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 7. Internal Management
Complex cooperative organisations can be organised by a powerful manager that is external to the individuals it manages. Management is obviously external in all the cases we have considered to this point: atoms and small molecules were managed by larger molecules to start the evolution of life; autocatalytic sets were managed by RNA to form early cells; mitochondria were managed within these cells to produce modern eukaryote cells; and human social systems were managed by external rulers and governments to form modern human societies.
But not all complex cooperative organisation includes a separate manager that supports cooperation and controls cheats and free riders. For example, multicellular organisms do not include a separate manager that controls other cells within the organism. Beehives, ant colonies and other insect societies also do not appear to include a dominant individual that uses its power over others to organise cooperation. And before the rise of managed agricultural communities about 10,000 years ago, small bands of humans cooperated in their hunting and other activities without any control by a separate ruler.
How has the barrier to the evolution of cooperation been overcome in these cases? How are cheating, theft and free riding prevented, and how are resources redistributed within the organisation to support cooperation? Has this been achieved without the use of power and control, or is the source of control just less obvious in these cases, compared with where the manager is separate?
We will see that these organisations without separate managers are able to evolve complex cooperation only because they are managed. But the source of the management is not external to the individuals being controlled. It is internal to each of them.
To see how this internal management can operate, consider a tribe of humans that includes a number of individuals who specialise in tasks useful to the tribe. Someone makes arrows and spears, another makes bows, others specialise in hunting, and so on. These specialists survive in the tribe by exchanging what they produce for the necessities that they do not produce. But as we have seen, cheating can undermine this cooperation. If cheating is not prevented, the specialists will not be able to survive as specialists, even though their cooperative division of labour is extremely beneficial to the tribe.
Cheating could be prevented in the tribe in either of two ways: first, cheating would not be profitable if a powerful tribal ruler punished individuals who undermined cooperation by cheating. Second, cooperation would not be disrupted by cheats if the tribe was made up of individuals who each had a genetic predisposition to be trustworthy and honest. Cheating would not occur if each individual carried genes that prevented him from being a cheat. For example, the genes might make the individual feel emotionally repelled by the idea of acting dishonestly. It would not be in the nature of such an individual to cheat.
The same result would be achieved if, instead of a genetic predisposition, all members had a learnt predisposition against cheating. There would be no cheating if all members of the tribe had been inculcated with a very strong belief that they should be trustworthy and not cheat in exchanges. For example, they might have been taught that cheating was fundamentally wrong, not the way for any respectable member of the tribe to behave, and against the wishes of the gods who created the tribe and its world. Norms, moral beliefs and other rules of behaviour that are supported by religious systems are examples of learnt predispositions that would be capable of controlling behaviour in these ways.
To control the members of the tribe successfully, the genetic or moral predispositions would have to be sufficiently strong to stop the individuals from doing what would otherwise be in their self-interest. Even when an individual could see that cheating was clearly in its interests, its predispositions would have to stop it from doing so. They would have to be strong enough to cause individuals to resist temptation successfully. Individuals would have to be left with no choice. Ordinary beliefs that changed as circumstances changed would not be sufficient. Beliefs that could not be changed by the individual such as sacred beliefs would be necessary. Strong genetic or learnt moral predispositions of this kind could control the tribe as completely as a powerful external manager. The tribe and its members would be hard wired to behave cooperatively.
To control all the members of the tribe, the hard-wired predispositions would also have to be contained in each and every member of the tribe. Any individual that escaped the hard wiring would not be restrained from pursuing its individual interests at the expense of the group. It would gain the benefits of being in a tribe of cooperators without having to contribute to the cooperation. It would be free to cheat and free ride, and would advance its own interests by doing so. If many members of the tribe did not contain the hard wiring, cooperation would be undermined, and cheating and free riding would take over.
A further condition would also have to be met if the control was to be maintained in a tribe over time: the hard wiring would have to be produced in all the members of the tribe across the generations; through time, the tribe would have to contain only individuals who included the hard wiring. This would be achieved if the hard wiring was produced in each new individual who was born into the tribe. The normal process of genetic inheritance would tend to do this for genetic predispositions. Hard wiring that was acquired by learning could also meet this condition if it included a predisposition to inculcate others with its beliefs and predispositions. For example, the learnt hard wiring might cause parents to inculcate their children with a set of moral and religious beliefs. One of these beliefs could be that parents must teach the entire set of beliefs to their children. If the inculcation was successful, the children would grow up to teach their children the set of beliefs, including the belief they should repeat the inculcation process. Each generation would inculcate the next with the set of beliefs. Each would teach the next the difference between right and wrong.
A religious cult within modern society is an example of a set of learnt predispositions that usually include a predisposition to inculcate the beliefs in others. The set of predispositions organise the members of the cult to behave in ways that help the predispositions to reproduce themselves in others. The cults that flourish are those that are based on a set of beliefs that are best at organising their own reproduction.
The processes of inheritance and inculcation would go a long way to ensuring that genetic and learnt predispositions were reproduced in all members of the tribe. But they would not be 100 per cent successful. Mutations would arise from time to time in both genes and in moral beliefs. However, such breakdowns in hard wiring could be repaired. The predispositions could be maintained universally in the tribe if those who were not hard wired to cooperate were punished and perhaps expelled from the tribe.
This punishment could be organised by the hard wiring itself. If this were the case, the hard wiring would not just include predispositions to cooperate in various ways and to refrain from cheating. The hard wiring would also include a predisposition to punish and expel members of the tribe who acted as if they did not include the full set of hard-wired predispositions (including the predisposition to punish). So individuals with the hard wiring would cooperate, and they would punish individuals who did not cooperate, or who failed to punish cooperators.
For example, where the hard wiring was genetic, it might cause members of the tribe to be emotionally outraged by the behaviour of non-cooperators and to be impulsively aggressive toward them. Where the hard wiring was learnt, members of the tribe might be inculcated with a strong moral belief that non-cooperators were deviants who deserved the strongest punishment because they had set themselves against all that was valued by the tribe and its gods.
In this way, a tribe could be managed by a set of hard-wired predispositions that could reproduce itself in most of the members of the group, and cause them to expel any remaining members who did not appear to contain the hard wiring. I will refer to such a manager as a distributed internal manager: the management is internal to the individuals being managed, and is distributed in the sense that it is contained in each of the members of the group.
A distributed internal manager is formed of a set of hard-wired predispositions that are reproduced in each of the members of a group of organisms. As we have seen, the predispositions can be genetic, or can be strong moral or other beliefs that are learnt during the life of the organism. As we shall see, internal managers are established by genes in insect societies and in the societies of cells that are multicellular organisms. And they were established by systems of moral and religious beliefs as well as genetic predispositions in the early tribal societies that preceded modern human societies managed by external rulers.
A distributed internal manager clearly has the potential to organise a group cooperatively. It can hard wire individuals so that they participate directly in cooperation, provide resources to maintain group activities, refrain from cheating, theft and free riding, and punish individuals who undermine cooperation. In this way it can cause individuals to act in the interests of the group as a whole, and to treat others in the group as self. An internal manager can organise cooperation directly by hard wiring specific cooperative behaviours in individuals. Or it can do it indirectly by controlling exchanges between individuals so that individuals capture the benefits of their cooperative acts on others. For example, individuals might be hard wired so that they do not cheat in reciprocal cooperative exchanges. This would enable the reciprocal altruism mechanism to operate successfully.
Internal management has the potential to organise cooperation in these ways, but will evolution produce management that does this? Will evolution favour the emergence of distributed internal management that organises a group to act cooperatively? More specifically, will cooperators organised by internal management out-compete non-cooperators within a population or groups of organisms? It turns out that just as evolution favours external managers that support cooperation, it will also favour internal managers that do so. This is because the manager is able to capture any benefits that it produces, and cooperation is capable of producing immense benefits.
There are a number of ways in which an internal manager can use its power to capture the benefits of any cooperation that it organises. As we have seen already, the manager can organise the individuals it controls to exclude non-cooperators, cheats and free riders from their group. If all individuals who do not contain the manager are expelled from a group, the manager will capture all the benefits it organises within the group. This is because each and every member of the group will contain a copy of the manager. So if an individual helps others in the group, it helps copies of the manager. The manager will also capture the harmful effects of any behaviour that it causes within the group. If an individual hurts others in the group, it hurts the copies of the manager that they contain. If the sum total of the effects of an action on others are harmful, the total effects on the manager will be harmful. And if they are beneficial, the manager will benefit overall. As a result, if the manager can cause an individual to act in ways that produce net benefits for members of the group, the manager will benefit, even though the behaviour might not benefit the individual. The manager will capture the benefits of whatever it can do to advance the position of the group. The evolutionary interests of the manager will be aligned with the evolutionary interests of the group as a whole. What is good for the group will be good for the manager.
There are other ways in which a manger can ensure that it captures the benefits of any cooperation that it creates. It can organise the individuals that it controls to direct their cooperation only toward others who contain copies of the manager. This will ensure that only individuals who contain a copy of the manager will benefit from the cooperation. For example, if the manager is a set of norms and moral codes, it will prevent individuals from cooperating with others who do not follow its codes and norms. If the manger is genetic, it will prevent individuals from cooperating with others who are unlikely to contain copies of itself, such as non-relatives. Cooperators organised by such a manager will be able to out-compete non-cooperators within a group or within a population. The manager may begin by controlling only a few individuals, but eventually it will take over the entire group or population.
As with external managers, the best way in which an internal manager can profit from its ability to control a group is by promoting cooperation. And the better a manager is at organising cooperation, the greater the benefits it can capture. If a number of managers are competing within a group, the manager that is best at organising cooperation will do best, provided it can capture the benefits of the cooperation it organises. If a number of groups are competing, the group with the manager that is best at organising cooperation will do best. Both within groups and between groups, evolution will favour managers who use their power to overcome the barrier to the evolution of cooperation. Evolution will produce distributed internal managers that use their power to establish complex cooperative organisations. We have seen already that evolution can progress by producing cooperative organisations managed by external managers. It can also progress by using internal managers to organise cooperation. As we shall see in detail in Part 4 of this book, evolution on this planet has used both internal and external management to progressively produce cooperative organisations of larger and larger scale.
The most successful internal managers will be those that are best at capturing the benefits created by their management. We have seen that some of these benefits will be lost to the manager if it is unable to suppress all cheats and free riders in the organisation. In general, benefits will be lost if any members of the organisation contain a different manager. And it is not just that these other managers will capture some of the benefits of cooperation. It will also be in their interests to compete within the organisation to extract whatever benefits they can get. Competition between different managers within an organisation will seriously undermine cooperation. As a result, evolution will favour managers that are better at suppressing competition from other mangers and at preventing them from taking any of the benefits of cooperation. It is no accident that the systems of moral beliefs that have survived to the present are usually highly intolerant of those with different beliefs.
Distributed internal management that is unable to suppress competition from other managers within the organisation will not be effective at organising cooperation. It will not be able to capture all the benefits of cooperation. An example is the process of genetical kin selection that we considered earlier. Kin selection begins to operate when an individual contains a genetic manager that predisposes it to cooperate with others. The normal process of reproduction will produce related individuals who also contain the same manager. If it is to be successful, the manager must capture the benefits of the cooperation it produces. So evolution will favour genetic managers that organise individuals to direct their cooperation only towards others who contain copies of itself. As we have seen, managers can use relatedness to target their cooperation in this way. But, because relatedness is not a perfect indicator of a shared manager, a manager will not capture all the benefits of the cooperation it organises. Some of the benefits will leak to individuals who do not include the manager. The manager will not find it profitable to organise some forms of cooperation even though it provides net benefits. As a result, the ability of genetical kin selection to organise cooperation is limited.
Genetical managers in some organisms have found better ways to capture the benefits of their management, and are better at establishing cooperation. The genetical managers of white-winged choughs provide an example. White-winged choughs are amongst the most cooperative of birds. All choughs in a group can be involved in feeding the young produced by the group, whether or not they are closely related. But although the members of the group might not be close relations, they are all likely to contain the genetical manager that produces this cooperation. And the manager will therefore capture the benefits the cooperation produces. This is because the manager organises the members of the group to punish and expel those who do not cooperate. Individuals who contain a different manager that causes them to attempt to free ride or cheat will be forced out of the group.
The internal genetic managers of the most complex insect societies have developed mechanisms that are even more effective at ensuring they capture the benefits of their management and suppress internal competition from other managers. The managers organise their society to increase the probability that only individuals that include the manager will contribute to the reproduction of the colony. To the extent they are successful, internal competition will be prevented, and cooperation will not be undermined.
Managers attempt to do this by organising the society so that it can reproduce only through the sons and daughters of the queen who originally founded the colony. As far as possible, all other members of the society will be prevented from producing offspring that can found a new colony. Managers have discovered numerous ways of organising a society to help achieve this. For example, in most ant colonies, queens are predisposed to produce a pheromone that suppresses egg production in worker ants. The workers themselves are hard wired to respond to the pheromone in this way. Queens will also attack any worker ants that develop ovaries, preventing them from reproducing. And in honeybees, workers are predisposed to eat any eggs that are laid by other workers.
If the manager is able to successfully organise the society in this way, any individual who does not contain the manager will have no opportunity to compete with the manager to contribute to the next generation. Destructive competition within the society will be suppressed. Within complex insect societies that include these management controls, there is little competition and conflict. But internal conflict is common in smaller, less complex insect societies, particularly where more than one manager can reproduce and found a new colony. Within these societies, the competing managers will attempt to organise the individuals they control to preferentially direct their cooperation toward the other members of the society who are most likely to include a copy of the manager.
In these examples, evolution has produced arrangements that suppress competition from other individuals and from other managers within the organisation, enabling cooperation to be established. But this suppression of competition amongst alternative managers produces another evolutionary problem. Evolution proceeds by producing and testing new arrangements that may prove to be better than the old. If no new possibilities are tested, evolution stops, and organisations are unable to discover new and better adaptations. Management can evolve and improve only by testing out new forms of management. But if individuals with mutant management arise within an organisation, they can compete with the existing manager, taking resources and undermining cooperation. This is similar to the destructive cooperation that can arise between RNA molecules within proto cells. It will be in the evolutionary interests of mangers to quickly suppress any new managers that arise within the organisation. But this will cause another problemthe suppression of new managers will also suppress evolutionary innovation.
In the case of the tribes of humans that were managed by sets of strongly held beliefs and other behavioural predispositions, this meant that new ideas and innovative beliefs could not be tolerated. Any deviant behaviour that conflicted with the set of moral beliefs that managed the group would be vigorously suppressed. Groups that were organised in this way would therefore have very stable systems of belief, but at the price of a stunted ability to adapt and innovate. The beliefs that managed the group would be very conservative, and disagreement would be strongly punished. As we will see later when we deal with human evolution in detail, this is a central reason why human societies managed by separate rulers have been far more successful in evolutionary terms.
But where distributed internal management has been able to establish complex cooperative organisation, it has done so because it has discovered ways to suppress internal competition without also suppressing innovation. The best example is multicellular organisms, which are organisations of cells managed by distributed internal management. Each cell generally contains a copy of the same genetic material. These genes can control the actions of each cell within the organisation, organising whatever cooperation is beneficial. This internal management has established the extraordinary level of cooperative specialisation and division of labour found within multicellular organisms, including ourselves.
Managers in complex multicellular organisms have discovered two sets of arrangements that combine together to suppress competition without putting a stop to evolutionary innovation. First, managers organise the development of the organism so that only a small group of cells, the germ cells, can produce the eggs or sperm that reproduce the organism. All other cells and their managers are locked out of reproduction. This is achieved early in development when the germ cells are separated out from the other cells of the organism, and are kept physically apart. So the overwhelming majority of the cells of the organism (and their managers) are prevented from competing to get into the next generation. There is no practical way for them to produce eggs or sperm to participate in reproduction. This greatly reduces the possibility of destructive competition between germ cells and other cells within the organism.
The potential for competition between germ cells is also suppressed. In principle, they could compete with each other to form the most eggs and sperm. But the germ cells are kept inactive and prevented from reproducing. They are therefore largely powerless within the organismany mutant genetic manager that arises within the germ cells is unable to produce a group of cells that it could manage to pursue its competitive interests against other germ cells.
The second way in which competition is suppressed is to have the multicellular organism reproduce by passing through a single cell stage formed by the union of egg and sperm. This single cell then divides repeatedly to form a new multi-celled organism. So only a single manager gets into a new organism. This greatly reduces the potential for disruptive competition: the only way new managers can arise in the new organism is through mutations that occur after the bottleneck of the single cell stage. And, as we have seen, competition amongst any fresh mutants is suppressed by separating out germ cells from other cells early in development.
These two sets of arrangements combine to enable new mutant managers to be produced and tested without destructive competition within organisms. Mutant cells and mangers are produced in the germ cells, but cannot compete destructively there. Neither can they compete with each other when a new organism is formed. The single cell bottleneck ensures that only one manager gets into each new organism that is produced. Fresh mutants could arise as the new organism grows, but competition from these is suppressed because they are prevented from reproducing. As a result, destructive competition within the organism is largely prevented. New mutant managers can arise, but the only way they can compete with other managers is through competition between the organisms that they manage. The manager that organises the best organism will win. So multicellular organisms are able to try out new and possibly better managers without unleashing destructive competition within the organism.
The evolution of insect societies and of multicellular organisms illustrate how internal management can control and organise a group of living processes to suppress competition and promote cooperation. It can be as effective as external management. But internal distributed control is not as visible or as obvious as external control. This is not just because it controls individuals from within. It is also because the control is distributed amongst each of the individuals that are being controlled. The operation and evolution of internal management cannot be understood by looking at only the behaviour and the evolutionary fate of just one individual. The behaviour and the evolutionary fate of all of the individuals who contain the manager must be taken into account. An internal manager organises its group as a coordinated whole, and it can be understood only as a whole. For example, it is not possible to understand the evolutionary significance of moral codes or systems of religious beliefs without recognising their group effects.
A group that is managed by distributed internal management is as much a vertical organisation as a group managed by a separate, external manager. It has an additional layer of organisation compared with a group that is not controlled by a manager. This additional layer is the set of predispositions that are reproduced in each of the members of the group through time. These predispositions control and coordinate the behaviour of members of the group. In the examples we have considered, the additional layer can be a set of genetic predispositions or a set of learnt beliefs such as moral codes.
The group cannot be said to include such an additional level of organisation unless a common set of predispositions is reproduced across the members of the group. If each of the individuals in the group has a different set of predispositions, there is no manager that controls and coordinates their behaviour as a group. Each individual will be controlled by its particular set of predispositions, and these will be a distinct level of organisation within the individual. But the group as a whole is not controlled and coordinated unless a common set of predispositions is reproduced across the group. Different sets of genetic or moral predispositions in different individuals do not create a new level of organisation across a group.
* * * * *
This brings us to a point where it is useful to summarise where we have got to in completing the task I set myself in Chapter 2. In that Chapter we identified what I would have to do to show that the evolution of life progresses toward increasing cooperation. How far have we got?
In Chapter 3 we saw that cooperation between living processes can be extremely advantageous in purely evolutionary terms. The benefits include specialisation, division of labour, the discovery and exploitation of new types of adaptation, coordinated adaptation over wider scales, and suppression of destructive competition. Potentially, these benefits will be produced by cooperation between any living processes, wherever they are in the universe. As a result, the benefits of cooperation will not be fully exhausted until all living processes participate in the one cooperative organisation. Until this occurs, the unexhausted benefits will continue to drive evolution that progressively exploits the benefits.
However, we saw in Chapter 4 that evolutionary mechanisms including gene-based natural selection cannot easily exploit the benefits of cooperation. There is a barrier to the evolution of cooperation that applies generally to all living processes, including to self-reproducing molecular processes, cells, multicellular organisms, and societies of organisms. Evolution favours individuals who pursue their own evolutionary interests. In general, individuals will not cooperate unless it pays them to do so. So despite the benefits of cooperation, evolutionary mechanisms cannot easily establish cooperative organisation. If they could, evolution would still have progressed toward increased cooperation between living processes, but it would have done so very quickly.
But in Chapter 5 we found that evolution could exploit the benefits of cooperation where it was able to establish a certain type of complex organisation. We saw that cooperation could evolve in organisations that include a manager who controls the other members of the organisation. Appropriate management could ensure that individuals would benefit in evolutionary terms from their cooperative effects on others, and would be disadvantaged by any harmful effects on others. Management could therefore align the adaptive interests of individuals and the organisation, ensuring that it paid for individuals to adapt cooperatively.
In Chapter 6 and in this Chapter we have seen that evolution will tend to produce managers whose interests are aligned with the interests of the group of individuals they manage. Managers will therefore do better in evolutionary terms when they use their power to support cooperation amongst the members of the group.
Putting these elements together, we can see that wherever life emerges in the universe, the immense potential benefits of cooperation will progressively drive the evolution of cooperative organisations of increasing scale in space and time. Life will begin when large molecules arise that are able to manage atoms and smaller molecules to produce new instances of the large molecules. If these larger molecules cooperate together, they can form organisations that will be even more successful in evolutionary terms. But cooperative organisations of molecules will not evolve unless it is in the evolutionary interests of the larger molecules to cooperate. A powerful manager can make it in the interests of a group of larger molecules to cooperate. The manager can use its power to organise cooperation.
Once a molecule that can act as a manager emerges, evolution will produce cooperative organisations of molecular processes. But the organisations will be small in scale. The scale of organisation that a particular manager is capable of controlling will always be limited. No manager can be all-powerful. Many organisations of limited scale will be formed. Within each managed organisation of molecular processes, the benefits of cooperation will be exploited. But there will be competition between organisations. The potential benefits of cooperation between the members of different organisations will continue to be unexploited. The barrier to the evolution of cooperation will have been overcome within organisations, but not between organisations.
This unexhausted potential for cooperation between organisations (and between the individuals in different organisations) will drive continued progressive evolution toward increasing cooperation. When powerful organisations emerge that can control other organisations, evolution will favour the formation of groups of organisations, each managed by a powerful organisation. The result will be cooperative organisations of organisations. But these larger organisations will still be of limited scale, the potential benefits of cooperation between the larger organisations will not have been exploited, and this unexhausted potential will drive a further repetition of this evolutionary sequence.
Continued repetitions of this process will progressively produce cooperative organisations of wider and wider scale. The potential benefits of cooperation will not be exhausted until all living processes in the universe are united in a single organisation of the largest possible scale. All the matter, energy and living processes of the universe will be managed into a single cooperative organisation. As we shall see in detail in Part 4 of this book, evolution on earth to date has organised molecular processes into small-scale prokaryote cells, prokaryote cells into larger-scale eukaryote cells, eukaryote cells into multicellular organisms, and organisms into societies. It is about to produce a unified cooperative organisation of living processes on the scale of the planet, managed by humans.
So the evolution of life has all the key features of a process that is fundamentally progressive: there are virtually unlimited potentials for improvement through the establishment of wider-scale cooperation between living processes. And these potentials can be exploited only through a step-by-step process in which each step necessarily builds on and improves on the previous step.
But there is another aspect of the progressive evolution toward increasing cooperation that we have not yet considered. We have seen that the formation of organisations in which a manager controls the other members of the organisation will enable cooperation to be exploited within the organisation. The alignment of the interests of the manager and the interests of the organisation as a whole ensure that the pursuit by managers of their own adaptive interests also serves the adaptive interests of the organisation. However, the extent to which the manager and other members of the organisation are actually successful in discovering the forms of cooperation that are best for the organisation depends on their adaptive ability. It depends on how good they are at evolving, how smart they are at searching out the best cooperative arrangements, and how innovative and creative they are at discovering new cooperative adaptations and adapting these to changing conditions.
Their evolvability will also determine how effective they are at discovering how to form cooperative organisations of wider and wider scale. Living processes that evolve by the genetic evolutionary mechanism will be slow to discover how to form cooperative organisations. The genetic mechanism has no foresight or ability to anticipate future evolution. It has no capacity to understand the progressive evolutionary sequence of increasing cooperation that I have outlined, and no ability to use the sequence to guide its evolution. The genetic mechanism will still exploit the potential benefits of cooperation, but only when it blindly stumbles on cooperation that proves successful. In contrast, when living processes arise that develop an understanding of the progressive nature of evolution, evolution can move far more rapidly and directly to produce cooperative organisations of increasing scale. Once organisms know where evolution is headed and develop the ability to use this knowledge to guide their own evolution, progressive evolution will accelerate.
In Part 3 of the book (Chapters 8 to 12 inclusive) we will look at how evolution has progressively improved the adaptive ability and evolvability of living processes on this planet. We will see where humans fit into this evolutionary sequence. Then we will look at what improvements in our adaptive ability are likely to be necessary for continued evolutionary success in the future. We will identify the adaptive abilities we will need as individuals and the abilities our societies will need if we are to contribute significantly to the future evolution of life in the universe. We will see that a significant improvement in our evolvability will come once we can use our knowledge of the direction of evolution to build cooperative human organisations of greater and greater scale.
. Buss, L. W. (1987) The Evolution of Individuality. Princeton: Princeton University Press.
. Wilson, E. O. (1975) Sociobiology: The New Synthesis. Cambridge, MA: Harvard University Press.
Boehm, C. (1993) Egalitarian Behaviour and Reverse Dominance Hierarchy. Current Anthropology
. Stewart, J. E. (1995) Metaevolution. Journal of Social and Evolutionary Systems 18:113-147; and Stewart, J. E. (1997) Evolutionary Progress. Journal of Social and Evolutionary Systems 20: 335-362.
. Boyd, R. and Richerson, P. (1992) Punishment Allows the Evolution of Cooperation (or anything else) in sizable groups. Journal of Ethology and Sociobiology 13: 171-195.
. Stewart, J. E. (1997) Evolutionary transitions and artificial life. Artificial Life 3: 101-120.
. Alexander, R. D. (1987) The biology of moral systems. New York: Aldine de Gruyter.
. Hamilton, W. (1964) The Genetical Evolution of Social Behaviour. Journal of Theoretical Biology.7: 1-52.
. Heinsohn, R. G. (1994) Helping is costly to young birds in cooperatively breeding white-winged choughs. Proc. R. Soc. Lond. B 256: 293-298.
. Stewart: Evolutionary transitions and artificial life. op. cit.
. Holldobler, B. and E. O. Wilson (1990) The Ants. Cambridge, MA: Harvard University Press.
. Heinze, J., Holldobler, B. and C. Peeters (1994) Conflict and cooperation in ant societies. Naturwissenschaften 81: 489-497.
. Ratnieks, F. L. and P. K. Visscher (1989) Worker policing in the honeybee. Nature, Lond. 342: 796-797.
. Heinze, Holldobler, and Peeters: Conflict and cooperation in ant societies. op. cit.
. Buss: The Evolution of Individuality. op. cit.
. Maynard Smith, J., and Szathmary, E. (1995) The Major Transitions in Evolution. Oxford: W. H. Freeman.