J. Doran, Artificial societies and cognitive archaeology, "Archeologia e Calcolatori" 3, 1992, 1232-1233.

At first sight any connection between human belief systems and computers seems remote. But on closer inspection important and useful linkages may be developed. It is quite possible to identify, in a consistent and defensible way, the beliefs that an agent in an artificial society created on a computer holds about its world. Furthermore, it is quite possible to discuss and to experiment with the formation of sets of collective beliefs common to members of a community of agents, and to relate such beliefs to the "truth" in the created world, so that it becomes possible to study the creation and impact on the society and its future of collective misbeliefs. It should be noted that by "belief" I here typically mean a descriptive belief about the environment.

There is a controversial issue here. As just implied, it is natural in computer-based artificial societies to distinguish sharply between what an agent believes, and what is actually there in the computer created world. Discrepancies are apparent. But many social scientists find a sharp distinction between belief and misbelief unconvincing and unacceptable as applied to our own social world, and therefore see here a major argument against the validity of computer based models of society.

Perhaps even more controversial is that we can begin to model affect. Worldwide, there have been several projects of this type and more are ongoing. It seems that it is possible to encompass within artificial societies analogues of certain types of emotional dynamics as well as aspects of indi-vidual rationality. Hence it may prove possible to address religious experience and its impact. I shall not attempt to explore this possibility here, important though it is. Before passing on, however, I should perhaps make quite clear that there is no suggestion that computers, or the agents created within them, actually feel emotions any more than there is real water inside a computer running a mathematical model of the Mediterranean Sea.