstances arise again. Learning avoids the repetition of costly trial-and-error. The behaviour of young animals generally includes a high level of trial-and-error until they learn to target their behaviour more accurately at their particular adaptive goals.
The change-and-test process can be targeted even more accurately if the organism is able to form mental representations or models of itself and of its environment, and is able to test possible changes against these models mentally, before trying them out in practice. Instead of using the change-and-test process to try out actual changes in the real world, possible adaptive changes are first tried out mentally.
On this planet, we humans have the most highly developed capacity to search for adaptations using mental processes. In some circumstances our modelling capacity is so effective that it can completely eliminate the need for external trial-and-error. When our mental model of a situation can accurately predict the consequences of our alternative acts, we can mentally design an action that will directly achieve our adaptive goal. We then simply implement the changes that we see will achieve our goal. No external change-and-test process is necessary.
When we set out to solve an adaptive problem, it is these mental modelling processes that we are conscious of using. When we plan how we our going to cook our evening meal, think about how to fix a car engine, or imagine what we might have to do to improve our career prospects, we are using mental representations to come up with behaviours that will achieve our adaptive goals.
In contrast, we are not conscious of the simpler change-and-test processes that are continually adapting our bodies and internal organs to variations in temperature and to changes in the availability and usage of food, water and oxygen. We continually experience our mental processes, but have no experience of the workings of these other adaptive mechanisms in our bodies. When we think of how we might solve an adaptive problem, we tend to think of the mental processes we could use.
As a result, we do not have a good mental feel for how change-and-test processes in other living processes successfully solve complex adaptive problems without the use of mental modelling. We find it hard to see how the genetic evolutionary mechanism, physiological systems in other animals, and the adaptive processes in economic markets and insect societies can discover and establish complex adaptation without using the mental processes we associate with intelligence.
This blind spot in our understanding of adaptive processes has been particularly limiting when we have set out to design and adapt our economic and other social systems. We have a tendency to think that these complex systems can be best designed and adapted by the use of human intelligence. We think that if we collect enough information about the system, we can understand it sufficiently to decide the course of action needed to produce the result we want. However, we can rarely have sufficient information about complex and rapidly changing systems to make them predictable enough for us to adapt them in this way. The experience of centrally-planned economies has made this increasingly clear in recent years. Attempts to use central planning to match production levels to the needs of consumers have been spectacularly unsuccessful.
The alternative is to produce economic and other social systems that include their own adaptive change-and-test processes. An example is the economic market that we briefly considered earlier. A market system uses change-and-test processes to adapt production to match the needs and preferences of consumers. Such a process will be more effective if it can take advantage of the mental capacities of the participants in the system to better target the changes that are tested. But ultimately it is the systemic change-and-test process that adapts the system. And such a process can work even if the mental modelling used by participants to target their behaviour is ineffective. We will return to these issues when we consider in detail the future evolution of human societies.
Armed with this broad understanding of the nature of the basic process that living things use to evolve and adapt, we will trace the evolution of the evolvability of living things on this planet. Over the next four Chapters we will see how the potential benefits of cooperation have driven an impressive sequence of improvements in these abilities, and how this progressive evolution can be expected to continue into the future.
We begin in Chapter 9 by looking at how living processes evolved before genetic systems emerged. We will see how autocatalytic sets could evolve without genes, and how their evolvability could improve. Then we will move on to consider the evolution of the genetic system itself. We will look in detail at the evolution of the evolvability of genetic systems. We will see how natural selection has improved the ability of genetic systems to discover adaptations. Genetic systems produce a pattern of mutations and other genetic changes that is far from random. The pattern is biased toward changes that are more likely to produce useful adaptation.
But genetic systems do not adapt organisms during their life. Genetic systems cannot discover adaptations by trying out genetic changes within individual organisms. New adaptive processes have had to evolve to adapt individuals. Our physiological, emotional and mental adaptive systems are examples of what evolution has produced to fill this vacuum. In Chapter 10 we trace in detail the progressive evolution of the internal mechanisms that adapt and evolve individual organisms during their life.
In humans, the internal adaptive processes are now evolutionary mechanisms in their own rightthrough language, discoveries made by our adaptive processes are passed and accumulated from generation to generation. In Chapters 11 and 12 we will look at how these internal adaptive processes are evolving in humans at present, and how they are likely to continue to evolve in the future. We will see how we must improve our evolvability by developing new psychological skills if we are to contribute to the future evolution of life in the universe.
. Stewart, J. E. (1995) Metaevolution. Journal of Social and Evolutionary Systems 18: 113-147.
. Stewart: Metaevolution. op. cit.; and Stewart, J. E. (1997) Evolutionary Progress. Journal of Social and Evolutionary Systems 20: 335-362.
. What I label here as a change-and-test process is essentially the same as what is called a generate-and-test process in Dennett, D. C. (1995) Darwins Dangerous Idea. New York: Simon and Schuster.
. McFarland, D. (1985) Animal behavior: psychobiology, ethology, and evolution. Menlo Park, California: Benjamin/Cummings Pub. Co.
. See Skinner, B. F. (1953) Science and Human Behaviour. New York: Knopf.
. Seeley, T. D. (1995) The Wisdom of the Hive. Cambridge, MA: Harvard University Press.
. For example, see Haveman, R. H. and K. A. Knopf (1966) The Market System. New York: John Wiley and Sons.
. Skinner: Science and Human Behaviour. op. cit.
. For example, see Popper, K. R. (1972) Objective knowledge - an evolutionary approach. Oxford: Clarendon; Dennett: Darwins Dangerous Idea. op. cit.; and Stewart: Metaevolution. op. cit.