Reviewed by
Rob Stocker
School of Engineering and Information Technology (SEIT), University of New South Wales at the Australian Defence Force Academy
It is not surprising that social science research should focus on the questions that surround the study of ethics and, in particular, the evolution of ethics in human society. That there are many questions unanswered, provides a fertile environment for exploring such a rich domain.
Given that such exploration in the real world is, at best limited, and at worst, difficult or impossible (due to the very constraints that control the ethics of human research), it is also logical that simulation technologies would emerge as likely candidates to apply to this domain. The improving availability of high-end computing power has seen the development of increasingly more sophisticated computer simulations over the past two decades.
Whilst the authors are at pains to categorise the book in terms of appealing to different audiences, the book is most certainly targeted towards those with much more than a basic understanding of scientific process, particularly the experimental chapters (5 and 6).
The book takes a structuralist perspective based on proven past research combined from several areas and applied to novel situations, and therefore makes a useful contribution to the fields of ethics, evolution and simulation.
The authors acknowledge that the book emerges from the combined efforts in two PhD theses and it certainly reads in that light. The structure broadly follows a pattern of literature review, methodology, experimentation, results and conclusions, that is, normal scientific process.
Chapters 1 to 4 (inclusive) form the 'literature review' and clearly establish the domain in which this research is conducted. From a brief discussion of the science of ethics in Chapter 1, where the authors introduce ethics, evolution, simulation and experimental philosophy, they discuss in turn, ethics and evolutionary psychology, simulation as experimentation and evolutionary artificial life in Chapters 2, 3 and 4 respectively.
Examination of Aristotle's virtue ethics, Kant's categorical imperitavism, Socrates' consequentialism, Epicurus' hedonism (egoism) and utilitarianism, identifies act utilitarianism as the best (in this case) underlying theory for developing simulations that can be applied to the evolutionary ethics domain.
Simulations and their application as experimental tools are carefully scrutinized to justify their suitability as an appropriate medium for developing the evolutionary artificial life environments used in the study of evolutionary ethics. The authors clearly make a strong case for the usefulness of simulations as an experimental methodology.
In Chapter 4, the authors focus on computational media and the examination of "the actual world and its actual denizens", using ALife, cellular automata, genetic algorithms and the Iterated Prisoner's Dilemma (IPD) to develop their artificial agent society. They provide an authoritative discussion on the components of the simulation environment including: the "world", time and food; and, agent properties of life span, health and utility, and behaviour. Simulating evolutionary processes describes the production rules, decision trees, mutation and meta-mutation for evolution and agent genotypes. ALife thus provides a "set of simulation methods" that allows for a population of agents to interact, reproduce and exhibit complex cognitive and social behaviours (e.g., decision making) and thus enabling researchers to explore concepts from evolutionary psychology and utilitarian ethics.
Chapters 5 ("Experiments in Evolution") and 6 ("Experiments in Ethics") describe specific experiments where the models explained in Chapter 4 are modified to suit each of the problems addressed.
Commencing with foundation ideas about individual, group, species and kin selection processes and the evolution of aging, Chapter 5 initially reports on four experiments applying the impact of genetic aging, accidental death and disease on a population to explore the evolution of short life spans. The results demonstrate the dependence of aging on diversity, that is, groups with shorter life spans have greater diversity. The chapter continues to explore the evolutionary stability of suicide, and the evolution of parental investment, with some preliminary discussion on the evolution of utility (part of a future work). In a series of experiments, the results demonstrate:
Chapter 6 takes a more application centric view than previous chapters. Specifically it looks at two problems of applied ethics, rape and abortion. It attempts to determine: when these behaviours "can and can't evolve"; what are the "consequences of their presence or absence"; and when present do they "convey utilitarian advantage to the population having them"? Initially reviewing work conducted on cooperation, altruism and altruistic suicide, the authors discuss rape and sexually dimorphic behaviour and identify common issues supported by two controversial theories:
Modifications to the simulation for this series of experiments assume the first of these theories. Here the principal cause of "sexually dimorphic behaviour" is health. Females invest more in the offspring than males and are therefore less healthy and less interested in consensual sex. Consequently males are more likely to rape than females.
Experimental results for the investigation of the evolution of abortion suggest that environmental conditions have impact. Unexpected change in food supply (i.e., negatively) provides abortion with a fitness benefit for the population, particularly where high post-natal investment is required.
In the final chapter, the authors continue to espouse the virtues of modelling complex issues with simulations, albeit recognizing the current limits of technology. Within the overall theme of the book, they open up a "Pandora's box" of potential issues to which simulation technologies may be applied and they identify some key questions for future research for ethicists, with the admonition to "Go forth and simulate better!".
From a practical point of view, the authors provide clear justification of their approach to their investigations, acknowledging limitations and assumptions both in their own ideas and the technologies they apply. They sensibly provide support for their arguments and discussion with appropriate and relevant references. The experimental chapters are quite complex and demand close and careful reading - but the science is strong.
Combining the "old science" of ethics with the "new science" of simulations, Mascaro et al. have taken a "path seldom trod". By selecting challenging and controversial topics in their investigation of the evolution of ethics, they have answered several questions and importantly opened new pathways for future research.
An interesting read, indeed!
Return to Contents
of this issue