Access Electronic Edition
Centre for Policy Modelling, Manchester Metropolitan University, Manchester, UK.
This is a thought provoking book and I feel the author is attempting to say a lot of interesting things - perhaps too many. Overall the book is composed of three kinds of material: a) uncontroversial overviews of evolutionary theories of co-operation; b) attempts at original theories for the evolution of co-operation that are applicable to cultural evolutionary systems in addition to biological systems; c) wild, grandiose and vague generalisations, which start from evolutionary theory but are essentially political ideology.
Given that there are already plenty of accessible books covering the first topic and the last would (in my opinion) best be kept out of a book which is aiming at the "popular science" level, we are left with b) as the unique and valuable contribution the book could make. However there is too little discussion of this topic and what there is has been covered lightly with the all-important detail glossed over. Perhaps this latter criticism is not quite fair since references are made to what would appear to be more technical works and the book is pitched at the general reader - someone who is not familiar with concepts like the "tragedy of the commons" or "reciprocal altruism".
The basic premise of the book - that evolution is "progressive" - appears unconvincing and is not presented as clearly as perhaps it could be. The biggest concern I have is that the political points made could be misinterpreted by the general reader as points which are backed-up by some kind of scientific knowledge - which of course they are not.
The book is structured into four parts comprising several chapters each. Part one presents the general thesis of the book - that evolution is "progressive" and that as humans we should understand this progression and embrace it in order to improve our reproductive survival. Part two concerns the evolution of co-operation among self-interested individuals. Part three examines how the process of evolution itself evolves and Part four examines the implications of the previous chapters for the future of human evolution.
In Part one (comprising chapters 1 and 2) it is argued that the force of evolution will ultimately produce organisms with the psychological capacity to comprehend the progressive direction of evolution and that the very text is a part of that process.
Here is a quote from chapter one:
The organism will also know that only organisms that choose to struggle to develop this psychological capacity are likely to make a significant contribution to the future evolution of life in the universe. Those who choose instead to continue to serve obsolete values and motivations will be irrelevant to life, and face eventual extinction. The organism will know that the choice that faces it is, in an evolutionary sense, a choice to be or not to be.
Later in the chapter the point is hammered home:
Once individuals become aware of the direction of evolution, if they decide to continue to serve the dictates of past evolution they are choosing evolutionary failure, in the full knowledge that they are doing so. Individuals that make such a decision are choosing a life that is meaningless, absurd and ridiculous from an evolutionary perspective, and know that they are making such a choice.
What's wrong with this? For me this all smacks of two things that evolutionary theory is not supposed to be about: teleology (that evolution is directed towards some future goal) combined with a naturalistic fallacy - that because evolution is going in some known direction, behaviour that appears counter to that direction is "wrong" and deserves contempt.
A major triumph of the modern evolutionary synthesis was to remove direction, to show how undirected forces, blind and unintelligent, could produce the order we observe in biological systems. Stewart has attempted to introduce directedness and to make that direction positive and progressive (based on his own meanings of those terms). By re-introducing the notion of a progressive direction, the next step is to identify those systems that are progressing and those that aren't. From this standpoint a naturalist fallacy starts to emerge and this is particularly dangerous when conclusions start to be drawn with respect to human social systems. Two steps from the modern synthesis and we arrive at a kind of prescriptive political agenda, here is a quote from chapter six:
As we will see in detail, our current systems of government fail to meet the evolutionary ideal. Democratic processes are not sufficient to completely align the interests of the governors with the governed. As a result, even highly democratic human societies do not harness the benefits of co-operation in the interests of society as a whole A major challenge for humans in the future is therefore to devise forms of organisation that overcome this limitation, and that are more likely to be successful in evolutionary terms.
More quotes could be given to make the point because such statements are scattered throughout the text. It might have been better to have collected all these kinds of statements into a single chapter at the end of the text and concentrated on the evolutionary theorising first.
In chapter two the notion of "progress" is explored. Human technological progress is introduced as a process that is obviously progressive. It is unclear however in what sense "progressive" is being used here. The general idea appears to be that new technologies are "better" than old ones (new computers are faster and cost less or some such). It is unclear what this concept of "progressive" has got to do with the term introduced in chapter one. The claim is then made that genetic evolution is progressive but that the fossil record might not always make this clear since genetic evolution is "blind" and not very smart. Some reference is made to cultural evolution but it was unclear where to situate this in the direction of the argument. The author seemed to be saying that cultural evolution is progressive but that genetic evolution was less so. However, this was not totally clear. By the end of chapter two the stated aim of the author was to make clear what was meant by "progressive evolution" and to convince the reader that it was indeed a real phenomena. I was left unconvinced that the author achieved either of these aims.
In chapter three it is argued that aggregates of entities can outperform individual entities via specialisation and the exploitation of new emergent properties - essentially, that thewhole is greater than the sum of the parts. Examples are given in biological and human social organisations. The general thrust of the chapter is that co-operation is a good thing since it reduces competition between individual entities. Co-operation is good for survival since the co-operating entities do better by exploiting the benefits.
In chapter four we get an outline of the "tragedy of the commons" and "free rider" problems. Again we get examples from biological systems and human social systems. Here the argument is that such examples offer a challenge to the idea of "progressive evolution". Perhaps here we get a better idea of what is meant by Stewart's "progressive evolution" - increases in co-operation.
In chapter five Hamiltonian kin selection and reciprocal altruism are introduced and the problems with these mechanisms as a general theory of "progressive evolution" (co-operation) are outlined in limited detail. The idea of a leader or manager that will enforce co-operation by punishing free riders and rewarding good guys is advanced as an additional possible co-operation producing mechanism. It is argued that the manager would need to be all-powerful, able to redistribute benefits and punish without restriction. This is interesting, and we are told that in following chapters we will see how such a manager mechanism might evolve.
In chapter six it is argued that a "dominant" that has power over others will ultimately evolve into a manager whose interests are identical to the group it controls. The idea being that the manager will use its power to increase the productive capacity of the group so it can skim off more for itself. Examples are given in a human social context. Here I have problems. What is the reproductive entity in such systems? This is not made clear. If managers are reproducing at much higher rates than "non-dominants" then won't we end up with a whole bunch of managers and no followers? These issues appear to be glossed over. It is also unclear just where the dominants get their absolute power. Indeed, it's hard to figure out here if the author is talking about the reproductive success of individuals or some other cultural entity (such as an institution or organisational structure). No real explanation is given of the mechanisms by which complex social structures might reproduce.
An example of a "manager" system is presented which draws on speculation about the emergence of life based on RNA molecules (managers) and autocatalytic sets of proteins (followers). An idea that emerges from these examples is that all entities that form a grouping (including the manager) have to be reproduced together as a single entity to stop free riding. Do human intuitions reproduce as single entities? Do they reproduce (make copies of themselves) at all? Although basically glossed over, there is some acknowledgement of the problem. Near the end of chapter six the author states:
If enough individuals band together, they have sometimes been able to develop sufficient power to overthrow ... a ruler. Collective action of this sort, and the threat of it, have been very significant in maintaining some alignment of interests between the governor and the governed in human societies as competition declined.
Here we have an interesting concept being proposed - namely, that some kind of internal dynamic within a social system might produce co-operative results. But where is the evolution here? What is replicating and competing? Is this really comparable to the RNA and cell examples we are given? In these examples we have clearly defined replicators in competition with many others. It wasn't clear (to me at least) how the two are related.
In chapter seven an alternative mechanism producing co-operation is introduced - the concept of an internal manager - some reproductive unit that all individuals within a group possess. Examples are given from the social insects in which elaborate mechanisms are employed to ensure all reproduced individuals share the "internal manager" genes. In the context of human societies shared learned norms (such as moral and religious beliefs) are advanced as internal managers. These are not new ideas of course and as a general overview, Stewart does a reasonable job of introducing them. However, I can't help feeling that he gets a little carried away:
Continued repetitions of this process will progressively produce co-operative organisations of wider and wider scale. The potential benefits of co-operation 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 co-operative organisation.
This appears to indicate that evolution will end, that evolution has a "sell by" date, and all entities in the universe will attain the ultimate evolutionary stable strategy. Then, no more change. But there is no real proof given that this will happen. Just that the various co-operation forming process outlined will somehow all come together in the future. But even given that such a "global maxima" does exists, no evidence is given that things can't get stuck in local maxima. Consequently what we have here is a kind of statement of faith that "everything will work out in the end".
In chapter eight the concept of local search is introduced in which individuals may try out various behaviours using smart heuristics. It is argued that market economies harness this "smart co-operation" whereas a planned economy does not.
In chapter nine the concept of directed, rather than completely random, mutation and recombination is presented. Here a substantive argument is made for the hypothesis that the form which mutation and recombination take in a population can itself evolve, producing smarter kinds of evolution than random generate-and-test search. The argument presented is convincing, with up-to-date references to empirical work. However, what is being argued is that past experience produces search heuristics in the form of non-random mutation and recombination. Although this may produce outcomes that look like directed evolution (over some time scale), does this really support the hypothesis that evolution actually has a general direction? Putting aside the teleology, this chapter was a valuable and interesting overview of the "directed mutation/recombination" concept. It would be of interest to any reader who wanted to understand possible mechanisms that produce such phenomena and get pointers to up-to-date detailed studies. In contrast to some of the previous chapters, grand generalisations are not made and claims are qualified.
In chapter ten the idea of organisms developing the ability to mentally model their environment is introduced. Such organisms have an advantage since they have greater ability to predict the results of their actions. Emotions are advanced as genetic hard-wired goals that organisms attempt to satisfy using complex learned behaviour. Cultural evolution via imitation and language is outlined as a superior evolutionary process that often challenges hard-wired goals. The argument then goes on to claim that as new more powerful mental models are constructed by human organisms of their own destiny (i.e. evolutionary models) new kinds of "psychological skills" will be needed in order to overcome hard-wired outdated emotional goals. Here again, the argument goes from a reasonable descriptive account of current evolutionary theories to a prelude to something very prescriptive of human behaviour.
In chapter eleven Stewart now advances a political agenda:
To be more specific, how ready are humans to put the interests of a planetary organisation ahead of those of our nation, ethnic community, and religion? Will we abandon any belief, prejudice and value that might stand in the way of our support for a truly global human society in which all individuals of all races and backgrounds are treated equally? What if this means a reduced standard of living? Will we accept and support the development of a capacity for the planetary society to adapt for the outside/future if this means a lower level of satisfaction of our immediate material, emotional and social needs?
The rest of the chapter advances aspirations that individuals will redistribute wealth and help one another to achieve Stewart's utopian vision. To achieve this, Stewart argues that by "internal modelling" of cognitive processes human beings can overcome the stumbling blocks to achievement of the utopian vision. I could not see how the evolutionary processes previously described justified any of this. Stewart is entitled to his views but they are not backed by any kind of evolutionary theory. In chapter 12, Stewart outlines a distinction between what he calls "Linear Modelling", "Systematic Modelling" and "Evolutionary Modelling". He attempts to apply these categories to individuals as kinds of "psychology".
Part four of the book (comprising chapters 13 to 19) outlines Stewart's manifesto. Summarising conclusions from previous chapters, he argues that democratic market economies are not sufficient to bring about his planetary utopia. His solution is a global government correcting the bad effects of the global market economy. The European Union is given as an example of a step in the right direction. Elective democracy is considered to be dysfunctional, to be replaced by referenda and "vertical markets" where groups buy and sell social policies. It would seem that the book would have been much better structured if all the socio-political interpretations and pronouncements from the previous parts had been put into this final section with an initial cautionary disclaimer that the author was speculating on the social and political implications of the work previously outlined. Personally I was unconvinced by Stewart's manifesto, it seemed naïve though possibly well intentioned. The view that his position is somehow backed up by evolutionary theory needs to be challenged because it is simply not true.
In conclusion, Stewart knows his evolutionary theory and he has some interesting ideas on novel mechanisms that may produce co-operative organisations within an evolutionary framework. He presents references to relevant literature and manages to present much of that literature in a form digestible to a popular audience. Stylistically however, the book contains much repetition, probably an indication that many of the chapters are based on previously published papers. He is committed to computational modelling as a tool to understand evolutionary processes though he only makes references to modelling work and does give details of any models. However, the fundamental thesis of the book is unconvincing and his constant mixing of evolutionary theory and social and political prescriptions (without appropriate caveats or caution) make it potentially a vehicle for the propagation a dangerous untruth. The untruth being that evolutionary theory is a science that indicates one form of social and political organisation over another. This flaw could be rectified if all the social and political prescriptions were moved into the final section along with appropriate caveats. There's nothing wrong with authors expressing political views but they should be clearly delineated from the science.
Return to Contents of this issue
© Copyright Journal of Artificial Societies and Social Simulation, 2002