Note in the above paragraph that I have not defined "fair" or "corrupt"; it is precisely the purpose of an ethical system to define such terms. So some ethically acceptable behaviors in one society might conceivably be judged unacceptable in another society. But I think most readers from an American or Western European background would understand the above paragraph, and they could point to societies or countries that we would deem "corrupt." This suggests that at least in some societies, ethics is somehow linked to the economy—to ways to make a living. Indeed, Bernard Shaw made that link explicit when he wrote [2], "The love of economy is the root of all virtue." I must confess that, for many reasons, I like Shaw's sentence; e.g., the word "economy" resonants there in all its meanings. But then, are there societies that have created an ethical system without an economic basis? If so, it would have to be a society that had no interest in prolonging its existence, or so it seems to me. Certainly, there are and have been societies that skew the ethical system to favor the economic well-being of selected groups in the society, but the economies of such societies, I submit, would not expand as fast as possible. (Consider, for example, the old Soviet Union.)

Having reasoned this far, I was anxious to see whether Kauffman confronts the role of ethics; he does, but in only two pages. He starts with a few historically-based comments to show that ethics, as a subject of study, is difficult to broach, surrounded by logic traps even for the wary. Kauffman then turns his attention to summarizing attempts to identify the prerequisites to the development of a system of ethics. He notes that ethics requires

1. the ability to act,
2. the ability to recognize alternative actions and choose from among them, and
3. the willingness to accept personal responsibility for the actions one chooses to take.

So ethics becomes possible only when an organism has the ability to act. Consider an E. coli floating in a medium. At some point hunger stimulates an action: the E. coli starts swimming, looking for supper. The E. coli has exercised the ability to act, which, Kauffman maintains, is the first step toward a system of ethics. Ok, fine. We do not believe that E. coli have consciousness, so the action to start looking for supper was not a conscious decision; nevertheless, an action was taken. More properly, the E. coli was reacting to a stimulus, much as a young plant grows towards sunlight. So, item (1) above might be reworded to say that an ability to react to a stimulus is necessarily a prerequisite to developing a system of ethics.

These are not abstract, arcane, philosophical issues; rather, they are assuming greater importance, for, as Kauffman notes, we are in the process of building electromechanical, autonomous agents—not merely conventional robots, but nanoscale devices that will have some measure of self-direction. They will be designed, at the very least, to react to certain stimuli. And there will be a host of ethical issues surrounding such devices: When should they be used? How should they be used? Should they be built at all? Should we try to limit the complexity of such devices? Will there come a time when such devices can devise an ethics of their own? Can we prevent them from developing an ethical system that conflicts with ours? (We can expect that this will require something more elaborate than Asimov's three laws of robotics [3]. For an unnerving prediction of one possible threat that might arise if we fail to establish ethical standards for interacting with nanoscale organisms, see Michael Crichton's novel Prey [4].)

So the motivation may be stronger than it first appears. But going further, I think the three steps, cited above, are incomplete. Ethics, it seems to me, refers to a set of rules that guide how a living thing (or a self-directed autonomous agent) interacts with other living things. For example, human ethics probably starts when people group themselves into a society; a single human living on a deserted isle has no motivation for developing a system of human ethics. Further, I suggest that ethics becomes linked to economics when a society institutes a division of labor, so that each individual no longer has to do all the things necessary to sustain life. Once individuals are expected to contribute to a society according to their (actual or perceived) talents, then ideas of "fairness", "justice", and "corruption" become meaningful. If these conjectures are close to true, then it seems that the point at which nanoscale autonomous agents could start an ethics is when several of them can communicate with one another and collaborate on a common goal.

This implies that "human" ethics—the kinds of ethics that guides relations among humans—generalizes beyond human societies. Consider, for example, the brown hyena, which inhabits the Kalahari desert in southern Africa [5]. Brown hyenas organize themselves into packs and use an impressive repertoire of vocalizations to plan and execute an attack on prey: they communicate to reach a common goal. Further, they exercise communal care for their young in which some females stay in the den to protect and suckle the young, while other females join the hunt: they employ a division of labor. This is a sophisticated set of social behaviors and I conjecture a system of hyenoid ethics operates to make these behaviors work. There must be a set of mutually-accepted rules by which a pack decides who hunts and who stays with the young, who plays which roles in an attack, and who gets what portions of the prey. Moreover, it seems these conjectures are testable.

But ethics is not limited to guiding relations among members of a species, ethical behavior extends to relations between one species and others in an environment. A single human on a deserted isle has no need of human ethics, but that person, to make a living, does need rules for how it interacts with other life in the environment. Consider the Buddhist caste system, in which members of the lowest caste are those, such as fishermen, who take life: the Buddhist ethical system recognizes the "right" of every living thing to try to make its living. Or consider how, over the last generation, American attitudes have changed towards the environment, so that behaviors toward the environment that were once considered ethically neutral are now viewed as unethical. This illustrates that an ethical system is open to change, and one motivation for change arises when the actions we choose lead to undesirable consequences—in particular, when those consequences limit the expansion of possible ways to make a living.

Finally, we come to the point at which I must take issue with Kauffman; he writes,

We scientists find the facts. You citizens across the globe can argue the ethics. (p. 117)

I find this statement troubling. It is sanctimonious—it has the air of Pontius Pilate washing his hands; it is dangerous, because other scientists (especially graduate students and post-docs) can appeal to this sentiment, either naÔvely or maliciously, to justify unethical behavior that might "advance" their science or their careers; and it is itself unethical, because it renounces personal responsibility.

Moreover, it is misleading. I am not convinced that scientists "find the facts," though admittedly many think that's what they do. I think science, like many other areas of human thought, tells stories. However, scientific stories differ from other kinds in that they are testable in the physical world. Stories told by religions are not scientific because they are not testable. This is not a value judgment about religion; religious stories may be valuable, necessary, crucial parts of human existence, perhaps because they are not scientific. But even though scientific stories are testable, those tests rarely establish facts; more often, they eliminate less plausible stories in favor of more plausible ones. Often the tests merely establish bounds, because every test of a scientific story takes place within a context. Every test is specific to a particular situation; none are general to all situations.

In this sense science differs from some areas of mathematics and logic in which absolute proofs are possible, thereby establishing unambiguous facts. Instead, those concepts and relations that are truly scientific are unprovable. They are testable within a context, but there is no guarantee that tomorrow a scientific story might be tested in a new context and found wanting.

Furthermore, there is a cultural element to scientific stories. For example, Newtonian mechanics is a product of Western European culture at a certain time in its history. What were the elements in that culture that made Newton's discoveries possible at that time? And what if those elements had been in place in, say, China, 100 years before Newton? Would the three laws of motion have been discovered then? If so, would they have taken the same form that Newton (an Englishman writing in Latin) gave them? Perhaps not.

Another example. I postulated in §2 of this review that no society can develop a sustainable technology without having discovered the first law of thermodynamics. For example, an alien people on the other side of the Milky Way cannot become technologically advanced without knowing that, in the absence of nuclear reactions, energy is conserved. But I do not think those aliens would necessarily be aware of the concepts that we call "energy" or "thermodynamics" or even "law." I suggest these are human constructs—products of a certain culture during a certain period in human history—and those constructs (never mind the words) are unlikely to be universal. Alien races must necessarily deal with such concepts in alien ways. My postulate is merely that an advanced alien society will have some formulation that someone, knowledgeable in both alien and human sciences, might be able to show is equivalent to what we call the first law of thermodynamics. But, to twist a phrase from Emerson [6], their two is not our two, and their four is not our four.

In other words, I'm suggesting that scientific stories provide interpretations or interpolations of reality; we test them against reality as best we can, though the tests themselves are imbued with our interpretations—our mind sets. But we don't know how accurate our interpretations are; our tests can only show that some interpretations are more realistic than others. At some stage in a sequence of tests we hope to arrive at an interpretation that allows us to engineer reality to add to our ways for making a living, but we remain ignorant of much of the reality that lies between our points of interpolation. And further, other minds, in other cultures, at other stages of intellectual development may apply different kinds of interpolations at different points of contact to reality and thereby attain wholly different, but equally useful, interpretations of reality. N. R. Hanson has written something similar [7]:

Laws of nature tell us nothing about the world, but only about
what procedures to follow in representing the world.

But these criticisms aside, I think the real value of what Kauffman has written is that he is hinting that a science of ethics might be possible. Is it possible that we can construct stories about an ethical system that are testable? Can we use those tests to develop "better" ethical systems? Conceivably.

• Introductory Page of This Review of Investigations
• §1 A Fourth Law of Thermodynamics?
• §2 Expanding Possibilities

(jmh 05 Sep 06) © 2006 by J. M. Haile. All rights reserved.

Literature Cited

[1] Atkins, P. W., The Second Law, Scientific American Library, W. H. Freeman and Co., New York, 1984.

[2] Shaw, G. Bernard, Maxims for Revolutionists, in Man and Superman: A Comedy and a Philosophy, Bretano's, New York, 1905.

[3] Asimov, Isaac, I, Robot, Gnome Press, New York,1950.

[4] Crichton, Michael, Prey, HarperCollins, New York,2002.

[5] Owens, Mark, and Delia Owens, Cry of the Kalahari, Houghton-Mifflin, Boston, 1984.

[6] What Emerson actually wrote was, "Their two is not the real two, their four not the real four," in "Self-Reliance," Ralph Waldo Emerson: Essays and Journals, International Collector's Library, Garden City, NY, 1968, p. 94.

[7] Hanson, N. R., Perception and Discovery: An Introduction to Scientific Inquiry, Freeman, Cooper, San Francisco, 1970, p. 344.